diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21043 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 114.50381679389314,
+  "eval_steps": 500,
+  "global_step": 30000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03816793893129771,
+      "grad_norm": 11.100768089294434,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 1.2659,
+      "step": 10
+    },
+    {
+      "epoch": 0.07633587786259542,
+      "grad_norm": 4.894315242767334,
+      "learning_rate": 1.2666666666666666e-07,
+      "loss": 1.1033,
+      "step": 20
+    },
+    {
+      "epoch": 0.11450381679389313,
+      "grad_norm": 8.250714302062988,
+      "learning_rate": 1.9333333333333337e-07,
+      "loss": 1.0279,
+      "step": 30
+    },
+    {
+      "epoch": 0.15267175572519084,
+      "grad_norm": 4.837605953216553,
+      "learning_rate": 2.6e-07,
+      "loss": 0.9319,
+      "step": 40
+    },
+    {
+      "epoch": 0.19083969465648856,
+      "grad_norm": 6.9533257484436035,
+      "learning_rate": 3.266666666666667e-07,
+      "loss": 1.3479,
+      "step": 50
+    },
+    {
+      "epoch": 0.22900763358778625,
+      "grad_norm": 7.103541851043701,
+      "learning_rate": 3.9333333333333336e-07,
+      "loss": 1.0561,
+      "step": 60
+    },
+    {
+      "epoch": 0.26717557251908397,
+      "grad_norm": 5.55342435836792,
+      "learning_rate": 4.6000000000000004e-07,
+      "loss": 1.0501,
+      "step": 70
+    },
+    {
+      "epoch": 0.3053435114503817,
+      "grad_norm": 6.5388264656066895,
+      "learning_rate": 5.266666666666667e-07,
+      "loss": 1.2378,
+      "step": 80
+    },
+    {
+      "epoch": 0.3435114503816794,
+      "grad_norm": 7.388347148895264,
+      "learning_rate": 5.933333333333334e-07,
+      "loss": 0.8909,
+      "step": 90
+    },
+    {
+      "epoch": 0.3816793893129771,
+      "grad_norm": 3.4540562629699707,
+      "learning_rate": 6.6e-07,
+      "loss": 1.1626,
+      "step": 100
+    },
+    {
+      "epoch": 0.4198473282442748,
+      "grad_norm": 5.60060977935791,
+      "learning_rate": 7.266666666666668e-07,
+      "loss": 1.0488,
+      "step": 110
+    },
+    {
+      "epoch": 0.4580152671755725,
+      "grad_norm": 4.210075378417969,
+      "learning_rate": 7.933333333333335e-07,
+      "loss": 1.0841,
+      "step": 120
+    },
+    {
+      "epoch": 0.4961832061068702,
+      "grad_norm": 10.231063842773438,
+      "learning_rate": 8.6e-07,
+      "loss": 1.0938,
+      "step": 130
+    },
+    {
+      "epoch": 0.5343511450381679,
+      "grad_norm": 5.16438102722168,
+      "learning_rate": 9.266666666666667e-07,
+      "loss": 0.875,
+      "step": 140
+    },
+    {
+      "epoch": 0.5725190839694656,
+      "grad_norm": 1.9672963619232178,
+      "learning_rate": 9.933333333333333e-07,
+      "loss": 0.7903,
+      "step": 150
+    },
+    {
+      "epoch": 0.6106870229007634,
+      "grad_norm": 2.4632840156555176,
+      "learning_rate": 1.06e-06,
+      "loss": 0.7718,
+      "step": 160
+    },
+    {
+      "epoch": 0.648854961832061,
+      "grad_norm": 2.09865140914917,
+      "learning_rate": 1.1266666666666667e-06,
+      "loss": 0.8049,
+      "step": 170
+    },
+    {
+      "epoch": 0.6870229007633588,
+      "grad_norm": 2.6679656505584717,
+      "learning_rate": 1.1933333333333335e-06,
+      "loss": 0.6023,
+      "step": 180
+    },
+    {
+      "epoch": 0.7251908396946565,
+      "grad_norm": 4.250058650970459,
+      "learning_rate": 1.26e-06,
+      "loss": 0.6673,
+      "step": 190
+    },
+    {
+      "epoch": 0.7633587786259542,
+      "grad_norm": 2.9157729148864746,
+      "learning_rate": 1.3266666666666667e-06,
+      "loss": 0.6071,
+      "step": 200
+    },
+    {
+      "epoch": 0.8015267175572519,
+      "grad_norm": 2.76187801361084,
+      "learning_rate": 1.3933333333333335e-06,
+      "loss": 0.6026,
+      "step": 210
+    },
+    {
+      "epoch": 0.8396946564885496,
+      "grad_norm": 1.7407684326171875,
+      "learning_rate": 1.46e-06,
+      "loss": 0.5992,
+      "step": 220
+    },
+    {
+      "epoch": 0.8778625954198473,
+      "grad_norm": 1.8290092945098877,
+      "learning_rate": 1.526666666666667e-06,
+      "loss": 0.4348,
+      "step": 230
+    },
+    {
+      "epoch": 0.916030534351145,
+      "grad_norm": 1.939551591873169,
+      "learning_rate": 1.5933333333333335e-06,
+      "loss": 0.5998,
+      "step": 240
+    },
+    {
+      "epoch": 0.9541984732824428,
+      "grad_norm": 1.6572309732437134,
+      "learning_rate": 1.6600000000000002e-06,
+      "loss": 0.3822,
+      "step": 250
+    },
+    {
+      "epoch": 0.9923664122137404,
+      "grad_norm": 2.4682672023773193,
+      "learning_rate": 1.7266666666666667e-06,
+      "loss": 0.4126,
+      "step": 260
+    },
+    {
+      "epoch": 1.0305343511450382,
+      "grad_norm": 2.106403112411499,
+      "learning_rate": 1.7933333333333337e-06,
+      "loss": 0.3898,
+      "step": 270
+    },
+    {
+      "epoch": 1.0687022900763359,
+      "grad_norm": 1.331116795539856,
+      "learning_rate": 1.8600000000000002e-06,
+      "loss": 0.3547,
+      "step": 280
+    },
+    {
+      "epoch": 1.1068702290076335,
+      "grad_norm": 1.4754210710525513,
+      "learning_rate": 1.926666666666667e-06,
+      "loss": 0.3567,
+      "step": 290
+    },
+    {
+      "epoch": 1.1450381679389312,
+      "grad_norm": 1.8555392026901245,
+      "learning_rate": 1.9933333333333334e-06,
+      "loss": 0.3293,
+      "step": 300
+    },
+    {
+      "epoch": 1.183206106870229,
+      "grad_norm": 1.6697174310684204,
+      "learning_rate": 2.06e-06,
+      "loss": 0.3181,
+      "step": 310
+    },
+    {
+      "epoch": 1.2213740458015268,
+      "grad_norm": 1.0834065675735474,
+      "learning_rate": 2.126666666666667e-06,
+      "loss": 0.2892,
+      "step": 320
+    },
+    {
+      "epoch": 1.2595419847328244,
+      "grad_norm": 1.046123743057251,
+      "learning_rate": 2.1933333333333332e-06,
+      "loss": 0.3114,
+      "step": 330
+    },
+    {
+      "epoch": 1.297709923664122,
+      "grad_norm": 3.9311656951904297,
+      "learning_rate": 2.2600000000000004e-06,
+      "loss": 0.2735,
+      "step": 340
+    },
+    {
+      "epoch": 1.33587786259542,
+      "grad_norm": 2.4286820888519287,
+      "learning_rate": 2.3266666666666667e-06,
+      "loss": 0.2951,
+      "step": 350
+    },
+    {
+      "epoch": 1.3740458015267176,
+      "grad_norm": 1.0940420627593994,
+      "learning_rate": 2.3933333333333334e-06,
+      "loss": 0.2591,
+      "step": 360
+    },
+    {
+      "epoch": 1.4122137404580153,
+      "grad_norm": 1.823198914527893,
+      "learning_rate": 2.46e-06,
+      "loss": 0.2793,
+      "step": 370
+    },
+    {
+      "epoch": 1.450381679389313,
+      "grad_norm": 2.960927724838257,
+      "learning_rate": 2.526666666666667e-06,
+      "loss": 0.2855,
+      "step": 380
+    },
+    {
+      "epoch": 1.4885496183206106,
+      "grad_norm": 1.0646835565567017,
+      "learning_rate": 2.5933333333333336e-06,
+      "loss": 0.2421,
+      "step": 390
+    },
+    {
+      "epoch": 1.5267175572519083,
+      "grad_norm": 1.0445008277893066,
+      "learning_rate": 2.6600000000000004e-06,
+      "loss": 0.2547,
+      "step": 400
+    },
+    {
+      "epoch": 1.5648854961832062,
+      "grad_norm": 1.456775188446045,
+      "learning_rate": 2.726666666666667e-06,
+      "loss": 0.2318,
+      "step": 410
+    },
+    {
+      "epoch": 1.6030534351145038,
+      "grad_norm": 1.3221451044082642,
+      "learning_rate": 2.7933333333333334e-06,
+      "loss": 0.2203,
+      "step": 420
+    },
+    {
+      "epoch": 1.6412213740458015,
+      "grad_norm": 0.8609516024589539,
+      "learning_rate": 2.86e-06,
+      "loss": 0.2129,
+      "step": 430
+    },
+    {
+      "epoch": 1.6793893129770994,
+      "grad_norm": 1.8033744096755981,
+      "learning_rate": 2.9266666666666673e-06,
+      "loss": 0.2442,
+      "step": 440
+    },
+    {
+      "epoch": 1.717557251908397,
+      "grad_norm": 1.3024567365646362,
+      "learning_rate": 2.9933333333333336e-06,
+      "loss": 0.2504,
+      "step": 450
+    },
+    {
+      "epoch": 1.7557251908396947,
+      "grad_norm": 1.1774224042892456,
+      "learning_rate": 3.0600000000000003e-06,
+      "loss": 0.2251,
+      "step": 460
+    },
+    {
+      "epoch": 1.7938931297709924,
+      "grad_norm": 0.9202947020530701,
+      "learning_rate": 3.1266666666666667e-06,
+      "loss": 0.1936,
+      "step": 470
+    },
+    {
+      "epoch": 1.83206106870229,
+      "grad_norm": 1.2168657779693604,
+      "learning_rate": 3.193333333333334e-06,
+      "loss": 0.2638,
+      "step": 480
+    },
+    {
+      "epoch": 1.8702290076335877,
+      "grad_norm": 1.1469149589538574,
+      "learning_rate": 3.2600000000000006e-06,
+      "loss": 0.1834,
+      "step": 490
+    },
+    {
+      "epoch": 1.9083969465648853,
+      "grad_norm": 1.923591136932373,
+      "learning_rate": 3.326666666666667e-06,
+      "loss": 0.1972,
+      "step": 500
+    },
+    {
+      "epoch": 1.9465648854961832,
+      "grad_norm": 0.7359837889671326,
+      "learning_rate": 3.3933333333333336e-06,
+      "loss": 0.2008,
+      "step": 510
+    },
+    {
+      "epoch": 1.984732824427481,
+      "grad_norm": 1.34367036819458,
+      "learning_rate": 3.46e-06,
+      "loss": 0.1836,
+      "step": 520
+    },
+    {
+      "epoch": 2.0229007633587788,
+      "grad_norm": 0.9186052680015564,
+      "learning_rate": 3.526666666666667e-06,
+      "loss": 0.19,
+      "step": 530
+    },
+    {
+      "epoch": 2.0610687022900764,
+      "grad_norm": 1.0746238231658936,
+      "learning_rate": 3.593333333333334e-06,
+      "loss": 0.1821,
+      "step": 540
+    },
+    {
+      "epoch": 2.099236641221374,
+      "grad_norm": 0.6840347051620483,
+      "learning_rate": 3.66e-06,
+      "loss": 0.1709,
+      "step": 550
+    },
+    {
+      "epoch": 2.1374045801526718,
+      "grad_norm": 1.209904670715332,
+      "learning_rate": 3.726666666666667e-06,
+      "loss": 0.1677,
+      "step": 560
+    },
+    {
+      "epoch": 2.1755725190839694,
+      "grad_norm": 1.1672343015670776,
+      "learning_rate": 3.793333333333334e-06,
+      "loss": 0.1811,
+      "step": 570
+    },
+    {
+      "epoch": 2.213740458015267,
+      "grad_norm": 1.155468225479126,
+      "learning_rate": 3.86e-06,
+      "loss": 0.2217,
+      "step": 580
+    },
+    {
+      "epoch": 2.2519083969465647,
+      "grad_norm": 1.3697214126586914,
+      "learning_rate": 3.926666666666667e-06,
+      "loss": 0.2069,
+      "step": 590
+    },
+    {
+      "epoch": 2.2900763358778624,
+      "grad_norm": 1.2069604396820068,
+      "learning_rate": 3.993333333333334e-06,
+      "loss": 0.1824,
+      "step": 600
+    },
+    {
+      "epoch": 2.32824427480916,
+      "grad_norm": 1.0572060346603394,
+      "learning_rate": 4.060000000000001e-06,
+      "loss": 0.2115,
+      "step": 610
+    },
+    {
+      "epoch": 2.366412213740458,
+      "grad_norm": 3.517552137374878,
+      "learning_rate": 4.126666666666667e-06,
+      "loss": 0.1802,
+      "step": 620
+    },
+    {
+      "epoch": 2.404580152671756,
+      "grad_norm": 0.6924391388893127,
+      "learning_rate": 4.1933333333333336e-06,
+      "loss": 0.1771,
+      "step": 630
+    },
+    {
+      "epoch": 2.4427480916030535,
+      "grad_norm": 1.402863621711731,
+      "learning_rate": 4.26e-06,
+      "loss": 0.1774,
+      "step": 640
+    },
+    {
+      "epoch": 2.480916030534351,
+      "grad_norm": 3.4374442100524902,
+      "learning_rate": 4.326666666666667e-06,
+      "loss": 0.1678,
+      "step": 650
+    },
+    {
+      "epoch": 2.519083969465649,
+      "grad_norm": 1.5495679378509521,
+      "learning_rate": 4.393333333333334e-06,
+      "loss": 0.1867,
+      "step": 660
+    },
+    {
+      "epoch": 2.5572519083969465,
+      "grad_norm": 0.8096297383308411,
+      "learning_rate": 4.4600000000000005e-06,
+      "loss": 0.1647,
+      "step": 670
+    },
+    {
+      "epoch": 2.595419847328244,
+      "grad_norm": 1.6879674196243286,
+      "learning_rate": 4.526666666666667e-06,
+      "loss": 0.1838,
+      "step": 680
+    },
+    {
+      "epoch": 2.633587786259542,
+      "grad_norm": 1.4142144918441772,
+      "learning_rate": 4.593333333333333e-06,
+      "loss": 0.15,
+      "step": 690
+    },
+    {
+      "epoch": 2.67175572519084,
+      "grad_norm": 1.1931599378585815,
+      "learning_rate": 4.66e-06,
+      "loss": 0.1637,
+      "step": 700
+    },
+    {
+      "epoch": 2.7099236641221376,
+      "grad_norm": 0.7196221947669983,
+      "learning_rate": 4.7266666666666674e-06,
+      "loss": 0.1746,
+      "step": 710
+    },
+    {
+      "epoch": 2.7480916030534353,
+      "grad_norm": 1.426375150680542,
+      "learning_rate": 4.793333333333334e-06,
+      "loss": 0.1618,
+      "step": 720
+    },
+    {
+      "epoch": 2.786259541984733,
+      "grad_norm": 1.2002971172332764,
+      "learning_rate": 4.86e-06,
+      "loss": 0.1647,
+      "step": 730
+    },
+    {
+      "epoch": 2.8244274809160306,
+      "grad_norm": 1.275342345237732,
+      "learning_rate": 4.926666666666667e-06,
+      "loss": 0.1531,
+      "step": 740
+    },
+    {
+      "epoch": 2.8625954198473282,
+      "grad_norm": 0.6153695583343506,
+      "learning_rate": 4.9933333333333335e-06,
+      "loss": 0.1747,
+      "step": 750
+    },
+    {
+      "epoch": 2.900763358778626,
+      "grad_norm": 0.6474863290786743,
+      "learning_rate": 5.060000000000001e-06,
+      "loss": 0.1724,
+      "step": 760
+    },
+    {
+      "epoch": 2.9389312977099236,
+      "grad_norm": 0.8682901263237,
+      "learning_rate": 5.126666666666668e-06,
+      "loss": 0.1724,
+      "step": 770
+    },
+    {
+      "epoch": 2.9770992366412212,
+      "grad_norm": 0.9209064841270447,
+      "learning_rate": 5.193333333333333e-06,
+      "loss": 0.1791,
+      "step": 780
+    },
+    {
+      "epoch": 3.015267175572519,
+      "grad_norm": 1.5052716732025146,
+      "learning_rate": 5.2600000000000005e-06,
+      "loss": 0.1536,
+      "step": 790
+    },
+    {
+      "epoch": 3.053435114503817,
+      "grad_norm": 0.6915569305419922,
+      "learning_rate": 5.326666666666667e-06,
+      "loss": 0.1644,
+      "step": 800
+    },
+    {
+      "epoch": 3.0916030534351147,
+      "grad_norm": 1.194627046585083,
+      "learning_rate": 5.393333333333334e-06,
+      "loss": 0.1673,
+      "step": 810
+    },
+    {
+      "epoch": 3.1297709923664123,
+      "grad_norm": 1.9529887437820435,
+      "learning_rate": 5.460000000000001e-06,
+      "loss": 0.1541,
+      "step": 820
+    },
+    {
+      "epoch": 3.16793893129771,
+      "grad_norm": 0.9561631679534912,
+      "learning_rate": 5.5266666666666666e-06,
+      "loss": 0.154,
+      "step": 830
+    },
+    {
+      "epoch": 3.2061068702290076,
+      "grad_norm": 0.7585867643356323,
+      "learning_rate": 5.593333333333334e-06,
+      "loss": 0.1443,
+      "step": 840
+    },
+    {
+      "epoch": 3.2442748091603053,
+      "grad_norm": 0.9479978084564209,
+      "learning_rate": 5.66e-06,
+      "loss": 0.1705,
+      "step": 850
+    },
+    {
+      "epoch": 3.282442748091603,
+      "grad_norm": 1.1037328243255615,
+      "learning_rate": 5.726666666666667e-06,
+      "loss": 0.1538,
+      "step": 860
+    },
+    {
+      "epoch": 3.3206106870229006,
+      "grad_norm": 1.2902960777282715,
+      "learning_rate": 5.793333333333334e-06,
+      "loss": 0.1739,
+      "step": 870
+    },
+    {
+      "epoch": 3.3587786259541983,
+      "grad_norm": 0.8667603135108948,
+      "learning_rate": 5.86e-06,
+      "loss": 0.1451,
+      "step": 880
+    },
+    {
+      "epoch": 3.3969465648854964,
+      "grad_norm": 0.6339547038078308,
+      "learning_rate": 5.926666666666667e-06,
+      "loss": 0.1514,
+      "step": 890
+    },
+    {
+      "epoch": 3.435114503816794,
+      "grad_norm": 0.6177077889442444,
+      "learning_rate": 5.993333333333334e-06,
+      "loss": 0.1513,
+      "step": 900
+    },
+    {
+      "epoch": 3.4732824427480917,
+      "grad_norm": 1.273289680480957,
+      "learning_rate": 6.0600000000000004e-06,
+      "loss": 0.1457,
+      "step": 910
+    },
+    {
+      "epoch": 3.5114503816793894,
+      "grad_norm": 0.7146188020706177,
+      "learning_rate": 6.126666666666668e-06,
+      "loss": 0.174,
+      "step": 920
+    },
+    {
+      "epoch": 3.549618320610687,
+      "grad_norm": 0.6311309337615967,
+      "learning_rate": 6.193333333333333e-06,
+      "loss": 0.1493,
+      "step": 930
+    },
+    {
+      "epoch": 3.5877862595419847,
+      "grad_norm": 0.916165292263031,
+      "learning_rate": 6.26e-06,
+      "loss": 0.1552,
+      "step": 940
+    },
+    {
+      "epoch": 3.6259541984732824,
+      "grad_norm": 1.2083804607391357,
+      "learning_rate": 6.326666666666667e-06,
+      "loss": 0.1533,
+      "step": 950
+    },
+    {
+      "epoch": 3.66412213740458,
+      "grad_norm": 0.7804637551307678,
+      "learning_rate": 6.393333333333334e-06,
+      "loss": 0.1527,
+      "step": 960
+    },
+    {
+      "epoch": 3.7022900763358777,
+      "grad_norm": 2.1257057189941406,
+      "learning_rate": 6.460000000000001e-06,
+      "loss": 0.1838,
+      "step": 970
+    },
+    {
+      "epoch": 3.7404580152671754,
+      "grad_norm": 0.7072081565856934,
+      "learning_rate": 6.526666666666666e-06,
+      "loss": 0.1574,
+      "step": 980
+    },
+    {
+      "epoch": 3.778625954198473,
+      "grad_norm": 1.387868046760559,
+      "learning_rate": 6.5933333333333335e-06,
+      "loss": 0.1409,
+      "step": 990
+    },
+    {
+      "epoch": 3.816793893129771,
+      "grad_norm": 2.3242297172546387,
+      "learning_rate": 6.660000000000001e-06,
+      "loss": 0.1384,
+      "step": 1000
+    },
+    {
+      "epoch": 3.854961832061069,
+      "grad_norm": 0.578836977481842,
+      "learning_rate": 6.726666666666667e-06,
+      "loss": 0.151,
+      "step": 1010
+    },
+    {
+      "epoch": 3.8931297709923665,
+      "grad_norm": 0.8613945841789246,
+      "learning_rate": 6.793333333333334e-06,
+      "loss": 0.1561,
+      "step": 1020
+    },
+    {
+      "epoch": 3.931297709923664,
+      "grad_norm": 0.8455677628517151,
+      "learning_rate": 6.860000000000001e-06,
+      "loss": 0.1561,
+      "step": 1030
+    },
+    {
+      "epoch": 3.969465648854962,
+      "grad_norm": 2.132455825805664,
+      "learning_rate": 6.926666666666667e-06,
+      "loss": 0.1589,
+      "step": 1040
+    },
+    {
+      "epoch": 4.00763358778626,
+      "grad_norm": 2.1734399795532227,
+      "learning_rate": 6.993333333333334e-06,
+      "loss": 0.1643,
+      "step": 1050
+    },
+    {
+      "epoch": 4.0458015267175576,
+      "grad_norm": 0.7833858132362366,
+      "learning_rate": 7.06e-06,
+      "loss": 0.1425,
+      "step": 1060
+    },
+    {
+      "epoch": 4.083969465648855,
+      "grad_norm": 1.160144567489624,
+      "learning_rate": 7.126666666666667e-06,
+      "loss": 0.1446,
+      "step": 1070
+    },
+    {
+      "epoch": 4.122137404580153,
+      "grad_norm": 0.5361282825469971,
+      "learning_rate": 7.1933333333333345e-06,
+      "loss": 0.13,
+      "step": 1080
+    },
+    {
+      "epoch": 4.1603053435114505,
+      "grad_norm": 1.3808670043945312,
+      "learning_rate": 7.260000000000001e-06,
+      "loss": 0.1692,
+      "step": 1090
+    },
+    {
+      "epoch": 4.198473282442748,
+      "grad_norm": 1.0317463874816895,
+      "learning_rate": 7.326666666666667e-06,
+      "loss": 0.1346,
+      "step": 1100
+    },
+    {
+      "epoch": 4.236641221374046,
+      "grad_norm": 0.7219449877738953,
+      "learning_rate": 7.393333333333333e-06,
+      "loss": 0.1508,
+      "step": 1110
+    },
+    {
+      "epoch": 4.2748091603053435,
+      "grad_norm": 4.125285625457764,
+      "learning_rate": 7.4600000000000006e-06,
+      "loss": 0.1658,
+      "step": 1120
+    },
+    {
+      "epoch": 4.312977099236641,
+      "grad_norm": 1.0689539909362793,
+      "learning_rate": 7.526666666666668e-06,
+      "loss": 0.1459,
+      "step": 1130
+    },
+    {
+      "epoch": 4.351145038167939,
+      "grad_norm": 0.8281168937683105,
+      "learning_rate": 7.593333333333334e-06,
+      "loss": 0.1327,
+      "step": 1140
+    },
+    {
+      "epoch": 4.3893129770992365,
+      "grad_norm": 0.4322452247142792,
+      "learning_rate": 7.660000000000001e-06,
+      "loss": 0.1469,
+      "step": 1150
+    },
+    {
+      "epoch": 4.427480916030534,
+      "grad_norm": 0.7191686630249023,
+      "learning_rate": 7.726666666666667e-06,
+      "loss": 0.1352,
+      "step": 1160
+    },
+    {
+      "epoch": 4.465648854961832,
+      "grad_norm": 0.676465630531311,
+      "learning_rate": 7.793333333333334e-06,
+      "loss": 0.1336,
+      "step": 1170
+    },
+    {
+      "epoch": 4.5038167938931295,
+      "grad_norm": 0.7372109293937683,
+      "learning_rate": 7.860000000000001e-06,
+      "loss": 0.1483,
+      "step": 1180
+    },
+    {
+      "epoch": 4.541984732824427,
+      "grad_norm": 1.089241862297058,
+      "learning_rate": 7.926666666666666e-06,
+      "loss": 0.142,
+      "step": 1190
+    },
+    {
+      "epoch": 4.580152671755725,
+      "grad_norm": 0.5060117840766907,
+      "learning_rate": 7.993333333333334e-06,
+      "loss": 0.1542,
+      "step": 1200
+    },
+    {
+      "epoch": 4.618320610687023,
+      "grad_norm": 0.5482001900672913,
+      "learning_rate": 8.06e-06,
+      "loss": 0.1326,
+      "step": 1210
+    },
+    {
+      "epoch": 4.65648854961832,
+      "grad_norm": 0.677850604057312,
+      "learning_rate": 8.126666666666668e-06,
+      "loss": 0.1377,
+      "step": 1220
+    },
+    {
+      "epoch": 4.694656488549619,
+      "grad_norm": 0.4879799783229828,
+      "learning_rate": 8.193333333333335e-06,
+      "loss": 0.1323,
+      "step": 1230
+    },
+    {
+      "epoch": 4.732824427480916,
+      "grad_norm": 0.8837642073631287,
+      "learning_rate": 8.26e-06,
+      "loss": 0.134,
+      "step": 1240
+    },
+    {
+      "epoch": 4.770992366412214,
+      "grad_norm": 0.6340999007225037,
+      "learning_rate": 8.326666666666668e-06,
+      "loss": 0.1267,
+      "step": 1250
+    },
+    {
+      "epoch": 4.809160305343512,
+      "grad_norm": 1.3912076950073242,
+      "learning_rate": 8.393333333333335e-06,
+      "loss": 0.1483,
+      "step": 1260
+    },
+    {
+      "epoch": 4.847328244274809,
+      "grad_norm": 0.44081875681877136,
+      "learning_rate": 8.46e-06,
+      "loss": 0.1514,
+      "step": 1270
+    },
+    {
+      "epoch": 4.885496183206107,
+      "grad_norm": 1.8900103569030762,
+      "learning_rate": 8.526666666666667e-06,
+      "loss": 0.1492,
+      "step": 1280
+    },
+    {
+      "epoch": 4.923664122137405,
+      "grad_norm": 1.0593299865722656,
+      "learning_rate": 8.593333333333333e-06,
+      "loss": 0.1339,
+      "step": 1290
+    },
+    {
+      "epoch": 4.961832061068702,
+      "grad_norm": 0.8677412271499634,
+      "learning_rate": 8.66e-06,
+      "loss": 0.1383,
+      "step": 1300
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.49994489550590515,
+      "learning_rate": 8.726666666666667e-06,
+      "loss": 0.1562,
+      "step": 1310
+    },
+    {
+      "epoch": 5.038167938931298,
+      "grad_norm": 0.9444197416305542,
+      "learning_rate": 8.793333333333334e-06,
+      "loss": 0.1584,
+      "step": 1320
+    },
+    {
+      "epoch": 5.076335877862595,
+      "grad_norm": 0.6566011309623718,
+      "learning_rate": 8.860000000000002e-06,
+      "loss": 0.1259,
+      "step": 1330
+    },
+    {
+      "epoch": 5.114503816793893,
+      "grad_norm": 0.574004590511322,
+      "learning_rate": 8.926666666666669e-06,
+      "loss": 0.1242,
+      "step": 1340
+    },
+    {
+      "epoch": 5.152671755725191,
+      "grad_norm": 0.4502525329589844,
+      "learning_rate": 8.993333333333334e-06,
+      "loss": 0.126,
+      "step": 1350
+    },
+    {
+      "epoch": 5.190839694656488,
+      "grad_norm": 0.6264625787734985,
+      "learning_rate": 9.060000000000001e-06,
+      "loss": 0.1458,
+      "step": 1360
+    },
+    {
+      "epoch": 5.229007633587786,
+      "grad_norm": 0.6364710927009583,
+      "learning_rate": 9.126666666666667e-06,
+      "loss": 0.1417,
+      "step": 1370
+    },
+    {
+      "epoch": 5.267175572519084,
+      "grad_norm": 0.8996373414993286,
+      "learning_rate": 9.193333333333334e-06,
+      "loss": 0.151,
+      "step": 1380
+    },
+    {
+      "epoch": 5.305343511450381,
+      "grad_norm": 1.3520936965942383,
+      "learning_rate": 9.260000000000001e-06,
+      "loss": 0.1447,
+      "step": 1390
+    },
+    {
+      "epoch": 5.34351145038168,
+      "grad_norm": 0.7287914752960205,
+      "learning_rate": 9.326666666666667e-06,
+      "loss": 0.1417,
+      "step": 1400
+    },
+    {
+      "epoch": 5.3816793893129775,
+      "grad_norm": 1.365167498588562,
+      "learning_rate": 9.393333333333334e-06,
+      "loss": 0.137,
+      "step": 1410
+    },
+    {
+      "epoch": 5.419847328244275,
+      "grad_norm": 0.889162540435791,
+      "learning_rate": 9.460000000000001e-06,
+      "loss": 0.1686,
+      "step": 1420
+    },
+    {
+      "epoch": 5.458015267175573,
+      "grad_norm": 1.228797197341919,
+      "learning_rate": 9.526666666666668e-06,
+      "loss": 0.1359,
+      "step": 1430
+    },
+    {
+      "epoch": 5.4961832061068705,
+      "grad_norm": 0.6598213911056519,
+      "learning_rate": 9.593333333333335e-06,
+      "loss": 0.135,
+      "step": 1440
+    },
+    {
+      "epoch": 5.534351145038168,
+      "grad_norm": 0.6431540846824646,
+      "learning_rate": 9.66e-06,
+      "loss": 0.1381,
+      "step": 1450
+    },
+    {
+      "epoch": 5.572519083969466,
+      "grad_norm": 0.731825053691864,
+      "learning_rate": 9.726666666666668e-06,
+      "loss": 0.1493,
+      "step": 1460
+    },
+    {
+      "epoch": 5.6106870229007635,
+      "grad_norm": 0.5904266238212585,
+      "learning_rate": 9.793333333333333e-06,
+      "loss": 0.1279,
+      "step": 1470
+    },
+    {
+      "epoch": 5.648854961832061,
+      "grad_norm": 1.9022412300109863,
+      "learning_rate": 9.86e-06,
+      "loss": 0.1275,
+      "step": 1480
+    },
+    {
+      "epoch": 5.687022900763359,
+      "grad_norm": 0.5651864409446716,
+      "learning_rate": 9.926666666666668e-06,
+      "loss": 0.1326,
+      "step": 1490
+    },
+    {
+      "epoch": 5.7251908396946565,
+      "grad_norm": 0.8466259241104126,
+      "learning_rate": 9.993333333333333e-06,
+      "loss": 0.141,
+      "step": 1500
+    },
+    {
+      "epoch": 5.763358778625954,
+      "grad_norm": 0.9748515486717224,
+      "learning_rate": 9.999997539434007e-06,
+      "loss": 0.1293,
+      "step": 1510
+    },
+    {
+      "epoch": 5.801526717557252,
+      "grad_norm": 0.6662544012069702,
+      "learning_rate": 9.999989033776898e-06,
+      "loss": 0.1321,
+      "step": 1520
+    },
+    {
+      "epoch": 5.8396946564885495,
+      "grad_norm": 0.4294588267803192,
+      "learning_rate": 9.999974452661642e-06,
+      "loss": 0.1387,
+      "step": 1530
+    },
+    {
+      "epoch": 5.877862595419847,
+      "grad_norm": 1.3427824974060059,
+      "learning_rate": 9.99995379610596e-06,
+      "loss": 0.144,
+      "step": 1540
+    },
+    {
+      "epoch": 5.916030534351145,
+      "grad_norm": 0.6164063215255737,
+      "learning_rate": 9.99992706413495e-06,
+      "loss": 0.1639,
+      "step": 1550
+    },
+    {
+      "epoch": 5.9541984732824424,
+      "grad_norm": 0.511205792427063,
+      "learning_rate": 9.999894256781095e-06,
+      "loss": 0.1401,
+      "step": 1560
+    },
+    {
+      "epoch": 5.99236641221374,
+      "grad_norm": 0.5297313928604126,
+      "learning_rate": 9.99985537408426e-06,
+      "loss": 0.1284,
+      "step": 1570
+    },
+    {
+      "epoch": 6.030534351145038,
+      "grad_norm": 0.4822237491607666,
+      "learning_rate": 9.999810416091689e-06,
+      "loss": 0.1251,
+      "step": 1580
+    },
+    {
+      "epoch": 6.068702290076336,
+      "grad_norm": 0.49136868119239807,
+      "learning_rate": 9.99975938285801e-06,
+      "loss": 0.135,
+      "step": 1590
+    },
+    {
+      "epoch": 6.106870229007634,
+      "grad_norm": 0.5754089951515198,
+      "learning_rate": 9.999702274445235e-06,
+      "loss": 0.1154,
+      "step": 1600
+    },
+    {
+      "epoch": 6.145038167938932,
+      "grad_norm": 0.5563395619392395,
+      "learning_rate": 9.999639090922758e-06,
+      "loss": 0.1229,
+      "step": 1610
+    },
+    {
+      "epoch": 6.183206106870229,
+      "grad_norm": 0.6620687246322632,
+      "learning_rate": 9.999569832367346e-06,
+      "loss": 0.1239,
+      "step": 1620
+    },
+    {
+      "epoch": 6.221374045801527,
+      "grad_norm": 0.7389596700668335,
+      "learning_rate": 9.999494498863163e-06,
+      "loss": 0.1308,
+      "step": 1630
+    },
+    {
+      "epoch": 6.259541984732825,
+      "grad_norm": 1.2152165174484253,
+      "learning_rate": 9.99941309050174e-06,
+      "loss": 0.1382,
+      "step": 1640
+    },
+    {
+      "epoch": 6.297709923664122,
+      "grad_norm": 0.7054020762443542,
+      "learning_rate": 9.999325607382e-06,
+      "loss": 0.1657,
+      "step": 1650
+    },
+    {
+      "epoch": 6.33587786259542,
+      "grad_norm": 0.48537522554397583,
+      "learning_rate": 9.999232049610239e-06,
+      "loss": 0.1295,
+      "step": 1660
+    },
+    {
+      "epoch": 6.374045801526718,
+      "grad_norm": 0.9919182062149048,
+      "learning_rate": 9.999132417300141e-06,
+      "loss": 0.1454,
+      "step": 1670
+    },
+    {
+      "epoch": 6.412213740458015,
+      "grad_norm": 0.5933136343955994,
+      "learning_rate": 9.99902671057277e-06,
+      "loss": 0.1393,
+      "step": 1680
+    },
+    {
+      "epoch": 6.450381679389313,
+      "grad_norm": 1.4632704257965088,
+      "learning_rate": 9.998914929556569e-06,
+      "loss": 0.1269,
+      "step": 1690
+    },
+    {
+      "epoch": 6.488549618320611,
+      "grad_norm": 0.8186988234519958,
+      "learning_rate": 9.99879707438736e-06,
+      "loss": 0.1328,
+      "step": 1700
+    },
+    {
+      "epoch": 6.526717557251908,
+      "grad_norm": 0.7318149209022522,
+      "learning_rate": 9.998673145208351e-06,
+      "loss": 0.1278,
+      "step": 1710
+    },
+    {
+      "epoch": 6.564885496183206,
+      "grad_norm": 0.6014979481697083,
+      "learning_rate": 9.998543142170127e-06,
+      "loss": 0.1331,
+      "step": 1720
+    },
+    {
+      "epoch": 6.603053435114504,
+      "grad_norm": 0.5737615823745728,
+      "learning_rate": 9.99840706543065e-06,
+      "loss": 0.1369,
+      "step": 1730
+    },
+    {
+      "epoch": 6.641221374045801,
+      "grad_norm": 1.0225228071212769,
+      "learning_rate": 9.998264915155274e-06,
+      "loss": 0.1429,
+      "step": 1740
+    },
+    {
+      "epoch": 6.679389312977099,
+      "grad_norm": 0.7193993330001831,
+      "learning_rate": 9.998116691516718e-06,
+      "loss": 0.1235,
+      "step": 1750
+    },
+    {
+      "epoch": 6.717557251908397,
+      "grad_norm": 2.232433795928955,
+      "learning_rate": 9.997962394695091e-06,
+      "loss": 0.1266,
+      "step": 1760
+    },
+    {
+      "epoch": 6.755725190839694,
+      "grad_norm": 0.7240338325500488,
+      "learning_rate": 9.997802024877876e-06,
+      "loss": 0.1358,
+      "step": 1770
+    },
+    {
+      "epoch": 6.793893129770993,
+      "grad_norm": 0.48704954981803894,
+      "learning_rate": 9.997635582259941e-06,
+      "loss": 0.1292,
+      "step": 1780
+    },
+    {
+      "epoch": 6.8320610687022905,
+      "grad_norm": 0.4361136257648468,
+      "learning_rate": 9.997463067043526e-06,
+      "loss": 0.1282,
+      "step": 1790
+    },
+    {
+      "epoch": 6.870229007633588,
+      "grad_norm": 0.8777408003807068,
+      "learning_rate": 9.997284479438254e-06,
+      "loss": 0.1289,
+      "step": 1800
+    },
+    {
+      "epoch": 6.908396946564886,
+      "grad_norm": 0.586004912853241,
+      "learning_rate": 9.997099819661127e-06,
+      "loss": 0.1308,
+      "step": 1810
+    },
+    {
+      "epoch": 6.9465648854961835,
+      "grad_norm": 0.34389933943748474,
+      "learning_rate": 9.996909087936523e-06,
+      "loss": 0.1323,
+      "step": 1820
+    },
+    {
+      "epoch": 6.984732824427481,
+      "grad_norm": 0.6134178638458252,
+      "learning_rate": 9.996712284496202e-06,
+      "loss": 0.1215,
+      "step": 1830
+    },
+    {
+      "epoch": 7.022900763358779,
+      "grad_norm": 0.5625924468040466,
+      "learning_rate": 9.996509409579293e-06,
+      "loss": 0.1316,
+      "step": 1840
+    },
+    {
+      "epoch": 7.061068702290076,
+      "grad_norm": 0.6711544394493103,
+      "learning_rate": 9.996300463432313e-06,
+      "loss": 0.1188,
+      "step": 1850
+    },
+    {
+      "epoch": 7.099236641221374,
+      "grad_norm": 0.46400946378707886,
+      "learning_rate": 9.996085446309149e-06,
+      "loss": 0.1356,
+      "step": 1860
+    },
+    {
+      "epoch": 7.137404580152672,
+      "grad_norm": 0.6120234727859497,
+      "learning_rate": 9.995864358471067e-06,
+      "loss": 0.1551,
+      "step": 1870
+    },
+    {
+      "epoch": 7.175572519083969,
+      "grad_norm": 0.8439425230026245,
+      "learning_rate": 9.99563720018671e-06,
+      "loss": 0.1382,
+      "step": 1880
+    },
+    {
+      "epoch": 7.213740458015267,
+      "grad_norm": 0.7507050037384033,
+      "learning_rate": 9.995403971732098e-06,
+      "loss": 0.1292,
+      "step": 1890
+    },
+    {
+      "epoch": 7.251908396946565,
+      "grad_norm": 1.3689290285110474,
+      "learning_rate": 9.995164673390624e-06,
+      "loss": 0.1244,
+      "step": 1900
+    },
+    {
+      "epoch": 7.290076335877862,
+      "grad_norm": 0.8372470140457153,
+      "learning_rate": 9.994919305453059e-06,
+      "loss": 0.1318,
+      "step": 1910
+    },
+    {
+      "epoch": 7.32824427480916,
+      "grad_norm": 0.5124345421791077,
+      "learning_rate": 9.994667868217549e-06,
+      "loss": 0.1198,
+      "step": 1920
+    },
+    {
+      "epoch": 7.366412213740458,
+      "grad_norm": 0.48633715510368347,
+      "learning_rate": 9.99441036198961e-06,
+      "loss": 0.1364,
+      "step": 1930
+    },
+    {
+      "epoch": 7.404580152671755,
+      "grad_norm": 0.800594687461853,
+      "learning_rate": 9.994146787082141e-06,
+      "loss": 0.1246,
+      "step": 1940
+    },
+    {
+      "epoch": 7.442748091603053,
+      "grad_norm": 0.5229592323303223,
+      "learning_rate": 9.993877143815406e-06,
+      "loss": 0.1236,
+      "step": 1950
+    },
+    {
+      "epoch": 7.480916030534351,
+      "grad_norm": 0.49684378504753113,
+      "learning_rate": 9.993601432517052e-06,
+      "loss": 0.1209,
+      "step": 1960
+    },
+    {
+      "epoch": 7.519083969465649,
+      "grad_norm": 0.4344727396965027,
+      "learning_rate": 9.99331965352209e-06,
+      "loss": 0.119,
+      "step": 1970
+    },
+    {
+      "epoch": 7.557251908396947,
+      "grad_norm": 1.0743569135665894,
+      "learning_rate": 9.993031807172912e-06,
+      "loss": 0.1304,
+      "step": 1980
+    },
+    {
+      "epoch": 7.595419847328245,
+      "grad_norm": 0.44631901383399963,
+      "learning_rate": 9.992737893819274e-06,
+      "loss": 0.1309,
+      "step": 1990
+    },
+    {
+      "epoch": 7.633587786259542,
+      "grad_norm": 0.8176268935203552,
+      "learning_rate": 9.992437913818311e-06,
+      "loss": 0.1245,
+      "step": 2000
+    },
+    {
+      "epoch": 7.67175572519084,
+      "grad_norm": 0.6087300181388855,
+      "learning_rate": 9.992131867534526e-06,
+      "loss": 0.1262,
+      "step": 2010
+    },
+    {
+      "epoch": 7.709923664122138,
+      "grad_norm": 0.5834507346153259,
+      "learning_rate": 9.991819755339796e-06,
+      "loss": 0.1215,
+      "step": 2020
+    },
+    {
+      "epoch": 7.748091603053435,
+      "grad_norm": 0.4410012364387512,
+      "learning_rate": 9.991501577613365e-06,
+      "loss": 0.1271,
+      "step": 2030
+    },
+    {
+      "epoch": 7.786259541984733,
+      "grad_norm": 0.8793311715126038,
+      "learning_rate": 9.991177334741851e-06,
+      "loss": 0.1199,
+      "step": 2040
+    },
+    {
+      "epoch": 7.824427480916031,
+      "grad_norm": 0.4199819564819336,
+      "learning_rate": 9.990847027119235e-06,
+      "loss": 0.1229,
+      "step": 2050
+    },
+    {
+      "epoch": 7.862595419847328,
+      "grad_norm": 0.5443435311317444,
+      "learning_rate": 9.990510655146877e-06,
+      "loss": 0.1155,
+      "step": 2060
+    },
+    {
+      "epoch": 7.900763358778626,
+      "grad_norm": 1.0193560123443604,
+      "learning_rate": 9.990168219233496e-06,
+      "loss": 0.1295,
+      "step": 2070
+    },
+    {
+      "epoch": 7.938931297709924,
+      "grad_norm": 0.5311276912689209,
+      "learning_rate": 9.989819719795189e-06,
+      "loss": 0.1274,
+      "step": 2080
+    },
+    {
+      "epoch": 7.977099236641221,
+      "grad_norm": 0.5494500994682312,
+      "learning_rate": 9.989465157255413e-06,
+      "loss": 0.1356,
+      "step": 2090
+    },
+    {
+      "epoch": 8.01526717557252,
+      "grad_norm": 0.4246234595775604,
+      "learning_rate": 9.989104532044994e-06,
+      "loss": 0.1213,
+      "step": 2100
+    },
+    {
+      "epoch": 8.053435114503817,
+      "grad_norm": 0.5394090414047241,
+      "learning_rate": 9.988737844602128e-06,
+      "loss": 0.1284,
+      "step": 2110
+    },
+    {
+      "epoch": 8.091603053435115,
+      "grad_norm": 0.6479999423027039,
+      "learning_rate": 9.988365095372373e-06,
+      "loss": 0.1278,
+      "step": 2120
+    },
+    {
+      "epoch": 8.129770992366412,
+      "grad_norm": 0.7949787378311157,
+      "learning_rate": 9.987986284808654e-06,
+      "loss": 0.1285,
+      "step": 2130
+    },
+    {
+      "epoch": 8.16793893129771,
+      "grad_norm": 0.37061432003974915,
+      "learning_rate": 9.987601413371264e-06,
+      "loss": 0.1218,
+      "step": 2140
+    },
+    {
+      "epoch": 8.206106870229007,
+      "grad_norm": 0.6125175952911377,
+      "learning_rate": 9.987210481527857e-06,
+      "loss": 0.1284,
+      "step": 2150
+    },
+    {
+      "epoch": 8.244274809160306,
+      "grad_norm": 1.3589234352111816,
+      "learning_rate": 9.98681348975345e-06,
+      "loss": 0.1281,
+      "step": 2160
+    },
+    {
+      "epoch": 8.282442748091603,
+      "grad_norm": 0.6536274552345276,
+      "learning_rate": 9.986410438530428e-06,
+      "loss": 0.1237,
+      "step": 2170
+    },
+    {
+      "epoch": 8.320610687022901,
+      "grad_norm": 0.5439227223396301,
+      "learning_rate": 9.986001328348534e-06,
+      "loss": 0.1292,
+      "step": 2180
+    },
+    {
+      "epoch": 8.358778625954198,
+      "grad_norm": 0.6106489896774292,
+      "learning_rate": 9.985586159704879e-06,
+      "loss": 0.1237,
+      "step": 2190
+    },
+    {
+      "epoch": 8.396946564885496,
+      "grad_norm": 0.422396719455719,
+      "learning_rate": 9.98516493310393e-06,
+      "loss": 0.1247,
+      "step": 2200
+    },
+    {
+      "epoch": 8.435114503816793,
+      "grad_norm": 0.4791705012321472,
+      "learning_rate": 9.984737649057514e-06,
+      "loss": 0.1291,
+      "step": 2210
+    },
+    {
+      "epoch": 8.473282442748092,
+      "grad_norm": 1.0834544897079468,
+      "learning_rate": 9.984304308084827e-06,
+      "loss": 0.1383,
+      "step": 2220
+    },
+    {
+      "epoch": 8.511450381679388,
+      "grad_norm": 0.5249638557434082,
+      "learning_rate": 9.983864910712416e-06,
+      "loss": 0.1249,
+      "step": 2230
+    },
+    {
+      "epoch": 8.549618320610687,
+      "grad_norm": 0.6899205446243286,
+      "learning_rate": 9.98341945747419e-06,
+      "loss": 0.1242,
+      "step": 2240
+    },
+    {
+      "epoch": 8.587786259541986,
+      "grad_norm": 0.6143678426742554,
+      "learning_rate": 9.98296794891142e-06,
+      "loss": 0.1394,
+      "step": 2250
+    },
+    {
+      "epoch": 8.625954198473282,
+      "grad_norm": 0.4065885543823242,
+      "learning_rate": 9.982510385572725e-06,
+      "loss": 0.1343,
+      "step": 2260
+    },
+    {
+      "epoch": 8.664122137404581,
+      "grad_norm": 1.2763944864273071,
+      "learning_rate": 9.982046768014093e-06,
+      "loss": 0.1448,
+      "step": 2270
+    },
+    {
+      "epoch": 8.702290076335878,
+      "grad_norm": 0.924741268157959,
+      "learning_rate": 9.981577096798864e-06,
+      "loss": 0.1442,
+      "step": 2280
+    },
+    {
+      "epoch": 8.740458015267176,
+      "grad_norm": 0.8587446212768555,
+      "learning_rate": 9.981101372497729e-06,
+      "loss": 0.13,
+      "step": 2290
+    },
+    {
+      "epoch": 8.778625954198473,
+      "grad_norm": 0.8710682392120361,
+      "learning_rate": 9.980619595688737e-06,
+      "loss": 0.1263,
+      "step": 2300
+    },
+    {
+      "epoch": 8.816793893129772,
+      "grad_norm": 0.7033365964889526,
+      "learning_rate": 9.980131766957296e-06,
+      "loss": 0.1207,
+      "step": 2310
+    },
+    {
+      "epoch": 8.854961832061068,
+      "grad_norm": 0.4609795808792114,
+      "learning_rate": 9.979637886896162e-06,
+      "loss": 0.1244,
+      "step": 2320
+    },
+    {
+      "epoch": 8.893129770992367,
+      "grad_norm": 0.7439802289009094,
+      "learning_rate": 9.979137956105446e-06,
+      "loss": 0.1202,
+      "step": 2330
+    },
+    {
+      "epoch": 8.931297709923664,
+      "grad_norm": 1.673240065574646,
+      "learning_rate": 9.978631975192613e-06,
+      "loss": 0.1258,
+      "step": 2340
+    },
+    {
+      "epoch": 8.969465648854962,
+      "grad_norm": 1.4078855514526367,
+      "learning_rate": 9.978119944772476e-06,
+      "loss": 0.1305,
+      "step": 2350
+    },
+    {
+      "epoch": 9.007633587786259,
+      "grad_norm": 0.703015923500061,
+      "learning_rate": 9.977601865467197e-06,
+      "loss": 0.1186,
+      "step": 2360
+    },
+    {
+      "epoch": 9.045801526717558,
+      "grad_norm": 0.973673939704895,
+      "learning_rate": 9.977077737906296e-06,
+      "loss": 0.129,
+      "step": 2370
+    },
+    {
+      "epoch": 9.083969465648854,
+      "grad_norm": 0.6487618684768677,
+      "learning_rate": 9.976547562726637e-06,
+      "loss": 0.1108,
+      "step": 2380
+    },
+    {
+      "epoch": 9.122137404580153,
+      "grad_norm": 0.4547242820262909,
+      "learning_rate": 9.97601134057243e-06,
+      "loss": 0.1211,
+      "step": 2390
+    },
+    {
+      "epoch": 9.16030534351145,
+      "grad_norm": 0.7747370600700378,
+      "learning_rate": 9.975469072095236e-06,
+      "loss": 0.122,
+      "step": 2400
+    },
+    {
+      "epoch": 9.198473282442748,
+      "grad_norm": 0.6227743029594421,
+      "learning_rate": 9.974920757953965e-06,
+      "loss": 0.1337,
+      "step": 2410
+    },
+    {
+      "epoch": 9.236641221374045,
+      "grad_norm": 1.6230658292770386,
+      "learning_rate": 9.97436639881487e-06,
+      "loss": 0.1333,
+      "step": 2420
+    },
+    {
+      "epoch": 9.274809160305344,
+      "grad_norm": 0.4734322130680084,
+      "learning_rate": 9.973805995351545e-06,
+      "loss": 0.1312,
+      "step": 2430
+    },
+    {
+      "epoch": 9.312977099236642,
+      "grad_norm": 1.1201488971710205,
+      "learning_rate": 9.973239548244938e-06,
+      "loss": 0.1332,
+      "step": 2440
+    },
+    {
+      "epoch": 9.351145038167939,
+      "grad_norm": 0.6548280119895935,
+      "learning_rate": 9.972667058183333e-06,
+      "loss": 0.1282,
+      "step": 2450
+    },
+    {
+      "epoch": 9.389312977099237,
+      "grad_norm": 0.9208749532699585,
+      "learning_rate": 9.972088525862363e-06,
+      "loss": 0.1397,
+      "step": 2460
+    },
+    {
+      "epoch": 9.427480916030534,
+      "grad_norm": 0.8364688158035278,
+      "learning_rate": 9.971503951984996e-06,
+      "loss": 0.1346,
+      "step": 2470
+    },
+    {
+      "epoch": 9.465648854961833,
+      "grad_norm": 0.5072993636131287,
+      "learning_rate": 9.970913337261544e-06,
+      "loss": 0.1216,
+      "step": 2480
+    },
+    {
+      "epoch": 9.50381679389313,
+      "grad_norm": 0.5047227144241333,
+      "learning_rate": 9.97031668240966e-06,
+      "loss": 0.1158,
+      "step": 2490
+    },
+    {
+      "epoch": 9.541984732824428,
+      "grad_norm": 0.6214285492897034,
+      "learning_rate": 9.96971398815434e-06,
+      "loss": 0.1275,
+      "step": 2500
+    },
+    {
+      "epoch": 9.580152671755725,
+      "grad_norm": 0.2843784987926483,
+      "learning_rate": 9.969105255227906e-06,
+      "loss": 0.1202,
+      "step": 2510
+    },
+    {
+      "epoch": 9.618320610687023,
+      "grad_norm": 0.9346826672554016,
+      "learning_rate": 9.968490484370035e-06,
+      "loss": 0.1257,
+      "step": 2520
+    },
+    {
+      "epoch": 9.65648854961832,
+      "grad_norm": 0.5003984570503235,
+      "learning_rate": 9.967869676327726e-06,
+      "loss": 0.1189,
+      "step": 2530
+    },
+    {
+      "epoch": 9.694656488549619,
+      "grad_norm": 0.535346508026123,
+      "learning_rate": 9.967242831855321e-06,
+      "loss": 0.1166,
+      "step": 2540
+    },
+    {
+      "epoch": 9.732824427480915,
+      "grad_norm": 2.279946804046631,
+      "learning_rate": 9.966609951714495e-06,
+      "loss": 0.1317,
+      "step": 2550
+    },
+    {
+      "epoch": 9.770992366412214,
+      "grad_norm": 0.4801623523235321,
+      "learning_rate": 9.965971036674256e-06,
+      "loss": 0.1233,
+      "step": 2560
+    },
+    {
+      "epoch": 9.80916030534351,
+      "grad_norm": 0.861571192741394,
+      "learning_rate": 9.965326087510947e-06,
+      "loss": 0.1238,
+      "step": 2570
+    },
+    {
+      "epoch": 9.84732824427481,
+      "grad_norm": 0.4469713568687439,
+      "learning_rate": 9.964675105008243e-06,
+      "loss": 0.1162,
+      "step": 2580
+    },
+    {
+      "epoch": 9.885496183206106,
+      "grad_norm": 0.7882781028747559,
+      "learning_rate": 9.964018089957148e-06,
+      "loss": 0.1225,
+      "step": 2590
+    },
+    {
+      "epoch": 9.923664122137405,
+      "grad_norm": 0.38498684763908386,
+      "learning_rate": 9.963355043155997e-06,
+      "loss": 0.1202,
+      "step": 2600
+    },
+    {
+      "epoch": 9.961832061068701,
+      "grad_norm": 0.5153051614761353,
+      "learning_rate": 9.962685965410455e-06,
+      "loss": 0.1202,
+      "step": 2610
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.4856826663017273,
+      "learning_rate": 9.962010857533514e-06,
+      "loss": 0.1152,
+      "step": 2620
+    },
+    {
+      "epoch": 10.038167938931299,
+      "grad_norm": 0.44617682695388794,
+      "learning_rate": 9.961329720345494e-06,
+      "loss": 0.1225,
+      "step": 2630
+    },
+    {
+      "epoch": 10.076335877862595,
+      "grad_norm": 0.5542685389518738,
+      "learning_rate": 9.96064255467404e-06,
+      "loss": 0.1271,
+      "step": 2640
+    },
+    {
+      "epoch": 10.114503816793894,
+      "grad_norm": 0.46713921427726746,
+      "learning_rate": 9.959949361354127e-06,
+      "loss": 0.1034,
+      "step": 2650
+    },
+    {
+      "epoch": 10.15267175572519,
+      "grad_norm": 0.3956896662712097,
+      "learning_rate": 9.959250141228046e-06,
+      "loss": 0.1086,
+      "step": 2660
+    },
+    {
+      "epoch": 10.19083969465649,
+      "grad_norm": 0.34693989157676697,
+      "learning_rate": 9.958544895145415e-06,
+      "loss": 0.1185,
+      "step": 2670
+    },
+    {
+      "epoch": 10.229007633587786,
+      "grad_norm": 0.5464341044425964,
+      "learning_rate": 9.957833623963178e-06,
+      "loss": 0.1213,
+      "step": 2680
+    },
+    {
+      "epoch": 10.267175572519085,
+      "grad_norm": 1.7212783098220825,
+      "learning_rate": 9.957116328545593e-06,
+      "loss": 0.1231,
+      "step": 2690
+    },
+    {
+      "epoch": 10.305343511450381,
+      "grad_norm": 0.5747115015983582,
+      "learning_rate": 9.956393009764244e-06,
+      "loss": 0.1436,
+      "step": 2700
+    },
+    {
+      "epoch": 10.34351145038168,
+      "grad_norm": 0.7528557181358337,
+      "learning_rate": 9.955663668498032e-06,
+      "loss": 0.1238,
+      "step": 2710
+    },
+    {
+      "epoch": 10.381679389312977,
+      "grad_norm": 0.7713605761528015,
+      "learning_rate": 9.954928305633174e-06,
+      "loss": 0.1309,
+      "step": 2720
+    },
+    {
+      "epoch": 10.419847328244275,
+      "grad_norm": 0.42558568716049194,
+      "learning_rate": 9.954186922063204e-06,
+      "loss": 0.1258,
+      "step": 2730
+    },
+    {
+      "epoch": 10.458015267175572,
+      "grad_norm": 0.44242027401924133,
+      "learning_rate": 9.953439518688974e-06,
+      "loss": 0.1097,
+      "step": 2740
+    },
+    {
+      "epoch": 10.49618320610687,
+      "grad_norm": 0.4029052257537842,
+      "learning_rate": 9.952686096418652e-06,
+      "loss": 0.1273,
+      "step": 2750
+    },
+    {
+      "epoch": 10.534351145038167,
+      "grad_norm": 0.38173234462738037,
+      "learning_rate": 9.951926656167715e-06,
+      "loss": 0.1163,
+      "step": 2760
+    },
+    {
+      "epoch": 10.572519083969466,
+      "grad_norm": 0.6817983984947205,
+      "learning_rate": 9.951161198858952e-06,
+      "loss": 0.1242,
+      "step": 2770
+    },
+    {
+      "epoch": 10.610687022900763,
+      "grad_norm": 0.47834619879722595,
+      "learning_rate": 9.95038972542247e-06,
+      "loss": 0.1173,
+      "step": 2780
+    },
+    {
+      "epoch": 10.648854961832061,
+      "grad_norm": 1.0968841314315796,
+      "learning_rate": 9.949612236795682e-06,
+      "loss": 0.1172,
+      "step": 2790
+    },
+    {
+      "epoch": 10.68702290076336,
+      "grad_norm": 0.4130131006240845,
+      "learning_rate": 9.948828733923305e-06,
+      "loss": 0.1202,
+      "step": 2800
+    },
+    {
+      "epoch": 10.725190839694656,
+      "grad_norm": 0.5207512974739075,
+      "learning_rate": 9.948039217757375e-06,
+      "loss": 0.1318,
+      "step": 2810
+    },
+    {
+      "epoch": 10.763358778625955,
+      "grad_norm": 0.5142215490341187,
+      "learning_rate": 9.947243689257226e-06,
+      "loss": 0.1175,
+      "step": 2820
+    },
+    {
+      "epoch": 10.801526717557252,
+      "grad_norm": 0.46614205837249756,
+      "learning_rate": 9.946442149389498e-06,
+      "loss": 0.1248,
+      "step": 2830
+    },
+    {
+      "epoch": 10.83969465648855,
+      "grad_norm": 0.6606496572494507,
+      "learning_rate": 9.94563459912814e-06,
+      "loss": 0.132,
+      "step": 2840
+    },
+    {
+      "epoch": 10.877862595419847,
+      "grad_norm": 0.7258016467094421,
+      "learning_rate": 9.944821039454403e-06,
+      "loss": 0.1153,
+      "step": 2850
+    },
+    {
+      "epoch": 10.916030534351146,
+      "grad_norm": 0.5228508114814758,
+      "learning_rate": 9.944001471356835e-06,
+      "loss": 0.134,
+      "step": 2860
+    },
+    {
+      "epoch": 10.954198473282442,
+      "grad_norm": 0.7532591223716736,
+      "learning_rate": 9.94317589583129e-06,
+      "loss": 0.1203,
+      "step": 2870
+    },
+    {
+      "epoch": 10.992366412213741,
+      "grad_norm": 0.8692820072174072,
+      "learning_rate": 9.942344313880922e-06,
+      "loss": 0.1177,
+      "step": 2880
+    },
+    {
+      "epoch": 11.030534351145038,
+      "grad_norm": 0.40786102414131165,
+      "learning_rate": 9.941506726516179e-06,
+      "loss": 0.1336,
+      "step": 2890
+    },
+    {
+      "epoch": 11.068702290076336,
+      "grad_norm": 0.4568503201007843,
+      "learning_rate": 9.94066313475481e-06,
+      "loss": 0.1187,
+      "step": 2900
+    },
+    {
+      "epoch": 11.106870229007633,
+      "grad_norm": 0.4302501678466797,
+      "learning_rate": 9.939813539621858e-06,
+      "loss": 0.1209,
+      "step": 2910
+    },
+    {
+      "epoch": 11.145038167938932,
+      "grad_norm": 0.32359543442726135,
+      "learning_rate": 9.93895794214966e-06,
+      "loss": 0.1077,
+      "step": 2920
+    },
+    {
+      "epoch": 11.183206106870228,
+      "grad_norm": 0.5436263680458069,
+      "learning_rate": 9.938096343377853e-06,
+      "loss": 0.1312,
+      "step": 2930
+    },
+    {
+      "epoch": 11.221374045801527,
+      "grad_norm": 0.3711869418621063,
+      "learning_rate": 9.937228744353354e-06,
+      "loss": 0.1173,
+      "step": 2940
+    },
+    {
+      "epoch": 11.259541984732824,
+      "grad_norm": 0.35187089443206787,
+      "learning_rate": 9.93635514613038e-06,
+      "loss": 0.1153,
+      "step": 2950
+    },
+    {
+      "epoch": 11.297709923664122,
+      "grad_norm": 0.5851736664772034,
+      "learning_rate": 9.935475549770436e-06,
+      "loss": 0.1169,
+      "step": 2960
+    },
+    {
+      "epoch": 11.335877862595419,
+      "grad_norm": 0.3298185467720032,
+      "learning_rate": 9.934589956342315e-06,
+      "loss": 0.1163,
+      "step": 2970
+    },
+    {
+      "epoch": 11.374045801526718,
+      "grad_norm": 0.6017898917198181,
+      "learning_rate": 9.933698366922093e-06,
+      "loss": 0.1216,
+      "step": 2980
+    },
+    {
+      "epoch": 11.412213740458014,
+      "grad_norm": 0.6507192254066467,
+      "learning_rate": 9.932800782593141e-06,
+      "loss": 0.1226,
+      "step": 2990
+    },
+    {
+      "epoch": 11.450381679389313,
+      "grad_norm": 0.3257594406604767,
+      "learning_rate": 9.931897204446104e-06,
+      "loss": 0.1308,
+      "step": 3000
+    },
+    {
+      "epoch": 11.488549618320612,
+      "grad_norm": 4.35601806640625,
+      "learning_rate": 9.930987633578916e-06,
+      "loss": 0.1232,
+      "step": 3010
+    },
+    {
+      "epoch": 11.526717557251908,
+      "grad_norm": 0.3781754970550537,
+      "learning_rate": 9.930072071096791e-06,
+      "loss": 0.1277,
+      "step": 3020
+    },
+    {
+      "epoch": 11.564885496183207,
+      "grad_norm": 0.8004752397537231,
+      "learning_rate": 9.929150518112225e-06,
+      "loss": 0.1216,
+      "step": 3030
+    },
+    {
+      "epoch": 11.603053435114504,
+      "grad_norm": 0.6701073050498962,
+      "learning_rate": 9.928222975744992e-06,
+      "loss": 0.1208,
+      "step": 3040
+    },
+    {
+      "epoch": 11.641221374045802,
+      "grad_norm": 0.39749905467033386,
+      "learning_rate": 9.92728944512214e-06,
+      "loss": 0.1079,
+      "step": 3050
+    },
+    {
+      "epoch": 11.679389312977099,
+      "grad_norm": 0.6213489174842834,
+      "learning_rate": 9.926349927378e-06,
+      "loss": 0.1297,
+      "step": 3060
+    },
+    {
+      "epoch": 11.717557251908397,
+      "grad_norm": 0.4827481508255005,
+      "learning_rate": 9.925404423654175e-06,
+      "loss": 0.1244,
+      "step": 3070
+    },
+    {
+      "epoch": 11.755725190839694,
+      "grad_norm": 0.7463303804397583,
+      "learning_rate": 9.924452935099537e-06,
+      "loss": 0.1128,
+      "step": 3080
+    },
+    {
+      "epoch": 11.793893129770993,
+      "grad_norm": 1.3172951936721802,
+      "learning_rate": 9.923495462870241e-06,
+      "loss": 0.1119,
+      "step": 3090
+    },
+    {
+      "epoch": 11.83206106870229,
+      "grad_norm": 0.8011961579322815,
+      "learning_rate": 9.9225320081297e-06,
+      "loss": 0.1115,
+      "step": 3100
+    },
+    {
+      "epoch": 11.870229007633588,
+      "grad_norm": 1.2579299211502075,
+      "learning_rate": 9.921562572048606e-06,
+      "loss": 0.1416,
+      "step": 3110
+    },
+    {
+      "epoch": 11.908396946564885,
+      "grad_norm": 0.6220617294311523,
+      "learning_rate": 9.920587155804913e-06,
+      "loss": 0.121,
+      "step": 3120
+    },
+    {
+      "epoch": 11.946564885496183,
+      "grad_norm": 0.4087788760662079,
+      "learning_rate": 9.919605760583846e-06,
+      "loss": 0.123,
+      "step": 3130
+    },
+    {
+      "epoch": 11.98473282442748,
+      "grad_norm": 0.3896748125553131,
+      "learning_rate": 9.91861838757789e-06,
+      "loss": 0.1182,
+      "step": 3140
+    },
+    {
+      "epoch": 12.022900763358779,
+      "grad_norm": 0.467168927192688,
+      "learning_rate": 9.917625037986798e-06,
+      "loss": 0.1145,
+      "step": 3150
+    },
+    {
+      "epoch": 12.061068702290076,
+      "grad_norm": 0.6059496998786926,
+      "learning_rate": 9.916625713017582e-06,
+      "loss": 0.1218,
+      "step": 3160
+    },
+    {
+      "epoch": 12.099236641221374,
+      "grad_norm": 0.579326868057251,
+      "learning_rate": 9.91562041388452e-06,
+      "loss": 0.114,
+      "step": 3170
+    },
+    {
+      "epoch": 12.137404580152673,
+      "grad_norm": 1.0949403047561646,
+      "learning_rate": 9.91460914180914e-06,
+      "loss": 0.1306,
+      "step": 3180
+    },
+    {
+      "epoch": 12.17557251908397,
+      "grad_norm": 0.4158419370651245,
+      "learning_rate": 9.913591898020234e-06,
+      "loss": 0.1169,
+      "step": 3190
+    },
+    {
+      "epoch": 12.213740458015268,
+      "grad_norm": 1.5080878734588623,
+      "learning_rate": 9.912568683753853e-06,
+      "loss": 0.1215,
+      "step": 3200
+    },
+    {
+      "epoch": 12.251908396946565,
+      "grad_norm": 0.5173312425613403,
+      "learning_rate": 9.911539500253295e-06,
+      "loss": 0.1037,
+      "step": 3210
+    },
+    {
+      "epoch": 12.290076335877863,
+      "grad_norm": 0.33992186188697815,
+      "learning_rate": 9.910504348769118e-06,
+      "loss": 0.1145,
+      "step": 3220
+    },
+    {
+      "epoch": 12.32824427480916,
+      "grad_norm": 0.881161093711853,
+      "learning_rate": 9.909463230559127e-06,
+      "loss": 0.1205,
+      "step": 3230
+    },
+    {
+      "epoch": 12.366412213740459,
+      "grad_norm": 0.376974493265152,
+      "learning_rate": 9.908416146888376e-06,
+      "loss": 0.1234,
+      "step": 3240
+    },
+    {
+      "epoch": 12.404580152671755,
+      "grad_norm": 0.4849688410758972,
+      "learning_rate": 9.907363099029175e-06,
+      "loss": 0.1221,
+      "step": 3250
+    },
+    {
+      "epoch": 12.442748091603054,
+      "grad_norm": 0.6830031871795654,
+      "learning_rate": 9.906304088261073e-06,
+      "loss": 0.1522,
+      "step": 3260
+    },
+    {
+      "epoch": 12.48091603053435,
+      "grad_norm": 0.450184166431427,
+      "learning_rate": 9.905239115870873e-06,
+      "loss": 0.1109,
+      "step": 3270
+    },
+    {
+      "epoch": 12.51908396946565,
+      "grad_norm": 0.32189562916755676,
+      "learning_rate": 9.904168183152611e-06,
+      "loss": 0.1344,
+      "step": 3280
+    },
+    {
+      "epoch": 12.557251908396946,
+      "grad_norm": 0.6055088043212891,
+      "learning_rate": 9.903091291407574e-06,
+      "loss": 0.1377,
+      "step": 3290
+    },
+    {
+      "epoch": 12.595419847328245,
+      "grad_norm": 0.833159863948822,
+      "learning_rate": 9.902008441944287e-06,
+      "loss": 0.1304,
+      "step": 3300
+    },
+    {
+      "epoch": 12.633587786259541,
+      "grad_norm": 0.6403981447219849,
+      "learning_rate": 9.900919636078511e-06,
+      "loss": 0.111,
+      "step": 3310
+    },
+    {
+      "epoch": 12.67175572519084,
+      "grad_norm": 1.0145337581634521,
+      "learning_rate": 9.899824875133255e-06,
+      "loss": 0.123,
+      "step": 3320
+    },
+    {
+      "epoch": 12.709923664122137,
+      "grad_norm": 0.8037505149841309,
+      "learning_rate": 9.89872416043875e-06,
+      "loss": 0.1188,
+      "step": 3330
+    },
+    {
+      "epoch": 12.748091603053435,
+      "grad_norm": 0.5243023633956909,
+      "learning_rate": 9.89761749333247e-06,
+      "loss": 0.1184,
+      "step": 3340
+    },
+    {
+      "epoch": 12.786259541984732,
+      "grad_norm": 0.3315357565879822,
+      "learning_rate": 9.896504875159122e-06,
+      "loss": 0.1168,
+      "step": 3350
+    },
+    {
+      "epoch": 12.82442748091603,
+      "grad_norm": 0.7898366451263428,
+      "learning_rate": 9.89538630727064e-06,
+      "loss": 0.1138,
+      "step": 3360
+    },
+    {
+      "epoch": 12.862595419847327,
+      "grad_norm": 0.3733002841472626,
+      "learning_rate": 9.89426179102619e-06,
+      "loss": 0.1168,
+      "step": 3370
+    },
+    {
+      "epoch": 12.900763358778626,
+      "grad_norm": 0.5741856098175049,
+      "learning_rate": 9.893131327792166e-06,
+      "loss": 0.1189,
+      "step": 3380
+    },
+    {
+      "epoch": 12.938931297709924,
+      "grad_norm": 0.5322036743164062,
+      "learning_rate": 9.891994918942183e-06,
+      "loss": 0.1182,
+      "step": 3390
+    },
+    {
+      "epoch": 12.977099236641221,
+      "grad_norm": 0.3387764096260071,
+      "learning_rate": 9.890852565857092e-06,
+      "loss": 0.1201,
+      "step": 3400
+    },
+    {
+      "epoch": 13.01526717557252,
+      "grad_norm": 0.6571619510650635,
+      "learning_rate": 9.889704269924955e-06,
+      "loss": 0.1217,
+      "step": 3410
+    },
+    {
+      "epoch": 13.053435114503817,
+      "grad_norm": 0.5754358172416687,
+      "learning_rate": 9.88855003254106e-06,
+      "loss": 0.1446,
+      "step": 3420
+    },
+    {
+      "epoch": 13.091603053435115,
+      "grad_norm": 0.5554888844490051,
+      "learning_rate": 9.887389855107917e-06,
+      "loss": 0.1107,
+      "step": 3430
+    },
+    {
+      "epoch": 13.129770992366412,
+      "grad_norm": 0.5900585055351257,
+      "learning_rate": 9.886223739035249e-06,
+      "loss": 0.1211,
+      "step": 3440
+    },
+    {
+      "epoch": 13.16793893129771,
+      "grad_norm": 0.42058074474334717,
+      "learning_rate": 9.885051685739997e-06,
+      "loss": 0.1208,
+      "step": 3450
+    },
+    {
+      "epoch": 13.206106870229007,
+      "grad_norm": 0.959195077419281,
+      "learning_rate": 9.883873696646316e-06,
+      "loss": 0.1127,
+      "step": 3460
+    },
+    {
+      "epoch": 13.244274809160306,
+      "grad_norm": 0.45505401492118835,
+      "learning_rate": 9.882689773185575e-06,
+      "loss": 0.1251,
+      "step": 3470
+    },
+    {
+      "epoch": 13.282442748091603,
+      "grad_norm": 0.7457808256149292,
+      "learning_rate": 9.881499916796354e-06,
+      "loss": 0.1189,
+      "step": 3480
+    },
+    {
+      "epoch": 13.320610687022901,
+      "grad_norm": 0.8850777745246887,
+      "learning_rate": 9.880304128924435e-06,
+      "loss": 0.1307,
+      "step": 3490
+    },
+    {
+      "epoch": 13.358778625954198,
+      "grad_norm": 0.7178351879119873,
+      "learning_rate": 9.879102411022818e-06,
+      "loss": 0.1172,
+      "step": 3500
+    },
+    {
+      "epoch": 13.396946564885496,
+      "grad_norm": 0.8748807311058044,
+      "learning_rate": 9.877894764551704e-06,
+      "loss": 0.1167,
+      "step": 3510
+    },
+    {
+      "epoch": 13.435114503816793,
+      "grad_norm": 1.205231785774231,
+      "learning_rate": 9.876681190978494e-06,
+      "loss": 0.1294,
+      "step": 3520
+    },
+    {
+      "epoch": 13.473282442748092,
+      "grad_norm": 0.4213230013847351,
+      "learning_rate": 9.875461691777797e-06,
+      "loss": 0.1131,
+      "step": 3530
+    },
+    {
+      "epoch": 13.511450381679388,
+      "grad_norm": 0.5431898832321167,
+      "learning_rate": 9.874236268431417e-06,
+      "loss": 0.1157,
+      "step": 3540
+    },
+    {
+      "epoch": 13.549618320610687,
+      "grad_norm": 0.8572608828544617,
+      "learning_rate": 9.873004922428362e-06,
+      "loss": 0.1259,
+      "step": 3550
+    },
+    {
+      "epoch": 13.587786259541986,
+      "grad_norm": 0.44893670082092285,
+      "learning_rate": 9.87176765526483e-06,
+      "loss": 0.129,
+      "step": 3560
+    },
+    {
+      "epoch": 13.625954198473282,
+      "grad_norm": 0.5084060430526733,
+      "learning_rate": 9.87052446844422e-06,
+      "loss": 0.1134,
+      "step": 3570
+    },
+    {
+      "epoch": 13.664122137404581,
+      "grad_norm": 1.1047592163085938,
+      "learning_rate": 9.869275363477122e-06,
+      "loss": 0.1176,
+      "step": 3580
+    },
+    {
+      "epoch": 13.702290076335878,
+      "grad_norm": 0.836169421672821,
+      "learning_rate": 9.868020341881313e-06,
+      "loss": 0.1186,
+      "step": 3590
+    },
+    {
+      "epoch": 13.740458015267176,
+      "grad_norm": 0.7019125819206238,
+      "learning_rate": 9.866759405181766e-06,
+      "loss": 0.1199,
+      "step": 3600
+    },
+    {
+      "epoch": 13.778625954198473,
+      "grad_norm": 0.5258054733276367,
+      "learning_rate": 9.865492554910634e-06,
+      "loss": 0.1227,
+      "step": 3610
+    },
+    {
+      "epoch": 13.816793893129772,
+      "grad_norm": 0.670691967010498,
+      "learning_rate": 9.864219792607262e-06,
+      "loss": 0.115,
+      "step": 3620
+    },
+    {
+      "epoch": 13.854961832061068,
+      "grad_norm": 0.5219152569770813,
+      "learning_rate": 9.862941119818176e-06,
+      "loss": 0.1249,
+      "step": 3630
+    },
+    {
+      "epoch": 13.893129770992367,
+      "grad_norm": 0.9201920032501221,
+      "learning_rate": 9.861656538097086e-06,
+      "loss": 0.1211,
+      "step": 3640
+    },
+    {
+      "epoch": 13.931297709923664,
+      "grad_norm": 0.4263167679309845,
+      "learning_rate": 9.860366049004878e-06,
+      "loss": 0.1115,
+      "step": 3650
+    },
+    {
+      "epoch": 13.969465648854962,
+      "grad_norm": 0.8185678124427795,
+      "learning_rate": 9.859069654109615e-06,
+      "loss": 0.133,
+      "step": 3660
+    },
+    {
+      "epoch": 14.007633587786259,
+      "grad_norm": 0.4073340594768524,
+      "learning_rate": 9.857767354986545e-06,
+      "loss": 0.1306,
+      "step": 3670
+    },
+    {
+      "epoch": 14.045801526717558,
+      "grad_norm": 0.6191373467445374,
+      "learning_rate": 9.856459153218078e-06,
+      "loss": 0.114,
+      "step": 3680
+    },
+    {
+      "epoch": 14.083969465648854,
+      "grad_norm": 0.9513982534408569,
+      "learning_rate": 9.855145050393808e-06,
+      "loss": 0.1276,
+      "step": 3690
+    },
+    {
+      "epoch": 14.122137404580153,
+      "grad_norm": 0.3555004894733429,
+      "learning_rate": 9.853825048110491e-06,
+      "loss": 0.1184,
+      "step": 3700
+    },
+    {
+      "epoch": 14.16030534351145,
+      "grad_norm": 0.4830268919467926,
+      "learning_rate": 9.852499147972055e-06,
+      "loss": 0.1063,
+      "step": 3710
+    },
+    {
+      "epoch": 14.198473282442748,
+      "grad_norm": 0.5373522043228149,
+      "learning_rate": 9.851167351589593e-06,
+      "loss": 0.1422,
+      "step": 3720
+    },
+    {
+      "epoch": 14.236641221374045,
+      "grad_norm": 0.36047184467315674,
+      "learning_rate": 9.849829660581364e-06,
+      "loss": 0.1151,
+      "step": 3730
+    },
+    {
+      "epoch": 14.274809160305344,
+      "grad_norm": 0.3627484142780304,
+      "learning_rate": 9.848486076572787e-06,
+      "loss": 0.1273,
+      "step": 3740
+    },
+    {
+      "epoch": 14.312977099236642,
+      "grad_norm": 0.37799179553985596,
+      "learning_rate": 9.847136601196445e-06,
+      "loss": 0.1098,
+      "step": 3750
+    },
+    {
+      "epoch": 14.351145038167939,
+      "grad_norm": 0.7704446315765381,
+      "learning_rate": 9.845781236092078e-06,
+      "loss": 0.1134,
+      "step": 3760
+    },
+    {
+      "epoch": 14.389312977099237,
+      "grad_norm": 0.36698824167251587,
+      "learning_rate": 9.844419982906584e-06,
+      "loss": 0.1395,
+      "step": 3770
+    },
+    {
+      "epoch": 14.427480916030534,
+      "grad_norm": 0.6548821926116943,
+      "learning_rate": 9.84305284329401e-06,
+      "loss": 0.1282,
+      "step": 3780
+    },
+    {
+      "epoch": 14.465648854961833,
+      "grad_norm": 0.3647122085094452,
+      "learning_rate": 9.84167981891556e-06,
+      "loss": 0.1082,
+      "step": 3790
+    },
+    {
+      "epoch": 14.50381679389313,
+      "grad_norm": 0.334097683429718,
+      "learning_rate": 9.84030091143959e-06,
+      "loss": 0.1065,
+      "step": 3800
+    },
+    {
+      "epoch": 14.541984732824428,
+      "grad_norm": 0.37364843487739563,
+      "learning_rate": 9.838916122541604e-06,
+      "loss": 0.1208,
+      "step": 3810
+    },
+    {
+      "epoch": 14.580152671755725,
+      "grad_norm": 0.40124377608299255,
+      "learning_rate": 9.837525453904247e-06,
+      "loss": 0.1315,
+      "step": 3820
+    },
+    {
+      "epoch": 14.618320610687023,
+      "grad_norm": 0.8429109454154968,
+      "learning_rate": 9.836128907217315e-06,
+      "loss": 0.1095,
+      "step": 3830
+    },
+    {
+      "epoch": 14.65648854961832,
+      "grad_norm": 0.441574364900589,
+      "learning_rate": 9.834726484177743e-06,
+      "loss": 0.1197,
+      "step": 3840
+    },
+    {
+      "epoch": 14.694656488549619,
+      "grad_norm": 0.8704144358634949,
+      "learning_rate": 9.833318186489608e-06,
+      "loss": 0.112,
+      "step": 3850
+    },
+    {
+      "epoch": 14.732824427480915,
+      "grad_norm": 0.5962323546409607,
+      "learning_rate": 9.831904015864127e-06,
+      "loss": 0.1105,
+      "step": 3860
+    },
+    {
+      "epoch": 14.770992366412214,
+      "grad_norm": 0.38795116543769836,
+      "learning_rate": 9.830483974019644e-06,
+      "loss": 0.1107,
+      "step": 3870
+    },
+    {
+      "epoch": 14.80916030534351,
+      "grad_norm": 0.847028911113739,
+      "learning_rate": 9.829058062681652e-06,
+      "loss": 0.1209,
+      "step": 3880
+    },
+    {
+      "epoch": 14.84732824427481,
+      "grad_norm": 0.3294160068035126,
+      "learning_rate": 9.82762628358276e-06,
+      "loss": 0.1082,
+      "step": 3890
+    },
+    {
+      "epoch": 14.885496183206106,
+      "grad_norm": 1.4115614891052246,
+      "learning_rate": 9.826188638462719e-06,
+      "loss": 0.1318,
+      "step": 3900
+    },
+    {
+      "epoch": 14.923664122137405,
+      "grad_norm": 0.47591233253479004,
+      "learning_rate": 9.824745129068403e-06,
+      "loss": 0.1409,
+      "step": 3910
+    },
+    {
+      "epoch": 14.961832061068701,
+      "grad_norm": 0.369427353143692,
+      "learning_rate": 9.82329575715381e-06,
+      "loss": 0.1184,
+      "step": 3920
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.4399472177028656,
+      "learning_rate": 9.821840524480066e-06,
+      "loss": 0.1095,
+      "step": 3930
+    },
+    {
+      "epoch": 15.038167938931299,
+      "grad_norm": 0.6284782886505127,
+      "learning_rate": 9.820379432815414e-06,
+      "loss": 0.1161,
+      "step": 3940
+    },
+    {
+      "epoch": 15.076335877862595,
+      "grad_norm": 0.4220380485057831,
+      "learning_rate": 9.81891248393522e-06,
+      "loss": 0.1151,
+      "step": 3950
+    },
+    {
+      "epoch": 15.114503816793894,
+      "grad_norm": 0.5931194424629211,
+      "learning_rate": 9.817439679621964e-06,
+      "loss": 0.1244,
+      "step": 3960
+    },
+    {
+      "epoch": 15.15267175572519,
+      "grad_norm": 0.31137576699256897,
+      "learning_rate": 9.815961021665243e-06,
+      "loss": 0.1194,
+      "step": 3970
+    },
+    {
+      "epoch": 15.19083969465649,
+      "grad_norm": 0.506864070892334,
+      "learning_rate": 9.814476511861764e-06,
+      "loss": 0.1112,
+      "step": 3980
+    },
+    {
+      "epoch": 15.229007633587786,
+      "grad_norm": 0.7712796330451965,
+      "learning_rate": 9.812986152015349e-06,
+      "loss": 0.1197,
+      "step": 3990
+    },
+    {
+      "epoch": 15.267175572519085,
+      "grad_norm": 0.49577248096466064,
+      "learning_rate": 9.811489943936922e-06,
+      "loss": 0.1148,
+      "step": 4000
+    },
+    {
+      "epoch": 15.305343511450381,
+      "grad_norm": 0.404431015253067,
+      "learning_rate": 9.809987889444523e-06,
+      "loss": 0.1186,
+      "step": 4010
+    },
+    {
+      "epoch": 15.34351145038168,
+      "grad_norm": 0.44168826937675476,
+      "learning_rate": 9.808479990363282e-06,
+      "loss": 0.1277,
+      "step": 4020
+    },
+    {
+      "epoch": 15.381679389312977,
+      "grad_norm": 0.2900712490081787,
+      "learning_rate": 9.806966248525447e-06,
+      "loss": 0.1116,
+      "step": 4030
+    },
+    {
+      "epoch": 15.419847328244275,
+      "grad_norm": 0.474483460187912,
+      "learning_rate": 9.805446665770348e-06,
+      "loss": 0.1249,
+      "step": 4040
+    },
+    {
+      "epoch": 15.458015267175572,
+      "grad_norm": 0.4446212947368622,
+      "learning_rate": 9.80392124394443e-06,
+      "loss": 0.1082,
+      "step": 4050
+    },
+    {
+      "epoch": 15.49618320610687,
+      "grad_norm": 0.3462320566177368,
+      "learning_rate": 9.802389984901218e-06,
+      "loss": 0.1263,
+      "step": 4060
+    },
+    {
+      "epoch": 15.534351145038167,
+      "grad_norm": 0.481645792722702,
+      "learning_rate": 9.80085289050134e-06,
+      "loss": 0.1164,
+      "step": 4070
+    },
+    {
+      "epoch": 15.572519083969466,
+      "grad_norm": 0.5727097988128662,
+      "learning_rate": 9.799309962612508e-06,
+      "loss": 0.1354,
+      "step": 4080
+    },
+    {
+      "epoch": 15.610687022900763,
+      "grad_norm": 0.9570332765579224,
+      "learning_rate": 9.797761203109527e-06,
+      "loss": 0.1192,
+      "step": 4090
+    },
+    {
+      "epoch": 15.648854961832061,
+      "grad_norm": 0.6497268080711365,
+      "learning_rate": 9.796206613874283e-06,
+      "loss": 0.1097,
+      "step": 4100
+    },
+    {
+      "epoch": 15.68702290076336,
+      "grad_norm": 0.4102814197540283,
+      "learning_rate": 9.794646196795754e-06,
+      "loss": 0.1041,
+      "step": 4110
+    },
+    {
+      "epoch": 15.725190839694656,
+      "grad_norm": 0.41581079363822937,
+      "learning_rate": 9.793079953769988e-06,
+      "loss": 0.1218,
+      "step": 4120
+    },
+    {
+      "epoch": 15.763358778625955,
+      "grad_norm": 0.5571273565292358,
+      "learning_rate": 9.79150788670012e-06,
+      "loss": 0.1083,
+      "step": 4130
+    },
+    {
+      "epoch": 15.801526717557252,
+      "grad_norm": 0.4825478792190552,
+      "learning_rate": 9.789929997496362e-06,
+      "loss": 0.1112,
+      "step": 4140
+    },
+    {
+      "epoch": 15.83969465648855,
+      "grad_norm": 0.35019156336784363,
+      "learning_rate": 9.788346288075993e-06,
+      "loss": 0.1187,
+      "step": 4150
+    },
+    {
+      "epoch": 15.877862595419847,
+      "grad_norm": 0.4018893539905548,
+      "learning_rate": 9.786756760363374e-06,
+      "loss": 0.1151,
+      "step": 4160
+    },
+    {
+      "epoch": 15.916030534351146,
+      "grad_norm": 0.6337506175041199,
+      "learning_rate": 9.78516141628993e-06,
+      "loss": 0.1114,
+      "step": 4170
+    },
+    {
+      "epoch": 15.954198473282442,
+      "grad_norm": 0.3153977394104004,
+      "learning_rate": 9.783560257794153e-06,
+      "loss": 0.1217,
+      "step": 4180
+    },
+    {
+      "epoch": 15.992366412213741,
+      "grad_norm": 0.8760756254196167,
+      "learning_rate": 9.781953286821604e-06,
+      "loss": 0.1148,
+      "step": 4190
+    },
+    {
+      "epoch": 16.03053435114504,
+      "grad_norm": 0.572750985622406,
+      "learning_rate": 9.7803405053249e-06,
+      "loss": 0.1161,
+      "step": 4200
+    },
+    {
+      "epoch": 16.068702290076335,
+      "grad_norm": 0.6838567852973938,
+      "learning_rate": 9.778721915263729e-06,
+      "loss": 0.1203,
+      "step": 4210
+    },
+    {
+      "epoch": 16.106870229007633,
+      "grad_norm": 0.4191221594810486,
+      "learning_rate": 9.777097518604824e-06,
+      "loss": 0.1159,
+      "step": 4220
+    },
+    {
+      "epoch": 16.14503816793893,
+      "grad_norm": 0.8830999135971069,
+      "learning_rate": 9.775467317321986e-06,
+      "loss": 0.1169,
+      "step": 4230
+    },
+    {
+      "epoch": 16.18320610687023,
+      "grad_norm": 0.4778836965560913,
+      "learning_rate": 9.773831313396056e-06,
+      "loss": 0.1149,
+      "step": 4240
+    },
+    {
+      "epoch": 16.221374045801525,
+      "grad_norm": 0.3479765057563782,
+      "learning_rate": 9.77218950881494e-06,
+      "loss": 0.1195,
+      "step": 4250
+    },
+    {
+      "epoch": 16.259541984732824,
+      "grad_norm": 0.39662837982177734,
+      "learning_rate": 9.770541905573583e-06,
+      "loss": 0.1275,
+      "step": 4260
+    },
+    {
+      "epoch": 16.297709923664122,
+      "grad_norm": 0.6865242719650269,
+      "learning_rate": 9.768888505673976e-06,
+      "loss": 0.1238,
+      "step": 4270
+    },
+    {
+      "epoch": 16.33587786259542,
+      "grad_norm": 0.3782191574573517,
+      "learning_rate": 9.767229311125162e-06,
+      "loss": 0.1133,
+      "step": 4280
+    },
+    {
+      "epoch": 16.374045801526716,
+      "grad_norm": 0.588668942451477,
+      "learning_rate": 9.76556432394321e-06,
+      "loss": 0.1362,
+      "step": 4290
+    },
+    {
+      "epoch": 16.412213740458014,
+      "grad_norm": 1.0646703243255615,
+      "learning_rate": 9.763893546151244e-06,
+      "loss": 0.1185,
+      "step": 4300
+    },
+    {
+      "epoch": 16.450381679389313,
+      "grad_norm": 0.4589475989341736,
+      "learning_rate": 9.762216979779412e-06,
+      "loss": 0.1173,
+      "step": 4310
+    },
+    {
+      "epoch": 16.48854961832061,
+      "grad_norm": 1.4795030355453491,
+      "learning_rate": 9.760534626864902e-06,
+      "loss": 0.125,
+      "step": 4320
+    },
+    {
+      "epoch": 16.52671755725191,
+      "grad_norm": 0.6526045799255371,
+      "learning_rate": 9.758846489451932e-06,
+      "loss": 0.126,
+      "step": 4330
+    },
+    {
+      "epoch": 16.564885496183205,
+      "grad_norm": 0.35587215423583984,
+      "learning_rate": 9.757152569591748e-06,
+      "loss": 0.1027,
+      "step": 4340
+    },
+    {
+      "epoch": 16.603053435114504,
+      "grad_norm": 0.8031662106513977,
+      "learning_rate": 9.755452869342621e-06,
+      "loss": 0.1163,
+      "step": 4350
+    },
+    {
+      "epoch": 16.641221374045802,
+      "grad_norm": 0.41244399547576904,
+      "learning_rate": 9.753747390769848e-06,
+      "loss": 0.1222,
+      "step": 4360
+    },
+    {
+      "epoch": 16.6793893129771,
+      "grad_norm": 0.4697263836860657,
+      "learning_rate": 9.752036135945743e-06,
+      "loss": 0.1078,
+      "step": 4370
+    },
+    {
+      "epoch": 16.717557251908396,
+      "grad_norm": 0.6083208322525024,
+      "learning_rate": 9.75031910694965e-06,
+      "loss": 0.1151,
+      "step": 4380
+    },
+    {
+      "epoch": 16.755725190839694,
+      "grad_norm": 0.4035137891769409,
+      "learning_rate": 9.748596305867913e-06,
+      "loss": 0.1085,
+      "step": 4390
+    },
+    {
+      "epoch": 16.793893129770993,
+      "grad_norm": 0.7176216244697571,
+      "learning_rate": 9.746867734793904e-06,
+      "loss": 0.1116,
+      "step": 4400
+    },
+    {
+      "epoch": 16.83206106870229,
+      "grad_norm": 0.4055471122264862,
+      "learning_rate": 9.745133395827994e-06,
+      "loss": 0.1248,
+      "step": 4410
+    },
+    {
+      "epoch": 16.870229007633586,
+      "grad_norm": 0.6174433827400208,
+      "learning_rate": 9.743393291077571e-06,
+      "loss": 0.1184,
+      "step": 4420
+    },
+    {
+      "epoch": 16.908396946564885,
+      "grad_norm": 0.5224699378013611,
+      "learning_rate": 9.741647422657029e-06,
+      "loss": 0.1114,
+      "step": 4430
+    },
+    {
+      "epoch": 16.946564885496183,
+      "grad_norm": 0.4283505976200104,
+      "learning_rate": 9.739895792687758e-06,
+      "loss": 0.1203,
+      "step": 4440
+    },
+    {
+      "epoch": 16.984732824427482,
+      "grad_norm": 1.0380663871765137,
+      "learning_rate": 9.738138403298158e-06,
+      "loss": 0.1119,
+      "step": 4450
+    },
+    {
+      "epoch": 17.022900763358777,
+      "grad_norm": 0.5753864645957947,
+      "learning_rate": 9.73637525662362e-06,
+      "loss": 0.1073,
+      "step": 4460
+    },
+    {
+      "epoch": 17.061068702290076,
+      "grad_norm": 0.8454107046127319,
+      "learning_rate": 9.734606354806533e-06,
+      "loss": 0.1148,
+      "step": 4470
+    },
+    {
+      "epoch": 17.099236641221374,
+      "grad_norm": 0.45036813616752625,
+      "learning_rate": 9.732831699996282e-06,
+      "loss": 0.1091,
+      "step": 4480
+    },
+    {
+      "epoch": 17.137404580152673,
+      "grad_norm": 0.3069184720516205,
+      "learning_rate": 9.731051294349238e-06,
+      "loss": 0.1298,
+      "step": 4490
+    },
+    {
+      "epoch": 17.17557251908397,
+      "grad_norm": 0.5899654626846313,
+      "learning_rate": 9.729265140028762e-06,
+      "loss": 0.1212,
+      "step": 4500
+    },
+    {
+      "epoch": 17.213740458015266,
+      "grad_norm": 0.5663384199142456,
+      "learning_rate": 9.727473239205202e-06,
+      "loss": 0.1163,
+      "step": 4510
+    },
+    {
+      "epoch": 17.251908396946565,
+      "grad_norm": 0.47053390741348267,
+      "learning_rate": 9.725675594055884e-06,
+      "loss": 0.1104,
+      "step": 4520
+    },
+    {
+      "epoch": 17.290076335877863,
+      "grad_norm": 0.5472087264060974,
+      "learning_rate": 9.723872206765115e-06,
+      "loss": 0.1186,
+      "step": 4530
+    },
+    {
+      "epoch": 17.328244274809162,
+      "grad_norm": 0.43163037300109863,
+      "learning_rate": 9.722063079524185e-06,
+      "loss": 0.118,
+      "step": 4540
+    },
+    {
+      "epoch": 17.366412213740457,
+      "grad_norm": 0.3856120705604553,
+      "learning_rate": 9.720248214531352e-06,
+      "loss": 0.1052,
+      "step": 4550
+    },
+    {
+      "epoch": 17.404580152671755,
+      "grad_norm": 0.47452834248542786,
+      "learning_rate": 9.718427613991848e-06,
+      "loss": 0.1026,
+      "step": 4560
+    },
+    {
+      "epoch": 17.442748091603054,
+      "grad_norm": 0.5274608731269836,
+      "learning_rate": 9.716601280117874e-06,
+      "loss": 0.118,
+      "step": 4570
+    },
+    {
+      "epoch": 17.480916030534353,
+      "grad_norm": 1.0746313333511353,
+      "learning_rate": 9.714769215128597e-06,
+      "loss": 0.1103,
+      "step": 4580
+    },
+    {
+      "epoch": 17.519083969465647,
+      "grad_norm": 0.3799738585948944,
+      "learning_rate": 9.712931421250152e-06,
+      "loss": 0.1026,
+      "step": 4590
+    },
+    {
+      "epoch": 17.557251908396946,
+      "grad_norm": 0.36720654368400574,
+      "learning_rate": 9.711087900715627e-06,
+      "loss": 0.1121,
+      "step": 4600
+    },
+    {
+      "epoch": 17.595419847328245,
+      "grad_norm": 0.4583773612976074,
+      "learning_rate": 9.709238655765079e-06,
+      "loss": 0.1129,
+      "step": 4610
+    },
+    {
+      "epoch": 17.633587786259543,
+      "grad_norm": 0.3917942941188812,
+      "learning_rate": 9.70738368864551e-06,
+      "loss": 0.1113,
+      "step": 4620
+    },
+    {
+      "epoch": 17.671755725190838,
+      "grad_norm": 0.4856705367565155,
+      "learning_rate": 9.705523001610884e-06,
+      "loss": 0.1059,
+      "step": 4630
+    },
+    {
+      "epoch": 17.709923664122137,
+      "grad_norm": 0.38836991786956787,
+      "learning_rate": 9.703656596922107e-06,
+      "loss": 0.1062,
+      "step": 4640
+    },
+    {
+      "epoch": 17.748091603053435,
+      "grad_norm": 0.77637779712677,
+      "learning_rate": 9.70178447684704e-06,
+      "loss": 0.1125,
+      "step": 4650
+    },
+    {
+      "epoch": 17.786259541984734,
+      "grad_norm": 1.447717547416687,
+      "learning_rate": 9.699906643660484e-06,
+      "loss": 0.1234,
+      "step": 4660
+    },
+    {
+      "epoch": 17.82442748091603,
+      "grad_norm": 0.4381782114505768,
+      "learning_rate": 9.698023099644186e-06,
+      "loss": 0.1179,
+      "step": 4670
+    },
+    {
+      "epoch": 17.862595419847327,
+      "grad_norm": 0.7676070928573608,
+      "learning_rate": 9.696133847086824e-06,
+      "loss": 0.1047,
+      "step": 4680
+    },
+    {
+      "epoch": 17.900763358778626,
+      "grad_norm": 0.36729001998901367,
+      "learning_rate": 9.694238888284023e-06,
+      "loss": 0.1246,
+      "step": 4690
+    },
+    {
+      "epoch": 17.938931297709924,
+      "grad_norm": 0.7202317714691162,
+      "learning_rate": 9.692338225538334e-06,
+      "loss": 0.1234,
+      "step": 4700
+    },
+    {
+      "epoch": 17.977099236641223,
+      "grad_norm": 0.8629083037376404,
+      "learning_rate": 9.690431861159242e-06,
+      "loss": 0.1153,
+      "step": 4710
+    },
+    {
+      "epoch": 18.015267175572518,
+      "grad_norm": 0.7108657956123352,
+      "learning_rate": 9.68851979746316e-06,
+      "loss": 0.1157,
+      "step": 4720
+    },
+    {
+      "epoch": 18.053435114503817,
+      "grad_norm": 0.3120957612991333,
+      "learning_rate": 9.686602036773427e-06,
+      "loss": 0.1134,
+      "step": 4730
+    },
+    {
+      "epoch": 18.091603053435115,
+      "grad_norm": 0.6133302450180054,
+      "learning_rate": 9.684678581420302e-06,
+      "loss": 0.1085,
+      "step": 4740
+    },
+    {
+      "epoch": 18.129770992366414,
+      "grad_norm": 0.8384220004081726,
+      "learning_rate": 9.682749433740963e-06,
+      "loss": 0.1116,
+      "step": 4750
+    },
+    {
+      "epoch": 18.16793893129771,
+      "grad_norm": 0.4204612672328949,
+      "learning_rate": 9.680814596079508e-06,
+      "loss": 0.1129,
+      "step": 4760
+    },
+    {
+      "epoch": 18.206106870229007,
+      "grad_norm": 0.8686299324035645,
+      "learning_rate": 9.678874070786945e-06,
+      "loss": 0.1226,
+      "step": 4770
+    },
+    {
+      "epoch": 18.244274809160306,
+      "grad_norm": 0.450674444437027,
+      "learning_rate": 9.676927860221199e-06,
+      "loss": 0.1189,
+      "step": 4780
+    },
+    {
+      "epoch": 18.282442748091604,
+      "grad_norm": 1.338226079940796,
+      "learning_rate": 9.674975966747098e-06,
+      "loss": 0.1275,
+      "step": 4790
+    },
+    {
+      "epoch": 18.3206106870229,
+      "grad_norm": 0.4029911160469055,
+      "learning_rate": 9.673018392736373e-06,
+      "loss": 0.1108,
+      "step": 4800
+    },
+    {
+      "epoch": 18.358778625954198,
+      "grad_norm": 0.6675863862037659,
+      "learning_rate": 9.671055140567667e-06,
+      "loss": 0.1127,
+      "step": 4810
+    },
+    {
+      "epoch": 18.396946564885496,
+      "grad_norm": 0.44830095767974854,
+      "learning_rate": 9.669086212626512e-06,
+      "loss": 0.1193,
+      "step": 4820
+    },
+    {
+      "epoch": 18.435114503816795,
+      "grad_norm": 0.5231329202651978,
+      "learning_rate": 9.667111611305343e-06,
+      "loss": 0.1326,
+      "step": 4830
+    },
+    {
+      "epoch": 18.47328244274809,
+      "grad_norm": 0.447273850440979,
+      "learning_rate": 9.665131339003487e-06,
+      "loss": 0.1144,
+      "step": 4840
+    },
+    {
+      "epoch": 18.51145038167939,
+      "grad_norm": 1.005131721496582,
+      "learning_rate": 9.663145398127158e-06,
+      "loss": 0.1227,
+      "step": 4850
+    },
+    {
+      "epoch": 18.549618320610687,
+      "grad_norm": 0.6508563160896301,
+      "learning_rate": 9.661153791089468e-06,
+      "loss": 0.117,
+      "step": 4860
+    },
+    {
+      "epoch": 18.587786259541986,
+      "grad_norm": 0.9407038688659668,
+      "learning_rate": 9.659156520310402e-06,
+      "loss": 0.1238,
+      "step": 4870
+    },
+    {
+      "epoch": 18.625954198473284,
+      "grad_norm": 0.356827050447464,
+      "learning_rate": 9.657153588216834e-06,
+      "loss": 0.1107,
+      "step": 4880
+    },
+    {
+      "epoch": 18.66412213740458,
+      "grad_norm": 0.4456828534603119,
+      "learning_rate": 9.655144997242516e-06,
+      "loss": 0.1157,
+      "step": 4890
+    },
+    {
+      "epoch": 18.702290076335878,
+      "grad_norm": 0.6558853387832642,
+      "learning_rate": 9.653130749828074e-06,
+      "loss": 0.1165,
+      "step": 4900
+    },
+    {
+      "epoch": 18.740458015267176,
+      "grad_norm": 0.5114887952804565,
+      "learning_rate": 9.651110848421011e-06,
+      "loss": 0.1238,
+      "step": 4910
+    },
+    {
+      "epoch": 18.778625954198475,
+      "grad_norm": 0.7068324089050293,
+      "learning_rate": 9.649085295475695e-06,
+      "loss": 0.1141,
+      "step": 4920
+    },
+    {
+      "epoch": 18.81679389312977,
+      "grad_norm": 0.47622790932655334,
+      "learning_rate": 9.647054093453366e-06,
+      "loss": 0.1117,
+      "step": 4930
+    },
+    {
+      "epoch": 18.85496183206107,
+      "grad_norm": 0.320960134267807,
+      "learning_rate": 9.645017244822124e-06,
+      "loss": 0.1023,
+      "step": 4940
+    },
+    {
+      "epoch": 18.893129770992367,
+      "grad_norm": 0.3788800835609436,
+      "learning_rate": 9.642974752056931e-06,
+      "loss": 0.119,
+      "step": 4950
+    },
+    {
+      "epoch": 18.931297709923665,
+      "grad_norm": 1.6596555709838867,
+      "learning_rate": 9.640926617639614e-06,
+      "loss": 0.1351,
+      "step": 4960
+    },
+    {
+      "epoch": 18.96946564885496,
+      "grad_norm": 0.5164493918418884,
+      "learning_rate": 9.638872844058844e-06,
+      "loss": 0.1051,
+      "step": 4970
+    },
+    {
+      "epoch": 19.00763358778626,
+      "grad_norm": 0.41121843457221985,
+      "learning_rate": 9.636813433810151e-06,
+      "loss": 0.1129,
+      "step": 4980
+    },
+    {
+      "epoch": 19.045801526717558,
+      "grad_norm": 0.3111073672771454,
+      "learning_rate": 9.634748389395914e-06,
+      "loss": 0.1084,
+      "step": 4990
+    },
+    {
+      "epoch": 19.083969465648856,
+      "grad_norm": 0.2994742691516876,
+      "learning_rate": 9.632677713325353e-06,
+      "loss": 0.1048,
+      "step": 5000
+    },
+    {
+      "epoch": 19.12213740458015,
+      "grad_norm": 0.32396626472473145,
+      "learning_rate": 9.63060140811454e-06,
+      "loss": 0.1049,
+      "step": 5010
+    },
+    {
+      "epoch": 19.16030534351145,
+      "grad_norm": 0.42831170558929443,
+      "learning_rate": 9.628519476286379e-06,
+      "loss": 0.1001,
+      "step": 5020
+    },
+    {
+      "epoch": 19.198473282442748,
+      "grad_norm": 0.46095895767211914,
+      "learning_rate": 9.626431920370613e-06,
+      "loss": 0.1203,
+      "step": 5030
+    },
+    {
+      "epoch": 19.236641221374047,
+      "grad_norm": 0.40226057171821594,
+      "learning_rate": 9.624338742903819e-06,
+      "loss": 0.1088,
+      "step": 5040
+    },
+    {
+      "epoch": 19.274809160305345,
+      "grad_norm": 0.5237085819244385,
+      "learning_rate": 9.622239946429407e-06,
+      "loss": 0.1111,
+      "step": 5050
+    },
+    {
+      "epoch": 19.31297709923664,
+      "grad_norm": 0.4355655014514923,
+      "learning_rate": 9.62013553349761e-06,
+      "loss": 0.1147,
+      "step": 5060
+    },
+    {
+      "epoch": 19.35114503816794,
+      "grad_norm": 0.7808547616004944,
+      "learning_rate": 9.61802550666549e-06,
+      "loss": 0.1117,
+      "step": 5070
+    },
+    {
+      "epoch": 19.389312977099237,
+      "grad_norm": 0.3424168527126312,
+      "learning_rate": 9.615909868496928e-06,
+      "loss": 0.1131,
+      "step": 5080
+    },
+    {
+      "epoch": 19.427480916030536,
+      "grad_norm": 0.38718533515930176,
+      "learning_rate": 9.613788621562622e-06,
+      "loss": 0.113,
+      "step": 5090
+    },
+    {
+      "epoch": 19.46564885496183,
+      "grad_norm": 0.3584721088409424,
+      "learning_rate": 9.611661768440092e-06,
+      "loss": 0.1029,
+      "step": 5100
+    },
+    {
+      "epoch": 19.50381679389313,
+      "grad_norm": 0.5123441815376282,
+      "learning_rate": 9.609529311713662e-06,
+      "loss": 0.1108,
+      "step": 5110
+    },
+    {
+      "epoch": 19.541984732824428,
+      "grad_norm": 0.37226274609565735,
+      "learning_rate": 9.607391253974467e-06,
+      "loss": 0.1103,
+      "step": 5120
+    },
+    {
+      "epoch": 19.580152671755727,
+      "grad_norm": 0.542453408241272,
+      "learning_rate": 9.605247597820448e-06,
+      "loss": 0.1119,
+      "step": 5130
+    },
+    {
+      "epoch": 19.61832061068702,
+      "grad_norm": 0.3321150839328766,
+      "learning_rate": 9.603098345856354e-06,
+      "loss": 0.1103,
+      "step": 5140
+    },
+    {
+      "epoch": 19.65648854961832,
+      "grad_norm": 0.4373333752155304,
+      "learning_rate": 9.600943500693724e-06,
+      "loss": 0.1147,
+      "step": 5150
+    },
+    {
+      "epoch": 19.69465648854962,
+      "grad_norm": 0.393868625164032,
+      "learning_rate": 9.598783064950902e-06,
+      "loss": 0.1079,
+      "step": 5160
+    },
+    {
+      "epoch": 19.732824427480917,
+      "grad_norm": 0.5854748487472534,
+      "learning_rate": 9.596617041253017e-06,
+      "loss": 0.116,
+      "step": 5170
+    },
+    {
+      "epoch": 19.770992366412212,
+      "grad_norm": 0.33527272939682007,
+      "learning_rate": 9.594445432231996e-06,
+      "loss": 0.1071,
+      "step": 5180
+    },
+    {
+      "epoch": 19.80916030534351,
+      "grad_norm": 0.38696524500846863,
+      "learning_rate": 9.592268240526546e-06,
+      "loss": 0.1237,
+      "step": 5190
+    },
+    {
+      "epoch": 19.84732824427481,
+      "grad_norm": 0.5272545218467712,
+      "learning_rate": 9.590085468782162e-06,
+      "loss": 0.122,
+      "step": 5200
+    },
+    {
+      "epoch": 19.885496183206108,
+      "grad_norm": 0.3966885209083557,
+      "learning_rate": 9.587897119651115e-06,
+      "loss": 0.1165,
+      "step": 5210
+    },
+    {
+      "epoch": 19.923664122137403,
+      "grad_norm": 0.6632593870162964,
+      "learning_rate": 9.585703195792459e-06,
+      "loss": 0.1279,
+      "step": 5220
+    },
+    {
+      "epoch": 19.9618320610687,
+      "grad_norm": 0.41774246096611023,
+      "learning_rate": 9.583503699872017e-06,
+      "loss": 0.1043,
+      "step": 5230
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 0.38030463457107544,
+      "learning_rate": 9.581298634562382e-06,
+      "loss": 0.1124,
+      "step": 5240
+    },
+    {
+      "epoch": 20.0381679389313,
+      "grad_norm": 0.314238965511322,
+      "learning_rate": 9.579088002542918e-06,
+      "loss": 0.1062,
+      "step": 5250
+    },
+    {
+      "epoch": 20.076335877862597,
+      "grad_norm": 0.3432968556880951,
+      "learning_rate": 9.57687180649975e-06,
+      "loss": 0.1182,
+      "step": 5260
+    },
+    {
+      "epoch": 20.114503816793892,
+      "grad_norm": 0.7172240614891052,
+      "learning_rate": 9.57465004912577e-06,
+      "loss": 0.1114,
+      "step": 5270
+    },
+    {
+      "epoch": 20.15267175572519,
+      "grad_norm": 0.3679429590702057,
+      "learning_rate": 9.572422733120614e-06,
+      "loss": 0.1146,
+      "step": 5280
+    },
+    {
+      "epoch": 20.19083969465649,
+      "grad_norm": 0.8204743266105652,
+      "learning_rate": 9.57018986119069e-06,
+      "loss": 0.1215,
+      "step": 5290
+    },
+    {
+      "epoch": 20.229007633587788,
+      "grad_norm": 0.5028970241546631,
+      "learning_rate": 9.56795143604914e-06,
+      "loss": 0.1157,
+      "step": 5300
+    },
+    {
+      "epoch": 20.267175572519083,
+      "grad_norm": 0.5000373125076294,
+      "learning_rate": 9.56570746041587e-06,
+      "loss": 0.116,
+      "step": 5310
+    },
+    {
+      "epoch": 20.30534351145038,
+      "grad_norm": 0.36823147535324097,
+      "learning_rate": 9.563457937017514e-06,
+      "loss": 0.1211,
+      "step": 5320
+    },
+    {
+      "epoch": 20.34351145038168,
+      "grad_norm": 0.39545708894729614,
+      "learning_rate": 9.56120286858746e-06,
+      "loss": 0.1164,
+      "step": 5330
+    },
+    {
+      "epoch": 20.38167938931298,
+      "grad_norm": 1.3910619020462036,
+      "learning_rate": 9.558942257865829e-06,
+      "loss": 0.124,
+      "step": 5340
+    },
+    {
+      "epoch": 20.419847328244273,
+      "grad_norm": 0.7138753533363342,
+      "learning_rate": 9.556676107599472e-06,
+      "loss": 0.1313,
+      "step": 5350
+    },
+    {
+      "epoch": 20.458015267175572,
+      "grad_norm": 0.36400383710861206,
+      "learning_rate": 9.554404420541979e-06,
+      "loss": 0.1193,
+      "step": 5360
+    },
+    {
+      "epoch": 20.49618320610687,
+      "grad_norm": 0.42554593086242676,
+      "learning_rate": 9.552127199453662e-06,
+      "loss": 0.1122,
+      "step": 5370
+    },
+    {
+      "epoch": 20.53435114503817,
+      "grad_norm": 0.3719650208950043,
+      "learning_rate": 9.549844447101559e-06,
+      "loss": 0.1112,
+      "step": 5380
+    },
+    {
+      "epoch": 20.572519083969464,
+      "grad_norm": 0.37084949016571045,
+      "learning_rate": 9.547556166259433e-06,
+      "loss": 0.1121,
+      "step": 5390
+    },
+    {
+      "epoch": 20.610687022900763,
+      "grad_norm": 0.4586028456687927,
+      "learning_rate": 9.545262359707756e-06,
+      "loss": 0.1044,
+      "step": 5400
+    },
+    {
+      "epoch": 20.64885496183206,
+      "grad_norm": 0.5926239490509033,
+      "learning_rate": 9.542963030233725e-06,
+      "loss": 0.1071,
+      "step": 5410
+    },
+    {
+      "epoch": 20.68702290076336,
+      "grad_norm": 0.4982014000415802,
+      "learning_rate": 9.540658180631236e-06,
+      "loss": 0.1117,
+      "step": 5420
+    },
+    {
+      "epoch": 20.725190839694655,
+      "grad_norm": 0.6193251013755798,
+      "learning_rate": 9.538347813700903e-06,
+      "loss": 0.1083,
+      "step": 5430
+    },
+    {
+      "epoch": 20.763358778625953,
+      "grad_norm": 0.3325391411781311,
+      "learning_rate": 9.536031932250037e-06,
+      "loss": 0.135,
+      "step": 5440
+    },
+    {
+      "epoch": 20.801526717557252,
+      "grad_norm": 0.2922006845474243,
+      "learning_rate": 9.533710539092653e-06,
+      "loss": 0.1128,
+      "step": 5450
+    },
+    {
+      "epoch": 20.83969465648855,
+      "grad_norm": 0.6511474251747131,
+      "learning_rate": 9.531383637049465e-06,
+      "loss": 0.119,
+      "step": 5460
+    },
+    {
+      "epoch": 20.87786259541985,
+      "grad_norm": 0.7577188014984131,
+      "learning_rate": 9.529051228947875e-06,
+      "loss": 0.1144,
+      "step": 5470
+    },
+    {
+      "epoch": 20.916030534351144,
+      "grad_norm": 0.8727062940597534,
+      "learning_rate": 9.52671331762198e-06,
+      "loss": 0.1243,
+      "step": 5480
+    },
+    {
+      "epoch": 20.954198473282442,
+      "grad_norm": 0.4881748557090759,
+      "learning_rate": 9.524369905912566e-06,
+      "loss": 0.1183,
+      "step": 5490
+    },
+    {
+      "epoch": 20.99236641221374,
+      "grad_norm": 1.3852527141571045,
+      "learning_rate": 9.522020996667092e-06,
+      "loss": 0.109,
+      "step": 5500
+    },
+    {
+      "epoch": 21.03053435114504,
+      "grad_norm": 0.3250916302204132,
+      "learning_rate": 9.51966659273971e-06,
+      "loss": 0.1099,
+      "step": 5510
+    },
+    {
+      "epoch": 21.068702290076335,
+      "grad_norm": 0.3848394453525543,
+      "learning_rate": 9.517306696991241e-06,
+      "loss": 0.1153,
+      "step": 5520
+    },
+    {
+      "epoch": 21.106870229007633,
+      "grad_norm": 0.2975100576877594,
+      "learning_rate": 9.51494131228918e-06,
+      "loss": 0.1031,
+      "step": 5530
+    },
+    {
+      "epoch": 21.14503816793893,
+      "grad_norm": 0.5803893804550171,
+      "learning_rate": 9.512570441507696e-06,
+      "loss": 0.1136,
+      "step": 5540
+    },
+    {
+      "epoch": 21.18320610687023,
+      "grad_norm": 0.5576812028884888,
+      "learning_rate": 9.510194087527615e-06,
+      "loss": 0.1202,
+      "step": 5550
+    },
+    {
+      "epoch": 21.221374045801525,
+      "grad_norm": 0.5307789444923401,
+      "learning_rate": 9.507812253236436e-06,
+      "loss": 0.1223,
+      "step": 5560
+    },
+    {
+      "epoch": 21.259541984732824,
+      "grad_norm": 0.6918527483940125,
+      "learning_rate": 9.50542494152831e-06,
+      "loss": 0.1177,
+      "step": 5570
+    },
+    {
+      "epoch": 21.297709923664122,
+      "grad_norm": 0.489572674036026,
+      "learning_rate": 9.503032155304046e-06,
+      "loss": 0.1104,
+      "step": 5580
+    },
+    {
+      "epoch": 21.33587786259542,
+      "grad_norm": 0.32068297266960144,
+      "learning_rate": 9.500633897471105e-06,
+      "loss": 0.109,
+      "step": 5590
+    },
+    {
+      "epoch": 21.374045801526716,
+      "grad_norm": 0.5839059948921204,
+      "learning_rate": 9.498230170943597e-06,
+      "loss": 0.1197,
+      "step": 5600
+    },
+    {
+      "epoch": 21.412213740458014,
+      "grad_norm": 0.8069677352905273,
+      "learning_rate": 9.495820978642275e-06,
+      "loss": 0.1146,
+      "step": 5610
+    },
+    {
+      "epoch": 21.450381679389313,
+      "grad_norm": 0.5091967582702637,
+      "learning_rate": 9.493406323494537e-06,
+      "loss": 0.1342,
+      "step": 5620
+    },
+    {
+      "epoch": 21.48854961832061,
+      "grad_norm": 0.44150835275650024,
+      "learning_rate": 9.490986208434415e-06,
+      "loss": 0.1161,
+      "step": 5630
+    },
+    {
+      "epoch": 21.52671755725191,
+      "grad_norm": 0.9089648127555847,
+      "learning_rate": 9.488560636402577e-06,
+      "loss": 0.1134,
+      "step": 5640
+    },
+    {
+      "epoch": 21.564885496183205,
+      "grad_norm": 0.6889301538467407,
+      "learning_rate": 9.486129610346322e-06,
+      "loss": 0.1129,
+      "step": 5650
+    },
+    {
+      "epoch": 21.603053435114504,
+      "grad_norm": 0.3800375461578369,
+      "learning_rate": 9.483693133219576e-06,
+      "loss": 0.1042,
+      "step": 5660
+    },
+    {
+      "epoch": 21.641221374045802,
+      "grad_norm": 0.4531780481338501,
+      "learning_rate": 9.481251207982888e-06,
+      "loss": 0.1137,
+      "step": 5670
+    },
+    {
+      "epoch": 21.6793893129771,
+      "grad_norm": 0.4850050210952759,
+      "learning_rate": 9.47880383760343e-06,
+      "loss": 0.115,
+      "step": 5680
+    },
+    {
+      "epoch": 21.717557251908396,
+      "grad_norm": 0.41983506083488464,
+      "learning_rate": 9.476351025054984e-06,
+      "loss": 0.1146,
+      "step": 5690
+    },
+    {
+      "epoch": 21.755725190839694,
+      "grad_norm": 0.47073015570640564,
+      "learning_rate": 9.473892773317952e-06,
+      "loss": 0.1142,
+      "step": 5700
+    },
+    {
+      "epoch": 21.793893129770993,
+      "grad_norm": 0.7770823836326599,
+      "learning_rate": 9.471429085379339e-06,
+      "loss": 0.1043,
+      "step": 5710
+    },
+    {
+      "epoch": 21.83206106870229,
+      "grad_norm": 0.7332136631011963,
+      "learning_rate": 9.468959964232757e-06,
+      "loss": 0.1048,
+      "step": 5720
+    },
+    {
+      "epoch": 21.870229007633586,
+      "grad_norm": 0.4616805911064148,
+      "learning_rate": 9.466485412878425e-06,
+      "loss": 0.1118,
+      "step": 5730
+    },
+    {
+      "epoch": 21.908396946564885,
+      "grad_norm": 0.3106837868690491,
+      "learning_rate": 9.464005434323154e-06,
+      "loss": 0.1116,
+      "step": 5740
+    },
+    {
+      "epoch": 21.946564885496183,
+      "grad_norm": 0.606867790222168,
+      "learning_rate": 9.461520031580352e-06,
+      "loss": 0.1116,
+      "step": 5750
+    },
+    {
+      "epoch": 21.984732824427482,
+      "grad_norm": 0.3522845506668091,
+      "learning_rate": 9.459029207670018e-06,
+      "loss": 0.1007,
+      "step": 5760
+    },
+    {
+      "epoch": 22.022900763358777,
+      "grad_norm": 0.5821218490600586,
+      "learning_rate": 9.456532965618738e-06,
+      "loss": 0.1169,
+      "step": 5770
+    },
+    {
+      "epoch": 22.061068702290076,
+      "grad_norm": 0.6198912262916565,
+      "learning_rate": 9.454031308459681e-06,
+      "loss": 0.1144,
+      "step": 5780
+    },
+    {
+      "epoch": 22.099236641221374,
+      "grad_norm": 0.396658331155777,
+      "learning_rate": 9.451524239232596e-06,
+      "loss": 0.1117,
+      "step": 5790
+    },
+    {
+      "epoch": 22.137404580152673,
+      "grad_norm": 0.4272678792476654,
+      "learning_rate": 9.44901176098381e-06,
+      "loss": 0.13,
+      "step": 5800
+    },
+    {
+      "epoch": 22.17557251908397,
+      "grad_norm": 0.5688971877098083,
+      "learning_rate": 9.446493876766219e-06,
+      "loss": 0.1147,
+      "step": 5810
+    },
+    {
+      "epoch": 22.213740458015266,
+      "grad_norm": 0.8139527440071106,
+      "learning_rate": 9.44397058963929e-06,
+      "loss": 0.1215,
+      "step": 5820
+    },
+    {
+      "epoch": 22.251908396946565,
+      "grad_norm": 0.9012984037399292,
+      "learning_rate": 9.441441902669057e-06,
+      "loss": 0.1144,
+      "step": 5830
+    },
+    {
+      "epoch": 22.290076335877863,
+      "grad_norm": 0.38339322805404663,
+      "learning_rate": 9.43890781892811e-06,
+      "loss": 0.1202,
+      "step": 5840
+    },
+    {
+      "epoch": 22.328244274809162,
+      "grad_norm": 0.34039992094039917,
+      "learning_rate": 9.436368341495603e-06,
+      "loss": 0.118,
+      "step": 5850
+    },
+    {
+      "epoch": 22.366412213740457,
+      "grad_norm": 1.0768190622329712,
+      "learning_rate": 9.43382347345724e-06,
+      "loss": 0.1057,
+      "step": 5860
+    },
+    {
+      "epoch": 22.404580152671755,
+      "grad_norm": 0.682651937007904,
+      "learning_rate": 9.431273217905272e-06,
+      "loss": 0.1111,
+      "step": 5870
+    },
+    {
+      "epoch": 22.442748091603054,
+      "grad_norm": 0.7133651375770569,
+      "learning_rate": 9.428717577938505e-06,
+      "loss": 0.105,
+      "step": 5880
+    },
+    {
+      "epoch": 22.480916030534353,
+      "grad_norm": 0.4730619192123413,
+      "learning_rate": 9.426156556662276e-06,
+      "loss": 0.1056,
+      "step": 5890
+    },
+    {
+      "epoch": 22.519083969465647,
+      "grad_norm": 0.37307435274124146,
+      "learning_rate": 9.423590157188475e-06,
+      "loss": 0.1122,
+      "step": 5900
+    },
+    {
+      "epoch": 22.557251908396946,
+      "grad_norm": 0.3784120976924896,
+      "learning_rate": 9.421018382635514e-06,
+      "loss": 0.1327,
+      "step": 5910
+    },
+    {
+      "epoch": 22.595419847328245,
+      "grad_norm": 0.6031510233879089,
+      "learning_rate": 9.418441236128344e-06,
+      "loss": 0.1074,
+      "step": 5920
+    },
+    {
+      "epoch": 22.633587786259543,
+      "grad_norm": 0.6121441721916199,
+      "learning_rate": 9.41585872079844e-06,
+      "loss": 0.1079,
+      "step": 5930
+    },
+    {
+      "epoch": 22.671755725190838,
+      "grad_norm": 0.5662669539451599,
+      "learning_rate": 9.413270839783802e-06,
+      "loss": 0.1096,
+      "step": 5940
+    },
+    {
+      "epoch": 22.709923664122137,
+      "grad_norm": 0.4588741064071655,
+      "learning_rate": 9.41067759622895e-06,
+      "loss": 0.1125,
+      "step": 5950
+    },
+    {
+      "epoch": 22.748091603053435,
+      "grad_norm": 0.5805822014808655,
+      "learning_rate": 9.408078993284917e-06,
+      "loss": 0.1143,
+      "step": 5960
+    },
+    {
+      "epoch": 22.786259541984734,
+      "grad_norm": 0.5425497889518738,
+      "learning_rate": 9.405475034109254e-06,
+      "loss": 0.1006,
+      "step": 5970
+    },
+    {
+      "epoch": 22.82442748091603,
+      "grad_norm": 0.43856534361839294,
+      "learning_rate": 9.402865721866017e-06,
+      "loss": 0.1156,
+      "step": 5980
+    },
+    {
+      "epoch": 22.862595419847327,
+      "grad_norm": 0.6085226535797119,
+      "learning_rate": 9.400251059725762e-06,
+      "loss": 0.1173,
+      "step": 5990
+    },
+    {
+      "epoch": 22.900763358778626,
+      "grad_norm": 0.41723042726516724,
+      "learning_rate": 9.397631050865554e-06,
+      "loss": 0.115,
+      "step": 6000
+    },
+    {
+      "epoch": 22.938931297709924,
+      "grad_norm": 0.47364330291748047,
+      "learning_rate": 9.395005698468948e-06,
+      "loss": 0.1096,
+      "step": 6010
+    },
+    {
+      "epoch": 22.977099236641223,
+      "grad_norm": 0.5531240701675415,
+      "learning_rate": 9.392375005726e-06,
+      "loss": 0.1115,
+      "step": 6020
+    },
+    {
+      "epoch": 23.015267175572518,
+      "grad_norm": 1.9053860902786255,
+      "learning_rate": 9.389738975833243e-06,
+      "loss": 0.1107,
+      "step": 6030
+    },
+    {
+      "epoch": 23.053435114503817,
+      "grad_norm": 0.4159781336784363,
+      "learning_rate": 9.387097611993707e-06,
+      "loss": 0.1061,
+      "step": 6040
+    },
+    {
+      "epoch": 23.091603053435115,
+      "grad_norm": 0.37060031294822693,
+      "learning_rate": 9.384450917416894e-06,
+      "loss": 0.1087,
+      "step": 6050
+    },
+    {
+      "epoch": 23.129770992366414,
+      "grad_norm": 0.3280550241470337,
+      "learning_rate": 9.381798895318792e-06,
+      "loss": 0.1082,
+      "step": 6060
+    },
+    {
+      "epoch": 23.16793893129771,
+      "grad_norm": 0.6817305088043213,
+      "learning_rate": 9.379141548921855e-06,
+      "loss": 0.1084,
+      "step": 6070
+    },
+    {
+      "epoch": 23.206106870229007,
+      "grad_norm": 1.089829683303833,
+      "learning_rate": 9.376478881455008e-06,
+      "loss": 0.1191,
+      "step": 6080
+    },
+    {
+      "epoch": 23.244274809160306,
+      "grad_norm": 0.514206051826477,
+      "learning_rate": 9.373810896153647e-06,
+      "loss": 0.108,
+      "step": 6090
+    },
+    {
+      "epoch": 23.282442748091604,
+      "grad_norm": 0.3831291198730469,
+      "learning_rate": 9.371137596259623e-06,
+      "loss": 0.1141,
+      "step": 6100
+    },
+    {
+      "epoch": 23.3206106870229,
+      "grad_norm": 0.4334690570831299,
+      "learning_rate": 9.368458985021249e-06,
+      "loss": 0.11,
+      "step": 6110
+    },
+    {
+      "epoch": 23.358778625954198,
+      "grad_norm": 0.43179851770401,
+      "learning_rate": 9.365775065693288e-06,
+      "loss": 0.1128,
+      "step": 6120
+    },
+    {
+      "epoch": 23.396946564885496,
+      "grad_norm": 0.43169301748275757,
+      "learning_rate": 9.363085841536958e-06,
+      "loss": 0.1113,
+      "step": 6130
+    },
+    {
+      "epoch": 23.435114503816795,
+      "grad_norm": 0.6322646737098694,
+      "learning_rate": 9.360391315819917e-06,
+      "loss": 0.1113,
+      "step": 6140
+    },
+    {
+      "epoch": 23.47328244274809,
+      "grad_norm": 0.37557345628738403,
+      "learning_rate": 9.35769149181627e-06,
+      "loss": 0.1107,
+      "step": 6150
+    },
+    {
+      "epoch": 23.51145038167939,
+      "grad_norm": 0.48005855083465576,
+      "learning_rate": 9.354986372806557e-06,
+      "loss": 0.1067,
+      "step": 6160
+    },
+    {
+      "epoch": 23.549618320610687,
+      "grad_norm": 0.5975325107574463,
+      "learning_rate": 9.352275962077752e-06,
+      "loss": 0.1199,
+      "step": 6170
+    },
+    {
+      "epoch": 23.587786259541986,
+      "grad_norm": 0.49410223960876465,
+      "learning_rate": 9.349560262923262e-06,
+      "loss": 0.1296,
+      "step": 6180
+    },
+    {
+      "epoch": 23.625954198473284,
+      "grad_norm": 0.3413020670413971,
+      "learning_rate": 9.346839278642915e-06,
+      "loss": 0.1102,
+      "step": 6190
+    },
+    {
+      "epoch": 23.66412213740458,
+      "grad_norm": 0.43629974126815796,
+      "learning_rate": 9.344113012542964e-06,
+      "loss": 0.1163,
+      "step": 6200
+    },
+    {
+      "epoch": 23.702290076335878,
+      "grad_norm": 0.4831625521183014,
+      "learning_rate": 9.341381467936079e-06,
+      "loss": 0.1056,
+      "step": 6210
+    },
+    {
+      "epoch": 23.740458015267176,
+      "grad_norm": 0.4807736575603485,
+      "learning_rate": 9.338644648141347e-06,
+      "loss": 0.1002,
+      "step": 6220
+    },
+    {
+      "epoch": 23.778625954198475,
+      "grad_norm": 0.6690664887428284,
+      "learning_rate": 9.335902556484258e-06,
+      "loss": 0.112,
+      "step": 6230
+    },
+    {
+      "epoch": 23.81679389312977,
+      "grad_norm": 0.7592681646347046,
+      "learning_rate": 9.333155196296712e-06,
+      "loss": 0.1051,
+      "step": 6240
+    },
+    {
+      "epoch": 23.85496183206107,
+      "grad_norm": 0.3155533969402313,
+      "learning_rate": 9.330402570917017e-06,
+      "loss": 0.1104,
+      "step": 6250
+    },
+    {
+      "epoch": 23.893129770992367,
+      "grad_norm": 0.43112990260124207,
+      "learning_rate": 9.327644683689866e-06,
+      "loss": 0.1098,
+      "step": 6260
+    },
+    {
+      "epoch": 23.931297709923665,
+      "grad_norm": 0.28097161650657654,
+      "learning_rate": 9.324881537966355e-06,
+      "loss": 0.1086,
+      "step": 6270
+    },
+    {
+      "epoch": 23.96946564885496,
+      "grad_norm": 0.48976218700408936,
+      "learning_rate": 9.322113137103964e-06,
+      "loss": 0.1106,
+      "step": 6280
+    },
+    {
+      "epoch": 24.00763358778626,
+      "grad_norm": 0.5290048718452454,
+      "learning_rate": 9.319339484466565e-06,
+      "loss": 0.1065,
+      "step": 6290
+    },
+    {
+      "epoch": 24.045801526717558,
+      "grad_norm": 0.48621800541877747,
+      "learning_rate": 9.316560583424404e-06,
+      "loss": 0.106,
+      "step": 6300
+    },
+    {
+      "epoch": 24.083969465648856,
+      "grad_norm": 0.3502595126628876,
+      "learning_rate": 9.313776437354109e-06,
+      "loss": 0.105,
+      "step": 6310
+    },
+    {
+      "epoch": 24.12213740458015,
+      "grad_norm": 0.3239123225212097,
+      "learning_rate": 9.310987049638681e-06,
+      "loss": 0.1121,
+      "step": 6320
+    },
+    {
+      "epoch": 24.16030534351145,
+      "grad_norm": 0.38720473647117615,
+      "learning_rate": 9.308192423667486e-06,
+      "loss": 0.1035,
+      "step": 6330
+    },
+    {
+      "epoch": 24.198473282442748,
+      "grad_norm": 0.382134348154068,
+      "learning_rate": 9.305392562836262e-06,
+      "loss": 0.1073,
+      "step": 6340
+    },
+    {
+      "epoch": 24.236641221374047,
+      "grad_norm": 0.37410208582878113,
+      "learning_rate": 9.302587470547101e-06,
+      "loss": 0.1247,
+      "step": 6350
+    },
+    {
+      "epoch": 24.274809160305345,
+      "grad_norm": 0.3606686592102051,
+      "learning_rate": 9.299777150208456e-06,
+      "loss": 0.1206,
+      "step": 6360
+    },
+    {
+      "epoch": 24.31297709923664,
+      "grad_norm": 0.5555996894836426,
+      "learning_rate": 9.296961605235133e-06,
+      "loss": 0.1219,
+      "step": 6370
+    },
+    {
+      "epoch": 24.35114503816794,
+      "grad_norm": 0.40895533561706543,
+      "learning_rate": 9.29414083904828e-06,
+      "loss": 0.1151,
+      "step": 6380
+    },
+    {
+      "epoch": 24.389312977099237,
+      "grad_norm": 0.3560793101787567,
+      "learning_rate": 9.291314855075397e-06,
+      "loss": 0.1171,
+      "step": 6390
+    },
+    {
+      "epoch": 24.427480916030536,
+      "grad_norm": 0.583595871925354,
+      "learning_rate": 9.288483656750322e-06,
+      "loss": 0.1203,
+      "step": 6400
+    },
+    {
+      "epoch": 24.46564885496183,
+      "grad_norm": 0.3709441125392914,
+      "learning_rate": 9.285647247513225e-06,
+      "loss": 0.1087,
+      "step": 6410
+    },
+    {
+      "epoch": 24.50381679389313,
+      "grad_norm": 0.5347302556037903,
+      "learning_rate": 9.282805630810614e-06,
+      "loss": 0.1089,
+      "step": 6420
+    },
+    {
+      "epoch": 24.541984732824428,
+      "grad_norm": 0.9013820290565491,
+      "learning_rate": 9.279958810095317e-06,
+      "loss": 0.1124,
+      "step": 6430
+    },
+    {
+      "epoch": 24.580152671755727,
+      "grad_norm": 0.32639890909194946,
+      "learning_rate": 9.277106788826494e-06,
+      "loss": 0.1224,
+      "step": 6440
+    },
+    {
+      "epoch": 24.61832061068702,
+      "grad_norm": 0.462785542011261,
+      "learning_rate": 9.274249570469618e-06,
+      "loss": 0.1157,
+      "step": 6450
+    },
+    {
+      "epoch": 24.65648854961832,
+      "grad_norm": 0.6265701651573181,
+      "learning_rate": 9.271387158496477e-06,
+      "loss": 0.1041,
+      "step": 6460
+    },
+    {
+      "epoch": 24.69465648854962,
+      "grad_norm": 0.4119669198989868,
+      "learning_rate": 9.268519556385173e-06,
+      "loss": 0.1091,
+      "step": 6470
+    },
+    {
+      "epoch": 24.732824427480917,
+      "grad_norm": 0.2758201062679291,
+      "learning_rate": 9.265646767620113e-06,
+      "loss": 0.1023,
+      "step": 6480
+    },
+    {
+      "epoch": 24.770992366412212,
+      "grad_norm": 0.4721677601337433,
+      "learning_rate": 9.262768795692006e-06,
+      "loss": 0.1004,
+      "step": 6490
+    },
+    {
+      "epoch": 24.80916030534351,
+      "grad_norm": 0.4377496838569641,
+      "learning_rate": 9.259885644097861e-06,
+      "loss": 0.1251,
+      "step": 6500
+    },
+    {
+      "epoch": 24.84732824427481,
+      "grad_norm": 0.3082864284515381,
+      "learning_rate": 9.256997316340976e-06,
+      "loss": 0.1116,
+      "step": 6510
+    },
+    {
+      "epoch": 24.885496183206108,
+      "grad_norm": 0.36191001534461975,
+      "learning_rate": 9.254103815930944e-06,
+      "loss": 0.1164,
+      "step": 6520
+    },
+    {
+      "epoch": 24.923664122137403,
+      "grad_norm": 0.34867507219314575,
+      "learning_rate": 9.251205146383637e-06,
+      "loss": 0.1155,
+      "step": 6530
+    },
+    {
+      "epoch": 24.9618320610687,
+      "grad_norm": 0.665786862373352,
+      "learning_rate": 9.248301311221216e-06,
+      "loss": 0.1075,
+      "step": 6540
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 0.5270240306854248,
+      "learning_rate": 9.245392313972116e-06,
+      "loss": 0.112,
+      "step": 6550
+    },
+    {
+      "epoch": 25.0381679389313,
+      "grad_norm": 0.5124590396881104,
+      "learning_rate": 9.242478158171038e-06,
+      "loss": 0.1091,
+      "step": 6560
+    },
+    {
+      "epoch": 25.076335877862597,
+      "grad_norm": 0.6183871626853943,
+      "learning_rate": 9.239558847358959e-06,
+      "loss": 0.1148,
+      "step": 6570
+    },
+    {
+      "epoch": 25.114503816793892,
+      "grad_norm": 0.4855692982673645,
+      "learning_rate": 9.236634385083115e-06,
+      "loss": 0.1086,
+      "step": 6580
+    },
+    {
+      "epoch": 25.15267175572519,
+      "grad_norm": 0.37100329995155334,
+      "learning_rate": 9.233704774897006e-06,
+      "loss": 0.1123,
+      "step": 6590
+    },
+    {
+      "epoch": 25.19083969465649,
+      "grad_norm": 0.5242245197296143,
+      "learning_rate": 9.230770020360383e-06,
+      "loss": 0.1155,
+      "step": 6600
+    },
+    {
+      "epoch": 25.229007633587788,
+      "grad_norm": 0.36837512254714966,
+      "learning_rate": 9.227830125039248e-06,
+      "loss": 0.1312,
+      "step": 6610
+    },
+    {
+      "epoch": 25.267175572519083,
+      "grad_norm": 0.4537639915943146,
+      "learning_rate": 9.224885092505853e-06,
+      "loss": 0.1111,
+      "step": 6620
+    },
+    {
+      "epoch": 25.30534351145038,
+      "grad_norm": 0.42237746715545654,
+      "learning_rate": 9.22193492633869e-06,
+      "loss": 0.1185,
+      "step": 6630
+    },
+    {
+      "epoch": 25.34351145038168,
+      "grad_norm": 0.6770774722099304,
+      "learning_rate": 9.21897963012249e-06,
+      "loss": 0.1094,
+      "step": 6640
+    },
+    {
+      "epoch": 25.38167938931298,
+      "grad_norm": 0.5275667905807495,
+      "learning_rate": 9.216019207448216e-06,
+      "loss": 0.1112,
+      "step": 6650
+    },
+    {
+      "epoch": 25.419847328244273,
+      "grad_norm": 0.5454118251800537,
+      "learning_rate": 9.213053661913061e-06,
+      "loss": 0.108,
+      "step": 6660
+    },
+    {
+      "epoch": 25.458015267175572,
+      "grad_norm": 0.4298250079154968,
+      "learning_rate": 9.210082997120439e-06,
+      "loss": 0.108,
+      "step": 6670
+    },
+    {
+      "epoch": 25.49618320610687,
+      "grad_norm": 0.5686354637145996,
+      "learning_rate": 9.207107216679994e-06,
+      "loss": 0.121,
+      "step": 6680
+    },
+    {
+      "epoch": 25.53435114503817,
+      "grad_norm": 0.7463754415512085,
+      "learning_rate": 9.204126324207575e-06,
+      "loss": 0.1155,
+      "step": 6690
+    },
+    {
+      "epoch": 25.572519083969464,
+      "grad_norm": 0.9898240566253662,
+      "learning_rate": 9.201140323325248e-06,
+      "loss": 0.1226,
+      "step": 6700
+    },
+    {
+      "epoch": 25.610687022900763,
+      "grad_norm": 0.40233856439590454,
+      "learning_rate": 9.198149217661287e-06,
+      "loss": 0.1073,
+      "step": 6710
+    },
+    {
+      "epoch": 25.64885496183206,
+      "grad_norm": 0.44068315625190735,
+      "learning_rate": 9.195153010850166e-06,
+      "loss": 0.1161,
+      "step": 6720
+    },
+    {
+      "epoch": 25.68702290076336,
+      "grad_norm": 0.5321930050849915,
+      "learning_rate": 9.192151706532562e-06,
+      "loss": 0.1041,
+      "step": 6730
+    },
+    {
+      "epoch": 25.725190839694655,
+      "grad_norm": 0.34386226534843445,
+      "learning_rate": 9.189145308355339e-06,
+      "loss": 0.111,
+      "step": 6740
+    },
+    {
+      "epoch": 25.763358778625953,
+      "grad_norm": 0.9037399291992188,
+      "learning_rate": 9.186133819971556e-06,
+      "loss": 0.1146,
+      "step": 6750
+    },
+    {
+      "epoch": 25.801526717557252,
+      "grad_norm": 0.2989799976348877,
+      "learning_rate": 9.183117245040455e-06,
+      "loss": 0.1291,
+      "step": 6760
+    },
+    {
+      "epoch": 25.83969465648855,
+      "grad_norm": 0.5754449367523193,
+      "learning_rate": 9.18009558722746e-06,
+      "loss": 0.1061,
+      "step": 6770
+    },
+    {
+      "epoch": 25.87786259541985,
+      "grad_norm": 0.3005794882774353,
+      "learning_rate": 9.177068850204167e-06,
+      "loss": 0.1062,
+      "step": 6780
+    },
+    {
+      "epoch": 25.916030534351144,
+      "grad_norm": 0.466094046831131,
+      "learning_rate": 9.174037037648351e-06,
+      "loss": 0.119,
+      "step": 6790
+    },
+    {
+      "epoch": 25.954198473282442,
+      "grad_norm": 0.44860929250717163,
+      "learning_rate": 9.171000153243948e-06,
+      "loss": 0.1121,
+      "step": 6800
+    },
+    {
+      "epoch": 25.99236641221374,
+      "grad_norm": 0.6850860714912415,
+      "learning_rate": 9.167958200681058e-06,
+      "loss": 0.1133,
+      "step": 6810
+    },
+    {
+      "epoch": 26.03053435114504,
+      "grad_norm": 0.5172090530395508,
+      "learning_rate": 9.164911183655943e-06,
+      "loss": 0.1096,
+      "step": 6820
+    },
+    {
+      "epoch": 26.068702290076335,
+      "grad_norm": 0.28496426343917847,
+      "learning_rate": 9.161859105871013e-06,
+      "loss": 0.1001,
+      "step": 6830
+    },
+    {
+      "epoch": 26.106870229007633,
+      "grad_norm": 0.4888004958629608,
+      "learning_rate": 9.158801971034832e-06,
+      "loss": 0.0996,
+      "step": 6840
+    },
+    {
+      "epoch": 26.14503816793893,
+      "grad_norm": 0.4661968946456909,
+      "learning_rate": 9.155739782862107e-06,
+      "loss": 0.1062,
+      "step": 6850
+    },
+    {
+      "epoch": 26.18320610687023,
+      "grad_norm": 0.6031951308250427,
+      "learning_rate": 9.152672545073687e-06,
+      "loss": 0.1138,
+      "step": 6860
+    },
+    {
+      "epoch": 26.221374045801525,
+      "grad_norm": 0.35957977175712585,
+      "learning_rate": 9.149600261396552e-06,
+      "loss": 0.1061,
+      "step": 6870
+    },
+    {
+      "epoch": 26.259541984732824,
+      "grad_norm": 0.3762602210044861,
+      "learning_rate": 9.146522935563816e-06,
+      "loss": 0.1143,
+      "step": 6880
+    },
+    {
+      "epoch": 26.297709923664122,
+      "grad_norm": 0.763342559337616,
+      "learning_rate": 9.143440571314723e-06,
+      "loss": 0.1143,
+      "step": 6890
+    },
+    {
+      "epoch": 26.33587786259542,
+      "grad_norm": 0.331662654876709,
+      "learning_rate": 9.140353172394637e-06,
+      "loss": 0.1009,
+      "step": 6900
+    },
+    {
+      "epoch": 26.374045801526716,
+      "grad_norm": 0.6234320402145386,
+      "learning_rate": 9.137260742555033e-06,
+      "loss": 0.1046,
+      "step": 6910
+    },
+    {
+      "epoch": 26.412213740458014,
+      "grad_norm": 0.2962917685508728,
+      "learning_rate": 9.134163285553511e-06,
+      "loss": 0.1058,
+      "step": 6920
+    },
+    {
+      "epoch": 26.450381679389313,
+      "grad_norm": 0.43455183506011963,
+      "learning_rate": 9.13106080515377e-06,
+      "loss": 0.116,
+      "step": 6930
+    },
+    {
+      "epoch": 26.48854961832061,
+      "grad_norm": 0.4187309145927429,
+      "learning_rate": 9.127953305125618e-06,
+      "loss": 0.1063,
+      "step": 6940
+    },
+    {
+      "epoch": 26.52671755725191,
+      "grad_norm": 0.5334097146987915,
+      "learning_rate": 9.124840789244958e-06,
+      "loss": 0.1191,
+      "step": 6950
+    },
+    {
+      "epoch": 26.564885496183205,
+      "grad_norm": 0.5690393447875977,
+      "learning_rate": 9.121723261293793e-06,
+      "loss": 0.1119,
+      "step": 6960
+    },
+    {
+      "epoch": 26.603053435114504,
+      "grad_norm": 0.41400864720344543,
+      "learning_rate": 9.118600725060214e-06,
+      "loss": 0.1023,
+      "step": 6970
+    },
+    {
+      "epoch": 26.641221374045802,
+      "grad_norm": 0.4753328561782837,
+      "learning_rate": 9.115473184338393e-06,
+      "loss": 0.1104,
+      "step": 6980
+    },
+    {
+      "epoch": 26.6793893129771,
+      "grad_norm": 0.29748255014419556,
+      "learning_rate": 9.11234064292859e-06,
+      "loss": 0.1003,
+      "step": 6990
+    },
+    {
+      "epoch": 26.717557251908396,
+      "grad_norm": 0.34052833914756775,
+      "learning_rate": 9.109203104637138e-06,
+      "loss": 0.1167,
+      "step": 7000
+    },
+    {
+      "epoch": 26.755725190839694,
+      "grad_norm": 0.39550068974494934,
+      "learning_rate": 9.10606057327644e-06,
+      "loss": 0.1128,
+      "step": 7010
+    },
+    {
+      "epoch": 26.793893129770993,
+      "grad_norm": 0.3241926431655884,
+      "learning_rate": 9.102913052664971e-06,
+      "loss": 0.1078,
+      "step": 7020
+    },
+    {
+      "epoch": 26.83206106870229,
+      "grad_norm": 0.3458397686481476,
+      "learning_rate": 9.099760546627262e-06,
+      "loss": 0.1104,
+      "step": 7030
+    },
+    {
+      "epoch": 26.870229007633586,
+      "grad_norm": 0.6425669193267822,
+      "learning_rate": 9.096603058993907e-06,
+      "loss": 0.1119,
+      "step": 7040
+    },
+    {
+      "epoch": 26.908396946564885,
+      "grad_norm": 0.39472290873527527,
+      "learning_rate": 9.093440593601553e-06,
+      "loss": 0.1109,
+      "step": 7050
+    },
+    {
+      "epoch": 26.946564885496183,
+      "grad_norm": 0.3740868866443634,
+      "learning_rate": 9.090273154292889e-06,
+      "loss": 0.1181,
+      "step": 7060
+    },
+    {
+      "epoch": 26.984732824427482,
+      "grad_norm": 0.38200682401657104,
+      "learning_rate": 9.087100744916656e-06,
+      "loss": 0.1065,
+      "step": 7070
+    },
+    {
+      "epoch": 27.022900763358777,
+      "grad_norm": 0.4096829295158386,
+      "learning_rate": 9.08392336932763e-06,
+      "loss": 0.1099,
+      "step": 7080
+    },
+    {
+      "epoch": 27.061068702290076,
+      "grad_norm": 0.6823604702949524,
+      "learning_rate": 9.08074103138662e-06,
+      "loss": 0.1167,
+      "step": 7090
+    },
+    {
+      "epoch": 27.099236641221374,
+      "grad_norm": 0.5257452726364136,
+      "learning_rate": 9.077553734960469e-06,
+      "loss": 0.1047,
+      "step": 7100
+    },
+    {
+      "epoch": 27.137404580152673,
+      "grad_norm": 0.4600130617618561,
+      "learning_rate": 9.074361483922041e-06,
+      "loss": 0.1006,
+      "step": 7110
+    },
+    {
+      "epoch": 27.17557251908397,
+      "grad_norm": 0.6489649415016174,
+      "learning_rate": 9.071164282150224e-06,
+      "loss": 0.1035,
+      "step": 7120
+    },
+    {
+      "epoch": 27.213740458015266,
+      "grad_norm": 1.3038793802261353,
+      "learning_rate": 9.067962133529919e-06,
+      "loss": 0.1039,
+      "step": 7130
+    },
+    {
+      "epoch": 27.251908396946565,
+      "grad_norm": 0.542229950428009,
+      "learning_rate": 9.064755041952036e-06,
+      "loss": 0.1162,
+      "step": 7140
+    },
+    {
+      "epoch": 27.290076335877863,
+      "grad_norm": 0.3620493412017822,
+      "learning_rate": 9.061543011313498e-06,
+      "loss": 0.1153,
+      "step": 7150
+    },
+    {
+      "epoch": 27.328244274809162,
+      "grad_norm": 0.44058114290237427,
+      "learning_rate": 9.05832604551722e-06,
+      "loss": 0.1148,
+      "step": 7160
+    },
+    {
+      "epoch": 27.366412213740457,
+      "grad_norm": 0.53098064661026,
+      "learning_rate": 9.055104148472123e-06,
+      "loss": 0.1158,
+      "step": 7170
+    },
+    {
+      "epoch": 27.404580152671755,
+      "grad_norm": 0.7122390270233154,
+      "learning_rate": 9.051877324093114e-06,
+      "loss": 0.1186,
+      "step": 7180
+    },
+    {
+      "epoch": 27.442748091603054,
+      "grad_norm": 0.6435628533363342,
+      "learning_rate": 9.04864557630109e-06,
+      "loss": 0.1176,
+      "step": 7190
+    },
+    {
+      "epoch": 27.480916030534353,
+      "grad_norm": 0.2909921705722809,
+      "learning_rate": 9.045408909022928e-06,
+      "loss": 0.1115,
+      "step": 7200
+    },
+    {
+      "epoch": 27.519083969465647,
+      "grad_norm": 0.39591115713119507,
+      "learning_rate": 9.042167326191484e-06,
+      "loss": 0.1077,
+      "step": 7210
+    },
+    {
+      "epoch": 27.557251908396946,
+      "grad_norm": 0.7656340003013611,
+      "learning_rate": 9.038920831745587e-06,
+      "loss": 0.1072,
+      "step": 7220
+    },
+    {
+      "epoch": 27.595419847328245,
+      "grad_norm": 0.5420284271240234,
+      "learning_rate": 9.035669429630036e-06,
+      "loss": 0.1113,
+      "step": 7230
+    },
+    {
+      "epoch": 27.633587786259543,
+      "grad_norm": 0.3711391091346741,
+      "learning_rate": 9.032413123795589e-06,
+      "loss": 0.1094,
+      "step": 7240
+    },
+    {
+      "epoch": 27.671755725190838,
+      "grad_norm": 0.6970059871673584,
+      "learning_rate": 9.029151918198962e-06,
+      "loss": 0.1083,
+      "step": 7250
+    },
+    {
+      "epoch": 27.709923664122137,
+      "grad_norm": 0.5760360956192017,
+      "learning_rate": 9.025885816802833e-06,
+      "loss": 0.1195,
+      "step": 7260
+    },
+    {
+      "epoch": 27.748091603053435,
+      "grad_norm": 0.3566513955593109,
+      "learning_rate": 9.022614823575819e-06,
+      "loss": 0.1091,
+      "step": 7270
+    },
+    {
+      "epoch": 27.786259541984734,
+      "grad_norm": 0.3161095380783081,
+      "learning_rate": 9.019338942492485e-06,
+      "loss": 0.1095,
+      "step": 7280
+    },
+    {
+      "epoch": 27.82442748091603,
+      "grad_norm": 0.3499198853969574,
+      "learning_rate": 9.01605817753334e-06,
+      "loss": 0.1113,
+      "step": 7290
+    },
+    {
+      "epoch": 27.862595419847327,
+      "grad_norm": 0.3654879927635193,
+      "learning_rate": 9.012772532684819e-06,
+      "loss": 0.111,
+      "step": 7300
+    },
+    {
+      "epoch": 27.900763358778626,
+      "grad_norm": 0.5612740516662598,
+      "learning_rate": 9.00948201193929e-06,
+      "loss": 0.107,
+      "step": 7310
+    },
+    {
+      "epoch": 27.938931297709924,
+      "grad_norm": 0.6282134652137756,
+      "learning_rate": 9.006186619295048e-06,
+      "loss": 0.117,
+      "step": 7320
+    },
+    {
+      "epoch": 27.977099236641223,
+      "grad_norm": 0.9098697900772095,
+      "learning_rate": 9.002886358756304e-06,
+      "loss": 0.115,
+      "step": 7330
+    },
+    {
+      "epoch": 28.015267175572518,
+      "grad_norm": 0.5928300619125366,
+      "learning_rate": 8.999581234333189e-06,
+      "loss": 0.1075,
+      "step": 7340
+    },
+    {
+      "epoch": 28.053435114503817,
+      "grad_norm": 0.39797747135162354,
+      "learning_rate": 8.996271250041735e-06,
+      "loss": 0.1185,
+      "step": 7350
+    },
+    {
+      "epoch": 28.091603053435115,
+      "grad_norm": 0.5730596780776978,
+      "learning_rate": 8.99295640990389e-06,
+      "loss": 0.1113,
+      "step": 7360
+    },
+    {
+      "epoch": 28.129770992366414,
+      "grad_norm": 0.40439486503601074,
+      "learning_rate": 8.989636717947496e-06,
+      "loss": 0.1126,
+      "step": 7370
+    },
+    {
+      "epoch": 28.16793893129771,
+      "grad_norm": 0.978628933429718,
+      "learning_rate": 8.986312178206291e-06,
+      "loss": 0.1018,
+      "step": 7380
+    },
+    {
+      "epoch": 28.206106870229007,
+      "grad_norm": 0.45754435658454895,
+      "learning_rate": 8.982982794719904e-06,
+      "loss": 0.1116,
+      "step": 7390
+    },
+    {
+      "epoch": 28.244274809160306,
+      "grad_norm": 0.3176497519016266,
+      "learning_rate": 8.979648571533852e-06,
+      "loss": 0.1055,
+      "step": 7400
+    },
+    {
+      "epoch": 28.282442748091604,
+      "grad_norm": 0.3644157946109772,
+      "learning_rate": 8.97630951269953e-06,
+      "loss": 0.1086,
+      "step": 7410
+    },
+    {
+      "epoch": 28.3206106870229,
+      "grad_norm": 0.4856671988964081,
+      "learning_rate": 8.972965622274206e-06,
+      "loss": 0.1148,
+      "step": 7420
+    },
+    {
+      "epoch": 28.358778625954198,
+      "grad_norm": 0.3878462314605713,
+      "learning_rate": 8.969616904321026e-06,
+      "loss": 0.1041,
+      "step": 7430
+    },
+    {
+      "epoch": 28.396946564885496,
+      "grad_norm": 0.4682347774505615,
+      "learning_rate": 8.966263362908998e-06,
+      "loss": 0.1167,
+      "step": 7440
+    },
+    {
+      "epoch": 28.435114503816795,
+      "grad_norm": 0.41615232825279236,
+      "learning_rate": 8.962905002112989e-06,
+      "loss": 0.1112,
+      "step": 7450
+    },
+    {
+      "epoch": 28.47328244274809,
+      "grad_norm": 0.39141789078712463,
+      "learning_rate": 8.959541826013725e-06,
+      "loss": 0.1099,
+      "step": 7460
+    },
+    {
+      "epoch": 28.51145038167939,
+      "grad_norm": 0.4517311453819275,
+      "learning_rate": 8.95617383869778e-06,
+      "loss": 0.1093,
+      "step": 7470
+    },
+    {
+      "epoch": 28.549618320610687,
+      "grad_norm": 0.3166011571884155,
+      "learning_rate": 8.952801044257581e-06,
+      "loss": 0.1008,
+      "step": 7480
+    },
+    {
+      "epoch": 28.587786259541986,
+      "grad_norm": 0.5795003771781921,
+      "learning_rate": 8.949423446791388e-06,
+      "loss": 0.1185,
+      "step": 7490
+    },
+    {
+      "epoch": 28.625954198473284,
+      "grad_norm": 0.7326614856719971,
+      "learning_rate": 8.9460410504033e-06,
+      "loss": 0.1177,
+      "step": 7500
+    },
+    {
+      "epoch": 28.66412213740458,
+      "grad_norm": 0.4169262647628784,
+      "learning_rate": 8.942653859203248e-06,
+      "loss": 0.1113,
+      "step": 7510
+    },
+    {
+      "epoch": 28.702290076335878,
+      "grad_norm": 0.43298423290252686,
+      "learning_rate": 8.93926187730699e-06,
+      "loss": 0.1072,
+      "step": 7520
+    },
+    {
+      "epoch": 28.740458015267176,
+      "grad_norm": 0.3784909248352051,
+      "learning_rate": 8.935865108836103e-06,
+      "loss": 0.1097,
+      "step": 7530
+    },
+    {
+      "epoch": 28.778625954198475,
+      "grad_norm": 0.7653591632843018,
+      "learning_rate": 8.932463557917982e-06,
+      "loss": 0.1127,
+      "step": 7540
+    },
+    {
+      "epoch": 28.81679389312977,
+      "grad_norm": 0.47900792956352234,
+      "learning_rate": 8.929057228685829e-06,
+      "loss": 0.1121,
+      "step": 7550
+    },
+    {
+      "epoch": 28.85496183206107,
+      "grad_norm": 0.6226459741592407,
+      "learning_rate": 8.925646125278657e-06,
+      "loss": 0.1082,
+      "step": 7560
+    },
+    {
+      "epoch": 28.893129770992367,
+      "grad_norm": 0.4821157455444336,
+      "learning_rate": 8.92223025184128e-06,
+      "loss": 0.1081,
+      "step": 7570
+    },
+    {
+      "epoch": 28.931297709923665,
+      "grad_norm": 0.3318312466144562,
+      "learning_rate": 8.918809612524305e-06,
+      "loss": 0.113,
+      "step": 7580
+    },
+    {
+      "epoch": 28.96946564885496,
+      "grad_norm": 0.4457385241985321,
+      "learning_rate": 8.91538421148413e-06,
+      "loss": 0.1162,
+      "step": 7590
+    },
+    {
+      "epoch": 29.00763358778626,
+      "grad_norm": 0.40070948004722595,
+      "learning_rate": 8.911954052882941e-06,
+      "loss": 0.1127,
+      "step": 7600
+    },
+    {
+      "epoch": 29.045801526717558,
+      "grad_norm": 0.4417935609817505,
+      "learning_rate": 8.908519140888704e-06,
+      "loss": 0.1035,
+      "step": 7610
+    },
+    {
+      "epoch": 29.083969465648856,
+      "grad_norm": 0.3581926226615906,
+      "learning_rate": 8.905079479675164e-06,
+      "loss": 0.1033,
+      "step": 7620
+    },
+    {
+      "epoch": 29.12213740458015,
+      "grad_norm": 0.4203307330608368,
+      "learning_rate": 8.901635073421831e-06,
+      "loss": 0.1021,
+      "step": 7630
+    },
+    {
+      "epoch": 29.16030534351145,
+      "grad_norm": 0.4549827575683594,
+      "learning_rate": 8.898185926313982e-06,
+      "loss": 0.1093,
+      "step": 7640
+    },
+    {
+      "epoch": 29.198473282442748,
+      "grad_norm": 0.3619544208049774,
+      "learning_rate": 8.894732042542659e-06,
+      "loss": 0.1126,
+      "step": 7650
+    },
+    {
+      "epoch": 29.236641221374047,
+      "grad_norm": 0.32612374424934387,
+      "learning_rate": 8.891273426304656e-06,
+      "loss": 0.1082,
+      "step": 7660
+    },
+    {
+      "epoch": 29.274809160305345,
+      "grad_norm": 0.44945961236953735,
+      "learning_rate": 8.887810081802514e-06,
+      "loss": 0.11,
+      "step": 7670
+    },
+    {
+      "epoch": 29.31297709923664,
+      "grad_norm": 0.5542881488800049,
+      "learning_rate": 8.88434201324453e-06,
+      "loss": 0.1066,
+      "step": 7680
+    },
+    {
+      "epoch": 29.35114503816794,
+      "grad_norm": 0.7978936433792114,
+      "learning_rate": 8.880869224844727e-06,
+      "loss": 0.1143,
+      "step": 7690
+    },
+    {
+      "epoch": 29.389312977099237,
+      "grad_norm": 0.6490185260772705,
+      "learning_rate": 8.877391720822874e-06,
+      "loss": 0.1155,
+      "step": 7700
+    },
+    {
+      "epoch": 29.427480916030536,
+      "grad_norm": 0.5318716168403625,
+      "learning_rate": 8.873909505404467e-06,
+      "loss": 0.1078,
+      "step": 7710
+    },
+    {
+      "epoch": 29.46564885496183,
+      "grad_norm": 0.40619373321533203,
+      "learning_rate": 8.870422582820726e-06,
+      "loss": 0.1147,
+      "step": 7720
+    },
+    {
+      "epoch": 29.50381679389313,
+      "grad_norm": 0.3707321286201477,
+      "learning_rate": 8.866930957308589e-06,
+      "loss": 0.1077,
+      "step": 7730
+    },
+    {
+      "epoch": 29.541984732824428,
+      "grad_norm": 0.5502585172653198,
+      "learning_rate": 8.863434633110711e-06,
+      "loss": 0.1046,
+      "step": 7740
+    },
+    {
+      "epoch": 29.580152671755727,
+      "grad_norm": 0.34364914894104004,
+      "learning_rate": 8.859933614475454e-06,
+      "loss": 0.1071,
+      "step": 7750
+    },
+    {
+      "epoch": 29.61832061068702,
+      "grad_norm": 0.302059531211853,
+      "learning_rate": 8.85642790565689e-06,
+      "loss": 0.1041,
+      "step": 7760
+    },
+    {
+      "epoch": 29.65648854961832,
+      "grad_norm": 0.5553315281867981,
+      "learning_rate": 8.852917510914783e-06,
+      "loss": 0.111,
+      "step": 7770
+    },
+    {
+      "epoch": 29.69465648854962,
+      "grad_norm": 0.3059275448322296,
+      "learning_rate": 8.84940243451459e-06,
+      "loss": 0.1057,
+      "step": 7780
+    },
+    {
+      "epoch": 29.732824427480917,
+      "grad_norm": 0.3817649483680725,
+      "learning_rate": 8.84588268072747e-06,
+      "loss": 0.1044,
+      "step": 7790
+    },
+    {
+      "epoch": 29.770992366412212,
+      "grad_norm": 0.3629932403564453,
+      "learning_rate": 8.842358253830245e-06,
+      "loss": 0.1092,
+      "step": 7800
+    },
+    {
+      "epoch": 29.80916030534351,
+      "grad_norm": 0.39645668864250183,
+      "learning_rate": 8.838829158105434e-06,
+      "loss": 0.1092,
+      "step": 7810
+    },
+    {
+      "epoch": 29.84732824427481,
+      "grad_norm": 0.3265850245952606,
+      "learning_rate": 8.835295397841217e-06,
+      "loss": 0.1039,
+      "step": 7820
+    },
+    {
+      "epoch": 29.885496183206108,
+      "grad_norm": 0.9192505478858948,
+      "learning_rate": 8.831756977331447e-06,
+      "loss": 0.1192,
+      "step": 7830
+    },
+    {
+      "epoch": 29.923664122137403,
+      "grad_norm": 1.0596504211425781,
+      "learning_rate": 8.828213900875639e-06,
+      "loss": 0.1154,
+      "step": 7840
+    },
+    {
+      "epoch": 29.9618320610687,
+      "grad_norm": 0.6525912880897522,
+      "learning_rate": 8.824666172778964e-06,
+      "loss": 0.113,
+      "step": 7850
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 0.4801918864250183,
+      "learning_rate": 8.821113797352246e-06,
+      "loss": 0.106,
+      "step": 7860
+    },
+    {
+      "epoch": 30.0381679389313,
+      "grad_norm": 0.8165585398674011,
+      "learning_rate": 8.817556778911957e-06,
+      "loss": 0.1145,
+      "step": 7870
+    },
+    {
+      "epoch": 30.076335877862597,
+      "grad_norm": 0.31382817029953003,
+      "learning_rate": 8.81399512178021e-06,
+      "loss": 0.1096,
+      "step": 7880
+    },
+    {
+      "epoch": 30.114503816793892,
+      "grad_norm": 0.38235539197921753,
+      "learning_rate": 8.810428830284752e-06,
+      "loss": 0.1064,
+      "step": 7890
+    },
+    {
+      "epoch": 30.15267175572519,
+      "grad_norm": 0.9553492069244385,
+      "learning_rate": 8.806857908758968e-06,
+      "loss": 0.11,
+      "step": 7900
+    },
+    {
+      "epoch": 30.19083969465649,
+      "grad_norm": 0.33214470744132996,
+      "learning_rate": 8.80328236154186e-06,
+      "loss": 0.1346,
+      "step": 7910
+    },
+    {
+      "epoch": 30.229007633587788,
+      "grad_norm": 0.5911892056465149,
+      "learning_rate": 8.799702192978056e-06,
+      "loss": 0.1101,
+      "step": 7920
+    },
+    {
+      "epoch": 30.267175572519083,
+      "grad_norm": 0.6221767067909241,
+      "learning_rate": 8.7961174074178e-06,
+      "loss": 0.1067,
+      "step": 7930
+    },
+    {
+      "epoch": 30.30534351145038,
+      "grad_norm": 0.9988340735435486,
+      "learning_rate": 8.792528009216942e-06,
+      "loss": 0.1043,
+      "step": 7940
+    },
+    {
+      "epoch": 30.34351145038168,
+      "grad_norm": 0.3819482624530792,
+      "learning_rate": 8.788934002736944e-06,
+      "loss": 0.1047,
+      "step": 7950
+    },
+    {
+      "epoch": 30.38167938931298,
+      "grad_norm": 0.36361387372016907,
+      "learning_rate": 8.785335392344858e-06,
+      "loss": 0.1092,
+      "step": 7960
+    },
+    {
+      "epoch": 30.419847328244273,
+      "grad_norm": 0.4566929340362549,
+      "learning_rate": 8.781732182413336e-06,
+      "loss": 0.1103,
+      "step": 7970
+    },
+    {
+      "epoch": 30.458015267175572,
+      "grad_norm": 0.6546979546546936,
+      "learning_rate": 8.778124377320619e-06,
+      "loss": 0.1107,
+      "step": 7980
+    },
+    {
+      "epoch": 30.49618320610687,
+      "grad_norm": 0.38045239448547363,
+      "learning_rate": 8.774511981450529e-06,
+      "loss": 0.1074,
+      "step": 7990
+    },
+    {
+      "epoch": 30.53435114503817,
+      "grad_norm": 0.6376705169677734,
+      "learning_rate": 8.770894999192468e-06,
+      "loss": 0.1162,
+      "step": 8000
+    },
+    {
+      "epoch": 30.572519083969464,
+      "grad_norm": 0.6514620184898376,
+      "learning_rate": 8.767273434941413e-06,
+      "loss": 0.1094,
+      "step": 8010
+    },
+    {
+      "epoch": 30.610687022900763,
+      "grad_norm": 0.49210497736930847,
+      "learning_rate": 8.763647293097902e-06,
+      "loss": 0.1143,
+      "step": 8020
+    },
+    {
+      "epoch": 30.64885496183206,
+      "grad_norm": 0.9991191029548645,
+      "learning_rate": 8.76001657806804e-06,
+      "loss": 0.108,
+      "step": 8030
+    },
+    {
+      "epoch": 30.68702290076336,
+      "grad_norm": 0.3525305688381195,
+      "learning_rate": 8.75638129426349e-06,
+      "loss": 0.1043,
+      "step": 8040
+    },
+    {
+      "epoch": 30.725190839694655,
+      "grad_norm": 0.3630237281322479,
+      "learning_rate": 8.752741446101464e-06,
+      "loss": 0.1033,
+      "step": 8050
+    },
+    {
+      "epoch": 30.763358778625953,
+      "grad_norm": 0.544648289680481,
+      "learning_rate": 8.749097038004722e-06,
+      "loss": 0.1059,
+      "step": 8060
+    },
+    {
+      "epoch": 30.801526717557252,
+      "grad_norm": 0.31798774003982544,
+      "learning_rate": 8.745448074401562e-06,
+      "loss": 0.1073,
+      "step": 8070
+    },
+    {
+      "epoch": 30.83969465648855,
+      "grad_norm": 0.35739681124687195,
+      "learning_rate": 8.741794559725818e-06,
+      "loss": 0.1084,
+      "step": 8080
+    },
+    {
+      "epoch": 30.87786259541985,
+      "grad_norm": 0.38239383697509766,
+      "learning_rate": 8.738136498416857e-06,
+      "loss": 0.11,
+      "step": 8090
+    },
+    {
+      "epoch": 30.916030534351144,
+      "grad_norm": 1.0929396152496338,
+      "learning_rate": 8.734473894919564e-06,
+      "loss": 0.1153,
+      "step": 8100
+    },
+    {
+      "epoch": 30.954198473282442,
+      "grad_norm": 0.4055207669734955,
+      "learning_rate": 8.730806753684354e-06,
+      "loss": 0.1021,
+      "step": 8110
+    },
+    {
+      "epoch": 30.99236641221374,
+      "grad_norm": 0.4326501190662384,
+      "learning_rate": 8.727135079167144e-06,
+      "loss": 0.1144,
+      "step": 8120
+    },
+    {
+      "epoch": 31.03053435114504,
+      "grad_norm": 0.39526522159576416,
+      "learning_rate": 8.723458875829368e-06,
+      "loss": 0.1018,
+      "step": 8130
+    },
+    {
+      "epoch": 31.068702290076335,
+      "grad_norm": 0.35406747460365295,
+      "learning_rate": 8.719778148137959e-06,
+      "loss": 0.1056,
+      "step": 8140
+    },
+    {
+      "epoch": 31.106870229007633,
+      "grad_norm": 0.380480021238327,
+      "learning_rate": 8.716092900565347e-06,
+      "loss": 0.1079,
+      "step": 8150
+    },
+    {
+      "epoch": 31.14503816793893,
+      "grad_norm": 0.3873949944972992,
+      "learning_rate": 8.712403137589455e-06,
+      "loss": 0.1065,
+      "step": 8160
+    },
+    {
+      "epoch": 31.18320610687023,
+      "grad_norm": 0.3808857202529907,
+      "learning_rate": 8.708708863693696e-06,
+      "loss": 0.1104,
+      "step": 8170
+    },
+    {
+      "epoch": 31.221374045801525,
+      "grad_norm": 0.4861178994178772,
+      "learning_rate": 8.705010083366961e-06,
+      "loss": 0.1064,
+      "step": 8180
+    },
+    {
+      "epoch": 31.259541984732824,
+      "grad_norm": 0.8248606324195862,
+      "learning_rate": 8.701306801103611e-06,
+      "loss": 0.1039,
+      "step": 8190
+    },
+    {
+      "epoch": 31.297709923664122,
+      "grad_norm": 0.5083164572715759,
+      "learning_rate": 8.69759902140349e-06,
+      "loss": 0.1093,
+      "step": 8200
+    },
+    {
+      "epoch": 31.33587786259542,
+      "grad_norm": 0.4453307092189789,
+      "learning_rate": 8.693886748771896e-06,
+      "loss": 0.1081,
+      "step": 8210
+    },
+    {
+      "epoch": 31.374045801526716,
+      "grad_norm": 0.4088967442512512,
+      "learning_rate": 8.690169987719593e-06,
+      "loss": 0.1193,
+      "step": 8220
+    },
+    {
+      "epoch": 31.412213740458014,
+      "grad_norm": 0.42236557602882385,
+      "learning_rate": 8.686448742762792e-06,
+      "loss": 0.1143,
+      "step": 8230
+    },
+    {
+      "epoch": 31.450381679389313,
+      "grad_norm": 0.9636140465736389,
+      "learning_rate": 8.68272301842316e-06,
+      "loss": 0.1119,
+      "step": 8240
+    },
+    {
+      "epoch": 31.48854961832061,
+      "grad_norm": 1.1265870332717896,
+      "learning_rate": 8.678992819227804e-06,
+      "loss": 0.1045,
+      "step": 8250
+    },
+    {
+      "epoch": 31.52671755725191,
+      "grad_norm": 0.7294782400131226,
+      "learning_rate": 8.675258149709265e-06,
+      "loss": 0.1139,
+      "step": 8260
+    },
+    {
+      "epoch": 31.564885496183205,
+      "grad_norm": 0.9136938452720642,
+      "learning_rate": 8.67151901440552e-06,
+      "loss": 0.1111,
+      "step": 8270
+    },
+    {
+      "epoch": 31.603053435114504,
+      "grad_norm": 0.4900221824645996,
+      "learning_rate": 8.667775417859971e-06,
+      "loss": 0.1077,
+      "step": 8280
+    },
+    {
+      "epoch": 31.641221374045802,
+      "grad_norm": 0.2998524010181427,
+      "learning_rate": 8.664027364621442e-06,
+      "loss": 0.1085,
+      "step": 8290
+    },
+    {
+      "epoch": 31.6793893129771,
+      "grad_norm": 0.5325132012367249,
+      "learning_rate": 8.660274859244167e-06,
+      "loss": 0.1093,
+      "step": 8300
+    },
+    {
+      "epoch": 31.717557251908396,
+      "grad_norm": 0.6447851657867432,
+      "learning_rate": 8.656517906287798e-06,
+      "loss": 0.111,
+      "step": 8310
+    },
+    {
+      "epoch": 31.755725190839694,
+      "grad_norm": 0.4099143147468567,
+      "learning_rate": 8.652756510317387e-06,
+      "loss": 0.1085,
+      "step": 8320
+    },
+    {
+      "epoch": 31.793893129770993,
+      "grad_norm": 0.30737578868865967,
+      "learning_rate": 8.648990675903382e-06,
+      "loss": 0.1077,
+      "step": 8330
+    },
+    {
+      "epoch": 31.83206106870229,
+      "grad_norm": 0.43916499614715576,
+      "learning_rate": 8.645220407621629e-06,
+      "loss": 0.1042,
+      "step": 8340
+    },
+    {
+      "epoch": 31.870229007633586,
+      "grad_norm": 0.3269640803337097,
+      "learning_rate": 8.64144571005336e-06,
+      "loss": 0.1038,
+      "step": 8350
+    },
+    {
+      "epoch": 31.908396946564885,
+      "grad_norm": 0.38933065533638,
+      "learning_rate": 8.637666587785185e-06,
+      "loss": 0.101,
+      "step": 8360
+    },
+    {
+      "epoch": 31.946564885496183,
+      "grad_norm": 0.42047321796417236,
+      "learning_rate": 8.633883045409096e-06,
+      "loss": 0.1107,
+      "step": 8370
+    },
+    {
+      "epoch": 31.984732824427482,
+      "grad_norm": 0.36545827984809875,
+      "learning_rate": 8.630095087522458e-06,
+      "loss": 0.1053,
+      "step": 8380
+    },
+    {
+      "epoch": 32.02290076335878,
+      "grad_norm": 0.5923290848731995,
+      "learning_rate": 8.62630271872799e-06,
+      "loss": 0.1096,
+      "step": 8390
+    },
+    {
+      "epoch": 32.06106870229008,
+      "grad_norm": 0.4061322510242462,
+      "learning_rate": 8.622505943633781e-06,
+      "loss": 0.1131,
+      "step": 8400
+    },
+    {
+      "epoch": 32.099236641221374,
+      "grad_norm": 0.3834056854248047,
+      "learning_rate": 8.618704766853271e-06,
+      "loss": 0.1003,
+      "step": 8410
+    },
+    {
+      "epoch": 32.13740458015267,
+      "grad_norm": 0.29086577892303467,
+      "learning_rate": 8.614899193005248e-06,
+      "loss": 0.1085,
+      "step": 8420
+    },
+    {
+      "epoch": 32.17557251908397,
+      "grad_norm": 0.48934558033943176,
+      "learning_rate": 8.611089226713843e-06,
+      "loss": 0.1089,
+      "step": 8430
+    },
+    {
+      "epoch": 32.213740458015266,
+      "grad_norm": 0.5814259648323059,
+      "learning_rate": 8.607274872608521e-06,
+      "loss": 0.1229,
+      "step": 8440
+    },
+    {
+      "epoch": 32.25190839694657,
+      "grad_norm": 0.36074042320251465,
+      "learning_rate": 8.603456135324089e-06,
+      "loss": 0.1099,
+      "step": 8450
+    },
+    {
+      "epoch": 32.29007633587786,
+      "grad_norm": 0.40568429231643677,
+      "learning_rate": 8.599633019500665e-06,
+      "loss": 0.1234,
+      "step": 8460
+    },
+    {
+      "epoch": 32.32824427480916,
+      "grad_norm": 0.4491671621799469,
+      "learning_rate": 8.595805529783703e-06,
+      "loss": 0.1156,
+      "step": 8470
+    },
+    {
+      "epoch": 32.36641221374046,
+      "grad_norm": 0.4605843126773834,
+      "learning_rate": 8.59197367082396e-06,
+      "loss": 0.1075,
+      "step": 8480
+    },
+    {
+      "epoch": 32.404580152671755,
+      "grad_norm": 0.5685547590255737,
+      "learning_rate": 8.588137447277502e-06,
+      "loss": 0.1205,
+      "step": 8490
+    },
+    {
+      "epoch": 32.44274809160305,
+      "grad_norm": 0.32008451223373413,
+      "learning_rate": 8.584296863805708e-06,
+      "loss": 0.1065,
+      "step": 8500
+    },
+    {
+      "epoch": 32.48091603053435,
+      "grad_norm": 0.3442102074623108,
+      "learning_rate": 8.580451925075249e-06,
+      "loss": 0.1123,
+      "step": 8510
+    },
+    {
+      "epoch": 32.51908396946565,
+      "grad_norm": 0.5090252161026001,
+      "learning_rate": 8.576602635758086e-06,
+      "loss": 0.1108,
+      "step": 8520
+    },
+    {
+      "epoch": 32.55725190839695,
+      "grad_norm": 0.7189807891845703,
+      "learning_rate": 8.572749000531468e-06,
+      "loss": 0.1109,
+      "step": 8530
+    },
+    {
+      "epoch": 32.595419847328245,
+      "grad_norm": 0.6325716972351074,
+      "learning_rate": 8.568891024077925e-06,
+      "loss": 0.1113,
+      "step": 8540
+    },
+    {
+      "epoch": 32.63358778625954,
+      "grad_norm": 0.28530317544937134,
+      "learning_rate": 8.565028711085266e-06,
+      "loss": 0.1014,
+      "step": 8550
+    },
+    {
+      "epoch": 32.67175572519084,
+      "grad_norm": 0.5091580152511597,
+      "learning_rate": 8.561162066246562e-06,
+      "loss": 0.1085,
+      "step": 8560
+    },
+    {
+      "epoch": 32.70992366412214,
+      "grad_norm": 0.39473584294319153,
+      "learning_rate": 8.557291094260151e-06,
+      "loss": 0.1041,
+      "step": 8570
+    },
+    {
+      "epoch": 32.74809160305343,
+      "grad_norm": 0.5815795660018921,
+      "learning_rate": 8.55341579982963e-06,
+      "loss": 0.1011,
+      "step": 8580
+    },
+    {
+      "epoch": 32.786259541984734,
+      "grad_norm": 0.46441009640693665,
+      "learning_rate": 8.549536187663848e-06,
+      "loss": 0.1027,
+      "step": 8590
+    },
+    {
+      "epoch": 32.82442748091603,
+      "grad_norm": 0.4190653860569,
+      "learning_rate": 8.545652262476898e-06,
+      "loss": 0.1075,
+      "step": 8600
+    },
+    {
+      "epoch": 32.86259541984733,
+      "grad_norm": 0.3595786392688751,
+      "learning_rate": 8.541764028988115e-06,
+      "loss": 0.097,
+      "step": 8610
+    },
+    {
+      "epoch": 32.900763358778626,
+      "grad_norm": 0.6225172281265259,
+      "learning_rate": 8.537871491922072e-06,
+      "loss": 0.1145,
+      "step": 8620
+    },
+    {
+      "epoch": 32.93893129770992,
+      "grad_norm": 1.1140257120132446,
+      "learning_rate": 8.533974656008566e-06,
+      "loss": 0.129,
+      "step": 8630
+    },
+    {
+      "epoch": 32.97709923664122,
+      "grad_norm": 1.0088690519332886,
+      "learning_rate": 8.530073525982621e-06,
+      "loss": 0.1171,
+      "step": 8640
+    },
+    {
+      "epoch": 33.01526717557252,
+      "grad_norm": 0.2912192940711975,
+      "learning_rate": 8.526168106584476e-06,
+      "loss": 0.1004,
+      "step": 8650
+    },
+    {
+      "epoch": 33.05343511450382,
+      "grad_norm": 0.337960422039032,
+      "learning_rate": 8.522258402559587e-06,
+      "loss": 0.1179,
+      "step": 8660
+    },
+    {
+      "epoch": 33.091603053435115,
+      "grad_norm": 0.30016374588012695,
+      "learning_rate": 8.518344418658612e-06,
+      "loss": 0.0991,
+      "step": 8670
+    },
+    {
+      "epoch": 33.12977099236641,
+      "grad_norm": 0.36123228073120117,
+      "learning_rate": 8.51442615963741e-06,
+      "loss": 0.1,
+      "step": 8680
+    },
+    {
+      "epoch": 33.16793893129771,
+      "grad_norm": 0.3525253236293793,
+      "learning_rate": 8.510503630257034e-06,
+      "loss": 0.0969,
+      "step": 8690
+    },
+    {
+      "epoch": 33.20610687022901,
+      "grad_norm": 0.3419656753540039,
+      "learning_rate": 8.506576835283731e-06,
+      "loss": 0.1014,
+      "step": 8700
+    },
+    {
+      "epoch": 33.2442748091603,
+      "grad_norm": 0.3661300837993622,
+      "learning_rate": 8.502645779488923e-06,
+      "loss": 0.1014,
+      "step": 8710
+    },
+    {
+      "epoch": 33.282442748091604,
+      "grad_norm": 0.6394238471984863,
+      "learning_rate": 8.498710467649214e-06,
+      "loss": 0.1046,
+      "step": 8720
+    },
+    {
+      "epoch": 33.3206106870229,
+      "grad_norm": 0.3386576473712921,
+      "learning_rate": 8.494770904546381e-06,
+      "loss": 0.1125,
+      "step": 8730
+    },
+    {
+      "epoch": 33.3587786259542,
+      "grad_norm": 0.5160825848579407,
+      "learning_rate": 8.490827094967364e-06,
+      "loss": 0.1072,
+      "step": 8740
+    },
+    {
+      "epoch": 33.396946564885496,
+      "grad_norm": 0.37867069244384766,
+      "learning_rate": 8.486879043704263e-06,
+      "loss": 0.1043,
+      "step": 8750
+    },
+    {
+      "epoch": 33.43511450381679,
+      "grad_norm": 0.30260443687438965,
+      "learning_rate": 8.482926755554333e-06,
+      "loss": 0.1074,
+      "step": 8760
+    },
+    {
+      "epoch": 33.47328244274809,
+      "grad_norm": 0.589114248752594,
+      "learning_rate": 8.478970235319975e-06,
+      "loss": 0.1072,
+      "step": 8770
+    },
+    {
+      "epoch": 33.51145038167939,
+      "grad_norm": 0.3591925799846649,
+      "learning_rate": 8.475009487808738e-06,
+      "loss": 0.1009,
+      "step": 8780
+    },
+    {
+      "epoch": 33.54961832061069,
+      "grad_norm": 0.6600471138954163,
+      "learning_rate": 8.471044517833299e-06,
+      "loss": 0.122,
+      "step": 8790
+    },
+    {
+      "epoch": 33.587786259541986,
+      "grad_norm": 0.36351725459098816,
+      "learning_rate": 8.467075330211474e-06,
+      "loss": 0.1099,
+      "step": 8800
+    },
+    {
+      "epoch": 33.62595419847328,
+      "grad_norm": 0.31405356526374817,
+      "learning_rate": 8.463101929766197e-06,
+      "loss": 0.1024,
+      "step": 8810
+    },
+    {
+      "epoch": 33.66412213740458,
+      "grad_norm": 0.34903696179389954,
+      "learning_rate": 8.459124321325529e-06,
+      "loss": 0.1087,
+      "step": 8820
+    },
+    {
+      "epoch": 33.70229007633588,
+      "grad_norm": 0.43546533584594727,
+      "learning_rate": 8.455142509722635e-06,
+      "loss": 0.1072,
+      "step": 8830
+    },
+    {
+      "epoch": 33.74045801526717,
+      "grad_norm": 0.43064677715301514,
+      "learning_rate": 8.451156499795791e-06,
+      "loss": 0.1159,
+      "step": 8840
+    },
+    {
+      "epoch": 33.778625954198475,
+      "grad_norm": 0.3457275331020355,
+      "learning_rate": 8.44716629638838e-06,
+      "loss": 0.1137,
+      "step": 8850
+    },
+    {
+      "epoch": 33.81679389312977,
+      "grad_norm": 0.4882639944553375,
+      "learning_rate": 8.443171904348873e-06,
+      "loss": 0.1035,
+      "step": 8860
+    },
+    {
+      "epoch": 33.85496183206107,
+      "grad_norm": 0.3734376132488251,
+      "learning_rate": 8.439173328530829e-06,
+      "loss": 0.1045,
+      "step": 8870
+    },
+    {
+      "epoch": 33.89312977099237,
+      "grad_norm": 0.3466688096523285,
+      "learning_rate": 8.435170573792902e-06,
+      "loss": 0.1041,
+      "step": 8880
+    },
+    {
+      "epoch": 33.93129770992366,
+      "grad_norm": 0.27618467807769775,
+      "learning_rate": 8.431163644998808e-06,
+      "loss": 0.1138,
+      "step": 8890
+    },
+    {
+      "epoch": 33.969465648854964,
+      "grad_norm": 0.3809553384780884,
+      "learning_rate": 8.42715254701735e-06,
+      "loss": 0.1111,
+      "step": 8900
+    },
+    {
+      "epoch": 34.00763358778626,
+      "grad_norm": 0.3887907564640045,
+      "learning_rate": 8.423137284722389e-06,
+      "loss": 0.1069,
+      "step": 8910
+    },
+    {
+      "epoch": 34.045801526717554,
+      "grad_norm": 0.3740697205066681,
+      "learning_rate": 8.419117862992846e-06,
+      "loss": 0.1063,
+      "step": 8920
+    },
+    {
+      "epoch": 34.083969465648856,
+      "grad_norm": 0.38636720180511475,
+      "learning_rate": 8.415094286712694e-06,
+      "loss": 0.1108,
+      "step": 8930
+    },
+    {
+      "epoch": 34.12213740458015,
+      "grad_norm": 0.2964230477809906,
+      "learning_rate": 8.411066560770965e-06,
+      "loss": 0.1062,
+      "step": 8940
+    },
+    {
+      "epoch": 34.16030534351145,
+      "grad_norm": 0.4712314009666443,
+      "learning_rate": 8.407034690061722e-06,
+      "loss": 0.0999,
+      "step": 8950
+    },
+    {
+      "epoch": 34.19847328244275,
+      "grad_norm": 0.7659854292869568,
+      "learning_rate": 8.402998679484067e-06,
+      "loss": 0.1045,
+      "step": 8960
+    },
+    {
+      "epoch": 34.23664122137404,
+      "grad_norm": 0.4793913960456848,
+      "learning_rate": 8.398958533942135e-06,
+      "loss": 0.1036,
+      "step": 8970
+    },
+    {
+      "epoch": 34.274809160305345,
+      "grad_norm": 0.37287986278533936,
+      "learning_rate": 8.394914258345084e-06,
+      "loss": 0.1144,
+      "step": 8980
+    },
+    {
+      "epoch": 34.31297709923664,
+      "grad_norm": 0.541438639163971,
+      "learning_rate": 8.390865857607089e-06,
+      "loss": 0.1055,
+      "step": 8990
+    },
+    {
+      "epoch": 34.35114503816794,
+      "grad_norm": 0.46863681077957153,
+      "learning_rate": 8.386813336647339e-06,
+      "loss": 0.1055,
+      "step": 9000
+    },
+    {
+      "epoch": 34.38931297709924,
+      "grad_norm": 0.7329716086387634,
+      "learning_rate": 8.38275670039003e-06,
+      "loss": 0.1179,
+      "step": 9010
+    },
+    {
+      "epoch": 34.42748091603053,
+      "grad_norm": 0.4788097143173218,
+      "learning_rate": 8.378695953764357e-06,
+      "loss": 0.1261,
+      "step": 9020
+    },
+    {
+      "epoch": 34.465648854961835,
+      "grad_norm": 0.4310746192932129,
+      "learning_rate": 8.374631101704509e-06,
+      "loss": 0.1175,
+      "step": 9030
+    },
+    {
+      "epoch": 34.50381679389313,
+      "grad_norm": 0.4793173372745514,
+      "learning_rate": 8.370562149149666e-06,
+      "loss": 0.1089,
+      "step": 9040
+    },
+    {
+      "epoch": 34.541984732824424,
+      "grad_norm": 0.5185418725013733,
+      "learning_rate": 8.366489101043989e-06,
+      "loss": 0.1043,
+      "step": 9050
+    },
+    {
+      "epoch": 34.58015267175573,
+      "grad_norm": 0.37807098031044006,
+      "learning_rate": 8.362411962336613e-06,
+      "loss": 0.1102,
+      "step": 9060
+    },
+    {
+      "epoch": 34.61832061068702,
+      "grad_norm": 0.5667879581451416,
+      "learning_rate": 8.358330737981651e-06,
+      "loss": 0.1137,
+      "step": 9070
+    },
+    {
+      "epoch": 34.656488549618324,
+      "grad_norm": 1.1132043600082397,
+      "learning_rate": 8.35424543293817e-06,
+      "loss": 0.1117,
+      "step": 9080
+    },
+    {
+      "epoch": 34.69465648854962,
+      "grad_norm": 0.3736783564090729,
+      "learning_rate": 8.350156052170206e-06,
+      "loss": 0.1099,
+      "step": 9090
+    },
+    {
+      "epoch": 34.732824427480914,
+      "grad_norm": 0.3822537362575531,
+      "learning_rate": 8.346062600646739e-06,
+      "loss": 0.0993,
+      "step": 9100
+    },
+    {
+      "epoch": 34.770992366412216,
+      "grad_norm": 0.32378950715065,
+      "learning_rate": 8.341965083341696e-06,
+      "loss": 0.1012,
+      "step": 9110
+    },
+    {
+      "epoch": 34.80916030534351,
+      "grad_norm": 0.3589276969432831,
+      "learning_rate": 8.337863505233954e-06,
+      "loss": 0.1064,
+      "step": 9120
+    },
+    {
+      "epoch": 34.847328244274806,
+      "grad_norm": 0.3442489802837372,
+      "learning_rate": 8.333757871307311e-06,
+      "loss": 0.1066,
+      "step": 9130
+    },
+    {
+      "epoch": 34.88549618320611,
+      "grad_norm": 0.43421897292137146,
+      "learning_rate": 8.329648186550501e-06,
+      "loss": 0.1107,
+      "step": 9140
+    },
+    {
+      "epoch": 34.9236641221374,
+      "grad_norm": 0.33839526772499084,
+      "learning_rate": 8.32553445595718e-06,
+      "loss": 0.1105,
+      "step": 9150
+    },
+    {
+      "epoch": 34.961832061068705,
+      "grad_norm": 0.3755687475204468,
+      "learning_rate": 8.321416684525917e-06,
+      "loss": 0.1102,
+      "step": 9160
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 0.3840588629245758,
+      "learning_rate": 8.317294877260193e-06,
+      "loss": 0.1087,
+      "step": 9170
+    },
+    {
+      "epoch": 35.038167938931295,
+      "grad_norm": 0.5461086630821228,
+      "learning_rate": 8.313169039168395e-06,
+      "loss": 0.1093,
+      "step": 9180
+    },
+    {
+      "epoch": 35.0763358778626,
+      "grad_norm": 0.42317765951156616,
+      "learning_rate": 8.3090391752638e-06,
+      "loss": 0.1133,
+      "step": 9190
+    },
+    {
+      "epoch": 35.11450381679389,
+      "grad_norm": 0.5004700422286987,
+      "learning_rate": 8.304905290564586e-06,
+      "loss": 0.0985,
+      "step": 9200
+    },
+    {
+      "epoch": 35.152671755725194,
+      "grad_norm": 0.42691975831985474,
+      "learning_rate": 8.300767390093814e-06,
+      "loss": 0.1029,
+      "step": 9210
+    },
+    {
+      "epoch": 35.19083969465649,
+      "grad_norm": 0.6003366708755493,
+      "learning_rate": 8.296625478879417e-06,
+      "loss": 0.1216,
+      "step": 9220
+    },
+    {
+      "epoch": 35.229007633587784,
+      "grad_norm": 0.31051206588745117,
+      "learning_rate": 8.292479561954214e-06,
+      "loss": 0.1022,
+      "step": 9230
+    },
+    {
+      "epoch": 35.267175572519086,
+      "grad_norm": 0.47282811999320984,
+      "learning_rate": 8.288329644355884e-06,
+      "loss": 0.1074,
+      "step": 9240
+    },
+    {
+      "epoch": 35.30534351145038,
+      "grad_norm": 0.7961530089378357,
+      "learning_rate": 8.284175731126964e-06,
+      "loss": 0.1041,
+      "step": 9250
+    },
+    {
+      "epoch": 35.343511450381676,
+      "grad_norm": 0.536555826663971,
+      "learning_rate": 8.280017827314854e-06,
+      "loss": 0.1102,
+      "step": 9260
+    },
+    {
+      "epoch": 35.38167938931298,
+      "grad_norm": 0.6327351331710815,
+      "learning_rate": 8.275855937971799e-06,
+      "loss": 0.1135,
+      "step": 9270
+    },
+    {
+      "epoch": 35.41984732824427,
+      "grad_norm": 0.39085444808006287,
+      "learning_rate": 8.271690068154887e-06,
+      "loss": 0.1086,
+      "step": 9280
+    },
+    {
+      "epoch": 35.458015267175576,
+      "grad_norm": 0.3909195065498352,
+      "learning_rate": 8.26752022292604e-06,
+      "loss": 0.1047,
+      "step": 9290
+    },
+    {
+      "epoch": 35.49618320610687,
+      "grad_norm": 0.44586417078971863,
+      "learning_rate": 8.263346407352017e-06,
+      "loss": 0.1016,
+      "step": 9300
+    },
+    {
+      "epoch": 35.534351145038165,
+      "grad_norm": 0.9170414805412292,
+      "learning_rate": 8.259168626504395e-06,
+      "loss": 0.1119,
+      "step": 9310
+    },
+    {
+      "epoch": 35.57251908396947,
+      "grad_norm": 0.5176953673362732,
+      "learning_rate": 8.25498688545957e-06,
+      "loss": 0.109,
+      "step": 9320
+    },
+    {
+      "epoch": 35.61068702290076,
+      "grad_norm": 0.3105170428752899,
+      "learning_rate": 8.250801189298759e-06,
+      "loss": 0.1089,
+      "step": 9330
+    },
+    {
+      "epoch": 35.64885496183206,
+      "grad_norm": 0.9050326943397522,
+      "learning_rate": 8.246611543107968e-06,
+      "loss": 0.1105,
+      "step": 9340
+    },
+    {
+      "epoch": 35.68702290076336,
+      "grad_norm": 0.36389780044555664,
+      "learning_rate": 8.24241795197802e-06,
+      "loss": 0.1075,
+      "step": 9350
+    },
+    {
+      "epoch": 35.725190839694655,
+      "grad_norm": 0.2961239814758301,
+      "learning_rate": 8.238220421004518e-06,
+      "loss": 0.1073,
+      "step": 9360
+    },
+    {
+      "epoch": 35.76335877862596,
+      "grad_norm": 0.4220345914363861,
+      "learning_rate": 8.23401895528786e-06,
+      "loss": 0.1004,
+      "step": 9370
+    },
+    {
+      "epoch": 35.80152671755725,
+      "grad_norm": 0.3124537765979767,
+      "learning_rate": 8.229813559933225e-06,
+      "loss": 0.1113,
+      "step": 9380
+    },
+    {
+      "epoch": 35.83969465648855,
+      "grad_norm": 0.4424472153186798,
+      "learning_rate": 8.22560424005056e-06,
+      "loss": 0.1074,
+      "step": 9390
+    },
+    {
+      "epoch": 35.87786259541985,
+      "grad_norm": 0.46610021591186523,
+      "learning_rate": 8.22139100075459e-06,
+      "loss": 0.1089,
+      "step": 9400
+    },
+    {
+      "epoch": 35.916030534351144,
+      "grad_norm": 0.5336711406707764,
+      "learning_rate": 8.217173847164799e-06,
+      "loss": 0.1031,
+      "step": 9410
+    },
+    {
+      "epoch": 35.954198473282446,
+      "grad_norm": 0.4767981171607971,
+      "learning_rate": 8.212952784405423e-06,
+      "loss": 0.1033,
+      "step": 9420
+    },
+    {
+      "epoch": 35.99236641221374,
+      "grad_norm": 0.3861401677131653,
+      "learning_rate": 8.208727817605453e-06,
+      "loss": 0.1031,
+      "step": 9430
+    },
+    {
+      "epoch": 36.030534351145036,
+      "grad_norm": 0.5217192769050598,
+      "learning_rate": 8.204498951898618e-06,
+      "loss": 0.1086,
+      "step": 9440
+    },
+    {
+      "epoch": 36.06870229007634,
+      "grad_norm": 0.41782206296920776,
+      "learning_rate": 8.200266192423396e-06,
+      "loss": 0.1091,
+      "step": 9450
+    },
+    {
+      "epoch": 36.10687022900763,
+      "grad_norm": 0.29525044560432434,
+      "learning_rate": 8.196029544322983e-06,
+      "loss": 0.1116,
+      "step": 9460
+    },
+    {
+      "epoch": 36.14503816793893,
+      "grad_norm": 0.4232509732246399,
+      "learning_rate": 8.191789012745306e-06,
+      "loss": 0.1101,
+      "step": 9470
+    },
+    {
+      "epoch": 36.18320610687023,
+      "grad_norm": 0.9247789978981018,
+      "learning_rate": 8.187544602843014e-06,
+      "loss": 0.0989,
+      "step": 9480
+    },
+    {
+      "epoch": 36.221374045801525,
+      "grad_norm": 0.3943084180355072,
+      "learning_rate": 8.183296319773466e-06,
+      "loss": 0.1031,
+      "step": 9490
+    },
+    {
+      "epoch": 36.25954198473283,
+      "grad_norm": 0.4537927806377411,
+      "learning_rate": 8.179044168698722e-06,
+      "loss": 0.1021,
+      "step": 9500
+    },
+    {
+      "epoch": 36.29770992366412,
+      "grad_norm": 0.5286827683448792,
+      "learning_rate": 8.174788154785548e-06,
+      "loss": 0.1085,
+      "step": 9510
+    },
+    {
+      "epoch": 36.33587786259542,
+      "grad_norm": 0.40371379256248474,
+      "learning_rate": 8.170528283205404e-06,
+      "loss": 0.1035,
+      "step": 9520
+    },
+    {
+      "epoch": 36.37404580152672,
+      "grad_norm": 0.4854196608066559,
+      "learning_rate": 8.166264559134434e-06,
+      "loss": 0.1115,
+      "step": 9530
+    },
+    {
+      "epoch": 36.412213740458014,
+      "grad_norm": 0.673309862613678,
+      "learning_rate": 8.161996987753466e-06,
+      "loss": 0.1044,
+      "step": 9540
+    },
+    {
+      "epoch": 36.45038167938931,
+      "grad_norm": 0.37787386775016785,
+      "learning_rate": 8.157725574248e-06,
+      "loss": 0.1136,
+      "step": 9550
+    },
+    {
+      "epoch": 36.48854961832061,
+      "grad_norm": 1.057741641998291,
+      "learning_rate": 8.15345032380821e-06,
+      "loss": 0.1129,
+      "step": 9560
+    },
+    {
+      "epoch": 36.52671755725191,
+      "grad_norm": 0.5287107229232788,
+      "learning_rate": 8.149171241628924e-06,
+      "loss": 0.0985,
+      "step": 9570
+    },
+    {
+      "epoch": 36.56488549618321,
+      "grad_norm": 0.4657130539417267,
+      "learning_rate": 8.144888332909631e-06,
+      "loss": 0.0994,
+      "step": 9580
+    },
+    {
+      "epoch": 36.603053435114504,
+      "grad_norm": 0.2887883484363556,
+      "learning_rate": 8.140601602854471e-06,
+      "loss": 0.1009,
+      "step": 9590
+    },
+    {
+      "epoch": 36.6412213740458,
+      "grad_norm": 0.5169790983200073,
+      "learning_rate": 8.136311056672224e-06,
+      "loss": 0.1026,
+      "step": 9600
+    },
+    {
+      "epoch": 36.6793893129771,
+      "grad_norm": 0.42216652631759644,
+      "learning_rate": 8.132016699576308e-06,
+      "loss": 0.1009,
+      "step": 9610
+    },
+    {
+      "epoch": 36.717557251908396,
+      "grad_norm": 0.39106106758117676,
+      "learning_rate": 8.127718536784771e-06,
+      "loss": 0.1056,
+      "step": 9620
+    },
+    {
+      "epoch": 36.7557251908397,
+      "grad_norm": 0.5721985697746277,
+      "learning_rate": 8.123416573520289e-06,
+      "loss": 0.1017,
+      "step": 9630
+    },
+    {
+      "epoch": 36.79389312977099,
+      "grad_norm": 0.5179575085639954,
+      "learning_rate": 8.119110815010152e-06,
+      "loss": 0.1127,
+      "step": 9640
+    },
+    {
+      "epoch": 36.83206106870229,
+      "grad_norm": 0.5956861972808838,
+      "learning_rate": 8.11480126648626e-06,
+      "loss": 0.1059,
+      "step": 9650
+    },
+    {
+      "epoch": 36.87022900763359,
+      "grad_norm": 0.37840545177459717,
+      "learning_rate": 8.110487933185123e-06,
+      "loss": 0.1052,
+      "step": 9660
+    },
+    {
+      "epoch": 36.908396946564885,
+      "grad_norm": 0.42260074615478516,
+      "learning_rate": 8.106170820347849e-06,
+      "loss": 0.1003,
+      "step": 9670
+    },
+    {
+      "epoch": 36.94656488549618,
+      "grad_norm": 0.7760491371154785,
+      "learning_rate": 8.101849933220134e-06,
+      "loss": 0.1081,
+      "step": 9680
+    },
+    {
+      "epoch": 36.98473282442748,
+      "grad_norm": 0.3848549723625183,
+      "learning_rate": 8.097525277052265e-06,
+      "loss": 0.1011,
+      "step": 9690
+    },
+    {
+      "epoch": 37.02290076335878,
+      "grad_norm": 0.72954261302948,
+      "learning_rate": 8.093196857099105e-06,
+      "loss": 0.1005,
+      "step": 9700
+    },
+    {
+      "epoch": 37.06106870229008,
+      "grad_norm": 0.515247642993927,
+      "learning_rate": 8.088864678620096e-06,
+      "loss": 0.1046,
+      "step": 9710
+    },
+    {
+      "epoch": 37.099236641221374,
+      "grad_norm": 0.6722749471664429,
+      "learning_rate": 8.084528746879243e-06,
+      "loss": 0.095,
+      "step": 9720
+    },
+    {
+      "epoch": 37.13740458015267,
+      "grad_norm": 0.3528861999511719,
+      "learning_rate": 8.080189067145107e-06,
+      "loss": 0.1046,
+      "step": 9730
+    },
+    {
+      "epoch": 37.17557251908397,
+      "grad_norm": 0.42519357800483704,
+      "learning_rate": 8.075845644690814e-06,
+      "loss": 0.0938,
+      "step": 9740
+    },
+    {
+      "epoch": 37.213740458015266,
+      "grad_norm": 0.5400734543800354,
+      "learning_rate": 8.07149848479403e-06,
+      "loss": 0.103,
+      "step": 9750
+    },
+    {
+      "epoch": 37.25190839694657,
+      "grad_norm": 0.6142025589942932,
+      "learning_rate": 8.067147592736963e-06,
+      "loss": 0.1027,
+      "step": 9760
+    },
+    {
+      "epoch": 37.29007633587786,
+      "grad_norm": 0.4489392936229706,
+      "learning_rate": 8.062792973806358e-06,
+      "loss": 0.1222,
+      "step": 9770
+    },
+    {
+      "epoch": 37.32824427480916,
+      "grad_norm": 0.576805830001831,
+      "learning_rate": 8.058434633293485e-06,
+      "loss": 0.1025,
+      "step": 9780
+    },
+    {
+      "epoch": 37.36641221374046,
+      "grad_norm": 0.38401392102241516,
+      "learning_rate": 8.054072576494142e-06,
+      "loss": 0.1,
+      "step": 9790
+    },
+    {
+      "epoch": 37.404580152671755,
+      "grad_norm": 0.4029317796230316,
+      "learning_rate": 8.04970680870864e-06,
+      "loss": 0.1047,
+      "step": 9800
+    },
+    {
+      "epoch": 37.44274809160305,
+      "grad_norm": 0.7393083572387695,
+      "learning_rate": 8.045337335241793e-06,
+      "loss": 0.0922,
+      "step": 9810
+    },
+    {
+      "epoch": 37.48091603053435,
+      "grad_norm": 0.4550725519657135,
+      "learning_rate": 8.040964161402932e-06,
+      "loss": 0.0988,
+      "step": 9820
+    },
+    {
+      "epoch": 37.51908396946565,
+      "grad_norm": 0.5443292260169983,
+      "learning_rate": 8.036587292505869e-06,
+      "loss": 0.0993,
+      "step": 9830
+    },
+    {
+      "epoch": 37.55725190839695,
+      "grad_norm": 0.6734360456466675,
+      "learning_rate": 8.032206733868912e-06,
+      "loss": 0.0907,
+      "step": 9840
+    },
+    {
+      "epoch": 37.595419847328245,
+      "grad_norm": 0.4426909387111664,
+      "learning_rate": 8.027822490814859e-06,
+      "loss": 0.0963,
+      "step": 9850
+    },
+    {
+      "epoch": 37.63358778625954,
+      "grad_norm": 0.7196950316429138,
+      "learning_rate": 8.023434568670971e-06,
+      "loss": 0.0974,
+      "step": 9860
+    },
+    {
+      "epoch": 37.67175572519084,
+      "grad_norm": 0.8929295539855957,
+      "learning_rate": 8.019042972768992e-06,
+      "loss": 0.0923,
+      "step": 9870
+    },
+    {
+      "epoch": 37.70992366412214,
+      "grad_norm": 0.4260941445827484,
+      "learning_rate": 8.014647708445124e-06,
+      "loss": 0.0862,
+      "step": 9880
+    },
+    {
+      "epoch": 37.74809160305343,
+      "grad_norm": 0.581256628036499,
+      "learning_rate": 8.010248781040027e-06,
+      "loss": 0.0962,
+      "step": 9890
+    },
+    {
+      "epoch": 37.786259541984734,
+      "grad_norm": 0.5044671893119812,
+      "learning_rate": 8.005846195898815e-06,
+      "loss": 0.0849,
+      "step": 9900
+    },
+    {
+      "epoch": 37.82442748091603,
+      "grad_norm": 0.5715168714523315,
+      "learning_rate": 8.00143995837104e-06,
+      "loss": 0.0864,
+      "step": 9910
+    },
+    {
+      "epoch": 37.86259541984733,
+      "grad_norm": 0.4902816116809845,
+      "learning_rate": 7.997030073810699e-06,
+      "loss": 0.09,
+      "step": 9920
+    },
+    {
+      "epoch": 37.900763358778626,
+      "grad_norm": 0.8679199814796448,
+      "learning_rate": 7.992616547576218e-06,
+      "loss": 0.1111,
+      "step": 9930
+    },
+    {
+      "epoch": 37.93893129770992,
+      "grad_norm": 0.5654365420341492,
+      "learning_rate": 7.988199385030446e-06,
+      "loss": 0.0981,
+      "step": 9940
+    },
+    {
+      "epoch": 37.97709923664122,
+      "grad_norm": 1.2245254516601562,
+      "learning_rate": 7.98377859154065e-06,
+      "loss": 0.0936,
+      "step": 9950
+    },
+    {
+      "epoch": 38.01526717557252,
+      "grad_norm": 0.9778199195861816,
+      "learning_rate": 7.979354172478516e-06,
+      "loss": 0.0862,
+      "step": 9960
+    },
+    {
+      "epoch": 38.05343511450382,
+      "grad_norm": 0.4366597831249237,
+      "learning_rate": 7.974926133220127e-06,
+      "loss": 0.0897,
+      "step": 9970
+    },
+    {
+      "epoch": 38.091603053435115,
+      "grad_norm": 0.5576930642127991,
+      "learning_rate": 7.970494479145968e-06,
+      "loss": 0.1004,
+      "step": 9980
+    },
+    {
+      "epoch": 38.12977099236641,
+      "grad_norm": 0.5354223251342773,
+      "learning_rate": 7.966059215640918e-06,
+      "loss": 0.0855,
+      "step": 9990
+    },
+    {
+      "epoch": 38.16793893129771,
+      "grad_norm": 0.4817463457584381,
+      "learning_rate": 7.96162034809424e-06,
+      "loss": 0.103,
+      "step": 10000
+    },
+    {
+      "epoch": 38.20610687022901,
+      "grad_norm": 1.0092859268188477,
+      "learning_rate": 7.957177881899579e-06,
+      "loss": 0.0936,
+      "step": 10010
+    },
+    {
+      "epoch": 38.2442748091603,
+      "grad_norm": 0.3715302050113678,
+      "learning_rate": 7.952731822454944e-06,
+      "loss": 0.0907,
+      "step": 10020
+    },
+    {
+      "epoch": 38.282442748091604,
+      "grad_norm": 0.7082487940788269,
+      "learning_rate": 7.948282175162723e-06,
+      "loss": 0.0889,
+      "step": 10030
+    },
+    {
+      "epoch": 38.3206106870229,
+      "grad_norm": 0.5549430251121521,
+      "learning_rate": 7.943828945429653e-06,
+      "loss": 0.1023,
+      "step": 10040
+    },
+    {
+      "epoch": 38.3587786259542,
+      "grad_norm": 0.41316139698028564,
+      "learning_rate": 7.939372138666828e-06,
+      "loss": 0.0843,
+      "step": 10050
+    },
+    {
+      "epoch": 38.396946564885496,
+      "grad_norm": 0.4269944429397583,
+      "learning_rate": 7.934911760289692e-06,
+      "loss": 0.0876,
+      "step": 10060
+    },
+    {
+      "epoch": 38.43511450381679,
+      "grad_norm": 0.35790082812309265,
+      "learning_rate": 7.930447815718022e-06,
+      "loss": 0.0835,
+      "step": 10070
+    },
+    {
+      "epoch": 38.47328244274809,
+      "grad_norm": 0.6934918165206909,
+      "learning_rate": 7.925980310375933e-06,
+      "loss": 0.0874,
+      "step": 10080
+    },
+    {
+      "epoch": 38.51145038167939,
+      "grad_norm": 0.36909425258636475,
+      "learning_rate": 7.921509249691865e-06,
+      "loss": 0.0937,
+      "step": 10090
+    },
+    {
+      "epoch": 38.54961832061069,
+      "grad_norm": 0.8461699485778809,
+      "learning_rate": 7.917034639098578e-06,
+      "loss": 0.0862,
+      "step": 10100
+    },
+    {
+      "epoch": 38.587786259541986,
+      "grad_norm": 0.45856791734695435,
+      "learning_rate": 7.912556484033146e-06,
+      "loss": 0.0915,
+      "step": 10110
+    },
+    {
+      "epoch": 38.62595419847328,
+      "grad_norm": 0.49433228373527527,
+      "learning_rate": 7.908074789936952e-06,
+      "loss": 0.0774,
+      "step": 10120
+    },
+    {
+      "epoch": 38.66412213740458,
+      "grad_norm": 0.3476361334323883,
+      "learning_rate": 7.903589562255673e-06,
+      "loss": 0.0933,
+      "step": 10130
+    },
+    {
+      "epoch": 38.70229007633588,
+      "grad_norm": 0.3297344744205475,
+      "learning_rate": 7.899100806439287e-06,
+      "loss": 0.0846,
+      "step": 10140
+    },
+    {
+      "epoch": 38.74045801526717,
+      "grad_norm": 0.36996668577194214,
+      "learning_rate": 7.894608527942049e-06,
+      "loss": 0.0948,
+      "step": 10150
+    },
+    {
+      "epoch": 38.778625954198475,
+      "grad_norm": 0.7153366804122925,
+      "learning_rate": 7.89011273222251e-06,
+      "loss": 0.0868,
+      "step": 10160
+    },
+    {
+      "epoch": 38.81679389312977,
+      "grad_norm": 0.571403443813324,
+      "learning_rate": 7.88561342474348e-06,
+      "loss": 0.0801,
+      "step": 10170
+    },
+    {
+      "epoch": 38.85496183206107,
+      "grad_norm": 0.5913143157958984,
+      "learning_rate": 7.881110610972045e-06,
+      "loss": 0.085,
+      "step": 10180
+    },
+    {
+      "epoch": 38.89312977099237,
+      "grad_norm": 0.36453530192375183,
+      "learning_rate": 7.876604296379545e-06,
+      "loss": 0.0777,
+      "step": 10190
+    },
+    {
+      "epoch": 38.93129770992366,
+      "grad_norm": 0.5440592169761658,
+      "learning_rate": 7.87209448644158e-06,
+      "loss": 0.0882,
+      "step": 10200
+    },
+    {
+      "epoch": 38.969465648854964,
+      "grad_norm": 0.5457788705825806,
+      "learning_rate": 7.867581186637991e-06,
+      "loss": 0.0779,
+      "step": 10210
+    },
+    {
+      "epoch": 39.00763358778626,
+      "grad_norm": 0.6085828542709351,
+      "learning_rate": 7.863064402452867e-06,
+      "loss": 0.088,
+      "step": 10220
+    },
+    {
+      "epoch": 39.045801526717554,
+      "grad_norm": 1.3797415494918823,
+      "learning_rate": 7.858544139374524e-06,
+      "loss": 0.0929,
+      "step": 10230
+    },
+    {
+      "epoch": 39.083969465648856,
+      "grad_norm": 0.5593680739402771,
+      "learning_rate": 7.854020402895508e-06,
+      "loss": 0.0891,
+      "step": 10240
+    },
+    {
+      "epoch": 39.12213740458015,
+      "grad_norm": 0.7378935813903809,
+      "learning_rate": 7.849493198512587e-06,
+      "loss": 0.085,
+      "step": 10250
+    },
+    {
+      "epoch": 39.16030534351145,
+      "grad_norm": 0.5513617992401123,
+      "learning_rate": 7.844962531726742e-06,
+      "loss": 0.0924,
+      "step": 10260
+    },
+    {
+      "epoch": 39.19847328244275,
+      "grad_norm": 0.47100234031677246,
+      "learning_rate": 7.840428408043156e-06,
+      "loss": 0.087,
+      "step": 10270
+    },
+    {
+      "epoch": 39.23664122137404,
+      "grad_norm": 0.7264495491981506,
+      "learning_rate": 7.835890832971218e-06,
+      "loss": 0.0797,
+      "step": 10280
+    },
+    {
+      "epoch": 39.274809160305345,
+      "grad_norm": 0.6747552752494812,
+      "learning_rate": 7.831349812024513e-06,
+      "loss": 0.0816,
+      "step": 10290
+    },
+    {
+      "epoch": 39.31297709923664,
+      "grad_norm": 0.568621814250946,
+      "learning_rate": 7.826805350720807e-06,
+      "loss": 0.0829,
+      "step": 10300
+    },
+    {
+      "epoch": 39.35114503816794,
+      "grad_norm": 0.8591163158416748,
+      "learning_rate": 7.82225745458205e-06,
+      "loss": 0.0945,
+      "step": 10310
+    },
+    {
+      "epoch": 39.38931297709924,
+      "grad_norm": 0.6126816272735596,
+      "learning_rate": 7.817706129134363e-06,
+      "loss": 0.09,
+      "step": 10320
+    },
+    {
+      "epoch": 39.42748091603053,
+      "grad_norm": 0.5068755745887756,
+      "learning_rate": 7.813151379908037e-06,
+      "loss": 0.085,
+      "step": 10330
+    },
+    {
+      "epoch": 39.465648854961835,
+      "grad_norm": 0.5643684267997742,
+      "learning_rate": 7.808593212437523e-06,
+      "loss": 0.0795,
+      "step": 10340
+    },
+    {
+      "epoch": 39.50381679389313,
+      "grad_norm": 0.2955627143383026,
+      "learning_rate": 7.804031632261421e-06,
+      "loss": 0.0744,
+      "step": 10350
+    },
+    {
+      "epoch": 39.541984732824424,
+      "grad_norm": 0.4418695569038391,
+      "learning_rate": 7.799466644922484e-06,
+      "loss": 0.0872,
+      "step": 10360
+    },
+    {
+      "epoch": 39.58015267175573,
+      "grad_norm": 0.5301413536071777,
+      "learning_rate": 7.794898255967602e-06,
+      "loss": 0.0899,
+      "step": 10370
+    },
+    {
+      "epoch": 39.61832061068702,
+      "grad_norm": 0.5224511623382568,
+      "learning_rate": 7.790326470947796e-06,
+      "loss": 0.0891,
+      "step": 10380
+    },
+    {
+      "epoch": 39.656488549618324,
+      "grad_norm": 0.447360634803772,
+      "learning_rate": 7.785751295418218e-06,
+      "loss": 0.0831,
+      "step": 10390
+    },
+    {
+      "epoch": 39.69465648854962,
+      "grad_norm": 0.6668679118156433,
+      "learning_rate": 7.781172734938136e-06,
+      "loss": 0.0857,
+      "step": 10400
+    },
+    {
+      "epoch": 39.732824427480914,
+      "grad_norm": 0.5265706777572632,
+      "learning_rate": 7.776590795070932e-06,
+      "loss": 0.0826,
+      "step": 10410
+    },
+    {
+      "epoch": 39.770992366412216,
+      "grad_norm": 0.44735923409461975,
+      "learning_rate": 7.772005481384099e-06,
+      "loss": 0.0889,
+      "step": 10420
+    },
+    {
+      "epoch": 39.80916030534351,
+      "grad_norm": 0.4243389070034027,
+      "learning_rate": 7.767416799449223e-06,
+      "loss": 0.0784,
+      "step": 10430
+    },
+    {
+      "epoch": 39.847328244274806,
+      "grad_norm": 0.5234137177467346,
+      "learning_rate": 7.762824754841985e-06,
+      "loss": 0.0805,
+      "step": 10440
+    },
+    {
+      "epoch": 39.88549618320611,
+      "grad_norm": 0.6243206858634949,
+      "learning_rate": 7.758229353142153e-06,
+      "loss": 0.08,
+      "step": 10450
+    },
+    {
+      "epoch": 39.9236641221374,
+      "grad_norm": 0.368437796831131,
+      "learning_rate": 7.753630599933572e-06,
+      "loss": 0.0857,
+      "step": 10460
+    },
+    {
+      "epoch": 39.961832061068705,
+      "grad_norm": 0.6152259707450867,
+      "learning_rate": 7.74902850080416e-06,
+      "loss": 0.0829,
+      "step": 10470
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 0.43827810883522034,
+      "learning_rate": 7.744423061345907e-06,
+      "loss": 0.0865,
+      "step": 10480
+    },
+    {
+      "epoch": 40.038167938931295,
+      "grad_norm": 0.39631178975105286,
+      "learning_rate": 7.73981428715485e-06,
+      "loss": 0.0917,
+      "step": 10490
+    },
+    {
+      "epoch": 40.0763358778626,
+      "grad_norm": 0.45038852095603943,
+      "learning_rate": 7.735202183831085e-06,
+      "loss": 0.0864,
+      "step": 10500
+    },
+    {
+      "epoch": 40.11450381679389,
+      "grad_norm": 0.451773077249527,
+      "learning_rate": 7.730586756978758e-06,
+      "loss": 0.0758,
+      "step": 10510
+    },
+    {
+      "epoch": 40.152671755725194,
+      "grad_norm": 0.48386844992637634,
+      "learning_rate": 7.72596801220604e-06,
+      "loss": 0.0782,
+      "step": 10520
+    },
+    {
+      "epoch": 40.19083969465649,
+      "grad_norm": 0.8366134166717529,
+      "learning_rate": 7.721345955125147e-06,
+      "loss": 0.0861,
+      "step": 10530
+    },
+    {
+      "epoch": 40.229007633587784,
+      "grad_norm": 0.4540402889251709,
+      "learning_rate": 7.716720591352311e-06,
+      "loss": 0.0814,
+      "step": 10540
+    },
+    {
+      "epoch": 40.267175572519086,
+      "grad_norm": 0.7006372213363647,
+      "learning_rate": 7.712091926507788e-06,
+      "loss": 0.0939,
+      "step": 10550
+    },
+    {
+      "epoch": 40.30534351145038,
+      "grad_norm": 0.4313132166862488,
+      "learning_rate": 7.70745996621584e-06,
+      "loss": 0.0869,
+      "step": 10560
+    },
+    {
+      "epoch": 40.343511450381676,
+      "grad_norm": 0.33567196130752563,
+      "learning_rate": 7.702824716104736e-06,
+      "loss": 0.0924,
+      "step": 10570
+    },
+    {
+      "epoch": 40.38167938931298,
+      "grad_norm": 0.5061486959457397,
+      "learning_rate": 7.698186181806744e-06,
+      "loss": 0.0839,
+      "step": 10580
+    },
+    {
+      "epoch": 40.41984732824427,
+      "grad_norm": 0.4450092017650604,
+      "learning_rate": 7.693544368958116e-06,
+      "loss": 0.0804,
+      "step": 10590
+    },
+    {
+      "epoch": 40.458015267175576,
+      "grad_norm": 0.8481007814407349,
+      "learning_rate": 7.688899283199097e-06,
+      "loss": 0.0778,
+      "step": 10600
+    },
+    {
+      "epoch": 40.49618320610687,
+      "grad_norm": 0.34229525923728943,
+      "learning_rate": 7.684250930173902e-06,
+      "loss": 0.0805,
+      "step": 10610
+    },
+    {
+      "epoch": 40.534351145038165,
+      "grad_norm": 0.39725250005722046,
+      "learning_rate": 7.679599315530717e-06,
+      "loss": 0.0836,
+      "step": 10620
+    },
+    {
+      "epoch": 40.57251908396947,
+      "grad_norm": 0.4148560166358948,
+      "learning_rate": 7.674944444921696e-06,
+      "loss": 0.0881,
+      "step": 10630
+    },
+    {
+      "epoch": 40.61068702290076,
+      "grad_norm": 0.3573581874370575,
+      "learning_rate": 7.670286324002943e-06,
+      "loss": 0.0754,
+      "step": 10640
+    },
+    {
+      "epoch": 40.64885496183206,
+      "grad_norm": 0.4562970995903015,
+      "learning_rate": 7.665624958434514e-06,
+      "loss": 0.0943,
+      "step": 10650
+    },
+    {
+      "epoch": 40.68702290076336,
+      "grad_norm": 0.7594497799873352,
+      "learning_rate": 7.66096035388041e-06,
+      "loss": 0.0772,
+      "step": 10660
+    },
+    {
+      "epoch": 40.725190839694655,
+      "grad_norm": 0.5188819169998169,
+      "learning_rate": 7.656292516008563e-06,
+      "loss": 0.079,
+      "step": 10670
+    },
+    {
+      "epoch": 40.76335877862596,
+      "grad_norm": 0.48357266187667847,
+      "learning_rate": 7.651621450490836e-06,
+      "loss": 0.0822,
+      "step": 10680
+    },
+    {
+      "epoch": 40.80152671755725,
+      "grad_norm": 0.46848195791244507,
+      "learning_rate": 7.646947163003017e-06,
+      "loss": 0.0769,
+      "step": 10690
+    },
+    {
+      "epoch": 40.83969465648855,
+      "grad_norm": 0.448741614818573,
+      "learning_rate": 7.642269659224804e-06,
+      "loss": 0.0817,
+      "step": 10700
+    },
+    {
+      "epoch": 40.87786259541985,
+      "grad_norm": 0.2872468829154968,
+      "learning_rate": 7.637588944839803e-06,
+      "loss": 0.0789,
+      "step": 10710
+    },
+    {
+      "epoch": 40.916030534351144,
+      "grad_norm": 0.641671359539032,
+      "learning_rate": 7.632905025535529e-06,
+      "loss": 0.0874,
+      "step": 10720
+    },
+    {
+      "epoch": 40.954198473282446,
+      "grad_norm": 0.39625340700149536,
+      "learning_rate": 7.628217907003379e-06,
+      "loss": 0.0761,
+      "step": 10730
+    },
+    {
+      "epoch": 40.99236641221374,
+      "grad_norm": 0.47728121280670166,
+      "learning_rate": 7.623527594938649e-06,
+      "loss": 0.0771,
+      "step": 10740
+    },
+    {
+      "epoch": 41.030534351145036,
+      "grad_norm": 0.42139771580696106,
+      "learning_rate": 7.618834095040508e-06,
+      "loss": 0.0797,
+      "step": 10750
+    },
+    {
+      "epoch": 41.06870229007634,
+      "grad_norm": 0.5375596284866333,
+      "learning_rate": 7.614137413012001e-06,
+      "loss": 0.0822,
+      "step": 10760
+    },
+    {
+      "epoch": 41.10687022900763,
+      "grad_norm": 0.26976653933525085,
+      "learning_rate": 7.609437554560042e-06,
+      "loss": 0.0758,
+      "step": 10770
+    },
+    {
+      "epoch": 41.14503816793893,
+      "grad_norm": 0.293875515460968,
+      "learning_rate": 7.604734525395398e-06,
+      "loss": 0.0801,
+      "step": 10780
+    },
+    {
+      "epoch": 41.18320610687023,
+      "grad_norm": 1.289528727531433,
+      "learning_rate": 7.600028331232698e-06,
+      "loss": 0.0854,
+      "step": 10790
+    },
+    {
+      "epoch": 41.221374045801525,
+      "grad_norm": 0.6084293723106384,
+      "learning_rate": 7.595318977790408e-06,
+      "loss": 0.0744,
+      "step": 10800
+    },
+    {
+      "epoch": 41.25954198473283,
+      "grad_norm": 0.660408079624176,
+      "learning_rate": 7.5906064707908355e-06,
+      "loss": 0.0777,
+      "step": 10810
+    },
+    {
+      "epoch": 41.29770992366412,
+      "grad_norm": 3.1526737213134766,
+      "learning_rate": 7.585890815960125e-06,
+      "loss": 0.0843,
+      "step": 10820
+    },
+    {
+      "epoch": 41.33587786259542,
+      "grad_norm": 0.5699805617332458,
+      "learning_rate": 7.581172019028238e-06,
+      "loss": 0.0812,
+      "step": 10830
+    },
+    {
+      "epoch": 41.37404580152672,
+      "grad_norm": 0.9830774068832397,
+      "learning_rate": 7.576450085728959e-06,
+      "loss": 0.0824,
+      "step": 10840
+    },
+    {
+      "epoch": 41.412213740458014,
+      "grad_norm": 0.4558395445346832,
+      "learning_rate": 7.571725021799885e-06,
+      "loss": 0.075,
+      "step": 10850
+    },
+    {
+      "epoch": 41.45038167938931,
+      "grad_norm": 0.4857986569404602,
+      "learning_rate": 7.566996832982409e-06,
+      "loss": 0.0835,
+      "step": 10860
+    },
+    {
+      "epoch": 41.48854961832061,
+      "grad_norm": 0.8208315968513489,
+      "learning_rate": 7.56226552502173e-06,
+      "loss": 0.0939,
+      "step": 10870
+    },
+    {
+      "epoch": 41.52671755725191,
+      "grad_norm": 0.6192477941513062,
+      "learning_rate": 7.557531103666833e-06,
+      "loss": 0.0813,
+      "step": 10880
+    },
+    {
+      "epoch": 41.56488549618321,
+      "grad_norm": 0.3482249081134796,
+      "learning_rate": 7.552793574670485e-06,
+      "loss": 0.0743,
+      "step": 10890
+    },
+    {
+      "epoch": 41.603053435114504,
+      "grad_norm": 0.35872194170951843,
+      "learning_rate": 7.5480529437892304e-06,
+      "loss": 0.0756,
+      "step": 10900
+    },
+    {
+      "epoch": 41.6412213740458,
+      "grad_norm": 0.38801950216293335,
+      "learning_rate": 7.543309216783384e-06,
+      "loss": 0.0712,
+      "step": 10910
+    },
+    {
+      "epoch": 41.6793893129771,
+      "grad_norm": 0.5348705053329468,
+      "learning_rate": 7.538562399417021e-06,
+      "loss": 0.077,
+      "step": 10920
+    },
+    {
+      "epoch": 41.717557251908396,
+      "grad_norm": 0.5075365304946899,
+      "learning_rate": 7.533812497457972e-06,
+      "loss": 0.0791,
+      "step": 10930
+    },
+    {
+      "epoch": 41.7557251908397,
+      "grad_norm": 0.32560238242149353,
+      "learning_rate": 7.529059516677815e-06,
+      "loss": 0.0706,
+      "step": 10940
+    },
+    {
+      "epoch": 41.79389312977099,
+      "grad_norm": 0.5091714859008789,
+      "learning_rate": 7.524303462851872e-06,
+      "loss": 0.0773,
+      "step": 10950
+    },
+    {
+      "epoch": 41.83206106870229,
+      "grad_norm": 0.5478523373603821,
+      "learning_rate": 7.519544341759193e-06,
+      "loss": 0.0739,
+      "step": 10960
+    },
+    {
+      "epoch": 41.87022900763359,
+      "grad_norm": 0.891545832157135,
+      "learning_rate": 7.514782159182562e-06,
+      "loss": 0.0806,
+      "step": 10970
+    },
+    {
+      "epoch": 41.908396946564885,
+      "grad_norm": 0.6090933680534363,
+      "learning_rate": 7.510016920908481e-06,
+      "loss": 0.0795,
+      "step": 10980
+    },
+    {
+      "epoch": 41.94656488549618,
+      "grad_norm": 0.9301139116287231,
+      "learning_rate": 7.505248632727159e-06,
+      "loss": 0.0815,
+      "step": 10990
+    },
+    {
+      "epoch": 41.98473282442748,
+      "grad_norm": 0.6181934475898743,
+      "learning_rate": 7.50047730043252e-06,
+      "loss": 0.0806,
+      "step": 11000
+    },
+    {
+      "epoch": 42.02290076335878,
+      "grad_norm": 0.5225881934165955,
+      "learning_rate": 7.495702929822183e-06,
+      "loss": 0.0746,
+      "step": 11010
+    },
+    {
+      "epoch": 42.06106870229008,
+      "grad_norm": 0.7685074806213379,
+      "learning_rate": 7.490925526697455e-06,
+      "loss": 0.0796,
+      "step": 11020
+    },
+    {
+      "epoch": 42.099236641221374,
+      "grad_norm": 0.475824773311615,
+      "learning_rate": 7.486145096863334e-06,
+      "loss": 0.0735,
+      "step": 11030
+    },
+    {
+      "epoch": 42.13740458015267,
+      "grad_norm": 0.3295309245586395,
+      "learning_rate": 7.481361646128491e-06,
+      "loss": 0.0705,
+      "step": 11040
+    },
+    {
+      "epoch": 42.17557251908397,
+      "grad_norm": 0.2535337805747986,
+      "learning_rate": 7.476575180305271e-06,
+      "loss": 0.0808,
+      "step": 11050
+    },
+    {
+      "epoch": 42.213740458015266,
+      "grad_norm": 0.4357345402240753,
+      "learning_rate": 7.471785705209682e-06,
+      "loss": 0.075,
+      "step": 11060
+    },
+    {
+      "epoch": 42.25190839694657,
+      "grad_norm": 0.9749108552932739,
+      "learning_rate": 7.4669932266613875e-06,
+      "loss": 0.0795,
+      "step": 11070
+    },
+    {
+      "epoch": 42.29007633587786,
+      "grad_norm": 0.428632527589798,
+      "learning_rate": 7.4621977504837e-06,
+      "loss": 0.0723,
+      "step": 11080
+    },
+    {
+      "epoch": 42.32824427480916,
+      "grad_norm": 0.26020991802215576,
+      "learning_rate": 7.457399282503574e-06,
+      "loss": 0.0853,
+      "step": 11090
+    },
+    {
+      "epoch": 42.36641221374046,
+      "grad_norm": 0.3331349492073059,
+      "learning_rate": 7.4525978285516046e-06,
+      "loss": 0.0757,
+      "step": 11100
+    },
+    {
+      "epoch": 42.404580152671755,
+      "grad_norm": 0.8639835119247437,
+      "learning_rate": 7.447793394462006e-06,
+      "loss": 0.078,
+      "step": 11110
+    },
+    {
+      "epoch": 42.44274809160305,
+      "grad_norm": 2.107271909713745,
+      "learning_rate": 7.442985986072624e-06,
+      "loss": 0.0769,
+      "step": 11120
+    },
+    {
+      "epoch": 42.48091603053435,
+      "grad_norm": 0.25591397285461426,
+      "learning_rate": 7.438175609224908e-06,
+      "loss": 0.0764,
+      "step": 11130
+    },
+    {
+      "epoch": 42.51908396946565,
+      "grad_norm": 0.346733957529068,
+      "learning_rate": 7.433362269763924e-06,
+      "loss": 0.0811,
+      "step": 11140
+    },
+    {
+      "epoch": 42.55725190839695,
+      "grad_norm": 0.29196739196777344,
+      "learning_rate": 7.428545973538329e-06,
+      "loss": 0.0797,
+      "step": 11150
+    },
+    {
+      "epoch": 42.595419847328245,
+      "grad_norm": 0.6709998846054077,
+      "learning_rate": 7.423726726400381e-06,
+      "loss": 0.084,
+      "step": 11160
+    },
+    {
+      "epoch": 42.63358778625954,
+      "grad_norm": 0.46684983372688293,
+      "learning_rate": 7.418904534205917e-06,
+      "loss": 0.0788,
+      "step": 11170
+    },
+    {
+      "epoch": 42.67175572519084,
+      "grad_norm": 0.39609742164611816,
+      "learning_rate": 7.414079402814356e-06,
+      "loss": 0.0734,
+      "step": 11180
+    },
+    {
+      "epoch": 42.70992366412214,
+      "grad_norm": 0.38470694422721863,
+      "learning_rate": 7.4092513380886876e-06,
+      "loss": 0.0797,
+      "step": 11190
+    },
+    {
+      "epoch": 42.74809160305343,
+      "grad_norm": 0.26760122179985046,
+      "learning_rate": 7.4044203458954665e-06,
+      "loss": 0.084,
+      "step": 11200
+    },
+    {
+      "epoch": 42.786259541984734,
+      "grad_norm": 0.8822160363197327,
+      "learning_rate": 7.3995864321048036e-06,
+      "loss": 0.0837,
+      "step": 11210
+    },
+    {
+      "epoch": 42.82442748091603,
+      "grad_norm": 0.42078760266304016,
+      "learning_rate": 7.394749602590359e-06,
+      "loss": 0.0775,
+      "step": 11220
+    },
+    {
+      "epoch": 42.86259541984733,
+      "grad_norm": 0.2612259089946747,
+      "learning_rate": 7.389909863229337e-06,
+      "loss": 0.0713,
+      "step": 11230
+    },
+    {
+      "epoch": 42.900763358778626,
+      "grad_norm": 0.6772575974464417,
+      "learning_rate": 7.385067219902478e-06,
+      "loss": 0.0813,
+      "step": 11240
+    },
+    {
+      "epoch": 42.93893129770992,
+      "grad_norm": 0.37130260467529297,
+      "learning_rate": 7.380221678494048e-06,
+      "loss": 0.081,
+      "step": 11250
+    },
+    {
+      "epoch": 42.97709923664122,
+      "grad_norm": 0.30204957723617554,
+      "learning_rate": 7.375373244891839e-06,
+      "loss": 0.0723,
+      "step": 11260
+    },
+    {
+      "epoch": 43.01526717557252,
+      "grad_norm": 0.515661358833313,
+      "learning_rate": 7.370521924987155e-06,
+      "loss": 0.078,
+      "step": 11270
+    },
+    {
+      "epoch": 43.05343511450382,
+      "grad_norm": 0.40739643573760986,
+      "learning_rate": 7.3656677246748064e-06,
+      "loss": 0.0858,
+      "step": 11280
+    },
+    {
+      "epoch": 43.091603053435115,
+      "grad_norm": 0.37059029936790466,
+      "learning_rate": 7.360810649853105e-06,
+      "loss": 0.073,
+      "step": 11290
+    },
+    {
+      "epoch": 43.12977099236641,
+      "grad_norm": 0.36438652873039246,
+      "learning_rate": 7.355950706423854e-06,
+      "loss": 0.1083,
+      "step": 11300
+    },
+    {
+      "epoch": 43.16793893129771,
+      "grad_norm": 0.4451878070831299,
+      "learning_rate": 7.351087900292342e-06,
+      "loss": 0.0794,
+      "step": 11310
+    },
+    {
+      "epoch": 43.20610687022901,
+      "grad_norm": 0.5628499984741211,
+      "learning_rate": 7.346222237367339e-06,
+      "loss": 0.0822,
+      "step": 11320
+    },
+    {
+      "epoch": 43.2442748091603,
+      "grad_norm": 0.374586820602417,
+      "learning_rate": 7.341353723561082e-06,
+      "loss": 0.0789,
+      "step": 11330
+    },
+    {
+      "epoch": 43.282442748091604,
+      "grad_norm": 0.8239962458610535,
+      "learning_rate": 7.336482364789277e-06,
+      "loss": 0.0809,
+      "step": 11340
+    },
+    {
+      "epoch": 43.3206106870229,
+      "grad_norm": 0.36449357867240906,
+      "learning_rate": 7.331608166971082e-06,
+      "loss": 0.0776,
+      "step": 11350
+    },
+    {
+      "epoch": 43.3587786259542,
+      "grad_norm": 0.5167990326881409,
+      "learning_rate": 7.326731136029108e-06,
+      "loss": 0.089,
+      "step": 11360
+    },
+    {
+      "epoch": 43.396946564885496,
+      "grad_norm": 0.6136403679847717,
+      "learning_rate": 7.321851277889408e-06,
+      "loss": 0.0766,
+      "step": 11370
+    },
+    {
+      "epoch": 43.43511450381679,
+      "grad_norm": 0.35888543725013733,
+      "learning_rate": 7.31696859848147e-06,
+      "loss": 0.0896,
+      "step": 11380
+    },
+    {
+      "epoch": 43.47328244274809,
+      "grad_norm": 0.4521687626838684,
+      "learning_rate": 7.312083103738207e-06,
+      "loss": 0.0749,
+      "step": 11390
+    },
+    {
+      "epoch": 43.51145038167939,
+      "grad_norm": 0.4348390996456146,
+      "learning_rate": 7.307194799595958e-06,
+      "loss": 0.0852,
+      "step": 11400
+    },
+    {
+      "epoch": 43.54961832061069,
+      "grad_norm": 0.7922037839889526,
+      "learning_rate": 7.302303691994474e-06,
+      "loss": 0.0793,
+      "step": 11410
+    },
+    {
+      "epoch": 43.587786259541986,
+      "grad_norm": 0.3498508334159851,
+      "learning_rate": 7.29740978687691e-06,
+      "loss": 0.0883,
+      "step": 11420
+    },
+    {
+      "epoch": 43.62595419847328,
+      "grad_norm": 0.3470660448074341,
+      "learning_rate": 7.292513090189824e-06,
+      "loss": 0.0693,
+      "step": 11430
+    },
+    {
+      "epoch": 43.66412213740458,
+      "grad_norm": 0.2837408483028412,
+      "learning_rate": 7.287613607883164e-06,
+      "loss": 0.0732,
+      "step": 11440
+    },
+    {
+      "epoch": 43.70229007633588,
+      "grad_norm": 0.5988171696662903,
+      "learning_rate": 7.282711345910263e-06,
+      "loss": 0.0734,
+      "step": 11450
+    },
+    {
+      "epoch": 43.74045801526717,
+      "grad_norm": 0.5002927780151367,
+      "learning_rate": 7.277806310227831e-06,
+      "loss": 0.0744,
+      "step": 11460
+    },
+    {
+      "epoch": 43.778625954198475,
+      "grad_norm": 0.4815424382686615,
+      "learning_rate": 7.272898506795948e-06,
+      "loss": 0.0866,
+      "step": 11470
+    },
+    {
+      "epoch": 43.81679389312977,
+      "grad_norm": 0.4680640697479248,
+      "learning_rate": 7.267987941578058e-06,
+      "loss": 0.0883,
+      "step": 11480
+    },
+    {
+      "epoch": 43.85496183206107,
+      "grad_norm": 0.3622831404209137,
+      "learning_rate": 7.263074620540963e-06,
+      "loss": 0.0812,
+      "step": 11490
+    },
+    {
+      "epoch": 43.89312977099237,
+      "grad_norm": 0.289620965719223,
+      "learning_rate": 7.25815854965481e-06,
+      "loss": 0.0788,
+      "step": 11500
+    },
+    {
+      "epoch": 43.93129770992366,
+      "grad_norm": 0.4161958396434784,
+      "learning_rate": 7.253239734893089e-06,
+      "loss": 0.0877,
+      "step": 11510
+    },
+    {
+      "epoch": 43.969465648854964,
+      "grad_norm": 0.27894461154937744,
+      "learning_rate": 7.248318182232623e-06,
+      "loss": 0.0706,
+      "step": 11520
+    },
+    {
+      "epoch": 44.00763358778626,
+      "grad_norm": 0.4374925494194031,
+      "learning_rate": 7.243393897653565e-06,
+      "loss": 0.0757,
+      "step": 11530
+    },
+    {
+      "epoch": 44.045801526717554,
+      "grad_norm": 0.4744294285774231,
+      "learning_rate": 7.23846688713938e-06,
+      "loss": 0.0805,
+      "step": 11540
+    },
+    {
+      "epoch": 44.083969465648856,
+      "grad_norm": 0.7247238755226135,
+      "learning_rate": 7.233537156676854e-06,
+      "loss": 0.0801,
+      "step": 11550
+    },
+    {
+      "epoch": 44.12213740458015,
+      "grad_norm": 0.5223896503448486,
+      "learning_rate": 7.228604712256076e-06,
+      "loss": 0.0763,
+      "step": 11560
+    },
+    {
+      "epoch": 44.16030534351145,
+      "grad_norm": 0.32578030228614807,
+      "learning_rate": 7.2236695598704265e-06,
+      "loss": 0.0667,
+      "step": 11570
+    },
+    {
+      "epoch": 44.19847328244275,
+      "grad_norm": 0.809870183467865,
+      "learning_rate": 7.218731705516585e-06,
+      "loss": 0.072,
+      "step": 11580
+    },
+    {
+      "epoch": 44.23664122137404,
+      "grad_norm": 0.722334623336792,
+      "learning_rate": 7.21379115519451e-06,
+      "loss": 0.093,
+      "step": 11590
+    },
+    {
+      "epoch": 44.274809160305345,
+      "grad_norm": 0.8618139028549194,
+      "learning_rate": 7.208847914907431e-06,
+      "loss": 0.0869,
+      "step": 11600
+    },
+    {
+      "epoch": 44.31297709923664,
+      "grad_norm": 0.40315625071525574,
+      "learning_rate": 7.203901990661857e-06,
+      "loss": 0.0805,
+      "step": 11610
+    },
+    {
+      "epoch": 44.35114503816794,
+      "grad_norm": 0.5582493543624878,
+      "learning_rate": 7.1989533884675486e-06,
+      "loss": 0.0782,
+      "step": 11620
+    },
+    {
+      "epoch": 44.38931297709924,
+      "grad_norm": 0.5148290395736694,
+      "learning_rate": 7.1940021143375264e-06,
+      "loss": 0.0817,
+      "step": 11630
+    },
+    {
+      "epoch": 44.42748091603053,
+      "grad_norm": 0.685417652130127,
+      "learning_rate": 7.189048174288054e-06,
+      "loss": 0.0824,
+      "step": 11640
+    },
+    {
+      "epoch": 44.465648854961835,
+      "grad_norm": 0.4359172284603119,
+      "learning_rate": 7.184091574338637e-06,
+      "loss": 0.0813,
+      "step": 11650
+    },
+    {
+      "epoch": 44.50381679389313,
+      "grad_norm": 0.37148523330688477,
+      "learning_rate": 7.179132320512009e-06,
+      "loss": 0.0835,
+      "step": 11660
+    },
+    {
+      "epoch": 44.541984732824424,
+      "grad_norm": 0.714996337890625,
+      "learning_rate": 7.174170418834134e-06,
+      "loss": 0.0835,
+      "step": 11670
+    },
+    {
+      "epoch": 44.58015267175573,
+      "grad_norm": 0.30582910776138306,
+      "learning_rate": 7.1692058753341885e-06,
+      "loss": 0.071,
+      "step": 11680
+    },
+    {
+      "epoch": 44.61832061068702,
+      "grad_norm": 0.4738406836986542,
+      "learning_rate": 7.164238696044562e-06,
+      "loss": 0.0722,
+      "step": 11690
+    },
+    {
+      "epoch": 44.656488549618324,
+      "grad_norm": 0.39357179403305054,
+      "learning_rate": 7.159268887000845e-06,
+      "loss": 0.075,
+      "step": 11700
+    },
+    {
+      "epoch": 44.69465648854962,
+      "grad_norm": 0.4723822772502899,
+      "learning_rate": 7.1542964542418265e-06,
+      "loss": 0.0876,
+      "step": 11710
+    },
+    {
+      "epoch": 44.732824427480914,
+      "grad_norm": 0.43717026710510254,
+      "learning_rate": 7.149321403809478e-06,
+      "loss": 0.0779,
+      "step": 11720
+    },
+    {
+      "epoch": 44.770992366412216,
+      "grad_norm": 0.40371444821357727,
+      "learning_rate": 7.1443437417489555e-06,
+      "loss": 0.073,
+      "step": 11730
+    },
+    {
+      "epoch": 44.80916030534351,
+      "grad_norm": 0.5036951899528503,
+      "learning_rate": 7.13936347410859e-06,
+      "loss": 0.0821,
+      "step": 11740
+    },
+    {
+      "epoch": 44.847328244274806,
+      "grad_norm": 0.4972344934940338,
+      "learning_rate": 7.1343806069398745e-06,
+      "loss": 0.0886,
+      "step": 11750
+    },
+    {
+      "epoch": 44.88549618320611,
+      "grad_norm": 0.4003848135471344,
+      "learning_rate": 7.12939514629746e-06,
+      "loss": 0.0782,
+      "step": 11760
+    },
+    {
+      "epoch": 44.9236641221374,
+      "grad_norm": 0.36399978399276733,
+      "learning_rate": 7.1244070982391556e-06,
+      "loss": 0.0918,
+      "step": 11770
+    },
+    {
+      "epoch": 44.961832061068705,
+      "grad_norm": 0.2910782992839813,
+      "learning_rate": 7.119416468825908e-06,
+      "loss": 0.0756,
+      "step": 11780
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 0.429004043340683,
+      "learning_rate": 7.114423264121804e-06,
+      "loss": 0.0846,
+      "step": 11790
+    },
+    {
+      "epoch": 45.038167938931295,
+      "grad_norm": 0.3626013994216919,
+      "learning_rate": 7.1094274901940566e-06,
+      "loss": 0.1058,
+      "step": 11800
+    },
+    {
+      "epoch": 45.0763358778626,
+      "grad_norm": 0.631897509098053,
+      "learning_rate": 7.104429153113001e-06,
+      "loss": 0.0802,
+      "step": 11810
+    },
+    {
+      "epoch": 45.11450381679389,
+      "grad_norm": 0.4777592122554779,
+      "learning_rate": 7.099428258952092e-06,
+      "loss": 0.0744,
+      "step": 11820
+    },
+    {
+      "epoch": 45.152671755725194,
+      "grad_norm": 0.4223073422908783,
+      "learning_rate": 7.094424813787883e-06,
+      "loss": 0.0749,
+      "step": 11830
+    },
+    {
+      "epoch": 45.19083969465649,
+      "grad_norm": 0.3444991707801819,
+      "learning_rate": 7.089418823700035e-06,
+      "loss": 0.0805,
+      "step": 11840
+    },
+    {
+      "epoch": 45.229007633587784,
+      "grad_norm": 0.41768884658813477,
+      "learning_rate": 7.084410294771298e-06,
+      "loss": 0.0725,
+      "step": 11850
+    },
+    {
+      "epoch": 45.267175572519086,
+      "grad_norm": 0.25872498750686646,
+      "learning_rate": 7.079399233087504e-06,
+      "loss": 0.076,
+      "step": 11860
+    },
+    {
+      "epoch": 45.30534351145038,
+      "grad_norm": 0.35184618830680847,
+      "learning_rate": 7.074385644737568e-06,
+      "loss": 0.0768,
+      "step": 11870
+    },
+    {
+      "epoch": 45.343511450381676,
+      "grad_norm": 0.6684393286705017,
+      "learning_rate": 7.069369535813473e-06,
+      "loss": 0.0831,
+      "step": 11880
+    },
+    {
+      "epoch": 45.38167938931298,
+      "grad_norm": 0.6411822438240051,
+      "learning_rate": 7.064350912410261e-06,
+      "loss": 0.0759,
+      "step": 11890
+    },
+    {
+      "epoch": 45.41984732824427,
+      "grad_norm": 0.363640695810318,
+      "learning_rate": 7.059329780626036e-06,
+      "loss": 0.0761,
+      "step": 11900
+    },
+    {
+      "epoch": 45.458015267175576,
+      "grad_norm": 0.5293934345245361,
+      "learning_rate": 7.054306146561944e-06,
+      "loss": 0.0751,
+      "step": 11910
+    },
+    {
+      "epoch": 45.49618320610687,
+      "grad_norm": 0.2698386609554291,
+      "learning_rate": 7.049280016322177e-06,
+      "loss": 0.0762,
+      "step": 11920
+    },
+    {
+      "epoch": 45.534351145038165,
+      "grad_norm": 0.39354628324508667,
+      "learning_rate": 7.044251396013957e-06,
+      "loss": 0.0678,
+      "step": 11930
+    },
+    {
+      "epoch": 45.57251908396947,
+      "grad_norm": 0.8730633854866028,
+      "learning_rate": 7.039220291747528e-06,
+      "loss": 0.0888,
+      "step": 11940
+    },
+    {
+      "epoch": 45.61068702290076,
+      "grad_norm": 0.4490543603897095,
+      "learning_rate": 7.034186709636159e-06,
+      "loss": 0.0761,
+      "step": 11950
+    },
+    {
+      "epoch": 45.64885496183206,
+      "grad_norm": 0.39466390013694763,
+      "learning_rate": 7.029150655796129e-06,
+      "loss": 0.0722,
+      "step": 11960
+    },
+    {
+      "epoch": 45.68702290076336,
+      "grad_norm": 0.3895810842514038,
+      "learning_rate": 7.024112136346713e-06,
+      "loss": 0.0774,
+      "step": 11970
+    },
+    {
+      "epoch": 45.725190839694655,
+      "grad_norm": 0.816741406917572,
+      "learning_rate": 7.019071157410191e-06,
+      "loss": 0.0879,
+      "step": 11980
+    },
+    {
+      "epoch": 45.76335877862596,
+      "grad_norm": 0.5164619088172913,
+      "learning_rate": 7.014027725111826e-06,
+      "loss": 0.0811,
+      "step": 11990
+    },
+    {
+      "epoch": 45.80152671755725,
+      "grad_norm": 0.5740540623664856,
+      "learning_rate": 7.0089818455798655e-06,
+      "loss": 0.1007,
+      "step": 12000
+    },
+    {
+      "epoch": 45.83969465648855,
+      "grad_norm": 0.5206035375595093,
+      "learning_rate": 7.0039335249455285e-06,
+      "loss": 0.0829,
+      "step": 12010
+    },
+    {
+      "epoch": 45.87786259541985,
+      "grad_norm": 0.7271233797073364,
+      "learning_rate": 6.998882769342998e-06,
+      "loss": 0.0747,
+      "step": 12020
+    },
+    {
+      "epoch": 45.916030534351144,
+      "grad_norm": 0.59157794713974,
+      "learning_rate": 6.993829584909423e-06,
+      "loss": 0.0856,
+      "step": 12030
+    },
+    {
+      "epoch": 45.954198473282446,
+      "grad_norm": 0.6484169363975525,
+      "learning_rate": 6.988773977784895e-06,
+      "loss": 0.0833,
+      "step": 12040
+    },
+    {
+      "epoch": 45.99236641221374,
+      "grad_norm": 0.494017094373703,
+      "learning_rate": 6.9837159541124544e-06,
+      "loss": 0.085,
+      "step": 12050
+    },
+    {
+      "epoch": 46.030534351145036,
+      "grad_norm": 0.5254559516906738,
+      "learning_rate": 6.978655520038079e-06,
+      "loss": 0.079,
+      "step": 12060
+    },
+    {
+      "epoch": 46.06870229007634,
+      "grad_norm": 0.5512828230857849,
+      "learning_rate": 6.9735926817106704e-06,
+      "loss": 0.0768,
+      "step": 12070
+    },
+    {
+      "epoch": 46.10687022900763,
+      "grad_norm": 0.48154160380363464,
+      "learning_rate": 6.968527445282056e-06,
+      "loss": 0.0768,
+      "step": 12080
+    },
+    {
+      "epoch": 46.14503816793893,
+      "grad_norm": 0.4283272624015808,
+      "learning_rate": 6.963459816906977e-06,
+      "loss": 0.0804,
+      "step": 12090
+    },
+    {
+      "epoch": 46.18320610687023,
+      "grad_norm": 1.275875449180603,
+      "learning_rate": 6.958389802743078e-06,
+      "loss": 0.0798,
+      "step": 12100
+    },
+    {
+      "epoch": 46.221374045801525,
+      "grad_norm": 0.565065324306488,
+      "learning_rate": 6.953317408950903e-06,
+      "loss": 0.0735,
+      "step": 12110
+    },
+    {
+      "epoch": 46.25954198473283,
+      "grad_norm": 0.7162991762161255,
+      "learning_rate": 6.948242641693891e-06,
+      "loss": 0.0741,
+      "step": 12120
+    },
+    {
+      "epoch": 46.29770992366412,
+      "grad_norm": 0.41520068049430847,
+      "learning_rate": 6.9431655071383605e-06,
+      "loss": 0.0735,
+      "step": 12130
+    },
+    {
+      "epoch": 46.33587786259542,
+      "grad_norm": 0.3980543613433838,
+      "learning_rate": 6.938086011453513e-06,
+      "loss": 0.0822,
+      "step": 12140
+    },
+    {
+      "epoch": 46.37404580152672,
+      "grad_norm": 0.28883951902389526,
+      "learning_rate": 6.93300416081141e-06,
+      "loss": 0.0836,
+      "step": 12150
+    },
+    {
+      "epoch": 46.412213740458014,
+      "grad_norm": 0.5937278270721436,
+      "learning_rate": 6.927919961386984e-06,
+      "loss": 0.0764,
+      "step": 12160
+    },
+    {
+      "epoch": 46.45038167938931,
+      "grad_norm": 0.40771183371543884,
+      "learning_rate": 6.922833419358012e-06,
+      "loss": 0.0831,
+      "step": 12170
+    },
+    {
+      "epoch": 46.48854961832061,
+      "grad_norm": 0.33484435081481934,
+      "learning_rate": 6.917744540905125e-06,
+      "loss": 0.0801,
+      "step": 12180
+    },
+    {
+      "epoch": 46.52671755725191,
+      "grad_norm": 0.5971487164497375,
+      "learning_rate": 6.9126533322117875e-06,
+      "loss": 0.0843,
+      "step": 12190
+    },
+    {
+      "epoch": 46.56488549618321,
+      "grad_norm": 0.612615704536438,
+      "learning_rate": 6.9075597994643e-06,
+      "loss": 0.0836,
+      "step": 12200
+    },
+    {
+      "epoch": 46.603053435114504,
+      "grad_norm": 0.2975360155105591,
+      "learning_rate": 6.902463948851786e-06,
+      "loss": 0.0734,
+      "step": 12210
+    },
+    {
+      "epoch": 46.6412213740458,
+      "grad_norm": 0.4995099604129791,
+      "learning_rate": 6.897365786566184e-06,
+      "loss": 0.0795,
+      "step": 12220
+    },
+    {
+      "epoch": 46.6793893129771,
+      "grad_norm": 0.3913675844669342,
+      "learning_rate": 6.892265318802242e-06,
+      "loss": 0.0782,
+      "step": 12230
+    },
+    {
+      "epoch": 46.717557251908396,
+      "grad_norm": 0.4445989429950714,
+      "learning_rate": 6.887162551757507e-06,
+      "loss": 0.0844,
+      "step": 12240
+    },
+    {
+      "epoch": 46.7557251908397,
+      "grad_norm": 0.3842538297176361,
+      "learning_rate": 6.882057491632326e-06,
+      "loss": 0.0721,
+      "step": 12250
+    },
+    {
+      "epoch": 46.79389312977099,
+      "grad_norm": 0.3080384135246277,
+      "learning_rate": 6.876950144629824e-06,
+      "loss": 0.0702,
+      "step": 12260
+    },
+    {
+      "epoch": 46.83206106870229,
+      "grad_norm": 0.5869445204734802,
+      "learning_rate": 6.8718405169559114e-06,
+      "loss": 0.0687,
+      "step": 12270
+    },
+    {
+      "epoch": 46.87022900763359,
+      "grad_norm": 0.4181062579154968,
+      "learning_rate": 6.866728614819268e-06,
+      "loss": 0.0786,
+      "step": 12280
+    },
+    {
+      "epoch": 46.908396946564885,
+      "grad_norm": 0.749423086643219,
+      "learning_rate": 6.861614444431337e-06,
+      "loss": 0.0862,
+      "step": 12290
+    },
+    {
+      "epoch": 46.94656488549618,
+      "grad_norm": 0.5488717555999756,
+      "learning_rate": 6.856498012006318e-06,
+      "loss": 0.0881,
+      "step": 12300
+    },
+    {
+      "epoch": 46.98473282442748,
+      "grad_norm": 0.8571105003356934,
+      "learning_rate": 6.851379323761157e-06,
+      "loss": 0.0773,
+      "step": 12310
+    },
+    {
+      "epoch": 47.02290076335878,
+      "grad_norm": 0.5261535048484802,
+      "learning_rate": 6.846258385915545e-06,
+      "loss": 0.0912,
+      "step": 12320
+    },
+    {
+      "epoch": 47.06106870229008,
+      "grad_norm": 0.895200252532959,
+      "learning_rate": 6.841135204691902e-06,
+      "loss": 0.0858,
+      "step": 12330
+    },
+    {
+      "epoch": 47.099236641221374,
+      "grad_norm": 0.49354881048202515,
+      "learning_rate": 6.8360097863153775e-06,
+      "loss": 0.0762,
+      "step": 12340
+    },
+    {
+      "epoch": 47.13740458015267,
+      "grad_norm": 0.42916223406791687,
+      "learning_rate": 6.830882137013839e-06,
+      "loss": 0.0749,
+      "step": 12350
+    },
+    {
+      "epoch": 47.17557251908397,
+      "grad_norm": 0.6174752116203308,
+      "learning_rate": 6.825752263017863e-06,
+      "loss": 0.0797,
+      "step": 12360
+    },
+    {
+      "epoch": 47.213740458015266,
+      "grad_norm": 0.4598800241947174,
+      "learning_rate": 6.820620170560731e-06,
+      "loss": 0.0828,
+      "step": 12370
+    },
+    {
+      "epoch": 47.25190839694657,
+      "grad_norm": 0.3219451904296875,
+      "learning_rate": 6.815485865878418e-06,
+      "loss": 0.0845,
+      "step": 12380
+    },
+    {
+      "epoch": 47.29007633587786,
+      "grad_norm": 0.4897594451904297,
+      "learning_rate": 6.8103493552095875e-06,
+      "loss": 0.0874,
+      "step": 12390
+    },
+    {
+      "epoch": 47.32824427480916,
+      "grad_norm": 0.35891905426979065,
+      "learning_rate": 6.805210644795588e-06,
+      "loss": 0.0715,
+      "step": 12400
+    },
+    {
+      "epoch": 47.36641221374046,
+      "grad_norm": 0.6998705267906189,
+      "learning_rate": 6.8000697408804326e-06,
+      "loss": 0.0799,
+      "step": 12410
+    },
+    {
+      "epoch": 47.404580152671755,
+      "grad_norm": 0.7352116703987122,
+      "learning_rate": 6.794926649710807e-06,
+      "loss": 0.0873,
+      "step": 12420
+    },
+    {
+      "epoch": 47.44274809160305,
+      "grad_norm": 0.34309205412864685,
+      "learning_rate": 6.7897813775360536e-06,
+      "loss": 0.0753,
+      "step": 12430
+    },
+    {
+      "epoch": 47.48091603053435,
+      "grad_norm": 0.6588733792304993,
+      "learning_rate": 6.784633930608158e-06,
+      "loss": 0.0777,
+      "step": 12440
+    },
+    {
+      "epoch": 47.51908396946565,
+      "grad_norm": 0.43710416555404663,
+      "learning_rate": 6.779484315181759e-06,
+      "loss": 0.0841,
+      "step": 12450
+    },
+    {
+      "epoch": 47.55725190839695,
+      "grad_norm": 0.4144386649131775,
+      "learning_rate": 6.774332537514122e-06,
+      "loss": 0.0727,
+      "step": 12460
+    },
+    {
+      "epoch": 47.595419847328245,
+      "grad_norm": 0.3395448625087738,
+      "learning_rate": 6.769178603865143e-06,
+      "loss": 0.08,
+      "step": 12470
+    },
+    {
+      "epoch": 47.63358778625954,
+      "grad_norm": 0.38854730129241943,
+      "learning_rate": 6.764022520497337e-06,
+      "loss": 0.0768,
+      "step": 12480
+    },
+    {
+      "epoch": 47.67175572519084,
+      "grad_norm": 0.43010884523391724,
+      "learning_rate": 6.758864293675833e-06,
+      "loss": 0.0757,
+      "step": 12490
+    },
+    {
+      "epoch": 47.70992366412214,
+      "grad_norm": 0.31416982412338257,
+      "learning_rate": 6.753703929668363e-06,
+      "loss": 0.0828,
+      "step": 12500
+    },
+    {
+      "epoch": 47.74809160305343,
+      "grad_norm": 0.3258754312992096,
+      "learning_rate": 6.7485414347452535e-06,
+      "loss": 0.0707,
+      "step": 12510
+    },
+    {
+      "epoch": 47.786259541984734,
+      "grad_norm": 0.4140434265136719,
+      "learning_rate": 6.743376815179424e-06,
+      "loss": 0.0717,
+      "step": 12520
+    },
+    {
+      "epoch": 47.82442748091603,
+      "grad_norm": 0.3196008801460266,
+      "learning_rate": 6.738210077246377e-06,
+      "loss": 0.0771,
+      "step": 12530
+    },
+    {
+      "epoch": 47.86259541984733,
+      "grad_norm": 0.36054739356040955,
+      "learning_rate": 6.733041227224182e-06,
+      "loss": 0.0739,
+      "step": 12540
+    },
+    {
+      "epoch": 47.900763358778626,
+      "grad_norm": 0.32305601239204407,
+      "learning_rate": 6.72787027139348e-06,
+      "loss": 0.0793,
+      "step": 12550
+    },
+    {
+      "epoch": 47.93893129770992,
+      "grad_norm": 0.41172558069229126,
+      "learning_rate": 6.72269721603747e-06,
+      "loss": 0.0746,
+      "step": 12560
+    },
+    {
+      "epoch": 47.97709923664122,
+      "grad_norm": 0.5158224701881409,
+      "learning_rate": 6.717522067441904e-06,
+      "loss": 0.0709,
+      "step": 12570
+    },
+    {
+      "epoch": 48.01526717557252,
+      "grad_norm": 0.33051443099975586,
+      "learning_rate": 6.712344831895075e-06,
+      "loss": 0.0799,
+      "step": 12580
+    },
+    {
+      "epoch": 48.05343511450382,
+      "grad_norm": 0.2499360889196396,
+      "learning_rate": 6.707165515687811e-06,
+      "loss": 0.0933,
+      "step": 12590
+    },
+    {
+      "epoch": 48.091603053435115,
+      "grad_norm": 0.3040105700492859,
+      "learning_rate": 6.70198412511347e-06,
+      "loss": 0.0684,
+      "step": 12600
+    },
+    {
+      "epoch": 48.12977099236641,
+      "grad_norm": 0.5028464794158936,
+      "learning_rate": 6.696800666467931e-06,
+      "loss": 0.0763,
+      "step": 12610
+    },
+    {
+      "epoch": 48.16793893129771,
+      "grad_norm": 0.6218233704566956,
+      "learning_rate": 6.691615146049584e-06,
+      "loss": 0.0713,
+      "step": 12620
+    },
+    {
+      "epoch": 48.20610687022901,
+      "grad_norm": 1.142249345779419,
+      "learning_rate": 6.686427570159324e-06,
+      "loss": 0.0881,
+      "step": 12630
+    },
+    {
+      "epoch": 48.2442748091603,
+      "grad_norm": 0.40421727299690247,
+      "learning_rate": 6.681237945100549e-06,
+      "loss": 0.0699,
+      "step": 12640
+    },
+    {
+      "epoch": 48.282442748091604,
+      "grad_norm": 0.4574754238128662,
+      "learning_rate": 6.676046277179139e-06,
+      "loss": 0.0794,
+      "step": 12650
+    },
+    {
+      "epoch": 48.3206106870229,
+      "grad_norm": 0.33654555678367615,
+      "learning_rate": 6.670852572703462e-06,
+      "loss": 0.0722,
+      "step": 12660
+    },
+    {
+      "epoch": 48.3587786259542,
+      "grad_norm": 0.2677624523639679,
+      "learning_rate": 6.665656837984359e-06,
+      "loss": 0.0855,
+      "step": 12670
+    },
+    {
+      "epoch": 48.396946564885496,
+      "grad_norm": 0.7722212672233582,
+      "learning_rate": 6.660459079335136e-06,
+      "loss": 0.0804,
+      "step": 12680
+    },
+    {
+      "epoch": 48.43511450381679,
+      "grad_norm": 0.32007500529289246,
+      "learning_rate": 6.655259303071558e-06,
+      "loss": 0.0786,
+      "step": 12690
+    },
+    {
+      "epoch": 48.47328244274809,
+      "grad_norm": 0.28395479917526245,
+      "learning_rate": 6.650057515511849e-06,
+      "loss": 0.0741,
+      "step": 12700
+    },
+    {
+      "epoch": 48.51145038167939,
+      "grad_norm": 0.5436563491821289,
+      "learning_rate": 6.644853722976667e-06,
+      "loss": 0.0879,
+      "step": 12710
+    },
+    {
+      "epoch": 48.54961832061069,
+      "grad_norm": 0.5980028510093689,
+      "learning_rate": 6.639647931789114e-06,
+      "loss": 0.0855,
+      "step": 12720
+    },
+    {
+      "epoch": 48.587786259541986,
+      "grad_norm": 0.345226913690567,
+      "learning_rate": 6.634440148274712e-06,
+      "loss": 0.0693,
+      "step": 12730
+    },
+    {
+      "epoch": 48.62595419847328,
+      "grad_norm": 0.559097409248352,
+      "learning_rate": 6.6292303787614156e-06,
+      "loss": 0.0704,
+      "step": 12740
+    },
+    {
+      "epoch": 48.66412213740458,
+      "grad_norm": 0.2557688355445862,
+      "learning_rate": 6.624018629579582e-06,
+      "loss": 0.0746,
+      "step": 12750
+    },
+    {
+      "epoch": 48.70229007633588,
+      "grad_norm": 0.38881349563598633,
+      "learning_rate": 6.618804907061977e-06,
+      "loss": 0.0727,
+      "step": 12760
+    },
+    {
+      "epoch": 48.74045801526717,
+      "grad_norm": 0.34416061639785767,
+      "learning_rate": 6.613589217543765e-06,
+      "loss": 0.0781,
+      "step": 12770
+    },
+    {
+      "epoch": 48.778625954198475,
+      "grad_norm": 0.6163748502731323,
+      "learning_rate": 6.608371567362505e-06,
+      "loss": 0.0785,
+      "step": 12780
+    },
+    {
+      "epoch": 48.81679389312977,
+      "grad_norm": 0.2579529583454132,
+      "learning_rate": 6.60315196285813e-06,
+      "loss": 0.064,
+      "step": 12790
+    },
+    {
+      "epoch": 48.85496183206107,
+      "grad_norm": 0.32075974345207214,
+      "learning_rate": 6.5979304103729545e-06,
+      "loss": 0.0831,
+      "step": 12800
+    },
+    {
+      "epoch": 48.89312977099237,
+      "grad_norm": 0.28072690963745117,
+      "learning_rate": 6.592706916251653e-06,
+      "loss": 0.068,
+      "step": 12810
+    },
+    {
+      "epoch": 48.93129770992366,
+      "grad_norm": 0.3512701690196991,
+      "learning_rate": 6.587481486841267e-06,
+      "loss": 0.0832,
+      "step": 12820
+    },
+    {
+      "epoch": 48.969465648854964,
+      "grad_norm": 0.3708653151988983,
+      "learning_rate": 6.582254128491184e-06,
+      "loss": 0.08,
+      "step": 12830
+    },
+    {
+      "epoch": 49.00763358778626,
+      "grad_norm": 0.5999414324760437,
+      "learning_rate": 6.577024847553139e-06,
+      "loss": 0.0707,
+      "step": 12840
+    },
+    {
+      "epoch": 49.045801526717554,
+      "grad_norm": 0.5251232385635376,
+      "learning_rate": 6.5717936503812e-06,
+      "loss": 0.078,
+      "step": 12850
+    },
+    {
+      "epoch": 49.083969465648856,
+      "grad_norm": 0.30655843019485474,
+      "learning_rate": 6.5665605433317655e-06,
+      "loss": 0.0761,
+      "step": 12860
+    },
+    {
+      "epoch": 49.12213740458015,
+      "grad_norm": 0.3946235775947571,
+      "learning_rate": 6.561325532763554e-06,
+      "loss": 0.0757,
+      "step": 12870
+    },
+    {
+      "epoch": 49.16030534351145,
+      "grad_norm": 0.27524781227111816,
+      "learning_rate": 6.556088625037598e-06,
+      "loss": 0.071,
+      "step": 12880
+    },
+    {
+      "epoch": 49.19847328244275,
+      "grad_norm": 0.27096670866012573,
+      "learning_rate": 6.550849826517231e-06,
+      "loss": 0.0781,
+      "step": 12890
+    },
+    {
+      "epoch": 49.23664122137404,
+      "grad_norm": 0.4801909029483795,
+      "learning_rate": 6.54560914356809e-06,
+      "loss": 0.0745,
+      "step": 12900
+    },
+    {
+      "epoch": 49.274809160305345,
+      "grad_norm": 0.4032978117465973,
+      "learning_rate": 6.5403665825580975e-06,
+      "loss": 0.0764,
+      "step": 12910
+    },
+    {
+      "epoch": 49.31297709923664,
+      "grad_norm": 0.46122097969055176,
+      "learning_rate": 6.53512214985746e-06,
+      "loss": 0.0733,
+      "step": 12920
+    },
+    {
+      "epoch": 49.35114503816794,
+      "grad_norm": 0.6307662129402161,
+      "learning_rate": 6.529875851838659e-06,
+      "loss": 0.0768,
+      "step": 12930
+    },
+    {
+      "epoch": 49.38931297709924,
+      "grad_norm": 0.28868502378463745,
+      "learning_rate": 6.5246276948764394e-06,
+      "loss": 0.0859,
+      "step": 12940
+    },
+    {
+      "epoch": 49.42748091603053,
+      "grad_norm": 0.463450163602829,
+      "learning_rate": 6.519377685347808e-06,
+      "loss": 0.0768,
+      "step": 12950
+    },
+    {
+      "epoch": 49.465648854961835,
+      "grad_norm": 0.4888218641281128,
+      "learning_rate": 6.514125829632021e-06,
+      "loss": 0.0741,
+      "step": 12960
+    },
+    {
+      "epoch": 49.50381679389313,
+      "grad_norm": 0.615373432636261,
+      "learning_rate": 6.508872134110578e-06,
+      "loss": 0.0736,
+      "step": 12970
+    },
+    {
+      "epoch": 49.541984732824424,
+      "grad_norm": 0.2985082268714905,
+      "learning_rate": 6.5036166051672135e-06,
+      "loss": 0.0885,
+      "step": 12980
+    },
+    {
+      "epoch": 49.58015267175573,
+      "grad_norm": 0.461047887802124,
+      "learning_rate": 6.498359249187893e-06,
+      "loss": 0.0756,
+      "step": 12990
+    },
+    {
+      "epoch": 49.61832061068702,
+      "grad_norm": 0.46043241024017334,
+      "learning_rate": 6.4931000725607985e-06,
+      "loss": 0.0753,
+      "step": 13000
+    },
+    {
+      "epoch": 49.656488549618324,
+      "grad_norm": 0.4072940945625305,
+      "learning_rate": 6.487839081676327e-06,
+      "loss": 0.0678,
+      "step": 13010
+    },
+    {
+      "epoch": 49.69465648854962,
+      "grad_norm": 0.6342893242835999,
+      "learning_rate": 6.482576282927076e-06,
+      "loss": 0.0719,
+      "step": 13020
+    },
+    {
+      "epoch": 49.732824427480914,
+      "grad_norm": 0.4662512242794037,
+      "learning_rate": 6.477311682707844e-06,
+      "loss": 0.0745,
+      "step": 13030
+    },
+    {
+      "epoch": 49.770992366412216,
+      "grad_norm": 0.2835651934146881,
+      "learning_rate": 6.472045287415616e-06,
+      "loss": 0.0689,
+      "step": 13040
+    },
+    {
+      "epoch": 49.80916030534351,
+      "grad_norm": 0.3895277678966522,
+      "learning_rate": 6.466777103449559e-06,
+      "loss": 0.0747,
+      "step": 13050
+    },
+    {
+      "epoch": 49.847328244274806,
+      "grad_norm": 0.32042866945266724,
+      "learning_rate": 6.461507137211012e-06,
+      "loss": 0.0688,
+      "step": 13060
+    },
+    {
+      "epoch": 49.88549618320611,
+      "grad_norm": 0.5186818838119507,
+      "learning_rate": 6.456235395103483e-06,
+      "loss": 0.0784,
+      "step": 13070
+    },
+    {
+      "epoch": 49.9236641221374,
+      "grad_norm": 0.28963205218315125,
+      "learning_rate": 6.450961883532635e-06,
+      "loss": 0.0769,
+      "step": 13080
+    },
+    {
+      "epoch": 49.961832061068705,
+      "grad_norm": 0.374496728181839,
+      "learning_rate": 6.445686608906283e-06,
+      "loss": 0.0674,
+      "step": 13090
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 0.35718169808387756,
+      "learning_rate": 6.44040957763438e-06,
+      "loss": 0.0773,
+      "step": 13100
+    },
+    {
+      "epoch": 50.038167938931295,
+      "grad_norm": 0.5534579753875732,
+      "learning_rate": 6.435130796129019e-06,
+      "loss": 0.0677,
+      "step": 13110
+    },
+    {
+      "epoch": 50.0763358778626,
+      "grad_norm": 0.5444230437278748,
+      "learning_rate": 6.4298502708044165e-06,
+      "loss": 0.08,
+      "step": 13120
+    },
+    {
+      "epoch": 50.11450381679389,
+      "grad_norm": 0.5170355439186096,
+      "learning_rate": 6.424568008076909e-06,
+      "loss": 0.07,
+      "step": 13130
+    },
+    {
+      "epoch": 50.152671755725194,
+      "grad_norm": 0.514159083366394,
+      "learning_rate": 6.419284014364944e-06,
+      "loss": 0.0907,
+      "step": 13140
+    },
+    {
+      "epoch": 50.19083969465649,
+      "grad_norm": 0.31660693883895874,
+      "learning_rate": 6.413998296089071e-06,
+      "loss": 0.0744,
+      "step": 13150
+    },
+    {
+      "epoch": 50.229007633587784,
+      "grad_norm": 0.2514859139919281,
+      "learning_rate": 6.408710859671938e-06,
+      "loss": 0.0733,
+      "step": 13160
+    },
+    {
+      "epoch": 50.267175572519086,
+      "grad_norm": 0.6438994407653809,
+      "learning_rate": 6.403421711538278e-06,
+      "loss": 0.0816,
+      "step": 13170
+    },
+    {
+      "epoch": 50.30534351145038,
+      "grad_norm": 0.5075148940086365,
+      "learning_rate": 6.398130858114904e-06,
+      "loss": 0.0808,
+      "step": 13180
+    },
+    {
+      "epoch": 50.343511450381676,
+      "grad_norm": 1.2176218032836914,
+      "learning_rate": 6.392838305830702e-06,
+      "loss": 0.0821,
+      "step": 13190
+    },
+    {
+      "epoch": 50.38167938931298,
+      "grad_norm": 0.32374969124794006,
+      "learning_rate": 6.387544061116622e-06,
+      "loss": 0.0789,
+      "step": 13200
+    },
+    {
+      "epoch": 50.41984732824427,
+      "grad_norm": 0.539487898349762,
+      "learning_rate": 6.382248130405671e-06,
+      "loss": 0.0717,
+      "step": 13210
+    },
+    {
+      "epoch": 50.458015267175576,
+      "grad_norm": 0.29437029361724854,
+      "learning_rate": 6.376950520132906e-06,
+      "loss": 0.0715,
+      "step": 13220
+    },
+    {
+      "epoch": 50.49618320610687,
+      "grad_norm": 0.2797829806804657,
+      "learning_rate": 6.371651236735418e-06,
+      "loss": 0.0691,
+      "step": 13230
+    },
+    {
+      "epoch": 50.534351145038165,
+      "grad_norm": 1.2593516111373901,
+      "learning_rate": 6.366350286652341e-06,
+      "loss": 0.0775,
+      "step": 13240
+    },
+    {
+      "epoch": 50.57251908396947,
+      "grad_norm": 0.5753695964813232,
+      "learning_rate": 6.361047676324827e-06,
+      "loss": 0.0787,
+      "step": 13250
+    },
+    {
+      "epoch": 50.61068702290076,
+      "grad_norm": 0.6826583743095398,
+      "learning_rate": 6.355743412196047e-06,
+      "loss": 0.071,
+      "step": 13260
+    },
+    {
+      "epoch": 50.64885496183206,
+      "grad_norm": 0.655083417892456,
+      "learning_rate": 6.350437500711184e-06,
+      "loss": 0.07,
+      "step": 13270
+    },
+    {
+      "epoch": 50.68702290076336,
+      "grad_norm": 0.3876498341560364,
+      "learning_rate": 6.345129948317419e-06,
+      "loss": 0.0641,
+      "step": 13280
+    },
+    {
+      "epoch": 50.725190839694655,
+      "grad_norm": 0.38112854957580566,
+      "learning_rate": 6.339820761463929e-06,
+      "loss": 0.0754,
+      "step": 13290
+    },
+    {
+      "epoch": 50.76335877862596,
+      "grad_norm": 0.402295857667923,
+      "learning_rate": 6.334509946601879e-06,
+      "loss": 0.0878,
+      "step": 13300
+    },
+    {
+      "epoch": 50.80152671755725,
+      "grad_norm": 0.3208377957344055,
+      "learning_rate": 6.329197510184406e-06,
+      "loss": 0.0746,
+      "step": 13310
+    },
+    {
+      "epoch": 50.83969465648855,
+      "grad_norm": 0.3299608528614044,
+      "learning_rate": 6.323883458666624e-06,
+      "loss": 0.0748,
+      "step": 13320
+    },
+    {
+      "epoch": 50.87786259541985,
+      "grad_norm": 0.3016846477985382,
+      "learning_rate": 6.318567798505605e-06,
+      "loss": 0.0708,
+      "step": 13330
+    },
+    {
+      "epoch": 50.916030534351144,
+      "grad_norm": 0.6639785766601562,
+      "learning_rate": 6.313250536160378e-06,
+      "loss": 0.0789,
+      "step": 13340
+    },
+    {
+      "epoch": 50.954198473282446,
+      "grad_norm": 0.3203372061252594,
+      "learning_rate": 6.307931678091918e-06,
+      "loss": 0.0794,
+      "step": 13350
+    },
+    {
+      "epoch": 50.99236641221374,
+      "grad_norm": 0.44946426153182983,
+      "learning_rate": 6.3026112307631385e-06,
+      "loss": 0.0792,
+      "step": 13360
+    },
+    {
+      "epoch": 51.030534351145036,
+      "grad_norm": 0.32271233201026917,
+      "learning_rate": 6.297289200638888e-06,
+      "loss": 0.0767,
+      "step": 13370
+    },
+    {
+      "epoch": 51.06870229007634,
+      "grad_norm": 0.32984182238578796,
+      "learning_rate": 6.29196559418593e-06,
+      "loss": 0.077,
+      "step": 13380
+    },
+    {
+      "epoch": 51.10687022900763,
+      "grad_norm": 0.27354347705841064,
+      "learning_rate": 6.286640417872951e-06,
+      "loss": 0.0725,
+      "step": 13390
+    },
+    {
+      "epoch": 51.14503816793893,
+      "grad_norm": 0.4007064700126648,
+      "learning_rate": 6.281313678170543e-06,
+      "loss": 0.0697,
+      "step": 13400
+    },
+    {
+      "epoch": 51.18320610687023,
+      "grad_norm": 0.2881835401058197,
+      "learning_rate": 6.275985381551195e-06,
+      "loss": 0.0816,
+      "step": 13410
+    },
+    {
+      "epoch": 51.221374045801525,
+      "grad_norm": 0.7041976451873779,
+      "learning_rate": 6.270655534489292e-06,
+      "loss": 0.0757,
+      "step": 13420
+    },
+    {
+      "epoch": 51.25954198473283,
+      "grad_norm": 0.28111717104911804,
+      "learning_rate": 6.265324143461098e-06,
+      "loss": 0.0808,
+      "step": 13430
+    },
+    {
+      "epoch": 51.29770992366412,
+      "grad_norm": 0.38881856203079224,
+      "learning_rate": 6.259991214944758e-06,
+      "loss": 0.0752,
+      "step": 13440
+    },
+    {
+      "epoch": 51.33587786259542,
+      "grad_norm": 0.25334322452545166,
+      "learning_rate": 6.254656755420283e-06,
+      "loss": 0.0712,
+      "step": 13450
+    },
+    {
+      "epoch": 51.37404580152672,
+      "grad_norm": 0.37281477451324463,
+      "learning_rate": 6.249320771369545e-06,
+      "loss": 0.0813,
+      "step": 13460
+    },
+    {
+      "epoch": 51.412213740458014,
+      "grad_norm": 1.6410506963729858,
+      "learning_rate": 6.243983269276263e-06,
+      "loss": 0.0907,
+      "step": 13470
+    },
+    {
+      "epoch": 51.45038167938931,
+      "grad_norm": 0.2578519582748413,
+      "learning_rate": 6.238644255626013e-06,
+      "loss": 0.0719,
+      "step": 13480
+    },
+    {
+      "epoch": 51.48854961832061,
+      "grad_norm": 0.912161648273468,
+      "learning_rate": 6.233303736906193e-06,
+      "loss": 0.0714,
+      "step": 13490
+    },
+    {
+      "epoch": 51.52671755725191,
+      "grad_norm": 0.5383570790290833,
+      "learning_rate": 6.2279617196060394e-06,
+      "loss": 0.0743,
+      "step": 13500
+    },
+    {
+      "epoch": 51.56488549618321,
+      "grad_norm": 0.6001830697059631,
+      "learning_rate": 6.2226182102166085e-06,
+      "loss": 0.0739,
+      "step": 13510
+    },
+    {
+      "epoch": 51.603053435114504,
+      "grad_norm": 0.40439608693122864,
+      "learning_rate": 6.217273215230767e-06,
+      "loss": 0.0749,
+      "step": 13520
+    },
+    {
+      "epoch": 51.6412213740458,
+      "grad_norm": 0.298961877822876,
+      "learning_rate": 6.2119267411431885e-06,
+      "loss": 0.0668,
+      "step": 13530
+    },
+    {
+      "epoch": 51.6793893129771,
+      "grad_norm": 0.4611186385154724,
+      "learning_rate": 6.206578794450339e-06,
+      "loss": 0.0789,
+      "step": 13540
+    },
+    {
+      "epoch": 51.717557251908396,
+      "grad_norm": 0.2702903747558594,
+      "learning_rate": 6.2012293816504855e-06,
+      "loss": 0.077,
+      "step": 13550
+    },
+    {
+      "epoch": 51.7557251908397,
+      "grad_norm": 0.38614171743392944,
+      "learning_rate": 6.195878509243661e-06,
+      "loss": 0.0801,
+      "step": 13560
+    },
+    {
+      "epoch": 51.79389312977099,
+      "grad_norm": 0.3832460343837738,
+      "learning_rate": 6.190526183731686e-06,
+      "loss": 0.0675,
+      "step": 13570
+    },
+    {
+      "epoch": 51.83206106870229,
+      "grad_norm": 0.7933033108711243,
+      "learning_rate": 6.185172411618138e-06,
+      "loss": 0.0703,
+      "step": 13580
+    },
+    {
+      "epoch": 51.87022900763359,
+      "grad_norm": 0.3744434714317322,
+      "learning_rate": 6.179817199408355e-06,
+      "loss": 0.0737,
+      "step": 13590
+    },
+    {
+      "epoch": 51.908396946564885,
+      "grad_norm": 0.49073082208633423,
+      "learning_rate": 6.174460553609426e-06,
+      "loss": 0.0843,
+      "step": 13600
+    },
+    {
+      "epoch": 51.94656488549618,
+      "grad_norm": 0.438060462474823,
+      "learning_rate": 6.16910248073018e-06,
+      "loss": 0.0729,
+      "step": 13610
+    },
+    {
+      "epoch": 51.98473282442748,
+      "grad_norm": 0.3065158724784851,
+      "learning_rate": 6.16374298728118e-06,
+      "loss": 0.075,
+      "step": 13620
+    },
+    {
+      "epoch": 52.02290076335878,
+      "grad_norm": 0.3208974301815033,
+      "learning_rate": 6.158382079774716e-06,
+      "loss": 0.0757,
+      "step": 13630
+    },
+    {
+      "epoch": 52.06106870229008,
+      "grad_norm": 0.36276963353157043,
+      "learning_rate": 6.153019764724799e-06,
+      "loss": 0.079,
+      "step": 13640
+    },
+    {
+      "epoch": 52.099236641221374,
+      "grad_norm": 0.36144378781318665,
+      "learning_rate": 6.147656048647144e-06,
+      "loss": 0.0665,
+      "step": 13650
+    },
+    {
+      "epoch": 52.13740458015267,
+      "grad_norm": 0.42839837074279785,
+      "learning_rate": 6.1422909380591724e-06,
+      "loss": 0.0732,
+      "step": 13660
+    },
+    {
+      "epoch": 52.17557251908397,
+      "grad_norm": 0.589056134223938,
+      "learning_rate": 6.136924439480001e-06,
+      "loss": 0.0813,
+      "step": 13670
+    },
+    {
+      "epoch": 52.213740458015266,
+      "grad_norm": 0.7054497599601746,
+      "learning_rate": 6.13155655943043e-06,
+      "loss": 0.0792,
+      "step": 13680
+    },
+    {
+      "epoch": 52.25190839694657,
+      "grad_norm": 0.5286399126052856,
+      "learning_rate": 6.126187304432941e-06,
+      "loss": 0.0732,
+      "step": 13690
+    },
+    {
+      "epoch": 52.29007633587786,
+      "grad_norm": 0.4254150986671448,
+      "learning_rate": 6.1208166810116846e-06,
+      "loss": 0.0774,
+      "step": 13700
+    },
+    {
+      "epoch": 52.32824427480916,
+      "grad_norm": 0.6747491955757141,
+      "learning_rate": 6.115444695692474e-06,
+      "loss": 0.0777,
+      "step": 13710
+    },
+    {
+      "epoch": 52.36641221374046,
+      "grad_norm": 0.3658086657524109,
+      "learning_rate": 6.110071355002779e-06,
+      "loss": 0.0655,
+      "step": 13720
+    },
+    {
+      "epoch": 52.404580152671755,
+      "grad_norm": 0.4212816655635834,
+      "learning_rate": 6.104696665471715e-06,
+      "loss": 0.0842,
+      "step": 13730
+    },
+    {
+      "epoch": 52.44274809160305,
+      "grad_norm": 0.6053048372268677,
+      "learning_rate": 6.099320633630036e-06,
+      "loss": 0.0804,
+      "step": 13740
+    },
+    {
+      "epoch": 52.48091603053435,
+      "grad_norm": 0.3349289894104004,
+      "learning_rate": 6.093943266010128e-06,
+      "loss": 0.0729,
+      "step": 13750
+    },
+    {
+      "epoch": 52.51908396946565,
+      "grad_norm": 0.3947334289550781,
+      "learning_rate": 6.088564569146001e-06,
+      "loss": 0.088,
+      "step": 13760
+    },
+    {
+      "epoch": 52.55725190839695,
+      "grad_norm": 0.4069903790950775,
+      "learning_rate": 6.083184549573275e-06,
+      "loss": 0.0761,
+      "step": 13770
+    },
+    {
+      "epoch": 52.595419847328245,
+      "grad_norm": 0.526199221611023,
+      "learning_rate": 6.0778032138291845e-06,
+      "loss": 0.0743,
+      "step": 13780
+    },
+    {
+      "epoch": 52.63358778625954,
+      "grad_norm": 0.37848567962646484,
+      "learning_rate": 6.072420568452559e-06,
+      "loss": 0.0722,
+      "step": 13790
+    },
+    {
+      "epoch": 52.67175572519084,
+      "grad_norm": 0.479997843503952,
+      "learning_rate": 6.067036619983822e-06,
+      "loss": 0.0813,
+      "step": 13800
+    },
+    {
+      "epoch": 52.70992366412214,
+      "grad_norm": 0.37879565358161926,
+      "learning_rate": 6.061651374964974e-06,
+      "loss": 0.0713,
+      "step": 13810
+    },
+    {
+      "epoch": 52.74809160305343,
+      "grad_norm": 0.2613222897052765,
+      "learning_rate": 6.056264839939601e-06,
+      "loss": 0.0733,
+      "step": 13820
+    },
+    {
+      "epoch": 52.786259541984734,
+      "grad_norm": 0.4881426990032196,
+      "learning_rate": 6.050877021452845e-06,
+      "loss": 0.0693,
+      "step": 13830
+    },
+    {
+      "epoch": 52.82442748091603,
+      "grad_norm": 0.8297280073165894,
+      "learning_rate": 6.0454879260514196e-06,
+      "loss": 0.0751,
+      "step": 13840
+    },
+    {
+      "epoch": 52.86259541984733,
+      "grad_norm": 0.8701691627502441,
+      "learning_rate": 6.0400975602835795e-06,
+      "loss": 0.0745,
+      "step": 13850
+    },
+    {
+      "epoch": 52.900763358778626,
+      "grad_norm": 0.40549200773239136,
+      "learning_rate": 6.034705930699129e-06,
+      "loss": 0.077,
+      "step": 13860
+    },
+    {
+      "epoch": 52.93893129770992,
+      "grad_norm": 0.5269019603729248,
+      "learning_rate": 6.029313043849407e-06,
+      "loss": 0.0732,
+      "step": 13870
+    },
+    {
+      "epoch": 52.97709923664122,
+      "grad_norm": 0.6514595746994019,
+      "learning_rate": 6.0239189062872795e-06,
+      "loss": 0.0683,
+      "step": 13880
+    },
+    {
+      "epoch": 53.01526717557252,
+      "grad_norm": 0.270759254693985,
+      "learning_rate": 6.01852352456713e-06,
+      "loss": 0.0774,
+      "step": 13890
+    },
+    {
+      "epoch": 53.05343511450382,
+      "grad_norm": 0.37720581889152527,
+      "learning_rate": 6.013126905244858e-06,
+      "loss": 0.0703,
+      "step": 13900
+    },
+    {
+      "epoch": 53.091603053435115,
+      "grad_norm": 0.45019468665122986,
+      "learning_rate": 6.007729054877865e-06,
+      "loss": 0.0709,
+      "step": 13910
+    },
+    {
+      "epoch": 53.12977099236641,
+      "grad_norm": 0.4115181565284729,
+      "learning_rate": 6.002329980025047e-06,
+      "loss": 0.0723,
+      "step": 13920
+    },
+    {
+      "epoch": 53.16793893129771,
+      "grad_norm": 0.5874790549278259,
+      "learning_rate": 5.99692968724679e-06,
+      "loss": 0.0708,
+      "step": 13930
+    },
+    {
+      "epoch": 53.20610687022901,
+      "grad_norm": 0.31998172402381897,
+      "learning_rate": 5.991528183104959e-06,
+      "loss": 0.0739,
+      "step": 13940
+    },
+    {
+      "epoch": 53.2442748091603,
+      "grad_norm": 0.2828161120414734,
+      "learning_rate": 5.98612547416289e-06,
+      "loss": 0.0739,
+      "step": 13950
+    },
+    {
+      "epoch": 53.282442748091604,
+      "grad_norm": 0.2661030888557434,
+      "learning_rate": 5.9807215669853855e-06,
+      "loss": 0.1,
+      "step": 13960
+    },
+    {
+      "epoch": 53.3206106870229,
+      "grad_norm": 1.1698861122131348,
+      "learning_rate": 5.9753164681387e-06,
+      "loss": 0.0806,
+      "step": 13970
+    },
+    {
+      "epoch": 53.3587786259542,
+      "grad_norm": 0.413404256105423,
+      "learning_rate": 5.969910184190539e-06,
+      "loss": 0.0695,
+      "step": 13980
+    },
+    {
+      "epoch": 53.396946564885496,
+      "grad_norm": 0.5563871264457703,
+      "learning_rate": 5.9645027217100485e-06,
+      "loss": 0.0831,
+      "step": 13990
+    },
+    {
+      "epoch": 53.43511450381679,
+      "grad_norm": 0.5594649314880371,
+      "learning_rate": 5.9590940872678035e-06,
+      "loss": 0.0806,
+      "step": 14000
+    },
+    {
+      "epoch": 53.47328244274809,
+      "grad_norm": 0.2644912898540497,
+      "learning_rate": 5.953684287435807e-06,
+      "loss": 0.0744,
+      "step": 14010
+    },
+    {
+      "epoch": 53.51145038167939,
+      "grad_norm": 0.31306540966033936,
+      "learning_rate": 5.948273328787474e-06,
+      "loss": 0.0806,
+      "step": 14020
+    },
+    {
+      "epoch": 53.54961832061069,
+      "grad_norm": 0.298922598361969,
+      "learning_rate": 5.942861217897631e-06,
+      "loss": 0.0722,
+      "step": 14030
+    },
+    {
+      "epoch": 53.587786259541986,
+      "grad_norm": 0.38911426067352295,
+      "learning_rate": 5.937447961342501e-06,
+      "loss": 0.0811,
+      "step": 14040
+    },
+    {
+      "epoch": 53.62595419847328,
+      "grad_norm": 0.44509273767471313,
+      "learning_rate": 5.932033565699704e-06,
+      "loss": 0.0824,
+      "step": 14050
+    },
+    {
+      "epoch": 53.66412213740458,
+      "grad_norm": 0.6681466102600098,
+      "learning_rate": 5.926618037548237e-06,
+      "loss": 0.0815,
+      "step": 14060
+    },
+    {
+      "epoch": 53.70229007633588,
+      "grad_norm": 0.5035837888717651,
+      "learning_rate": 5.921201383468483e-06,
+      "loss": 0.0693,
+      "step": 14070
+    },
+    {
+      "epoch": 53.74045801526717,
+      "grad_norm": 0.4998972713947296,
+      "learning_rate": 5.915783610042183e-06,
+      "loss": 0.0758,
+      "step": 14080
+    },
+    {
+      "epoch": 53.778625954198475,
+      "grad_norm": 0.36603638529777527,
+      "learning_rate": 5.910364723852444e-06,
+      "loss": 0.0701,
+      "step": 14090
+    },
+    {
+      "epoch": 53.81679389312977,
+      "grad_norm": 0.4405144453048706,
+      "learning_rate": 5.904944731483724e-06,
+      "loss": 0.073,
+      "step": 14100
+    },
+    {
+      "epoch": 53.85496183206107,
+      "grad_norm": 0.4353579878807068,
+      "learning_rate": 5.899523639521825e-06,
+      "loss": 0.0771,
+      "step": 14110
+    },
+    {
+      "epoch": 53.89312977099237,
+      "grad_norm": 0.2582431435585022,
+      "learning_rate": 5.894101454553883e-06,
+      "loss": 0.0665,
+      "step": 14120
+    },
+    {
+      "epoch": 53.93129770992366,
+      "grad_norm": 0.2779502868652344,
+      "learning_rate": 5.888678183168368e-06,
+      "loss": 0.0805,
+      "step": 14130
+    },
+    {
+      "epoch": 53.969465648854964,
+      "grad_norm": 0.35140860080718994,
+      "learning_rate": 5.883253831955061e-06,
+      "loss": 0.0758,
+      "step": 14140
+    },
+    {
+      "epoch": 54.00763358778626,
+      "grad_norm": 0.7238326668739319,
+      "learning_rate": 5.877828407505063e-06,
+      "loss": 0.0818,
+      "step": 14150
+    },
+    {
+      "epoch": 54.045801526717554,
+      "grad_norm": 0.43056339025497437,
+      "learning_rate": 5.872401916410777e-06,
+      "loss": 0.0942,
+      "step": 14160
+    },
+    {
+      "epoch": 54.083969465648856,
+      "grad_norm": 0.5613008141517639,
+      "learning_rate": 5.866974365265901e-06,
+      "loss": 0.0642,
+      "step": 14170
+    },
+    {
+      "epoch": 54.12213740458015,
+      "grad_norm": 0.3026123046875,
+      "learning_rate": 5.86154576066542e-06,
+      "loss": 0.0776,
+      "step": 14180
+    },
+    {
+      "epoch": 54.16030534351145,
+      "grad_norm": 0.3152710199356079,
+      "learning_rate": 5.856116109205602e-06,
+      "loss": 0.0674,
+      "step": 14190
+    },
+    {
+      "epoch": 54.19847328244275,
+      "grad_norm": 0.8080897331237793,
+      "learning_rate": 5.850685417483983e-06,
+      "loss": 0.0859,
+      "step": 14200
+    },
+    {
+      "epoch": 54.23664122137404,
+      "grad_norm": 0.3381716310977936,
+      "learning_rate": 5.845253692099369e-06,
+      "loss": 0.0822,
+      "step": 14210
+    },
+    {
+      "epoch": 54.274809160305345,
+      "grad_norm": 0.25009527802467346,
+      "learning_rate": 5.839820939651817e-06,
+      "loss": 0.0705,
+      "step": 14220
+    },
+    {
+      "epoch": 54.31297709923664,
+      "grad_norm": 0.24691715836524963,
+      "learning_rate": 5.8343871667426326e-06,
+      "loss": 0.0736,
+      "step": 14230
+    },
+    {
+      "epoch": 54.35114503816794,
+      "grad_norm": 0.32508522272109985,
+      "learning_rate": 5.828952379974364e-06,
+      "loss": 0.0692,
+      "step": 14240
+    },
+    {
+      "epoch": 54.38931297709924,
+      "grad_norm": 0.3096218407154083,
+      "learning_rate": 5.823516585950787e-06,
+      "loss": 0.0691,
+      "step": 14250
+    },
+    {
+      "epoch": 54.42748091603053,
+      "grad_norm": 0.3918653726577759,
+      "learning_rate": 5.818079791276907e-06,
+      "loss": 0.0761,
+      "step": 14260
+    },
+    {
+      "epoch": 54.465648854961835,
+      "grad_norm": 0.4393802881240845,
+      "learning_rate": 5.8126420025589415e-06,
+      "loss": 0.0816,
+      "step": 14270
+    },
+    {
+      "epoch": 54.50381679389313,
+      "grad_norm": 0.3444501459598541,
+      "learning_rate": 5.807203226404313e-06,
+      "loss": 0.0983,
+      "step": 14280
+    },
+    {
+      "epoch": 54.541984732824424,
+      "grad_norm": 0.6188817024230957,
+      "learning_rate": 5.801763469421652e-06,
+      "loss": 0.0816,
+      "step": 14290
+    },
+    {
+      "epoch": 54.58015267175573,
+      "grad_norm": 0.39169570803642273,
+      "learning_rate": 5.796322738220774e-06,
+      "loss": 0.0833,
+      "step": 14300
+    },
+    {
+      "epoch": 54.61832061068702,
+      "grad_norm": 0.5085362792015076,
+      "learning_rate": 5.79088103941268e-06,
+      "loss": 0.0685,
+      "step": 14310
+    },
+    {
+      "epoch": 54.656488549618324,
+      "grad_norm": 1.19497549533844,
+      "learning_rate": 5.78543837960955e-06,
+      "loss": 0.086,
+      "step": 14320
+    },
+    {
+      "epoch": 54.69465648854962,
+      "grad_norm": 0.8479430079460144,
+      "learning_rate": 5.7799947654247244e-06,
+      "loss": 0.0815,
+      "step": 14330
+    },
+    {
+      "epoch": 54.732824427480914,
+      "grad_norm": 0.4414825439453125,
+      "learning_rate": 5.774550203472712e-06,
+      "loss": 0.076,
+      "step": 14340
+    },
+    {
+      "epoch": 54.770992366412216,
+      "grad_norm": 0.3391081690788269,
+      "learning_rate": 5.769104700369165e-06,
+      "loss": 0.0733,
+      "step": 14350
+    },
+    {
+      "epoch": 54.80916030534351,
+      "grad_norm": 0.6143276691436768,
+      "learning_rate": 5.763658262730886e-06,
+      "loss": 0.0741,
+      "step": 14360
+    },
+    {
+      "epoch": 54.847328244274806,
+      "grad_norm": 0.42588314414024353,
+      "learning_rate": 5.7582108971758095e-06,
+      "loss": 0.0688,
+      "step": 14370
+    },
+    {
+      "epoch": 54.88549618320611,
+      "grad_norm": 0.7601977586746216,
+      "learning_rate": 5.752762610323e-06,
+      "loss": 0.0862,
+      "step": 14380
+    },
+    {
+      "epoch": 54.9236641221374,
+      "grad_norm": 0.30980366468429565,
+      "learning_rate": 5.747313408792636e-06,
+      "loss": 0.072,
+      "step": 14390
+    },
+    {
+      "epoch": 54.961832061068705,
+      "grad_norm": 0.8562023639678955,
+      "learning_rate": 5.7418632992060145e-06,
+      "loss": 0.0743,
+      "step": 14400
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 0.4307748079299927,
+      "learning_rate": 5.73641228818553e-06,
+      "loss": 0.0679,
+      "step": 14410
+    },
+    {
+      "epoch": 55.038167938931295,
+      "grad_norm": 0.2995448708534241,
+      "learning_rate": 5.730960382354677e-06,
+      "loss": 0.0749,
+      "step": 14420
+    },
+    {
+      "epoch": 55.0763358778626,
+      "grad_norm": 0.5709684491157532,
+      "learning_rate": 5.725507588338035e-06,
+      "loss": 0.0749,
+      "step": 14430
+    },
+    {
+      "epoch": 55.11450381679389,
+      "grad_norm": 0.5325103402137756,
+      "learning_rate": 5.720053912761261e-06,
+      "loss": 0.0696,
+      "step": 14440
+    },
+    {
+      "epoch": 55.152671755725194,
+      "grad_norm": 0.40759608149528503,
+      "learning_rate": 5.714599362251088e-06,
+      "loss": 0.0725,
+      "step": 14450
+    },
+    {
+      "epoch": 55.19083969465649,
+      "grad_norm": 0.3020817041397095,
+      "learning_rate": 5.709143943435307e-06,
+      "loss": 0.0776,
+      "step": 14460
+    },
+    {
+      "epoch": 55.229007633587784,
+      "grad_norm": 0.4237860441207886,
+      "learning_rate": 5.703687662942765e-06,
+      "loss": 0.0763,
+      "step": 14470
+    },
+    {
+      "epoch": 55.267175572519086,
+      "grad_norm": 0.8698391914367676,
+      "learning_rate": 5.6982305274033616e-06,
+      "loss": 0.0735,
+      "step": 14480
+    },
+    {
+      "epoch": 55.30534351145038,
+      "grad_norm": 0.5781353116035461,
+      "learning_rate": 5.692772543448027e-06,
+      "loss": 0.0779,
+      "step": 14490
+    },
+    {
+      "epoch": 55.343511450381676,
+      "grad_norm": 0.27626991271972656,
+      "learning_rate": 5.687313717708728e-06,
+      "loss": 0.0703,
+      "step": 14500
+    },
+    {
+      "epoch": 55.38167938931298,
+      "grad_norm": 0.28378692269325256,
+      "learning_rate": 5.681854056818453e-06,
+      "loss": 0.0795,
+      "step": 14510
+    },
+    {
+      "epoch": 55.41984732824427,
+      "grad_norm": 0.5798959732055664,
+      "learning_rate": 5.6763935674112045e-06,
+      "loss": 0.0706,
+      "step": 14520
+    },
+    {
+      "epoch": 55.458015267175576,
+      "grad_norm": 0.288756400346756,
+      "learning_rate": 5.670932256121992e-06,
+      "loss": 0.0691,
+      "step": 14530
+    },
+    {
+      "epoch": 55.49618320610687,
+      "grad_norm": 0.25545430183410645,
+      "learning_rate": 5.665470129586822e-06,
+      "loss": 0.074,
+      "step": 14540
+    },
+    {
+      "epoch": 55.534351145038165,
+      "grad_norm": 0.3490813374519348,
+      "learning_rate": 5.660007194442697e-06,
+      "loss": 0.078,
+      "step": 14550
+    },
+    {
+      "epoch": 55.57251908396947,
+      "grad_norm": 0.31008848547935486,
+      "learning_rate": 5.6545434573275945e-06,
+      "loss": 0.0772,
+      "step": 14560
+    },
+    {
+      "epoch": 55.61068702290076,
+      "grad_norm": 0.3187563717365265,
+      "learning_rate": 5.649078924880472e-06,
+      "loss": 0.0724,
+      "step": 14570
+    },
+    {
+      "epoch": 55.64885496183206,
+      "grad_norm": 0.4331674575805664,
+      "learning_rate": 5.643613603741253e-06,
+      "loss": 0.0677,
+      "step": 14580
+    },
+    {
+      "epoch": 55.68702290076336,
+      "grad_norm": 0.354324609041214,
+      "learning_rate": 5.6381475005508156e-06,
+      "loss": 0.0759,
+      "step": 14590
+    },
+    {
+      "epoch": 55.725190839694655,
+      "grad_norm": 0.4229590594768524,
+      "learning_rate": 5.632680621950992e-06,
+      "loss": 0.0783,
+      "step": 14600
+    },
+    {
+      "epoch": 55.76335877862596,
+      "grad_norm": 0.5442618727684021,
+      "learning_rate": 5.627212974584555e-06,
+      "loss": 0.0709,
+      "step": 14610
+    },
+    {
+      "epoch": 55.80152671755725,
+      "grad_norm": 0.31692686676979065,
+      "learning_rate": 5.62174456509521e-06,
+      "loss": 0.0718,
+      "step": 14620
+    },
+    {
+      "epoch": 55.83969465648855,
+      "grad_norm": 0.3284524977207184,
+      "learning_rate": 5.616275400127594e-06,
+      "loss": 0.0688,
+      "step": 14630
+    },
+    {
+      "epoch": 55.87786259541985,
+      "grad_norm": 0.3089140057563782,
+      "learning_rate": 5.610805486327254e-06,
+      "loss": 0.0686,
+      "step": 14640
+    },
+    {
+      "epoch": 55.916030534351144,
+      "grad_norm": 0.4001990556716919,
+      "learning_rate": 5.6053348303406545e-06,
+      "loss": 0.0766,
+      "step": 14650
+    },
+    {
+      "epoch": 55.954198473282446,
+      "grad_norm": 0.5436351299285889,
+      "learning_rate": 5.599863438815157e-06,
+      "loss": 0.0682,
+      "step": 14660
+    },
+    {
+      "epoch": 55.99236641221374,
+      "grad_norm": 0.47523510456085205,
+      "learning_rate": 5.594391318399017e-06,
+      "loss": 0.0724,
+      "step": 14670
+    },
+    {
+      "epoch": 56.030534351145036,
+      "grad_norm": 0.2652002274990082,
+      "learning_rate": 5.588918475741378e-06,
+      "loss": 0.0714,
+      "step": 14680
+    },
+    {
+      "epoch": 56.06870229007634,
+      "grad_norm": 0.2382209599018097,
+      "learning_rate": 5.583444917492259e-06,
+      "loss": 0.0711,
+      "step": 14690
+    },
+    {
+      "epoch": 56.10687022900763,
+      "grad_norm": 0.4424055516719818,
+      "learning_rate": 5.57797065030255e-06,
+      "loss": 0.0698,
+      "step": 14700
+    },
+    {
+      "epoch": 56.14503816793893,
+      "grad_norm": 0.25786203145980835,
+      "learning_rate": 5.572495680824001e-06,
+      "loss": 0.0745,
+      "step": 14710
+    },
+    {
+      "epoch": 56.18320610687023,
+      "grad_norm": 0.7077968716621399,
+      "learning_rate": 5.5670200157092195e-06,
+      "loss": 0.0747,
+      "step": 14720
+    },
+    {
+      "epoch": 56.221374045801525,
+      "grad_norm": 0.32011500000953674,
+      "learning_rate": 5.561543661611649e-06,
+      "loss": 0.0728,
+      "step": 14730
+    },
+    {
+      "epoch": 56.25954198473283,
+      "grad_norm": 0.6672192811965942,
+      "learning_rate": 5.556066625185584e-06,
+      "loss": 0.0744,
+      "step": 14740
+    },
+    {
+      "epoch": 56.29770992366412,
+      "grad_norm": 1.5504484176635742,
+      "learning_rate": 5.550588913086131e-06,
+      "loss": 0.0781,
+      "step": 14750
+    },
+    {
+      "epoch": 56.33587786259542,
+      "grad_norm": 0.32770517468452454,
+      "learning_rate": 5.545110531969234e-06,
+      "loss": 0.0737,
+      "step": 14760
+    },
+    {
+      "epoch": 56.37404580152672,
+      "grad_norm": 0.42965736985206604,
+      "learning_rate": 5.539631488491641e-06,
+      "loss": 0.0736,
+      "step": 14770
+    },
+    {
+      "epoch": 56.412213740458014,
+      "grad_norm": 0.277537077665329,
+      "learning_rate": 5.534151789310904e-06,
+      "loss": 0.0774,
+      "step": 14780
+    },
+    {
+      "epoch": 56.45038167938931,
+      "grad_norm": 0.5046352744102478,
+      "learning_rate": 5.528671441085376e-06,
+      "loss": 0.0713,
+      "step": 14790
+    },
+    {
+      "epoch": 56.48854961832061,
+      "grad_norm": 1.0368688106536865,
+      "learning_rate": 5.523190450474198e-06,
+      "loss": 0.0766,
+      "step": 14800
+    },
+    {
+      "epoch": 56.52671755725191,
+      "grad_norm": 0.36909282207489014,
+      "learning_rate": 5.517708824137287e-06,
+      "loss": 0.0755,
+      "step": 14810
+    },
+    {
+      "epoch": 56.56488549618321,
+      "grad_norm": 0.5959531664848328,
+      "learning_rate": 5.512226568735338e-06,
+      "loss": 0.0817,
+      "step": 14820
+    },
+    {
+      "epoch": 56.603053435114504,
+      "grad_norm": 0.2804238796234131,
+      "learning_rate": 5.506743690929809e-06,
+      "loss": 0.0779,
+      "step": 14830
+    },
+    {
+      "epoch": 56.6412213740458,
+      "grad_norm": 0.6354565620422363,
+      "learning_rate": 5.501260197382913e-06,
+      "loss": 0.0833,
+      "step": 14840
+    },
+    {
+      "epoch": 56.6793893129771,
+      "grad_norm": 0.6478805541992188,
+      "learning_rate": 5.49577609475761e-06,
+      "loss": 0.0808,
+      "step": 14850
+    },
+    {
+      "epoch": 56.717557251908396,
+      "grad_norm": 0.39894139766693115,
+      "learning_rate": 5.4902913897176035e-06,
+      "loss": 0.0729,
+      "step": 14860
+    },
+    {
+      "epoch": 56.7557251908397,
+      "grad_norm": 0.346055805683136,
+      "learning_rate": 5.484806088927329e-06,
+      "loss": 0.0772,
+      "step": 14870
+    },
+    {
+      "epoch": 56.79389312977099,
+      "grad_norm": 0.3596509099006653,
+      "learning_rate": 5.479320199051942e-06,
+      "loss": 0.0701,
+      "step": 14880
+    },
+    {
+      "epoch": 56.83206106870229,
+      "grad_norm": 0.46233484148979187,
+      "learning_rate": 5.473833726757314e-06,
+      "loss": 0.0734,
+      "step": 14890
+    },
+    {
+      "epoch": 56.87022900763359,
+      "grad_norm": 0.7055529952049255,
+      "learning_rate": 5.46834667871003e-06,
+      "loss": 0.0718,
+      "step": 14900
+    },
+    {
+      "epoch": 56.908396946564885,
+      "grad_norm": 0.9955065846443176,
+      "learning_rate": 5.462859061577369e-06,
+      "loss": 0.0794,
+      "step": 14910
+    },
+    {
+      "epoch": 56.94656488549618,
+      "grad_norm": 0.34203898906707764,
+      "learning_rate": 5.457370882027303e-06,
+      "loss": 0.0739,
+      "step": 14920
+    },
+    {
+      "epoch": 56.98473282442748,
+      "grad_norm": 0.21919408440589905,
+      "learning_rate": 5.451882146728489e-06,
+      "loss": 0.077,
+      "step": 14930
+    },
+    {
+      "epoch": 57.02290076335878,
+      "grad_norm": 0.4146571457386017,
+      "learning_rate": 5.446392862350255e-06,
+      "loss": 0.0791,
+      "step": 14940
+    },
+    {
+      "epoch": 57.06106870229008,
+      "grad_norm": 0.4011072516441345,
+      "learning_rate": 5.4409030355626035e-06,
+      "loss": 0.0835,
+      "step": 14950
+    },
+    {
+      "epoch": 57.099236641221374,
+      "grad_norm": 0.2836225628852844,
+      "learning_rate": 5.435412673036188e-06,
+      "loss": 0.0835,
+      "step": 14960
+    },
+    {
+      "epoch": 57.13740458015267,
+      "grad_norm": 0.41876521706581116,
+      "learning_rate": 5.429921781442318e-06,
+      "loss": 0.0754,
+      "step": 14970
+    },
+    {
+      "epoch": 57.17557251908397,
+      "grad_norm": 0.24527506530284882,
+      "learning_rate": 5.424430367452946e-06,
+      "loss": 0.0717,
+      "step": 14980
+    },
+    {
+      "epoch": 57.213740458015266,
+      "grad_norm": 0.5479999780654907,
+      "learning_rate": 5.418938437740655e-06,
+      "loss": 0.073,
+      "step": 14990
+    },
+    {
+      "epoch": 57.25190839694657,
+      "grad_norm": 1.0041180849075317,
+      "learning_rate": 5.413445998978658e-06,
+      "loss": 0.0765,
+      "step": 15000
+    },
+    {
+      "epoch": 57.29007633587786,
+      "grad_norm": 0.34665971994400024,
+      "learning_rate": 5.4079530578407895e-06,
+      "loss": 0.0738,
+      "step": 15010
+    },
+    {
+      "epoch": 57.32824427480916,
+      "grad_norm": 0.3036256730556488,
+      "learning_rate": 5.402459621001486e-06,
+      "loss": 0.0675,
+      "step": 15020
+    },
+    {
+      "epoch": 57.36641221374046,
+      "grad_norm": 0.36312854290008545,
+      "learning_rate": 5.396965695135794e-06,
+      "loss": 0.08,
+      "step": 15030
+    },
+    {
+      "epoch": 57.404580152671755,
+      "grad_norm": 0.2859637439250946,
+      "learning_rate": 5.391471286919351e-06,
+      "loss": 0.0661,
+      "step": 15040
+    },
+    {
+      "epoch": 57.44274809160305,
+      "grad_norm": 0.4589112401008606,
+      "learning_rate": 5.385976403028381e-06,
+      "loss": 0.0697,
+      "step": 15050
+    },
+    {
+      "epoch": 57.48091603053435,
+      "grad_norm": 0.3610994517803192,
+      "learning_rate": 5.380481050139688e-06,
+      "loss": 0.0742,
+      "step": 15060
+    },
+    {
+      "epoch": 57.51908396946565,
+      "grad_norm": 0.37091097235679626,
+      "learning_rate": 5.37498523493064e-06,
+      "loss": 0.0716,
+      "step": 15070
+    },
+    {
+      "epoch": 57.55725190839695,
+      "grad_norm": 0.30698439478874207,
+      "learning_rate": 5.369488964079172e-06,
+      "loss": 0.0838,
+      "step": 15080
+    },
+    {
+      "epoch": 57.595419847328245,
+      "grad_norm": 0.31963449716567993,
+      "learning_rate": 5.363992244263774e-06,
+      "loss": 0.0736,
+      "step": 15090
+    },
+    {
+      "epoch": 57.63358778625954,
+      "grad_norm": 0.3956395089626312,
+      "learning_rate": 5.358495082163476e-06,
+      "loss": 0.0678,
+      "step": 15100
+    },
+    {
+      "epoch": 57.67175572519084,
+      "grad_norm": 0.4276552200317383,
+      "learning_rate": 5.35299748445785e-06,
+      "loss": 0.0768,
+      "step": 15110
+    },
+    {
+      "epoch": 57.70992366412214,
+      "grad_norm": 0.3108503818511963,
+      "learning_rate": 5.347499457826995e-06,
+      "loss": 0.0803,
+      "step": 15120
+    },
+    {
+      "epoch": 57.74809160305343,
+      "grad_norm": 0.35856226086616516,
+      "learning_rate": 5.342001008951531e-06,
+      "loss": 0.0794,
+      "step": 15130
+    },
+    {
+      "epoch": 57.786259541984734,
+      "grad_norm": 0.4197726547718048,
+      "learning_rate": 5.336502144512592e-06,
+      "loss": 0.0731,
+      "step": 15140
+    },
+    {
+      "epoch": 57.82442748091603,
+      "grad_norm": 0.3682864308357239,
+      "learning_rate": 5.331002871191817e-06,
+      "loss": 0.0723,
+      "step": 15150
+    },
+    {
+      "epoch": 57.86259541984733,
+      "grad_norm": 0.7807731628417969,
+      "learning_rate": 5.325503195671345e-06,
+      "loss": 0.0747,
+      "step": 15160
+    },
+    {
+      "epoch": 57.900763358778626,
+      "grad_norm": 0.3009580373764038,
+      "learning_rate": 5.320003124633795e-06,
+      "loss": 0.0732,
+      "step": 15170
+    },
+    {
+      "epoch": 57.93893129770992,
+      "grad_norm": 0.5685849189758301,
+      "learning_rate": 5.314502664762275e-06,
+      "loss": 0.0753,
+      "step": 15180
+    },
+    {
+      "epoch": 57.97709923664122,
+      "grad_norm": 0.32221153378486633,
+      "learning_rate": 5.3090018227403605e-06,
+      "loss": 0.0841,
+      "step": 15190
+    },
+    {
+      "epoch": 58.01526717557252,
+      "grad_norm": 0.4299134314060211,
+      "learning_rate": 5.303500605252095e-06,
+      "loss": 0.076,
+      "step": 15200
+    },
+    {
+      "epoch": 58.05343511450382,
+      "grad_norm": 0.604964017868042,
+      "learning_rate": 5.297999018981977e-06,
+      "loss": 0.079,
+      "step": 15210
+    },
+    {
+      "epoch": 58.091603053435115,
+      "grad_norm": 0.3654126226902008,
+      "learning_rate": 5.2924970706149505e-06,
+      "loss": 0.071,
+      "step": 15220
+    },
+    {
+      "epoch": 58.12977099236641,
+      "grad_norm": 0.4091419279575348,
+      "learning_rate": 5.286994766836402e-06,
+      "loss": 0.0785,
+      "step": 15230
+    },
+    {
+      "epoch": 58.16793893129771,
+      "grad_norm": 0.45684048533439636,
+      "learning_rate": 5.2814921143321506e-06,
+      "loss": 0.0678,
+      "step": 15240
+    },
+    {
+      "epoch": 58.20610687022901,
+      "grad_norm": 0.32718759775161743,
+      "learning_rate": 5.275989119788436e-06,
+      "loss": 0.0739,
+      "step": 15250
+    },
+    {
+      "epoch": 58.2442748091603,
+      "grad_norm": 0.36803168058395386,
+      "learning_rate": 5.2704857898919195e-06,
+      "loss": 0.0699,
+      "step": 15260
+    },
+    {
+      "epoch": 58.282442748091604,
+      "grad_norm": 0.38092291355133057,
+      "learning_rate": 5.264982131329661e-06,
+      "loss": 0.0712,
+      "step": 15270
+    },
+    {
+      "epoch": 58.3206106870229,
+      "grad_norm": 0.3395163118839264,
+      "learning_rate": 5.259478150789128e-06,
+      "loss": 0.0717,
+      "step": 15280
+    },
+    {
+      "epoch": 58.3587786259542,
+      "grad_norm": 0.27850109338760376,
+      "learning_rate": 5.253973854958173e-06,
+      "loss": 0.0668,
+      "step": 15290
+    },
+    {
+      "epoch": 58.396946564885496,
+      "grad_norm": 0.28005483746528625,
+      "learning_rate": 5.2484692505250375e-06,
+      "loss": 0.0762,
+      "step": 15300
+    },
+    {
+      "epoch": 58.43511450381679,
+      "grad_norm": 0.3931730091571808,
+      "learning_rate": 5.2429643441783325e-06,
+      "loss": 0.0768,
+      "step": 15310
+    },
+    {
+      "epoch": 58.47328244274809,
+      "grad_norm": 0.33013564348220825,
+      "learning_rate": 5.237459142607041e-06,
+      "loss": 0.0764,
+      "step": 15320
+    },
+    {
+      "epoch": 58.51145038167939,
+      "grad_norm": 0.3280738592147827,
+      "learning_rate": 5.2319536525004974e-06,
+      "loss": 0.0749,
+      "step": 15330
+    },
+    {
+      "epoch": 58.54961832061069,
+      "grad_norm": 0.3545534610748291,
+      "learning_rate": 5.226447880548398e-06,
+      "loss": 0.0801,
+      "step": 15340
+    },
+    {
+      "epoch": 58.587786259541986,
+      "grad_norm": 0.4157971739768982,
+      "learning_rate": 5.220941833440768e-06,
+      "loss": 0.0791,
+      "step": 15350
+    },
+    {
+      "epoch": 58.62595419847328,
+      "grad_norm": 0.4566729962825775,
+      "learning_rate": 5.215435517867978e-06,
+      "loss": 0.0721,
+      "step": 15360
+    },
+    {
+      "epoch": 58.66412213740458,
+      "grad_norm": 0.5043272972106934,
+      "learning_rate": 5.209928940520719e-06,
+      "loss": 0.0723,
+      "step": 15370
+    },
+    {
+      "epoch": 58.70229007633588,
+      "grad_norm": 0.5831508636474609,
+      "learning_rate": 5.204422108090004e-06,
+      "loss": 0.0728,
+      "step": 15380
+    },
+    {
+      "epoch": 58.74045801526717,
+      "grad_norm": 0.30468279123306274,
+      "learning_rate": 5.19891502726715e-06,
+      "loss": 0.0735,
+      "step": 15390
+    },
+    {
+      "epoch": 58.778625954198475,
+      "grad_norm": 0.6569890975952148,
+      "learning_rate": 5.193407704743782e-06,
+      "loss": 0.0751,
+      "step": 15400
+    },
+    {
+      "epoch": 58.81679389312977,
+      "grad_norm": 0.4217868745326996,
+      "learning_rate": 5.1879001472118155e-06,
+      "loss": 0.0742,
+      "step": 15410
+    },
+    {
+      "epoch": 58.85496183206107,
+      "grad_norm": 0.3305950164794922,
+      "learning_rate": 5.182392361363453e-06,
+      "loss": 0.08,
+      "step": 15420
+    },
+    {
+      "epoch": 58.89312977099237,
+      "grad_norm": 0.2556851804256439,
+      "learning_rate": 5.176884353891172e-06,
+      "loss": 0.0835,
+      "step": 15430
+    },
+    {
+      "epoch": 58.93129770992366,
+      "grad_norm": 0.5155206322669983,
+      "learning_rate": 5.171376131487722e-06,
+      "loss": 0.0758,
+      "step": 15440
+    },
+    {
+      "epoch": 58.969465648854964,
+      "grad_norm": 0.3321559429168701,
+      "learning_rate": 5.165867700846113e-06,
+      "loss": 0.071,
+      "step": 15450
+    },
+    {
+      "epoch": 59.00763358778626,
+      "grad_norm": 0.49137699604034424,
+      "learning_rate": 5.1603590686596065e-06,
+      "loss": 0.08,
+      "step": 15460
+    },
+    {
+      "epoch": 59.045801526717554,
+      "grad_norm": 0.4781742990016937,
+      "learning_rate": 5.154850241621712e-06,
+      "loss": 0.0768,
+      "step": 15470
+    },
+    {
+      "epoch": 59.083969465648856,
+      "grad_norm": 0.5240231156349182,
+      "learning_rate": 5.149341226426172e-06,
+      "loss": 0.0742,
+      "step": 15480
+    },
+    {
+      "epoch": 59.12213740458015,
+      "grad_norm": 0.5776994824409485,
+      "learning_rate": 5.143832029766959e-06,
+      "loss": 0.0685,
+      "step": 15490
+    },
+    {
+      "epoch": 59.16030534351145,
+      "grad_norm": 0.23645877838134766,
+      "learning_rate": 5.138322658338269e-06,
+      "loss": 0.0668,
+      "step": 15500
+    },
+    {
+      "epoch": 59.19847328244275,
+      "grad_norm": 0.3662410378456116,
+      "learning_rate": 5.132813118834504e-06,
+      "loss": 0.0765,
+      "step": 15510
+    },
+    {
+      "epoch": 59.23664122137404,
+      "grad_norm": 0.8370821475982666,
+      "learning_rate": 5.127303417950278e-06,
+      "loss": 0.0758,
+      "step": 15520
+    },
+    {
+      "epoch": 59.274809160305345,
+      "grad_norm": 0.29751116037368774,
+      "learning_rate": 5.121793562380395e-06,
+      "loss": 0.0799,
+      "step": 15530
+    },
+    {
+      "epoch": 59.31297709923664,
+      "grad_norm": 0.3275250792503357,
+      "learning_rate": 5.116283558819848e-06,
+      "loss": 0.0758,
+      "step": 15540
+    },
+    {
+      "epoch": 59.35114503816794,
+      "grad_norm": 0.36758387088775635,
+      "learning_rate": 5.110773413963813e-06,
+      "loss": 0.071,
+      "step": 15550
+    },
+    {
+      "epoch": 59.38931297709924,
+      "grad_norm": 0.4558661878108978,
+      "learning_rate": 5.1052631345076365e-06,
+      "loss": 0.0754,
+      "step": 15560
+    },
+    {
+      "epoch": 59.42748091603053,
+      "grad_norm": 0.7922174334526062,
+      "learning_rate": 5.099752727146824e-06,
+      "loss": 0.0802,
+      "step": 15570
+    },
+    {
+      "epoch": 59.465648854961835,
+      "grad_norm": 0.5909420847892761,
+      "learning_rate": 5.0942421985770415e-06,
+      "loss": 0.0734,
+      "step": 15580
+    },
+    {
+      "epoch": 59.50381679389313,
+      "grad_norm": 0.3770803213119507,
+      "learning_rate": 5.088731555494102e-06,
+      "loss": 0.0726,
+      "step": 15590
+    },
+    {
+      "epoch": 59.541984732824424,
+      "grad_norm": 0.4959609806537628,
+      "learning_rate": 5.083220804593956e-06,
+      "loss": 0.0751,
+      "step": 15600
+    },
+    {
+      "epoch": 59.58015267175573,
+      "grad_norm": 0.8414520621299744,
+      "learning_rate": 5.077709952572685e-06,
+      "loss": 0.0779,
+      "step": 15610
+    },
+    {
+      "epoch": 59.61832061068702,
+      "grad_norm": 0.43522128462791443,
+      "learning_rate": 5.072199006126494e-06,
+      "loss": 0.072,
+      "step": 15620
+    },
+    {
+      "epoch": 59.656488549618324,
+      "grad_norm": 0.3183342218399048,
+      "learning_rate": 5.066687971951702e-06,
+      "loss": 0.0743,
+      "step": 15630
+    },
+    {
+      "epoch": 59.69465648854962,
+      "grad_norm": 0.25728127360343933,
+      "learning_rate": 5.0611768567447375e-06,
+      "loss": 0.071,
+      "step": 15640
+    },
+    {
+      "epoch": 59.732824427480914,
+      "grad_norm": 0.24269990622997284,
+      "learning_rate": 5.055665667202121e-06,
+      "loss": 0.0682,
+      "step": 15650
+    },
+    {
+      "epoch": 59.770992366412216,
+      "grad_norm": 0.3147321343421936,
+      "learning_rate": 5.050154410020473e-06,
+      "loss": 0.072,
+      "step": 15660
+    },
+    {
+      "epoch": 59.80916030534351,
+      "grad_norm": 0.259830504655838,
+      "learning_rate": 5.044643091896485e-06,
+      "loss": 0.0728,
+      "step": 15670
+    },
+    {
+      "epoch": 59.847328244274806,
+      "grad_norm": 0.31697461009025574,
+      "learning_rate": 5.039131719526932e-06,
+      "loss": 0.0709,
+      "step": 15680
+    },
+    {
+      "epoch": 59.88549618320611,
+      "grad_norm": 0.6029514670372009,
+      "learning_rate": 5.03362029960865e-06,
+      "loss": 0.0728,
+      "step": 15690
+    },
+    {
+      "epoch": 59.9236641221374,
+      "grad_norm": 0.33722543716430664,
+      "learning_rate": 5.028108838838533e-06,
+      "loss": 0.0742,
+      "step": 15700
+    },
+    {
+      "epoch": 59.961832061068705,
+      "grad_norm": 0.6031141877174377,
+      "learning_rate": 5.022597343913528e-06,
+      "loss": 0.073,
+      "step": 15710
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 0.7514012455940247,
+      "learning_rate": 5.017085821530617e-06,
+      "loss": 0.0768,
+      "step": 15720
+    },
+    {
+      "epoch": 60.038167938931295,
+      "grad_norm": 0.3948346972465515,
+      "learning_rate": 5.011574278386823e-06,
+      "loss": 0.0766,
+      "step": 15730
+    },
+    {
+      "epoch": 60.0763358778626,
+      "grad_norm": 0.3414020538330078,
+      "learning_rate": 5.006062721179189e-06,
+      "loss": 0.0766,
+      "step": 15740
+    },
+    {
+      "epoch": 60.11450381679389,
+      "grad_norm": 0.2455693781375885,
+      "learning_rate": 5.000551156604777e-06,
+      "loss": 0.0718,
+      "step": 15750
+    },
+    {
+      "epoch": 60.152671755725194,
+      "grad_norm": 0.7095368504524231,
+      "learning_rate": 4.99503959136066e-06,
+      "loss": 0.07,
+      "step": 15760
+    },
+    {
+      "epoch": 60.19083969465649,
+      "grad_norm": 0.37839195132255554,
+      "learning_rate": 4.9895280321439036e-06,
+      "loss": 0.0733,
+      "step": 15770
+    },
+    {
+      "epoch": 60.229007633587784,
+      "grad_norm": 0.3145567774772644,
+      "learning_rate": 4.984016485651578e-06,
+      "loss": 0.0784,
+      "step": 15780
+    },
+    {
+      "epoch": 60.267175572519086,
+      "grad_norm": 0.24902135133743286,
+      "learning_rate": 4.978504958580728e-06,
+      "loss": 0.0767,
+      "step": 15790
+    },
+    {
+      "epoch": 60.30534351145038,
+      "grad_norm": 0.5785359144210815,
+      "learning_rate": 4.9729934576283815e-06,
+      "loss": 0.0916,
+      "step": 15800
+    },
+    {
+      "epoch": 60.343511450381676,
+      "grad_norm": 0.6354800462722778,
+      "learning_rate": 4.967481989491531e-06,
+      "loss": 0.0777,
+      "step": 15810
+    },
+    {
+      "epoch": 60.38167938931298,
+      "grad_norm": 0.6582098603248596,
+      "learning_rate": 4.961970560867126e-06,
+      "loss": 0.0743,
+      "step": 15820
+    },
+    {
+      "epoch": 60.41984732824427,
+      "grad_norm": 0.5716459155082703,
+      "learning_rate": 4.956459178452079e-06,
+      "loss": 0.0869,
+      "step": 15830
+    },
+    {
+      "epoch": 60.458015267175576,
+      "grad_norm": 0.3027888536453247,
+      "learning_rate": 4.950947848943235e-06,
+      "loss": 0.0897,
+      "step": 15840
+    },
+    {
+      "epoch": 60.49618320610687,
+      "grad_norm": 0.4717438519001007,
+      "learning_rate": 4.94543657903738e-06,
+      "loss": 0.0734,
+      "step": 15850
+    },
+    {
+      "epoch": 60.534351145038165,
+      "grad_norm": 0.3485773503780365,
+      "learning_rate": 4.939925375431226e-06,
+      "loss": 0.0764,
+      "step": 15860
+    },
+    {
+      "epoch": 60.57251908396947,
+      "grad_norm": 0.46527671813964844,
+      "learning_rate": 4.934414244821405e-06,
+      "loss": 0.0734,
+      "step": 15870
+    },
+    {
+      "epoch": 60.61068702290076,
+      "grad_norm": 0.2664990723133087,
+      "learning_rate": 4.928903193904461e-06,
+      "loss": 0.0738,
+      "step": 15880
+    },
+    {
+      "epoch": 60.64885496183206,
+      "grad_norm": 0.23172251880168915,
+      "learning_rate": 4.923392229376841e-06,
+      "loss": 0.0695,
+      "step": 15890
+    },
+    {
+      "epoch": 60.68702290076336,
+      "grad_norm": 0.6194725632667542,
+      "learning_rate": 4.917881357934886e-06,
+      "loss": 0.0733,
+      "step": 15900
+    },
+    {
+      "epoch": 60.725190839694655,
+      "grad_norm": 0.5128942131996155,
+      "learning_rate": 4.912370586274825e-06,
+      "loss": 0.0675,
+      "step": 15910
+    },
+    {
+      "epoch": 60.76335877862596,
+      "grad_norm": 0.3195113241672516,
+      "learning_rate": 4.906859921092763e-06,
+      "loss": 0.0694,
+      "step": 15920
+    },
+    {
+      "epoch": 60.80152671755725,
+      "grad_norm": 0.4648715555667877,
+      "learning_rate": 4.901349369084681e-06,
+      "loss": 0.0752,
+      "step": 15930
+    },
+    {
+      "epoch": 60.83969465648855,
+      "grad_norm": 0.37270602583885193,
+      "learning_rate": 4.895838936946416e-06,
+      "loss": 0.0813,
+      "step": 15940
+    },
+    {
+      "epoch": 60.87786259541985,
+      "grad_norm": 0.3094925284385681,
+      "learning_rate": 4.890328631373666e-06,
+      "loss": 0.0787,
+      "step": 15950
+    },
+    {
+      "epoch": 60.916030534351144,
+      "grad_norm": 0.8085688352584839,
+      "learning_rate": 4.88481845906197e-06,
+      "loss": 0.077,
+      "step": 15960
+    },
+    {
+      "epoch": 60.954198473282446,
+      "grad_norm": 0.45573312044143677,
+      "learning_rate": 4.879308426706707e-06,
+      "loss": 0.0752,
+      "step": 15970
+    },
+    {
+      "epoch": 60.99236641221374,
+      "grad_norm": 0.29330453276634216,
+      "learning_rate": 4.873798541003084e-06,
+      "loss": 0.0805,
+      "step": 15980
+    },
+    {
+      "epoch": 61.030534351145036,
+      "grad_norm": 0.26652470231056213,
+      "learning_rate": 4.868288808646136e-06,
+      "loss": 0.074,
+      "step": 15990
+    },
+    {
+      "epoch": 61.06870229007634,
+      "grad_norm": 0.2818049490451813,
+      "learning_rate": 4.862779236330705e-06,
+      "loss": 0.0706,
+      "step": 16000
+    },
+    {
+      "epoch": 61.10687022900763,
+      "grad_norm": 0.32146692276000977,
+      "learning_rate": 4.8572698307514395e-06,
+      "loss": 0.0704,
+      "step": 16010
+    },
+    {
+      "epoch": 61.14503816793893,
+      "grad_norm": 0.3116312325000763,
+      "learning_rate": 4.8517605986027904e-06,
+      "loss": 0.0671,
+      "step": 16020
+    },
+    {
+      "epoch": 61.18320610687023,
+      "grad_norm": 0.3970802128314972,
+      "learning_rate": 4.846251546578989e-06,
+      "loss": 0.0745,
+      "step": 16030
+    },
+    {
+      "epoch": 61.221374045801525,
+      "grad_norm": 0.35653743147850037,
+      "learning_rate": 4.8407426813740584e-06,
+      "loss": 0.0737,
+      "step": 16040
+    },
+    {
+      "epoch": 61.25954198473283,
+      "grad_norm": 0.3067050576210022,
+      "learning_rate": 4.835234009681787e-06,
+      "loss": 0.0702,
+      "step": 16050
+    },
+    {
+      "epoch": 61.29770992366412,
+      "grad_norm": 0.5067670941352844,
+      "learning_rate": 4.82972553819573e-06,
+      "loss": 0.0713,
+      "step": 16060
+    },
+    {
+      "epoch": 61.33587786259542,
+      "grad_norm": 0.31573083996772766,
+      "learning_rate": 4.824217273609199e-06,
+      "loss": 0.071,
+      "step": 16070
+    },
+    {
+      "epoch": 61.37404580152672,
+      "grad_norm": 0.29241570830345154,
+      "learning_rate": 4.818709222615255e-06,
+      "loss": 0.0712,
+      "step": 16080
+    },
+    {
+      "epoch": 61.412213740458014,
+      "grad_norm": 0.29491475224494934,
+      "learning_rate": 4.813201391906702e-06,
+      "loss": 0.0783,
+      "step": 16090
+    },
+    {
+      "epoch": 61.45038167938931,
+      "grad_norm": 0.5220329761505127,
+      "learning_rate": 4.807693788176071e-06,
+      "loss": 0.072,
+      "step": 16100
+    },
+    {
+      "epoch": 61.48854961832061,
+      "grad_norm": 0.27246901392936707,
+      "learning_rate": 4.802186418115622e-06,
+      "loss": 0.0756,
+      "step": 16110
+    },
+    {
+      "epoch": 61.52671755725191,
+      "grad_norm": 0.7175832390785217,
+      "learning_rate": 4.796679288417326e-06,
+      "loss": 0.0762,
+      "step": 16120
+    },
+    {
+      "epoch": 61.56488549618321,
+      "grad_norm": 0.4071796238422394,
+      "learning_rate": 4.791172405772866e-06,
+      "loss": 0.0787,
+      "step": 16130
+    },
+    {
+      "epoch": 61.603053435114504,
+      "grad_norm": 0.2860810458660126,
+      "learning_rate": 4.785665776873626e-06,
+      "loss": 0.0705,
+      "step": 16140
+    },
+    {
+      "epoch": 61.6412213740458,
+      "grad_norm": 0.34713274240493774,
+      "learning_rate": 4.780159408410677e-06,
+      "loss": 0.0759,
+      "step": 16150
+    },
+    {
+      "epoch": 61.6793893129771,
+      "grad_norm": 0.2859640419483185,
+      "learning_rate": 4.774653307074775e-06,
+      "loss": 0.0855,
+      "step": 16160
+    },
+    {
+      "epoch": 61.717557251908396,
+      "grad_norm": 0.30147480964660645,
+      "learning_rate": 4.7691474795563556e-06,
+      "loss": 0.0885,
+      "step": 16170
+    },
+    {
+      "epoch": 61.7557251908397,
+      "grad_norm": 0.516560435295105,
+      "learning_rate": 4.763641932545516e-06,
+      "loss": 0.0729,
+      "step": 16180
+    },
+    {
+      "epoch": 61.79389312977099,
+      "grad_norm": 0.29214778542518616,
+      "learning_rate": 4.758136672732013e-06,
+      "loss": 0.0762,
+      "step": 16190
+    },
+    {
+      "epoch": 61.83206106870229,
+      "grad_norm": 0.3169700503349304,
+      "learning_rate": 4.752631706805261e-06,
+      "loss": 0.069,
+      "step": 16200
+    },
+    {
+      "epoch": 61.87022900763359,
+      "grad_norm": 0.4579392969608307,
+      "learning_rate": 4.747127041454311e-06,
+      "loss": 0.0845,
+      "step": 16210
+    },
+    {
+      "epoch": 61.908396946564885,
+      "grad_norm": 0.4441590905189514,
+      "learning_rate": 4.741622683367849e-06,
+      "loss": 0.0681,
+      "step": 16220
+    },
+    {
+      "epoch": 61.94656488549618,
+      "grad_norm": 0.3080080449581146,
+      "learning_rate": 4.736118639234191e-06,
+      "loss": 0.0759,
+      "step": 16230
+    },
+    {
+      "epoch": 61.98473282442748,
+      "grad_norm": 0.23667652904987335,
+      "learning_rate": 4.7306149157412666e-06,
+      "loss": 0.0729,
+      "step": 16240
+    },
+    {
+      "epoch": 62.02290076335878,
+      "grad_norm": 0.3574734926223755,
+      "learning_rate": 4.7251115195766234e-06,
+      "loss": 0.0759,
+      "step": 16250
+    },
+    {
+      "epoch": 62.06106870229008,
+      "grad_norm": 1.1783217191696167,
+      "learning_rate": 4.719608457427404e-06,
+      "loss": 0.0943,
+      "step": 16260
+    },
+    {
+      "epoch": 62.099236641221374,
+      "grad_norm": 0.46935534477233887,
+      "learning_rate": 4.714105735980348e-06,
+      "loss": 0.0772,
+      "step": 16270
+    },
+    {
+      "epoch": 62.13740458015267,
+      "grad_norm": 0.32659921050071716,
+      "learning_rate": 4.708603361921779e-06,
+      "loss": 0.0761,
+      "step": 16280
+    },
+    {
+      "epoch": 62.17557251908397,
+      "grad_norm": 0.37154969573020935,
+      "learning_rate": 4.703101341937604e-06,
+      "loss": 0.0764,
+      "step": 16290
+    },
+    {
+      "epoch": 62.213740458015266,
+      "grad_norm": 0.5954548716545105,
+      "learning_rate": 4.697599682713292e-06,
+      "loss": 0.0725,
+      "step": 16300
+    },
+    {
+      "epoch": 62.25190839694657,
+      "grad_norm": 0.43994778394699097,
+      "learning_rate": 4.692098390933883e-06,
+      "loss": 0.0798,
+      "step": 16310
+    },
+    {
+      "epoch": 62.29007633587786,
+      "grad_norm": 0.42407238483428955,
+      "learning_rate": 4.686597473283962e-06,
+      "loss": 0.0674,
+      "step": 16320
+    },
+    {
+      "epoch": 62.32824427480916,
+      "grad_norm": 0.31318119168281555,
+      "learning_rate": 4.681096936447662e-06,
+      "loss": 0.0801,
+      "step": 16330
+    },
+    {
+      "epoch": 62.36641221374046,
+      "grad_norm": 0.4887228310108185,
+      "learning_rate": 4.675596787108652e-06,
+      "loss": 0.0717,
+      "step": 16340
+    },
+    {
+      "epoch": 62.404580152671755,
+      "grad_norm": 0.2779092788696289,
+      "learning_rate": 4.670097031950138e-06,
+      "loss": 0.0712,
+      "step": 16350
+    },
+    {
+      "epoch": 62.44274809160305,
+      "grad_norm": 0.839206874370575,
+      "learning_rate": 4.664597677654839e-06,
+      "loss": 0.0699,
+      "step": 16360
+    },
+    {
+      "epoch": 62.48091603053435,
+      "grad_norm": 0.25964847207069397,
+      "learning_rate": 4.6590987309049855e-06,
+      "loss": 0.0636,
+      "step": 16370
+    },
+    {
+      "epoch": 62.51908396946565,
+      "grad_norm": 0.34289413690567017,
+      "learning_rate": 4.65360019838232e-06,
+      "loss": 0.0779,
+      "step": 16380
+    },
+    {
+      "epoch": 62.55725190839695,
+      "grad_norm": 0.33554473519325256,
+      "learning_rate": 4.648102086768077e-06,
+      "loss": 0.0791,
+      "step": 16390
+    },
+    {
+      "epoch": 62.595419847328245,
+      "grad_norm": 0.4923129677772522,
+      "learning_rate": 4.642604402742979e-06,
+      "loss": 0.073,
+      "step": 16400
+    },
+    {
+      "epoch": 62.63358778625954,
+      "grad_norm": 0.44560110569000244,
+      "learning_rate": 4.6371071529872336e-06,
+      "loss": 0.0735,
+      "step": 16410
+    },
+    {
+      "epoch": 62.67175572519084,
+      "grad_norm": 0.47609812021255493,
+      "learning_rate": 4.6316103441805155e-06,
+      "loss": 0.071,
+      "step": 16420
+    },
+    {
+      "epoch": 62.70992366412214,
+      "grad_norm": 0.6052244901657104,
+      "learning_rate": 4.626113983001965e-06,
+      "loss": 0.0764,
+      "step": 16430
+    },
+    {
+      "epoch": 62.74809160305343,
+      "grad_norm": 0.3735920786857605,
+      "learning_rate": 4.620618076130182e-06,
+      "loss": 0.0671,
+      "step": 16440
+    },
+    {
+      "epoch": 62.786259541984734,
+      "grad_norm": 0.279205858707428,
+      "learning_rate": 4.615122630243207e-06,
+      "loss": 0.0767,
+      "step": 16450
+    },
+    {
+      "epoch": 62.82442748091603,
+      "grad_norm": 0.7630006670951843,
+      "learning_rate": 4.60962765201853e-06,
+      "loss": 0.0739,
+      "step": 16460
+    },
+    {
+      "epoch": 62.86259541984733,
+      "grad_norm": 0.22555769979953766,
+      "learning_rate": 4.604133148133066e-06,
+      "loss": 0.071,
+      "step": 16470
+    },
+    {
+      "epoch": 62.900763358778626,
+      "grad_norm": 0.42524755001068115,
+      "learning_rate": 4.598639125263155e-06,
+      "loss": 0.0794,
+      "step": 16480
+    },
+    {
+      "epoch": 62.93893129770992,
+      "grad_norm": 0.38487425446510315,
+      "learning_rate": 4.593145590084553e-06,
+      "loss": 0.0764,
+      "step": 16490
+    },
+    {
+      "epoch": 62.97709923664122,
+      "grad_norm": 0.3085957467556,
+      "learning_rate": 4.58765254927242e-06,
+      "loss": 0.0691,
+      "step": 16500
+    },
+    {
+      "epoch": 63.01526717557252,
+      "grad_norm": 0.4019497334957123,
+      "learning_rate": 4.582160009501323e-06,
+      "loss": 0.0692,
+      "step": 16510
+    },
+    {
+      "epoch": 63.05343511450382,
+      "grad_norm": 0.2319970428943634,
+      "learning_rate": 4.576667977445214e-06,
+      "loss": 0.0677,
+      "step": 16520
+    },
+    {
+      "epoch": 63.091603053435115,
+      "grad_norm": 0.4725869596004486,
+      "learning_rate": 4.571176459777431e-06,
+      "loss": 0.0739,
+      "step": 16530
+    },
+    {
+      "epoch": 63.12977099236641,
+      "grad_norm": 0.3336828351020813,
+      "learning_rate": 4.565685463170685e-06,
+      "loss": 0.0676,
+      "step": 16540
+    },
+    {
+      "epoch": 63.16793893129771,
+      "grad_norm": 0.6508419513702393,
+      "learning_rate": 4.560194994297054e-06,
+      "loss": 0.0765,
+      "step": 16550
+    },
+    {
+      "epoch": 63.20610687022901,
+      "grad_norm": 1.2349802255630493,
+      "learning_rate": 4.554705059827974e-06,
+      "loss": 0.0974,
+      "step": 16560
+    },
+    {
+      "epoch": 63.2442748091603,
+      "grad_norm": 0.23870617151260376,
+      "learning_rate": 4.549215666434237e-06,
+      "loss": 0.0632,
+      "step": 16570
+    },
+    {
+      "epoch": 63.282442748091604,
+      "grad_norm": 0.45109471678733826,
+      "learning_rate": 4.54372682078597e-06,
+      "loss": 0.0704,
+      "step": 16580
+    },
+    {
+      "epoch": 63.3206106870229,
+      "grad_norm": 0.9649439454078674,
+      "learning_rate": 4.538238529552641e-06,
+      "loss": 0.0808,
+      "step": 16590
+    },
+    {
+      "epoch": 63.3587786259542,
+      "grad_norm": 0.35490062832832336,
+      "learning_rate": 4.532750799403039e-06,
+      "loss": 0.0762,
+      "step": 16600
+    },
+    {
+      "epoch": 63.396946564885496,
+      "grad_norm": 0.453528493642807,
+      "learning_rate": 4.527263637005274e-06,
+      "loss": 0.0764,
+      "step": 16610
+    },
+    {
+      "epoch": 63.43511450381679,
+      "grad_norm": 0.4732442796230316,
+      "learning_rate": 4.521777049026767e-06,
+      "loss": 0.0772,
+      "step": 16620
+    },
+    {
+      "epoch": 63.47328244274809,
+      "grad_norm": 0.4358898401260376,
+      "learning_rate": 4.516291042134238e-06,
+      "loss": 0.0697,
+      "step": 16630
+    },
+    {
+      "epoch": 63.51145038167939,
+      "grad_norm": 0.3423472046852112,
+      "learning_rate": 4.5108056229937055e-06,
+      "loss": 0.0741,
+      "step": 16640
+    },
+    {
+      "epoch": 63.54961832061069,
+      "grad_norm": 0.5359134078025818,
+      "learning_rate": 4.505320798270467e-06,
+      "loss": 0.072,
+      "step": 16650
+    },
+    {
+      "epoch": 63.587786259541986,
+      "grad_norm": 0.2593790590763092,
+      "learning_rate": 4.4998365746291045e-06,
+      "loss": 0.0761,
+      "step": 16660
+    },
+    {
+      "epoch": 63.62595419847328,
+      "grad_norm": 0.24456137418746948,
+      "learning_rate": 4.494352958733466e-06,
+      "loss": 0.0652,
+      "step": 16670
+    },
+    {
+      "epoch": 63.66412213740458,
+      "grad_norm": 1.0902378559112549,
+      "learning_rate": 4.4888699572466624e-06,
+      "loss": 0.0741,
+      "step": 16680
+    },
+    {
+      "epoch": 63.70229007633588,
+      "grad_norm": 0.3199808895587921,
+      "learning_rate": 4.483387576831058e-06,
+      "loss": 0.0774,
+      "step": 16690
+    },
+    {
+      "epoch": 63.74045801526717,
+      "grad_norm": 1.09122896194458,
+      "learning_rate": 4.47790582414826e-06,
+      "loss": 0.0737,
+      "step": 16700
+    },
+    {
+      "epoch": 63.778625954198475,
+      "grad_norm": 0.3936939537525177,
+      "learning_rate": 4.472424705859115e-06,
+      "loss": 0.0705,
+      "step": 16710
+    },
+    {
+      "epoch": 63.81679389312977,
+      "grad_norm": 0.3976505398750305,
+      "learning_rate": 4.466944228623701e-06,
+      "loss": 0.073,
+      "step": 16720
+    },
+    {
+      "epoch": 63.85496183206107,
+      "grad_norm": 0.2906537652015686,
+      "learning_rate": 4.4614643991013125e-06,
+      "loss": 0.0748,
+      "step": 16730
+    },
+    {
+      "epoch": 63.89312977099237,
+      "grad_norm": 0.31357595324516296,
+      "learning_rate": 4.45598522395046e-06,
+      "loss": 0.0729,
+      "step": 16740
+    },
+    {
+      "epoch": 63.93129770992366,
+      "grad_norm": 0.30560722947120667,
+      "learning_rate": 4.450506709828858e-06,
+      "loss": 0.0751,
+      "step": 16750
+    },
+    {
+      "epoch": 63.969465648854964,
+      "grad_norm": 0.3903256356716156,
+      "learning_rate": 4.445028863393417e-06,
+      "loss": 0.0737,
+      "step": 16760
+    },
+    {
+      "epoch": 64.00763358778626,
+      "grad_norm": 0.5252776741981506,
+      "learning_rate": 4.439551691300236e-06,
+      "loss": 0.0748,
+      "step": 16770
+    },
+    {
+      "epoch": 64.04580152671755,
+      "grad_norm": 0.389498233795166,
+      "learning_rate": 4.4340752002045985e-06,
+      "loss": 0.0749,
+      "step": 16780
+    },
+    {
+      "epoch": 64.08396946564885,
+      "grad_norm": 0.3666820228099823,
+      "learning_rate": 4.428599396760957e-06,
+      "loss": 0.0691,
+      "step": 16790
+    },
+    {
+      "epoch": 64.12213740458016,
+      "grad_norm": 0.30693432688713074,
+      "learning_rate": 4.4231242876229256e-06,
+      "loss": 0.0736,
+      "step": 16800
+    },
+    {
+      "epoch": 64.16030534351145,
+      "grad_norm": 0.44508853554725647,
+      "learning_rate": 4.417649879443282e-06,
+      "loss": 0.0683,
+      "step": 16810
+    },
+    {
+      "epoch": 64.19847328244275,
+      "grad_norm": 0.3270077109336853,
+      "learning_rate": 4.4121761788739445e-06,
+      "loss": 0.0672,
+      "step": 16820
+    },
+    {
+      "epoch": 64.23664122137404,
+      "grad_norm": 0.2393168956041336,
+      "learning_rate": 4.406703192565981e-06,
+      "loss": 0.0708,
+      "step": 16830
+    },
+    {
+      "epoch": 64.27480916030534,
+      "grad_norm": 0.4205631613731384,
+      "learning_rate": 4.401230927169582e-06,
+      "loss": 0.0735,
+      "step": 16840
+    },
+    {
+      "epoch": 64.31297709923665,
+      "grad_norm": 0.7220420241355896,
+      "learning_rate": 4.395759389334067e-06,
+      "loss": 0.0704,
+      "step": 16850
+    },
+    {
+      "epoch": 64.35114503816794,
+      "grad_norm": 0.4618149697780609,
+      "learning_rate": 4.3902885857078685e-06,
+      "loss": 0.0716,
+      "step": 16860
+    },
+    {
+      "epoch": 64.38931297709924,
+      "grad_norm": 0.304708868265152,
+      "learning_rate": 4.384818522938531e-06,
+      "loss": 0.074,
+      "step": 16870
+    },
+    {
+      "epoch": 64.42748091603053,
+      "grad_norm": 0.6205267310142517,
+      "learning_rate": 4.379349207672696e-06,
+      "loss": 0.0716,
+      "step": 16880
+    },
+    {
+      "epoch": 64.46564885496183,
+      "grad_norm": 0.4849410057067871,
+      "learning_rate": 4.3738806465560975e-06,
+      "loss": 0.0744,
+      "step": 16890
+    },
+    {
+      "epoch": 64.50381679389314,
+      "grad_norm": 0.2881622612476349,
+      "learning_rate": 4.368412846233554e-06,
+      "loss": 0.0694,
+      "step": 16900
+    },
+    {
+      "epoch": 64.54198473282443,
+      "grad_norm": 0.5529125928878784,
+      "learning_rate": 4.362945813348956e-06,
+      "loss": 0.0724,
+      "step": 16910
+    },
+    {
+      "epoch": 64.58015267175573,
+      "grad_norm": 0.41142237186431885,
+      "learning_rate": 4.357479554545263e-06,
+      "loss": 0.0715,
+      "step": 16920
+    },
+    {
+      "epoch": 64.61832061068702,
+      "grad_norm": 0.6459675431251526,
+      "learning_rate": 4.352014076464499e-06,
+      "loss": 0.0723,
+      "step": 16930
+    },
+    {
+      "epoch": 64.65648854961832,
+      "grad_norm": 0.5050190091133118,
+      "learning_rate": 4.346549385747733e-06,
+      "loss": 0.0754,
+      "step": 16940
+    },
+    {
+      "epoch": 64.69465648854961,
+      "grad_norm": 0.3293865919113159,
+      "learning_rate": 4.34108548903508e-06,
+      "loss": 0.0696,
+      "step": 16950
+    },
+    {
+      "epoch": 64.73282442748092,
+      "grad_norm": 0.43672609329223633,
+      "learning_rate": 4.335622392965689e-06,
+      "loss": 0.0745,
+      "step": 16960
+    },
+    {
+      "epoch": 64.77099236641222,
+      "grad_norm": 0.43421950936317444,
+      "learning_rate": 4.330160104177738e-06,
+      "loss": 0.0691,
+      "step": 16970
+    },
+    {
+      "epoch": 64.80916030534351,
+      "grad_norm": 1.2267439365386963,
+      "learning_rate": 4.324698629308419e-06,
+      "loss": 0.0748,
+      "step": 16980
+    },
+    {
+      "epoch": 64.8473282442748,
+      "grad_norm": 0.27527856826782227,
+      "learning_rate": 4.3192379749939466e-06,
+      "loss": 0.0736,
+      "step": 16990
+    },
+    {
+      "epoch": 64.8854961832061,
+      "grad_norm": 0.84715735912323,
+      "learning_rate": 4.313778147869524e-06,
+      "loss": 0.0794,
+      "step": 17000
+    },
+    {
+      "epoch": 64.92366412213741,
+      "grad_norm": 0.49666497111320496,
+      "learning_rate": 4.308319154569358e-06,
+      "loss": 0.0752,
+      "step": 17010
+    },
+    {
+      "epoch": 64.9618320610687,
+      "grad_norm": 0.6374955773353577,
+      "learning_rate": 4.302861001726642e-06,
+      "loss": 0.0796,
+      "step": 17020
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 0.31495794653892517,
+      "learning_rate": 4.297403695973542e-06,
+      "loss": 0.0733,
+      "step": 17030
+    },
+    {
+      "epoch": 65.0381679389313,
+      "grad_norm": 0.4066680073738098,
+      "learning_rate": 4.291947243941203e-06,
+      "loss": 0.0755,
+      "step": 17040
+    },
+    {
+      "epoch": 65.07633587786259,
+      "grad_norm": 0.33870530128479004,
+      "learning_rate": 4.286491652259729e-06,
+      "loss": 0.0676,
+      "step": 17050
+    },
+    {
+      "epoch": 65.1145038167939,
+      "grad_norm": 0.2708606421947479,
+      "learning_rate": 4.281036927558178e-06,
+      "loss": 0.0708,
+      "step": 17060
+    },
+    {
+      "epoch": 65.1526717557252,
+      "grad_norm": 0.2700878083705902,
+      "learning_rate": 4.275583076464552e-06,
+      "loss": 0.071,
+      "step": 17070
+    },
+    {
+      "epoch": 65.19083969465649,
+      "grad_norm": 0.3311876952648163,
+      "learning_rate": 4.270130105605794e-06,
+      "loss": 0.0732,
+      "step": 17080
+    },
+    {
+      "epoch": 65.22900763358778,
+      "grad_norm": 0.2695978581905365,
+      "learning_rate": 4.264678021607782e-06,
+      "loss": 0.0736,
+      "step": 17090
+    },
+    {
+      "epoch": 65.26717557251908,
+      "grad_norm": 0.3300071954727173,
+      "learning_rate": 4.259226831095311e-06,
+      "loss": 0.0724,
+      "step": 17100
+    },
+    {
+      "epoch": 65.30534351145039,
+      "grad_norm": 0.2677368223667145,
+      "learning_rate": 4.25377654069209e-06,
+      "loss": 0.0703,
+      "step": 17110
+    },
+    {
+      "epoch": 65.34351145038168,
+      "grad_norm": 0.29270052909851074,
+      "learning_rate": 4.248327157020737e-06,
+      "loss": 0.0724,
+      "step": 17120
+    },
+    {
+      "epoch": 65.38167938931298,
+      "grad_norm": 0.6969024538993835,
+      "learning_rate": 4.242878686702763e-06,
+      "loss": 0.0909,
+      "step": 17130
+    },
+    {
+      "epoch": 65.41984732824427,
+      "grad_norm": 0.3158918023109436,
+      "learning_rate": 4.237431136358579e-06,
+      "loss": 0.0732,
+      "step": 17140
+    },
+    {
+      "epoch": 65.45801526717557,
+      "grad_norm": 0.4324190318584442,
+      "learning_rate": 4.231984512607471e-06,
+      "loss": 0.0729,
+      "step": 17150
+    },
+    {
+      "epoch": 65.49618320610686,
+      "grad_norm": 0.522308886051178,
+      "learning_rate": 4.226538822067598e-06,
+      "loss": 0.0781,
+      "step": 17160
+    },
+    {
+      "epoch": 65.53435114503817,
+      "grad_norm": 0.7167192697525024,
+      "learning_rate": 4.22109407135599e-06,
+      "loss": 0.0677,
+      "step": 17170
+    },
+    {
+      "epoch": 65.57251908396947,
+      "grad_norm": 0.4006265103816986,
+      "learning_rate": 4.2156502670885304e-06,
+      "loss": 0.0681,
+      "step": 17180
+    },
+    {
+      "epoch": 65.61068702290076,
+      "grad_norm": 0.246806800365448,
+      "learning_rate": 4.210207415879953e-06,
+      "loss": 0.0702,
+      "step": 17190
+    },
+    {
+      "epoch": 65.64885496183206,
+      "grad_norm": 0.3185785710811615,
+      "learning_rate": 4.204765524343841e-06,
+      "loss": 0.0648,
+      "step": 17200
+    },
+    {
+      "epoch": 65.68702290076335,
+      "grad_norm": 0.2700112760066986,
+      "learning_rate": 4.199324599092603e-06,
+      "loss": 0.0742,
+      "step": 17210
+    },
+    {
+      "epoch": 65.72519083969466,
+      "grad_norm": 0.7986395955085754,
+      "learning_rate": 4.1938846467374745e-06,
+      "loss": 0.0727,
+      "step": 17220
+    },
+    {
+      "epoch": 65.76335877862596,
+      "grad_norm": 0.34733664989471436,
+      "learning_rate": 4.1884456738885125e-06,
+      "loss": 0.0647,
+      "step": 17230
+    },
+    {
+      "epoch": 65.80152671755725,
+      "grad_norm": 0.35277488827705383,
+      "learning_rate": 4.18300768715458e-06,
+      "loss": 0.0784,
+      "step": 17240
+    },
+    {
+      "epoch": 65.83969465648855,
+      "grad_norm": 0.31571856141090393,
+      "learning_rate": 4.177570693143347e-06,
+      "loss": 0.0736,
+      "step": 17250
+    },
+    {
+      "epoch": 65.87786259541984,
+      "grad_norm": 0.47372332215309143,
+      "learning_rate": 4.172134698461271e-06,
+      "loss": 0.0749,
+      "step": 17260
+    },
+    {
+      "epoch": 65.91603053435115,
+      "grad_norm": 0.3981868624687195,
+      "learning_rate": 4.166699709713599e-06,
+      "loss": 0.0677,
+      "step": 17270
+    },
+    {
+      "epoch": 65.95419847328245,
+      "grad_norm": 0.4929441511631012,
+      "learning_rate": 4.161265733504355e-06,
+      "loss": 0.0733,
+      "step": 17280
+    },
+    {
+      "epoch": 65.99236641221374,
+      "grad_norm": 0.24368400871753693,
+      "learning_rate": 4.155832776436331e-06,
+      "loss": 0.0687,
+      "step": 17290
+    },
+    {
+      "epoch": 66.03053435114504,
+      "grad_norm": 0.3198246359825134,
+      "learning_rate": 4.150400845111085e-06,
+      "loss": 0.0813,
+      "step": 17300
+    },
+    {
+      "epoch": 66.06870229007633,
+      "grad_norm": 0.288518488407135,
+      "learning_rate": 4.144969946128923e-06,
+      "loss": 0.072,
+      "step": 17310
+    },
+    {
+      "epoch": 66.10687022900764,
+      "grad_norm": 0.2651676833629608,
+      "learning_rate": 4.139540086088901e-06,
+      "loss": 0.0696,
+      "step": 17320
+    },
+    {
+      "epoch": 66.14503816793894,
+      "grad_norm": 0.21474707126617432,
+      "learning_rate": 4.1341112715888106e-06,
+      "loss": 0.0711,
+      "step": 17330
+    },
+    {
+      "epoch": 66.18320610687023,
+      "grad_norm": 0.3536648452281952,
+      "learning_rate": 4.128683509225172e-06,
+      "loss": 0.0783,
+      "step": 17340
+    },
+    {
+      "epoch": 66.22137404580153,
+      "grad_norm": 0.3147779703140259,
+      "learning_rate": 4.123256805593231e-06,
+      "loss": 0.0691,
+      "step": 17350
+    },
+    {
+      "epoch": 66.25954198473282,
+      "grad_norm": 0.45088210701942444,
+      "learning_rate": 4.117831167286943e-06,
+      "loss": 0.0716,
+      "step": 17360
+    },
+    {
+      "epoch": 66.29770992366412,
+      "grad_norm": 0.4173077940940857,
+      "learning_rate": 4.112406600898968e-06,
+      "loss": 0.0756,
+      "step": 17370
+    },
+    {
+      "epoch": 66.33587786259542,
+      "grad_norm": 0.2509874403476715,
+      "learning_rate": 4.106983113020669e-06,
+      "loss": 0.0699,
+      "step": 17380
+    },
+    {
+      "epoch": 66.37404580152672,
+      "grad_norm": 0.5545008778572083,
+      "learning_rate": 4.101560710242094e-06,
+      "loss": 0.0757,
+      "step": 17390
+    },
+    {
+      "epoch": 66.41221374045801,
+      "grad_norm": 0.3343713879585266,
+      "learning_rate": 4.096139399151971e-06,
+      "loss": 0.0713,
+      "step": 17400
+    },
+    {
+      "epoch": 66.45038167938931,
+      "grad_norm": 0.40432122349739075,
+      "learning_rate": 4.090719186337709e-06,
+      "loss": 0.0685,
+      "step": 17410
+    },
+    {
+      "epoch": 66.4885496183206,
+      "grad_norm": 0.8056888580322266,
+      "learning_rate": 4.085300078385375e-06,
+      "loss": 0.0753,
+      "step": 17420
+    },
+    {
+      "epoch": 66.52671755725191,
+      "grad_norm": 0.43857723474502563,
+      "learning_rate": 4.079882081879696e-06,
+      "loss": 0.0714,
+      "step": 17430
+    },
+    {
+      "epoch": 66.56488549618321,
+      "grad_norm": 0.8525163531303406,
+      "learning_rate": 4.074465203404048e-06,
+      "loss": 0.0833,
+      "step": 17440
+    },
+    {
+      "epoch": 66.6030534351145,
+      "grad_norm": 0.7485587000846863,
+      "learning_rate": 4.06904944954045e-06,
+      "loss": 0.07,
+      "step": 17450
+    },
+    {
+      "epoch": 66.6412213740458,
+      "grad_norm": 0.22170186042785645,
+      "learning_rate": 4.063634826869553e-06,
+      "loss": 0.0746,
+      "step": 17460
+    },
+    {
+      "epoch": 66.6793893129771,
+      "grad_norm": 0.40687039494514465,
+      "learning_rate": 4.058221341970638e-06,
+      "loss": 0.0709,
+      "step": 17470
+    },
+    {
+      "epoch": 66.7175572519084,
+      "grad_norm": 0.3101550340652466,
+      "learning_rate": 4.052809001421595e-06,
+      "loss": 0.0653,
+      "step": 17480
+    },
+    {
+      "epoch": 66.7557251908397,
+      "grad_norm": 0.4546465277671814,
+      "learning_rate": 4.047397811798929e-06,
+      "loss": 0.0698,
+      "step": 17490
+    },
+    {
+      "epoch": 66.79389312977099,
+      "grad_norm": 0.35028156638145447,
+      "learning_rate": 4.041987779677745e-06,
+      "loss": 0.0774,
+      "step": 17500
+    },
+    {
+      "epoch": 66.83206106870229,
+      "grad_norm": 0.47379425168037415,
+      "learning_rate": 4.036578911631746e-06,
+      "loss": 0.0726,
+      "step": 17510
+    },
+    {
+      "epoch": 66.87022900763358,
+      "grad_norm": 0.6644511222839355,
+      "learning_rate": 4.0311712142332115e-06,
+      "loss": 0.069,
+      "step": 17520
+    },
+    {
+      "epoch": 66.90839694656489,
+      "grad_norm": 0.6375014185905457,
+      "learning_rate": 4.025764694053008e-06,
+      "loss": 0.0676,
+      "step": 17530
+    },
+    {
+      "epoch": 66.94656488549619,
+      "grad_norm": 0.39866891503334045,
+      "learning_rate": 4.020359357660566e-06,
+      "loss": 0.0773,
+      "step": 17540
+    },
+    {
+      "epoch": 66.98473282442748,
+      "grad_norm": 0.31311362981796265,
+      "learning_rate": 4.014955211623875e-06,
+      "loss": 0.0661,
+      "step": 17550
+    },
+    {
+      "epoch": 67.02290076335878,
+      "grad_norm": 0.42248404026031494,
+      "learning_rate": 4.00955226250949e-06,
+      "loss": 0.0787,
+      "step": 17560
+    },
+    {
+      "epoch": 67.06106870229007,
+      "grad_norm": 0.7026179432868958,
+      "learning_rate": 4.0041505168824976e-06,
+      "loss": 0.0755,
+      "step": 17570
+    },
+    {
+      "epoch": 67.09923664122137,
+      "grad_norm": 0.32166993618011475,
+      "learning_rate": 3.99874998130653e-06,
+      "loss": 0.0659,
+      "step": 17580
+    },
+    {
+      "epoch": 67.13740458015268,
+      "grad_norm": 0.2971125543117523,
+      "learning_rate": 3.993350662343746e-06,
+      "loss": 0.0675,
+      "step": 17590
+    },
+    {
+      "epoch": 67.17557251908397,
+      "grad_norm": 0.3091699182987213,
+      "learning_rate": 3.987952566554828e-06,
+      "loss": 0.0719,
+      "step": 17600
+    },
+    {
+      "epoch": 67.21374045801527,
+      "grad_norm": 0.24681086838245392,
+      "learning_rate": 3.982555700498971e-06,
+      "loss": 0.0732,
+      "step": 17610
+    },
+    {
+      "epoch": 67.25190839694656,
+      "grad_norm": 0.4795871078968048,
+      "learning_rate": 3.977160070733878e-06,
+      "loss": 0.0691,
+      "step": 17620
+    },
+    {
+      "epoch": 67.29007633587786,
+      "grad_norm": 0.6124231815338135,
+      "learning_rate": 3.971765683815746e-06,
+      "loss": 0.0702,
+      "step": 17630
+    },
+    {
+      "epoch": 67.32824427480917,
+      "grad_norm": 0.659125030040741,
+      "learning_rate": 3.966372546299264e-06,
+      "loss": 0.0744,
+      "step": 17640
+    },
+    {
+      "epoch": 67.36641221374046,
+      "grad_norm": 0.6344574689865112,
+      "learning_rate": 3.960980664737604e-06,
+      "loss": 0.0723,
+      "step": 17650
+    },
+    {
+      "epoch": 67.40458015267176,
+      "grad_norm": 0.3194618225097656,
+      "learning_rate": 3.955590045682408e-06,
+      "loss": 0.0714,
+      "step": 17660
+    },
+    {
+      "epoch": 67.44274809160305,
+      "grad_norm": 0.44845250248908997,
+      "learning_rate": 3.950200695683788e-06,
+      "loss": 0.0723,
+      "step": 17670
+    },
+    {
+      "epoch": 67.48091603053435,
+      "grad_norm": 1.0037466287612915,
+      "learning_rate": 3.944812621290314e-06,
+      "loss": 0.0691,
+      "step": 17680
+    },
+    {
+      "epoch": 67.51908396946565,
+      "grad_norm": 0.2766275107860565,
+      "learning_rate": 3.939425829049002e-06,
+      "loss": 0.0717,
+      "step": 17690
+    },
+    {
+      "epoch": 67.55725190839695,
+      "grad_norm": 0.654982328414917,
+      "learning_rate": 3.934040325505313e-06,
+      "loss": 0.0672,
+      "step": 17700
+    },
+    {
+      "epoch": 67.59541984732824,
+      "grad_norm": 0.36373141407966614,
+      "learning_rate": 3.928656117203141e-06,
+      "loss": 0.0713,
+      "step": 17710
+    },
+    {
+      "epoch": 67.63358778625954,
+      "grad_norm": 0.34860584139823914,
+      "learning_rate": 3.923273210684809e-06,
+      "loss": 0.0757,
+      "step": 17720
+    },
+    {
+      "epoch": 67.67175572519083,
+      "grad_norm": 0.5683857202529907,
+      "learning_rate": 3.917891612491055e-06,
+      "loss": 0.0675,
+      "step": 17730
+    },
+    {
+      "epoch": 67.70992366412214,
+      "grad_norm": 0.5951647162437439,
+      "learning_rate": 3.912511329161027e-06,
+      "loss": 0.0702,
+      "step": 17740
+    },
+    {
+      "epoch": 67.74809160305344,
+      "grad_norm": 0.27941733598709106,
+      "learning_rate": 3.907132367232279e-06,
+      "loss": 0.0696,
+      "step": 17750
+    },
+    {
+      "epoch": 67.78625954198473,
+      "grad_norm": 0.27847856283187866,
+      "learning_rate": 3.901754733240753e-06,
+      "loss": 0.0727,
+      "step": 17760
+    },
+    {
+      "epoch": 67.82442748091603,
+      "grad_norm": 0.5478169322013855,
+      "learning_rate": 3.896378433720786e-06,
+      "loss": 0.0719,
+      "step": 17770
+    },
+    {
+      "epoch": 67.86259541984732,
+      "grad_norm": 0.5233021378517151,
+      "learning_rate": 3.891003475205086e-06,
+      "loss": 0.065,
+      "step": 17780
+    },
+    {
+      "epoch": 67.90076335877862,
+      "grad_norm": 0.2591749131679535,
+      "learning_rate": 3.885629864224736e-06,
+      "loss": 0.0731,
+      "step": 17790
+    },
+    {
+      "epoch": 67.93893129770993,
+      "grad_norm": 0.39864397048950195,
+      "learning_rate": 3.880257607309178e-06,
+      "loss": 0.0697,
+      "step": 17800
+    },
+    {
+      "epoch": 67.97709923664122,
+      "grad_norm": 0.46604353189468384,
+      "learning_rate": 3.874886710986213e-06,
+      "loss": 0.0708,
+      "step": 17810
+    },
+    {
+      "epoch": 68.01526717557252,
+      "grad_norm": 0.35567206144332886,
+      "learning_rate": 3.869517181781983e-06,
+      "loss": 0.0679,
+      "step": 17820
+    },
+    {
+      "epoch": 68.05343511450381,
+      "grad_norm": 0.8951524496078491,
+      "learning_rate": 3.864149026220977e-06,
+      "loss": 0.076,
+      "step": 17830
+    },
+    {
+      "epoch": 68.09160305343511,
+      "grad_norm": 0.43323707580566406,
+      "learning_rate": 3.858782250826009e-06,
+      "loss": 0.0781,
+      "step": 17840
+    },
+    {
+      "epoch": 68.12977099236642,
+      "grad_norm": 0.3977680206298828,
+      "learning_rate": 3.853416862118214e-06,
+      "loss": 0.0766,
+      "step": 17850
+    },
+    {
+      "epoch": 68.16793893129771,
+      "grad_norm": 0.7726107239723206,
+      "learning_rate": 3.8480528666170495e-06,
+      "loss": 0.075,
+      "step": 17860
+    },
+    {
+      "epoch": 68.20610687022901,
+      "grad_norm": 0.40981167554855347,
+      "learning_rate": 3.8426902708402695e-06,
+      "loss": 0.0697,
+      "step": 17870
+    },
+    {
+      "epoch": 68.2442748091603,
+      "grad_norm": 0.23430369794368744,
+      "learning_rate": 3.8373290813039404e-06,
+      "loss": 0.0744,
+      "step": 17880
+    },
+    {
+      "epoch": 68.2824427480916,
+      "grad_norm": 0.33602553606033325,
+      "learning_rate": 3.83196930452241e-06,
+      "loss": 0.0729,
+      "step": 17890
+    },
+    {
+      "epoch": 68.3206106870229,
+      "grad_norm": 0.49401628971099854,
+      "learning_rate": 3.826610947008313e-06,
+      "loss": 0.0703,
+      "step": 17900
+    },
+    {
+      "epoch": 68.3587786259542,
+      "grad_norm": 0.4614551067352295,
+      "learning_rate": 3.821254015272558e-06,
+      "loss": 0.0699,
+      "step": 17910
+    },
+    {
+      "epoch": 68.3969465648855,
+      "grad_norm": 0.39021456241607666,
+      "learning_rate": 3.8158985158243214e-06,
+      "loss": 0.0722,
+      "step": 17920
+    },
+    {
+      "epoch": 68.43511450381679,
+      "grad_norm": 0.2698560059070587,
+      "learning_rate": 3.810544455171044e-06,
+      "loss": 0.0725,
+      "step": 17930
+    },
+    {
+      "epoch": 68.47328244274809,
+      "grad_norm": 0.339725136756897,
+      "learning_rate": 3.805191839818412e-06,
+      "loss": 0.0667,
+      "step": 17940
+    },
+    {
+      "epoch": 68.5114503816794,
+      "grad_norm": 0.36522993445396423,
+      "learning_rate": 3.7998406762703566e-06,
+      "loss": 0.0783,
+      "step": 17950
+    },
+    {
+      "epoch": 68.54961832061069,
+      "grad_norm": 0.4435858726501465,
+      "learning_rate": 3.794490971029048e-06,
+      "loss": 0.0769,
+      "step": 17960
+    },
+    {
+      "epoch": 68.58778625954199,
+      "grad_norm": 0.3447798490524292,
+      "learning_rate": 3.7891427305948815e-06,
+      "loss": 0.0898,
+      "step": 17970
+    },
+    {
+      "epoch": 68.62595419847328,
+      "grad_norm": 0.2673998177051544,
+      "learning_rate": 3.7837959614664714e-06,
+      "loss": 0.0806,
+      "step": 17980
+    },
+    {
+      "epoch": 68.66412213740458,
+      "grad_norm": 0.3040778636932373,
+      "learning_rate": 3.778450670140651e-06,
+      "loss": 0.0745,
+      "step": 17990
+    },
+    {
+      "epoch": 68.70229007633588,
+      "grad_norm": 0.3017938733100891,
+      "learning_rate": 3.773106863112451e-06,
+      "loss": 0.0692,
+      "step": 18000
+    },
+    {
+      "epoch": 68.74045801526718,
+      "grad_norm": 0.28814712166786194,
+      "learning_rate": 3.7677645468751e-06,
+      "loss": 0.0656,
+      "step": 18010
+    },
+    {
+      "epoch": 68.77862595419847,
+      "grad_norm": 0.5016555190086365,
+      "learning_rate": 3.7624237279200175e-06,
+      "loss": 0.0721,
+      "step": 18020
+    },
+    {
+      "epoch": 68.81679389312977,
+      "grad_norm": 0.23837906122207642,
+      "learning_rate": 3.7570844127367994e-06,
+      "loss": 0.0642,
+      "step": 18030
+    },
+    {
+      "epoch": 68.85496183206106,
+      "grad_norm": 0.5186642408370972,
+      "learning_rate": 3.7517466078132213e-06,
+      "loss": 0.0689,
+      "step": 18040
+    },
+    {
+      "epoch": 68.89312977099236,
+      "grad_norm": 0.4285561442375183,
+      "learning_rate": 3.7464103196352176e-06,
+      "loss": 0.0816,
+      "step": 18050
+    },
+    {
+      "epoch": 68.93129770992367,
+      "grad_norm": 0.3345412015914917,
+      "learning_rate": 3.7410755546868803e-06,
+      "loss": 0.0794,
+      "step": 18060
+    },
+    {
+      "epoch": 68.96946564885496,
+      "grad_norm": 0.3829120695590973,
+      "learning_rate": 3.7357423194504538e-06,
+      "loss": 0.0698,
+      "step": 18070
+    },
+    {
+      "epoch": 69.00763358778626,
+      "grad_norm": 0.327454537153244,
+      "learning_rate": 3.7304106204063186e-06,
+      "loss": 0.0763,
+      "step": 18080
+    },
+    {
+      "epoch": 69.04580152671755,
+      "grad_norm": 0.4839470088481903,
+      "learning_rate": 3.725080464032996e-06,
+      "loss": 0.0722,
+      "step": 18090
+    },
+    {
+      "epoch": 69.08396946564885,
+      "grad_norm": 0.2581571340560913,
+      "learning_rate": 3.7197518568071256e-06,
+      "loss": 0.0678,
+      "step": 18100
+    },
+    {
+      "epoch": 69.12213740458016,
+      "grad_norm": 0.41950273513793945,
+      "learning_rate": 3.7144248052034696e-06,
+      "loss": 0.0769,
+      "step": 18110
+    },
+    {
+      "epoch": 69.16030534351145,
+      "grad_norm": 0.35857993364334106,
+      "learning_rate": 3.7090993156948973e-06,
+      "loss": 0.069,
+      "step": 18120
+    },
+    {
+      "epoch": 69.19847328244275,
+      "grad_norm": 0.5615084171295166,
+      "learning_rate": 3.7037753947523786e-06,
+      "loss": 0.0648,
+      "step": 18130
+    },
+    {
+      "epoch": 69.23664122137404,
+      "grad_norm": 0.32720375061035156,
+      "learning_rate": 3.6984530488449833e-06,
+      "loss": 0.0669,
+      "step": 18140
+    },
+    {
+      "epoch": 69.27480916030534,
+      "grad_norm": 0.38296082615852356,
+      "learning_rate": 3.693132284439861e-06,
+      "loss": 0.0698,
+      "step": 18150
+    },
+    {
+      "epoch": 69.31297709923665,
+      "grad_norm": 0.39868733286857605,
+      "learning_rate": 3.687813108002242e-06,
+      "loss": 0.0693,
+      "step": 18160
+    },
+    {
+      "epoch": 69.35114503816794,
+      "grad_norm": 0.2485743910074234,
+      "learning_rate": 3.6824955259954285e-06,
+      "loss": 0.0765,
+      "step": 18170
+    },
+    {
+      "epoch": 69.38931297709924,
+      "grad_norm": 0.3294006884098053,
+      "learning_rate": 3.6771795448807847e-06,
+      "loss": 0.0684,
+      "step": 18180
+    },
+    {
+      "epoch": 69.42748091603053,
+      "grad_norm": 0.584002673625946,
+      "learning_rate": 3.6718651711177244e-06,
+      "loss": 0.0775,
+      "step": 18190
+    },
+    {
+      "epoch": 69.46564885496183,
+      "grad_norm": 0.3454589545726776,
+      "learning_rate": 3.6665524111637184e-06,
+      "loss": 0.0703,
+      "step": 18200
+    },
+    {
+      "epoch": 69.50381679389314,
+      "grad_norm": 0.4234253466129303,
+      "learning_rate": 3.6612412714742695e-06,
+      "loss": 0.07,
+      "step": 18210
+    },
+    {
+      "epoch": 69.54198473282443,
+      "grad_norm": 0.8012011051177979,
+      "learning_rate": 3.655931758502912e-06,
+      "loss": 0.0753,
+      "step": 18220
+    },
+    {
+      "epoch": 69.58015267175573,
+      "grad_norm": 0.25509560108184814,
+      "learning_rate": 3.6506238787012038e-06,
+      "loss": 0.0704,
+      "step": 18230
+    },
+    {
+      "epoch": 69.61832061068702,
+      "grad_norm": 0.4144178628921509,
+      "learning_rate": 3.645317638518721e-06,
+      "loss": 0.0743,
+      "step": 18240
+    },
+    {
+      "epoch": 69.65648854961832,
+      "grad_norm": 0.2962321937084198,
+      "learning_rate": 3.6400130444030456e-06,
+      "loss": 0.0815,
+      "step": 18250
+    },
+    {
+      "epoch": 69.69465648854961,
+      "grad_norm": 0.6895132660865784,
+      "learning_rate": 3.634710102799761e-06,
+      "loss": 0.0696,
+      "step": 18260
+    },
+    {
+      "epoch": 69.73282442748092,
+      "grad_norm": 0.2801784873008728,
+      "learning_rate": 3.62940882015244e-06,
+      "loss": 0.0733,
+      "step": 18270
+    },
+    {
+      "epoch": 69.77099236641222,
+      "grad_norm": 0.24803809821605682,
+      "learning_rate": 3.6241092029026405e-06,
+      "loss": 0.0778,
+      "step": 18280
+    },
+    {
+      "epoch": 69.80916030534351,
+      "grad_norm": 0.7623275518417358,
+      "learning_rate": 3.6188112574898955e-06,
+      "loss": 0.0768,
+      "step": 18290
+    },
+    {
+      "epoch": 69.8473282442748,
+      "grad_norm": 0.2937699556350708,
+      "learning_rate": 3.613514990351712e-06,
+      "loss": 0.0761,
+      "step": 18300
+    },
+    {
+      "epoch": 69.8854961832061,
+      "grad_norm": 0.3946186900138855,
+      "learning_rate": 3.608220407923552e-06,
+      "loss": 0.0707,
+      "step": 18310
+    },
+    {
+      "epoch": 69.92366412213741,
+      "grad_norm": 0.28563305735588074,
+      "learning_rate": 3.602927516638833e-06,
+      "loss": 0.0698,
+      "step": 18320
+    },
+    {
+      "epoch": 69.9618320610687,
+      "grad_norm": 0.25289323925971985,
+      "learning_rate": 3.597636322928917e-06,
+      "loss": 0.0726,
+      "step": 18330
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 0.21828551590442657,
+      "learning_rate": 3.5923468332231003e-06,
+      "loss": 0.0706,
+      "step": 18340
+    },
+    {
+      "epoch": 70.0381679389313,
+      "grad_norm": 0.2835104763507843,
+      "learning_rate": 3.5870590539486163e-06,
+      "loss": 0.0672,
+      "step": 18350
+    },
+    {
+      "epoch": 70.07633587786259,
+      "grad_norm": 0.2114870399236679,
+      "learning_rate": 3.5817729915306138e-06,
+      "loss": 0.0672,
+      "step": 18360
+    },
+    {
+      "epoch": 70.1145038167939,
+      "grad_norm": 0.33038604259490967,
+      "learning_rate": 3.5764886523921567e-06,
+      "loss": 0.062,
+      "step": 18370
+    },
+    {
+      "epoch": 70.1526717557252,
+      "grad_norm": 0.25041481852531433,
+      "learning_rate": 3.571206042954214e-06,
+      "loss": 0.0711,
+      "step": 18380
+    },
+    {
+      "epoch": 70.19083969465649,
+      "grad_norm": 0.30329805612564087,
+      "learning_rate": 3.565925169635657e-06,
+      "loss": 0.0814,
+      "step": 18390
+    },
+    {
+      "epoch": 70.22900763358778,
+      "grad_norm": 0.6908921599388123,
+      "learning_rate": 3.5606460388532406e-06,
+      "loss": 0.0786,
+      "step": 18400
+    },
+    {
+      "epoch": 70.26717557251908,
+      "grad_norm": 0.8706730008125305,
+      "learning_rate": 3.5553686570216116e-06,
+      "loss": 0.0758,
+      "step": 18410
+    },
+    {
+      "epoch": 70.30534351145039,
+      "grad_norm": 0.6713736057281494,
+      "learning_rate": 3.5500930305532845e-06,
+      "loss": 0.0827,
+      "step": 18420
+    },
+    {
+      "epoch": 70.34351145038168,
+      "grad_norm": 0.4993293583393097,
+      "learning_rate": 3.5448191658586423e-06,
+      "loss": 0.0777,
+      "step": 18430
+    },
+    {
+      "epoch": 70.38167938931298,
+      "grad_norm": 0.7732614874839783,
+      "learning_rate": 3.5395470693459267e-06,
+      "loss": 0.0762,
+      "step": 18440
+    },
+    {
+      "epoch": 70.41984732824427,
+      "grad_norm": 0.37474754452705383,
+      "learning_rate": 3.5342767474212344e-06,
+      "loss": 0.0741,
+      "step": 18450
+    },
+    {
+      "epoch": 70.45801526717557,
+      "grad_norm": 0.6375201344490051,
+      "learning_rate": 3.5290082064885025e-06,
+      "loss": 0.0704,
+      "step": 18460
+    },
+    {
+      "epoch": 70.49618320610686,
+      "grad_norm": 0.3896820545196533,
+      "learning_rate": 3.5237414529495056e-06,
+      "loss": 0.0739,
+      "step": 18470
+    },
+    {
+      "epoch": 70.53435114503817,
+      "grad_norm": 0.26550376415252686,
+      "learning_rate": 3.5184764932038457e-06,
+      "loss": 0.0941,
+      "step": 18480
+    },
+    {
+      "epoch": 70.57251908396947,
+      "grad_norm": 0.41228294372558594,
+      "learning_rate": 3.513213333648945e-06,
+      "loss": 0.0693,
+      "step": 18490
+    },
+    {
+      "epoch": 70.61068702290076,
+      "grad_norm": 0.3219086825847626,
+      "learning_rate": 3.507951980680037e-06,
+      "loss": 0.0706,
+      "step": 18500
+    },
+    {
+      "epoch": 70.64885496183206,
+      "grad_norm": 0.2977215051651001,
+      "learning_rate": 3.502692440690165e-06,
+      "loss": 0.0728,
+      "step": 18510
+    },
+    {
+      "epoch": 70.68702290076335,
+      "grad_norm": 0.45944830775260925,
+      "learning_rate": 3.497434720070165e-06,
+      "loss": 0.0658,
+      "step": 18520
+    },
+    {
+      "epoch": 70.72519083969466,
+      "grad_norm": 0.45942363142967224,
+      "learning_rate": 3.492178825208662e-06,
+      "loss": 0.0708,
+      "step": 18530
+    },
+    {
+      "epoch": 70.76335877862596,
+      "grad_norm": 0.401991069316864,
+      "learning_rate": 3.486924762492065e-06,
+      "loss": 0.0772,
+      "step": 18540
+    },
+    {
+      "epoch": 70.80152671755725,
+      "grad_norm": 0.390428751707077,
+      "learning_rate": 3.4816725383045534e-06,
+      "loss": 0.0712,
+      "step": 18550
+    },
+    {
+      "epoch": 70.83969465648855,
+      "grad_norm": 0.36165851354599,
+      "learning_rate": 3.476422159028079e-06,
+      "loss": 0.0714,
+      "step": 18560
+    },
+    {
+      "epoch": 70.87786259541984,
+      "grad_norm": 0.3357897996902466,
+      "learning_rate": 3.471173631042345e-06,
+      "loss": 0.0746,
+      "step": 18570
+    },
+    {
+      "epoch": 70.91603053435115,
+      "grad_norm": 0.9048196077346802,
+      "learning_rate": 3.465926960724808e-06,
+      "loss": 0.0731,
+      "step": 18580
+    },
+    {
+      "epoch": 70.95419847328245,
+      "grad_norm": 0.34917864203453064,
+      "learning_rate": 3.4606821544506664e-06,
+      "loss": 0.0722,
+      "step": 18590
+    },
+    {
+      "epoch": 70.99236641221374,
+      "grad_norm": 0.2705357074737549,
+      "learning_rate": 3.4554392185928563e-06,
+      "loss": 0.0867,
+      "step": 18600
+    },
+    {
+      "epoch": 71.03053435114504,
+      "grad_norm": 0.27343451976776123,
+      "learning_rate": 3.450198159522037e-06,
+      "loss": 0.0635,
+      "step": 18610
+    },
+    {
+      "epoch": 71.06870229007633,
+      "grad_norm": 0.48020657896995544,
+      "learning_rate": 3.444958983606592e-06,
+      "loss": 0.0731,
+      "step": 18620
+    },
+    {
+      "epoch": 71.10687022900764,
+      "grad_norm": 0.25576311349868774,
+      "learning_rate": 3.4397216972126126e-06,
+      "loss": 0.0738,
+      "step": 18630
+    },
+    {
+      "epoch": 71.14503816793894,
+      "grad_norm": 0.3990323543548584,
+      "learning_rate": 3.434486306703896e-06,
+      "loss": 0.0861,
+      "step": 18640
+    },
+    {
+      "epoch": 71.18320610687023,
+      "grad_norm": 0.25410884618759155,
+      "learning_rate": 3.429252818441935e-06,
+      "loss": 0.0759,
+      "step": 18650
+    },
+    {
+      "epoch": 71.22137404580153,
+      "grad_norm": 0.2940288782119751,
+      "learning_rate": 3.4240212387859097e-06,
+      "loss": 0.0745,
+      "step": 18660
+    },
+    {
+      "epoch": 71.25954198473282,
+      "grad_norm": 0.5581989288330078,
+      "learning_rate": 3.4187915740926856e-06,
+      "loss": 0.0738,
+      "step": 18670
+    },
+    {
+      "epoch": 71.29770992366412,
+      "grad_norm": 0.3819045424461365,
+      "learning_rate": 3.4135638307167962e-06,
+      "loss": 0.0729,
+      "step": 18680
+    },
+    {
+      "epoch": 71.33587786259542,
+      "grad_norm": 0.34871965646743774,
+      "learning_rate": 3.408338015010445e-06,
+      "loss": 0.0729,
+      "step": 18690
+    },
+    {
+      "epoch": 71.37404580152672,
+      "grad_norm": 0.3207714557647705,
+      "learning_rate": 3.4031141333234895e-06,
+      "loss": 0.0706,
+      "step": 18700
+    },
+    {
+      "epoch": 71.41221374045801,
+      "grad_norm": 0.27491647005081177,
+      "learning_rate": 3.397892192003437e-06,
+      "loss": 0.0704,
+      "step": 18710
+    },
+    {
+      "epoch": 71.45038167938931,
+      "grad_norm": 0.3418438732624054,
+      "learning_rate": 3.392672197395441e-06,
+      "loss": 0.0685,
+      "step": 18720
+    },
+    {
+      "epoch": 71.4885496183206,
+      "grad_norm": 0.8138360381126404,
+      "learning_rate": 3.3874541558422874e-06,
+      "loss": 0.0704,
+      "step": 18730
+    },
+    {
+      "epoch": 71.52671755725191,
+      "grad_norm": 0.2727760672569275,
+      "learning_rate": 3.3822380736843865e-06,
+      "loss": 0.087,
+      "step": 18740
+    },
+    {
+      "epoch": 71.56488549618321,
+      "grad_norm": 0.5229569673538208,
+      "learning_rate": 3.3770239572597715e-06,
+      "loss": 0.0794,
+      "step": 18750
+    },
+    {
+      "epoch": 71.6030534351145,
+      "grad_norm": 0.3278786540031433,
+      "learning_rate": 3.3718118129040833e-06,
+      "loss": 0.0712,
+      "step": 18760
+    },
+    {
+      "epoch": 71.6412213740458,
+      "grad_norm": 0.2960107624530792,
+      "learning_rate": 3.3666016469505725e-06,
+      "loss": 0.0671,
+      "step": 18770
+    },
+    {
+      "epoch": 71.6793893129771,
+      "grad_norm": 0.5897856950759888,
+      "learning_rate": 3.3613934657300793e-06,
+      "loss": 0.0704,
+      "step": 18780
+    },
+    {
+      "epoch": 71.7175572519084,
+      "grad_norm": 0.30159324407577515,
+      "learning_rate": 3.356187275571037e-06,
+      "loss": 0.0683,
+      "step": 18790
+    },
+    {
+      "epoch": 71.7557251908397,
+      "grad_norm": 0.32948631048202515,
+      "learning_rate": 3.350983082799456e-06,
+      "loss": 0.0779,
+      "step": 18800
+    },
+    {
+      "epoch": 71.79389312977099,
+      "grad_norm": 0.3158397078514099,
+      "learning_rate": 3.34578089373892e-06,
+      "loss": 0.0723,
+      "step": 18810
+    },
+    {
+      "epoch": 71.83206106870229,
+      "grad_norm": 0.32471826672554016,
+      "learning_rate": 3.3405807147105814e-06,
+      "loss": 0.0667,
+      "step": 18820
+    },
+    {
+      "epoch": 71.87022900763358,
+      "grad_norm": 0.5005679130554199,
+      "learning_rate": 3.3353825520331466e-06,
+      "loss": 0.0684,
+      "step": 18830
+    },
+    {
+      "epoch": 71.90839694656489,
+      "grad_norm": 0.9290713667869568,
+      "learning_rate": 3.330186412022876e-06,
+      "loss": 0.0772,
+      "step": 18840
+    },
+    {
+      "epoch": 71.94656488549619,
+      "grad_norm": 0.9230539202690125,
+      "learning_rate": 3.324992300993568e-06,
+      "loss": 0.0702,
+      "step": 18850
+    },
+    {
+      "epoch": 71.98473282442748,
+      "grad_norm": 0.2002493143081665,
+      "learning_rate": 3.3198002252565564e-06,
+      "loss": 0.0724,
+      "step": 18860
+    },
+    {
+      "epoch": 72.02290076335878,
+      "grad_norm": 0.7362732887268066,
+      "learning_rate": 3.3146101911207024e-06,
+      "loss": 0.0693,
+      "step": 18870
+    },
+    {
+      "epoch": 72.06106870229007,
+      "grad_norm": 0.6342445611953735,
+      "learning_rate": 3.3094222048923895e-06,
+      "loss": 0.0692,
+      "step": 18880
+    },
+    {
+      "epoch": 72.09923664122137,
+      "grad_norm": 0.997917652130127,
+      "learning_rate": 3.3042362728755084e-06,
+      "loss": 0.0727,
+      "step": 18890
+    },
+    {
+      "epoch": 72.13740458015268,
+      "grad_norm": 0.3136437237262726,
+      "learning_rate": 3.2990524013714565e-06,
+      "loss": 0.0626,
+      "step": 18900
+    },
+    {
+      "epoch": 72.17557251908397,
+      "grad_norm": 0.5527306795120239,
+      "learning_rate": 3.293870596679125e-06,
+      "loss": 0.067,
+      "step": 18910
+    },
+    {
+      "epoch": 72.21374045801527,
+      "grad_norm": 0.3831974267959595,
+      "learning_rate": 3.288690865094895e-06,
+      "loss": 0.0655,
+      "step": 18920
+    },
+    {
+      "epoch": 72.25190839694656,
+      "grad_norm": 0.297125905752182,
+      "learning_rate": 3.283513212912632e-06,
+      "loss": 0.0747,
+      "step": 18930
+    },
+    {
+      "epoch": 72.29007633587786,
+      "grad_norm": 0.30272427201271057,
+      "learning_rate": 3.278337646423669e-06,
+      "loss": 0.0821,
+      "step": 18940
+    },
+    {
+      "epoch": 72.32824427480917,
+      "grad_norm": 0.4911503493785858,
+      "learning_rate": 3.273164171916806e-06,
+      "loss": 0.0715,
+      "step": 18950
+    },
+    {
+      "epoch": 72.36641221374046,
+      "grad_norm": 0.8124505281448364,
+      "learning_rate": 3.267992795678306e-06,
+      "loss": 0.0676,
+      "step": 18960
+    },
+    {
+      "epoch": 72.40458015267176,
+      "grad_norm": 0.2938520610332489,
+      "learning_rate": 3.262823523991875e-06,
+      "loss": 0.0754,
+      "step": 18970
+    },
+    {
+      "epoch": 72.44274809160305,
+      "grad_norm": 0.26553019881248474,
+      "learning_rate": 3.2576563631386694e-06,
+      "loss": 0.0733,
+      "step": 18980
+    },
+    {
+      "epoch": 72.48091603053435,
+      "grad_norm": 0.3607490062713623,
+      "learning_rate": 3.2524913193972747e-06,
+      "loss": 0.074,
+      "step": 18990
+    },
+    {
+      "epoch": 72.51908396946565,
+      "grad_norm": 0.5048254132270813,
+      "learning_rate": 3.247328399043706e-06,
+      "loss": 0.0707,
+      "step": 19000
+    },
+    {
+      "epoch": 72.55725190839695,
+      "grad_norm": 0.3066476583480835,
+      "learning_rate": 3.2421676083513987e-06,
+      "loss": 0.0728,
+      "step": 19010
+    },
+    {
+      "epoch": 72.59541984732824,
+      "grad_norm": 0.2978527843952179,
+      "learning_rate": 3.2370089535911988e-06,
+      "loss": 0.0721,
+      "step": 19020
+    },
+    {
+      "epoch": 72.63358778625954,
+      "grad_norm": 0.2315007746219635,
+      "learning_rate": 3.2318524410313602e-06,
+      "loss": 0.0722,
+      "step": 19030
+    },
+    {
+      "epoch": 72.67175572519083,
+      "grad_norm": 0.5974384546279907,
+      "learning_rate": 3.22669807693753e-06,
+      "loss": 0.0716,
+      "step": 19040
+    },
+    {
+      "epoch": 72.70992366412214,
+      "grad_norm": 0.2158258557319641,
+      "learning_rate": 3.2215458675727497e-06,
+      "loss": 0.0687,
+      "step": 19050
+    },
+    {
+      "epoch": 72.74809160305344,
+      "grad_norm": 0.23273561894893646,
+      "learning_rate": 3.2163958191974375e-06,
+      "loss": 0.0676,
+      "step": 19060
+    },
+    {
+      "epoch": 72.78625954198473,
+      "grad_norm": 0.9219455122947693,
+      "learning_rate": 3.211247938069387e-06,
+      "loss": 0.0708,
+      "step": 19070
+    },
+    {
+      "epoch": 72.82442748091603,
+      "grad_norm": 0.949112057685852,
+      "learning_rate": 3.2061022304437596e-06,
+      "loss": 0.0779,
+      "step": 19080
+    },
+    {
+      "epoch": 72.86259541984732,
+      "grad_norm": 0.3178614675998688,
+      "learning_rate": 3.2009587025730765e-06,
+      "loss": 0.0762,
+      "step": 19090
+    },
+    {
+      "epoch": 72.90076335877862,
+      "grad_norm": 0.3184674084186554,
+      "learning_rate": 3.1958173607072075e-06,
+      "loss": 0.071,
+      "step": 19100
+    },
+    {
+      "epoch": 72.93893129770993,
+      "grad_norm": 0.6181210875511169,
+      "learning_rate": 3.1906782110933698e-06,
+      "loss": 0.0842,
+      "step": 19110
+    },
+    {
+      "epoch": 72.97709923664122,
+      "grad_norm": 0.4641569256782532,
+      "learning_rate": 3.1855412599761137e-06,
+      "loss": 0.0761,
+      "step": 19120
+    },
+    {
+      "epoch": 73.01526717557252,
+      "grad_norm": 0.8154666423797607,
+      "learning_rate": 3.1804065135973165e-06,
+      "loss": 0.0765,
+      "step": 19130
+    },
+    {
+      "epoch": 73.05343511450381,
+      "grad_norm": 0.4169461131095886,
+      "learning_rate": 3.175273978196184e-06,
+      "loss": 0.0796,
+      "step": 19140
+    },
+    {
+      "epoch": 73.09160305343511,
+      "grad_norm": 0.3701060712337494,
+      "learning_rate": 3.1701436600092283e-06,
+      "loss": 0.0755,
+      "step": 19150
+    },
+    {
+      "epoch": 73.12977099236642,
+      "grad_norm": 0.49110186100006104,
+      "learning_rate": 3.16501556527027e-06,
+      "loss": 0.0661,
+      "step": 19160
+    },
+    {
+      "epoch": 73.16793893129771,
+      "grad_norm": 0.2956679165363312,
+      "learning_rate": 3.1598897002104266e-06,
+      "loss": 0.0703,
+      "step": 19170
+    },
+    {
+      "epoch": 73.20610687022901,
+      "grad_norm": 0.35779961943626404,
+      "learning_rate": 3.1547660710581087e-06,
+      "loss": 0.074,
+      "step": 19180
+    },
+    {
+      "epoch": 73.2442748091603,
+      "grad_norm": 0.3238672614097595,
+      "learning_rate": 3.149644684039008e-06,
+      "loss": 0.0714,
+      "step": 19190
+    },
+    {
+      "epoch": 73.2824427480916,
+      "grad_norm": 0.3601135015487671,
+      "learning_rate": 3.144525545376095e-06,
+      "loss": 0.0733,
+      "step": 19200
+    },
+    {
+      "epoch": 73.3206106870229,
+      "grad_norm": 0.3360394537448883,
+      "learning_rate": 3.1394086612896035e-06,
+      "loss": 0.0708,
+      "step": 19210
+    },
+    {
+      "epoch": 73.3587786259542,
+      "grad_norm": 0.6372358202934265,
+      "learning_rate": 3.1342940379970315e-06,
+      "loss": 0.0712,
+      "step": 19220
+    },
+    {
+      "epoch": 73.3969465648855,
+      "grad_norm": 0.44446098804473877,
+      "learning_rate": 3.129181681713127e-06,
+      "loss": 0.0729,
+      "step": 19230
+    },
+    {
+      "epoch": 73.43511450381679,
+      "grad_norm": 0.2544264793395996,
+      "learning_rate": 3.1240715986498856e-06,
+      "loss": 0.0775,
+      "step": 19240
+    },
+    {
+      "epoch": 73.47328244274809,
+      "grad_norm": 0.30455854535102844,
+      "learning_rate": 3.1189637950165398e-06,
+      "loss": 0.0663,
+      "step": 19250
+    },
+    {
+      "epoch": 73.5114503816794,
+      "grad_norm": 0.28932657837867737,
+      "learning_rate": 3.1138582770195547e-06,
+      "loss": 0.0689,
+      "step": 19260
+    },
+    {
+      "epoch": 73.54961832061069,
+      "grad_norm": 0.40639564394950867,
+      "learning_rate": 3.1087550508626145e-06,
+      "loss": 0.0699,
+      "step": 19270
+    },
+    {
+      "epoch": 73.58778625954199,
+      "grad_norm": 0.3635731041431427,
+      "learning_rate": 3.1036541227466204e-06,
+      "loss": 0.0744,
+      "step": 19280
+    },
+    {
+      "epoch": 73.62595419847328,
+      "grad_norm": 0.25109219551086426,
+      "learning_rate": 3.098555498869679e-06,
+      "loss": 0.0765,
+      "step": 19290
+    },
+    {
+      "epoch": 73.66412213740458,
+      "grad_norm": 0.3934338390827179,
+      "learning_rate": 3.093459185427102e-06,
+      "loss": 0.0739,
+      "step": 19300
+    },
+    {
+      "epoch": 73.70229007633588,
+      "grad_norm": 0.2998042106628418,
+      "learning_rate": 3.088365188611391e-06,
+      "loss": 0.0724,
+      "step": 19310
+    },
+    {
+      "epoch": 73.74045801526718,
+      "grad_norm": 0.36657044291496277,
+      "learning_rate": 3.0832735146122295e-06,
+      "loss": 0.0676,
+      "step": 19320
+    },
+    {
+      "epoch": 73.77862595419847,
+      "grad_norm": 0.6263729929924011,
+      "learning_rate": 3.078184169616485e-06,
+      "loss": 0.0705,
+      "step": 19330
+    },
+    {
+      "epoch": 73.81679389312977,
+      "grad_norm": 0.6651374101638794,
+      "learning_rate": 3.073097159808187e-06,
+      "loss": 0.0761,
+      "step": 19340
+    },
+    {
+      "epoch": 73.85496183206106,
+      "grad_norm": 0.33448323607444763,
+      "learning_rate": 3.068012491368537e-06,
+      "loss": 0.0698,
+      "step": 19350
+    },
+    {
+      "epoch": 73.89312977099236,
+      "grad_norm": 0.2528843879699707,
+      "learning_rate": 3.0629301704758846e-06,
+      "loss": 0.0692,
+      "step": 19360
+    },
+    {
+      "epoch": 73.93129770992367,
+      "grad_norm": 0.36234578490257263,
+      "learning_rate": 3.0578502033057288e-06,
+      "loss": 0.0729,
+      "step": 19370
+    },
+    {
+      "epoch": 73.96946564885496,
+      "grad_norm": 0.2573143243789673,
+      "learning_rate": 3.0527725960307083e-06,
+      "loss": 0.0715,
+      "step": 19380
+    },
+    {
+      "epoch": 74.00763358778626,
+      "grad_norm": 0.3500208556652069,
+      "learning_rate": 3.0476973548205945e-06,
+      "loss": 0.0688,
+      "step": 19390
+    },
+    {
+      "epoch": 74.04580152671755,
+      "grad_norm": 0.3336540460586548,
+      "learning_rate": 3.042624485842285e-06,
+      "loss": 0.0645,
+      "step": 19400
+    },
+    {
+      "epoch": 74.08396946564885,
+      "grad_norm": 0.2523120641708374,
+      "learning_rate": 3.0375539952597943e-06,
+      "loss": 0.0718,
+      "step": 19410
+    },
+    {
+      "epoch": 74.12213740458016,
+      "grad_norm": 0.38191351294517517,
+      "learning_rate": 3.0324858892342467e-06,
+      "loss": 0.0698,
+      "step": 19420
+    },
+    {
+      "epoch": 74.16030534351145,
+      "grad_norm": 0.2716592252254486,
+      "learning_rate": 3.027420173923867e-06,
+      "loss": 0.0686,
+      "step": 19430
+    },
+    {
+      "epoch": 74.19847328244275,
+      "grad_norm": 0.4997287094593048,
+      "learning_rate": 3.022356855483979e-06,
+      "loss": 0.0691,
+      "step": 19440
+    },
+    {
+      "epoch": 74.23664122137404,
+      "grad_norm": 0.2641426622867584,
+      "learning_rate": 3.017295940066989e-06,
+      "loss": 0.0725,
+      "step": 19450
+    },
+    {
+      "epoch": 74.27480916030534,
+      "grad_norm": 0.23110733926296234,
+      "learning_rate": 3.0122374338223905e-06,
+      "loss": 0.0732,
+      "step": 19460
+    },
+    {
+      "epoch": 74.31297709923665,
+      "grad_norm": 0.31251829862594604,
+      "learning_rate": 3.007181342896743e-06,
+      "loss": 0.068,
+      "step": 19470
+    },
+    {
+      "epoch": 74.35114503816794,
+      "grad_norm": 0.39099785685539246,
+      "learning_rate": 3.0021276734336744e-06,
+      "loss": 0.0712,
+      "step": 19480
+    },
+    {
+      "epoch": 74.38931297709924,
+      "grad_norm": 0.26635631918907166,
+      "learning_rate": 2.997076431573871e-06,
+      "loss": 0.0722,
+      "step": 19490
+    },
+    {
+      "epoch": 74.42748091603053,
+      "grad_norm": 0.3487379848957062,
+      "learning_rate": 2.9920276234550636e-06,
+      "loss": 0.0711,
+      "step": 19500
+    },
+    {
+      "epoch": 74.46564885496183,
+      "grad_norm": 0.33060070872306824,
+      "learning_rate": 2.9869812552120355e-06,
+      "loss": 0.0738,
+      "step": 19510
+    },
+    {
+      "epoch": 74.50381679389314,
+      "grad_norm": 0.3587673604488373,
+      "learning_rate": 2.9819373329765977e-06,
+      "loss": 0.0679,
+      "step": 19520
+    },
+    {
+      "epoch": 74.54198473282443,
+      "grad_norm": 0.33571016788482666,
+      "learning_rate": 2.97689586287759e-06,
+      "loss": 0.0735,
+      "step": 19530
+    },
+    {
+      "epoch": 74.58015267175573,
+      "grad_norm": 0.5554803609848022,
+      "learning_rate": 2.9718568510408763e-06,
+      "loss": 0.0672,
+      "step": 19540
+    },
+    {
+      "epoch": 74.61832061068702,
+      "grad_norm": 0.35926058888435364,
+      "learning_rate": 2.966820303589327e-06,
+      "loss": 0.0747,
+      "step": 19550
+    },
+    {
+      "epoch": 74.65648854961832,
+      "grad_norm": 0.2982501685619354,
+      "learning_rate": 2.961786226642829e-06,
+      "loss": 0.0716,
+      "step": 19560
+    },
+    {
+      "epoch": 74.69465648854961,
+      "grad_norm": 0.35031527280807495,
+      "learning_rate": 2.9567546263182554e-06,
+      "loss": 0.0742,
+      "step": 19570
+    },
+    {
+      "epoch": 74.73282442748092,
+      "grad_norm": 0.6128705143928528,
+      "learning_rate": 2.951725508729476e-06,
+      "loss": 0.0731,
+      "step": 19580
+    },
+    {
+      "epoch": 74.77099236641222,
+      "grad_norm": 0.2246423214673996,
+      "learning_rate": 2.9466988799873443e-06,
+      "loss": 0.0675,
+      "step": 19590
+    },
+    {
+      "epoch": 74.80916030534351,
+      "grad_norm": 0.2989789843559265,
+      "learning_rate": 2.9416747461996853e-06,
+      "loss": 0.0762,
+      "step": 19600
+    },
+    {
+      "epoch": 74.8473282442748,
+      "grad_norm": 0.34754475951194763,
+      "learning_rate": 2.9366531134712974e-06,
+      "loss": 0.0699,
+      "step": 19610
+    },
+    {
+      "epoch": 74.8854961832061,
+      "grad_norm": 0.37204280495643616,
+      "learning_rate": 2.931633987903937e-06,
+      "loss": 0.0711,
+      "step": 19620
+    },
+    {
+      "epoch": 74.92366412213741,
+      "grad_norm": 0.5289718508720398,
+      "learning_rate": 2.926617375596317e-06,
+      "loss": 0.0818,
+      "step": 19630
+    },
+    {
+      "epoch": 74.9618320610687,
+      "grad_norm": 0.3462826907634735,
+      "learning_rate": 2.9216032826440927e-06,
+      "loss": 0.0702,
+      "step": 19640
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 0.29999420046806335,
+      "learning_rate": 2.9165917151398594e-06,
+      "loss": 0.072,
+      "step": 19650
+    },
+    {
+      "epoch": 75.0381679389313,
+      "grad_norm": 0.28326013684272766,
+      "learning_rate": 2.9115826791731426e-06,
+      "loss": 0.0695,
+      "step": 19660
+    },
+    {
+      "epoch": 75.07633587786259,
+      "grad_norm": 0.7487019300460815,
+      "learning_rate": 2.9065761808303983e-06,
+      "loss": 0.0744,
+      "step": 19670
+    },
+    {
+      "epoch": 75.1145038167939,
+      "grad_norm": 0.24250933527946472,
+      "learning_rate": 2.9015722261949918e-06,
+      "loss": 0.0716,
+      "step": 19680
+    },
+    {
+      "epoch": 75.1526717557252,
+      "grad_norm": 0.272839218378067,
+      "learning_rate": 2.8965708213471987e-06,
+      "loss": 0.0738,
+      "step": 19690
+    },
+    {
+      "epoch": 75.19083969465649,
+      "grad_norm": 0.26355302333831787,
+      "learning_rate": 2.891571972364198e-06,
+      "loss": 0.0717,
+      "step": 19700
+    },
+    {
+      "epoch": 75.22900763358778,
+      "grad_norm": 0.28697460889816284,
+      "learning_rate": 2.8865756853200605e-06,
+      "loss": 0.0708,
+      "step": 19710
+    },
+    {
+      "epoch": 75.26717557251908,
+      "grad_norm": 0.5366164445877075,
+      "learning_rate": 2.8815819662857505e-06,
+      "loss": 0.0696,
+      "step": 19720
+    },
+    {
+      "epoch": 75.30534351145039,
+      "grad_norm": 0.5985289812088013,
+      "learning_rate": 2.876590821329105e-06,
+      "loss": 0.0767,
+      "step": 19730
+    },
+    {
+      "epoch": 75.34351145038168,
+      "grad_norm": 1.137639045715332,
+      "learning_rate": 2.8716022565148362e-06,
+      "loss": 0.0733,
+      "step": 19740
+    },
+    {
+      "epoch": 75.38167938931298,
+      "grad_norm": 0.27952316403388977,
+      "learning_rate": 2.8666162779045205e-06,
+      "loss": 0.0733,
+      "step": 19750
+    },
+    {
+      "epoch": 75.41984732824427,
+      "grad_norm": 1.1592620611190796,
+      "learning_rate": 2.8616328915565907e-06,
+      "loss": 0.0744,
+      "step": 19760
+    },
+    {
+      "epoch": 75.45801526717557,
+      "grad_norm": 0.5299224257469177,
+      "learning_rate": 2.856652103526334e-06,
+      "loss": 0.0706,
+      "step": 19770
+    },
+    {
+      "epoch": 75.49618320610686,
+      "grad_norm": 0.2791510224342346,
+      "learning_rate": 2.8516739198658753e-06,
+      "loss": 0.0666,
+      "step": 19780
+    },
+    {
+      "epoch": 75.53435114503817,
+      "grad_norm": 0.32924363017082214,
+      "learning_rate": 2.8466983466241772e-06,
+      "loss": 0.0777,
+      "step": 19790
+    },
+    {
+      "epoch": 75.57251908396947,
+      "grad_norm": 0.4635617434978485,
+      "learning_rate": 2.841725389847032e-06,
+      "loss": 0.0685,
+      "step": 19800
+    },
+    {
+      "epoch": 75.61068702290076,
+      "grad_norm": 0.42956969141960144,
+      "learning_rate": 2.8367550555770507e-06,
+      "loss": 0.0747,
+      "step": 19810
+    },
+    {
+      "epoch": 75.64885496183206,
+      "grad_norm": 0.3257053792476654,
+      "learning_rate": 2.8317873498536554e-06,
+      "loss": 0.0705,
+      "step": 19820
+    },
+    {
+      "epoch": 75.68702290076335,
+      "grad_norm": 0.3546193540096283,
+      "learning_rate": 2.8268222787130805e-06,
+      "loss": 0.0747,
+      "step": 19830
+    },
+    {
+      "epoch": 75.72519083969466,
+      "grad_norm": 0.23434428870677948,
+      "learning_rate": 2.8218598481883552e-06,
+      "loss": 0.0728,
+      "step": 19840
+    },
+    {
+      "epoch": 75.76335877862596,
+      "grad_norm": 0.3509843647480011,
+      "learning_rate": 2.816900064309299e-06,
+      "loss": 0.0685,
+      "step": 19850
+    },
+    {
+      "epoch": 75.80152671755725,
+      "grad_norm": 0.350881963968277,
+      "learning_rate": 2.811942933102517e-06,
+      "loss": 0.0668,
+      "step": 19860
+    },
+    {
+      "epoch": 75.83969465648855,
+      "grad_norm": 0.2302703559398651,
+      "learning_rate": 2.8069884605913912e-06,
+      "loss": 0.0679,
+      "step": 19870
+    },
+    {
+      "epoch": 75.87786259541984,
+      "grad_norm": 0.39022761583328247,
+      "learning_rate": 2.802036652796074e-06,
+      "loss": 0.0718,
+      "step": 19880
+    },
+    {
+      "epoch": 75.91603053435115,
+      "grad_norm": 0.536963939666748,
+      "learning_rate": 2.797087515733478e-06,
+      "loss": 0.0688,
+      "step": 19890
+    },
+    {
+      "epoch": 75.95419847328245,
+      "grad_norm": 0.33239343762397766,
+      "learning_rate": 2.7921410554172724e-06,
+      "loss": 0.0714,
+      "step": 19900
+    },
+    {
+      "epoch": 75.99236641221374,
+      "grad_norm": 0.2767488658428192,
+      "learning_rate": 2.787197277857871e-06,
+      "loss": 0.0702,
+      "step": 19910
+    },
+    {
+      "epoch": 76.03053435114504,
+      "grad_norm": 0.30666476488113403,
+      "learning_rate": 2.7822561890624287e-06,
+      "loss": 0.0674,
+      "step": 19920
+    },
+    {
+      "epoch": 76.06870229007633,
+      "grad_norm": 0.2637414038181305,
+      "learning_rate": 2.777317795034839e-06,
+      "loss": 0.0684,
+      "step": 19930
+    },
+    {
+      "epoch": 76.10687022900764,
+      "grad_norm": 0.23923492431640625,
+      "learning_rate": 2.772382101775711e-06,
+      "loss": 0.0719,
+      "step": 19940
+    },
+    {
+      "epoch": 76.14503816793894,
+      "grad_norm": 0.8826484084129333,
+      "learning_rate": 2.7674491152823825e-06,
+      "loss": 0.0727,
+      "step": 19950
+    },
+    {
+      "epoch": 76.18320610687023,
+      "grad_norm": 0.6208845973014832,
+      "learning_rate": 2.7625188415488946e-06,
+      "loss": 0.0715,
+      "step": 19960
+    },
+    {
+      "epoch": 76.22137404580153,
+      "grad_norm": 0.3055213987827301,
+      "learning_rate": 2.7575912865659925e-06,
+      "loss": 0.0719,
+      "step": 19970
+    },
+    {
+      "epoch": 76.25954198473282,
+      "grad_norm": 0.3592512011528015,
+      "learning_rate": 2.752666456321125e-06,
+      "loss": 0.0659,
+      "step": 19980
+    },
+    {
+      "epoch": 76.29770992366412,
+      "grad_norm": 0.38851675391197205,
+      "learning_rate": 2.7477443567984225e-06,
+      "loss": 0.0701,
+      "step": 19990
+    },
+    {
+      "epoch": 76.33587786259542,
+      "grad_norm": 0.6826211810112,
+      "learning_rate": 2.7428249939787e-06,
+      "loss": 0.0745,
+      "step": 20000
+    },
+    {
+      "epoch": 76.37404580152672,
+      "grad_norm": 0.26521873474121094,
+      "learning_rate": 2.7379083738394485e-06,
+      "loss": 0.0772,
+      "step": 20010
+    },
+    {
+      "epoch": 76.41221374045801,
+      "grad_norm": 0.5465177893638611,
+      "learning_rate": 2.732994502354823e-06,
+      "loss": 0.0699,
+      "step": 20020
+    },
+    {
+      "epoch": 76.45038167938931,
+      "grad_norm": 0.22223325073719025,
+      "learning_rate": 2.72808338549564e-06,
+      "loss": 0.0703,
+      "step": 20030
+    },
+    {
+      "epoch": 76.4885496183206,
+      "grad_norm": 0.3529285788536072,
+      "learning_rate": 2.723175029229374e-06,
+      "loss": 0.0709,
+      "step": 20040
+    },
+    {
+      "epoch": 76.52671755725191,
+      "grad_norm": 0.686523973941803,
+      "learning_rate": 2.718269439520138e-06,
+      "loss": 0.0706,
+      "step": 20050
+    },
+    {
+      "epoch": 76.56488549618321,
+      "grad_norm": 0.31421926617622375,
+      "learning_rate": 2.713366622328686e-06,
+      "loss": 0.0698,
+      "step": 20060
+    },
+    {
+      "epoch": 76.6030534351145,
+      "grad_norm": 0.5857685804367065,
+      "learning_rate": 2.7084665836124006e-06,
+      "loss": 0.0771,
+      "step": 20070
+    },
+    {
+      "epoch": 76.6412213740458,
+      "grad_norm": 0.3716685473918915,
+      "learning_rate": 2.703569329325296e-06,
+      "loss": 0.0697,
+      "step": 20080
+    },
+    {
+      "epoch": 76.6793893129771,
+      "grad_norm": 0.3061921298503876,
+      "learning_rate": 2.698674865417994e-06,
+      "loss": 0.0723,
+      "step": 20090
+    },
+    {
+      "epoch": 76.7175572519084,
+      "grad_norm": 0.23903951048851013,
+      "learning_rate": 2.693783197837733e-06,
+      "loss": 0.0714,
+      "step": 20100
+    },
+    {
+      "epoch": 76.7557251908397,
+      "grad_norm": 0.22559994459152222,
+      "learning_rate": 2.6888943325283482e-06,
+      "loss": 0.0774,
+      "step": 20110
+    },
+    {
+      "epoch": 76.79389312977099,
+      "grad_norm": 0.48428580164909363,
+      "learning_rate": 2.6840082754302734e-06,
+      "loss": 0.0686,
+      "step": 20120
+    },
+    {
+      "epoch": 76.83206106870229,
+      "grad_norm": 0.3806080222129822,
+      "learning_rate": 2.6791250324805252e-06,
+      "loss": 0.0689,
+      "step": 20130
+    },
+    {
+      "epoch": 76.87022900763358,
+      "grad_norm": 0.49289458990097046,
+      "learning_rate": 2.6742446096127086e-06,
+      "loss": 0.068,
+      "step": 20140
+    },
+    {
+      "epoch": 76.90839694656489,
+      "grad_norm": 0.5932338833808899,
+      "learning_rate": 2.669367012756996e-06,
+      "loss": 0.068,
+      "step": 20150
+    },
+    {
+      "epoch": 76.94656488549619,
+      "grad_norm": 0.2798921465873718,
+      "learning_rate": 2.664492247840127e-06,
+      "loss": 0.0731,
+      "step": 20160
+    },
+    {
+      "epoch": 76.98473282442748,
+      "grad_norm": 0.25367864966392517,
+      "learning_rate": 2.6596203207854006e-06,
+      "loss": 0.0678,
+      "step": 20170
+    },
+    {
+      "epoch": 77.02290076335878,
+      "grad_norm": 0.28277409076690674,
+      "learning_rate": 2.654751237512666e-06,
+      "loss": 0.0713,
+      "step": 20180
+    },
+    {
+      "epoch": 77.06106870229007,
+      "grad_norm": 0.3937804102897644,
+      "learning_rate": 2.649885003938323e-06,
+      "loss": 0.0698,
+      "step": 20190
+    },
+    {
+      "epoch": 77.09923664122137,
+      "grad_norm": 0.35799962282180786,
+      "learning_rate": 2.6450216259753005e-06,
+      "loss": 0.0761,
+      "step": 20200
+    },
+    {
+      "epoch": 77.13740458015268,
+      "grad_norm": 0.26287925243377686,
+      "learning_rate": 2.6401611095330632e-06,
+      "loss": 0.0654,
+      "step": 20210
+    },
+    {
+      "epoch": 77.17557251908397,
+      "grad_norm": 0.46736228466033936,
+      "learning_rate": 2.6353034605175937e-06,
+      "loss": 0.0715,
+      "step": 20220
+    },
+    {
+      "epoch": 77.21374045801527,
+      "grad_norm": 0.46314504742622375,
+      "learning_rate": 2.6304486848313982e-06,
+      "loss": 0.0633,
+      "step": 20230
+    },
+    {
+      "epoch": 77.25190839694656,
+      "grad_norm": 0.5143646597862244,
+      "learning_rate": 2.6255967883734823e-06,
+      "loss": 0.0706,
+      "step": 20240
+    },
+    {
+      "epoch": 77.29007633587786,
+      "grad_norm": 0.2608904242515564,
+      "learning_rate": 2.620747777039363e-06,
+      "loss": 0.068,
+      "step": 20250
+    },
+    {
+      "epoch": 77.32824427480917,
+      "grad_norm": 0.8567964434623718,
+      "learning_rate": 2.6159016567210426e-06,
+      "loss": 0.086,
+      "step": 20260
+    },
+    {
+      "epoch": 77.36641221374046,
+      "grad_norm": 0.3883376717567444,
+      "learning_rate": 2.6110584333070153e-06,
+      "loss": 0.075,
+      "step": 20270
+    },
+    {
+      "epoch": 77.40458015267176,
+      "grad_norm": 0.27981120347976685,
+      "learning_rate": 2.606218112682254e-06,
+      "loss": 0.0666,
+      "step": 20280
+    },
+    {
+      "epoch": 77.44274809160305,
+      "grad_norm": 0.2423625886440277,
+      "learning_rate": 2.601380700728203e-06,
+      "loss": 0.0745,
+      "step": 20290
+    },
+    {
+      "epoch": 77.48091603053435,
+      "grad_norm": 0.23799222707748413,
+      "learning_rate": 2.596546203322777e-06,
+      "loss": 0.0709,
+      "step": 20300
+    },
+    {
+      "epoch": 77.51908396946565,
+      "grad_norm": 0.2332644760608673,
+      "learning_rate": 2.591714626340346e-06,
+      "loss": 0.0746,
+      "step": 20310
+    },
+    {
+      "epoch": 77.55725190839695,
+      "grad_norm": 0.29361745715141296,
+      "learning_rate": 2.5868859756517294e-06,
+      "loss": 0.0765,
+      "step": 20320
+    },
+    {
+      "epoch": 77.59541984732824,
+      "grad_norm": 0.28440865874290466,
+      "learning_rate": 2.582060257124195e-06,
+      "loss": 0.068,
+      "step": 20330
+    },
+    {
+      "epoch": 77.63358778625954,
+      "grad_norm": 0.3904019296169281,
+      "learning_rate": 2.577237476621442e-06,
+      "loss": 0.072,
+      "step": 20340
+    },
+    {
+      "epoch": 77.67175572519083,
+      "grad_norm": 0.833919882774353,
+      "learning_rate": 2.5724176400036094e-06,
+      "loss": 0.0713,
+      "step": 20350
+    },
+    {
+      "epoch": 77.70992366412214,
+      "grad_norm": 0.2217206060886383,
+      "learning_rate": 2.5676007531272475e-06,
+      "loss": 0.0812,
+      "step": 20360
+    },
+    {
+      "epoch": 77.74809160305344,
+      "grad_norm": 0.307720810174942,
+      "learning_rate": 2.562786821845333e-06,
+      "loss": 0.072,
+      "step": 20370
+    },
+    {
+      "epoch": 77.78625954198473,
+      "grad_norm": 0.28931522369384766,
+      "learning_rate": 2.5579758520072446e-06,
+      "loss": 0.0757,
+      "step": 20380
+    },
+    {
+      "epoch": 77.82442748091603,
+      "grad_norm": 0.39654994010925293,
+      "learning_rate": 2.5531678494587612e-06,
+      "loss": 0.0693,
+      "step": 20390
+    },
+    {
+      "epoch": 77.86259541984732,
+      "grad_norm": 0.55362868309021,
+      "learning_rate": 2.5483628200420648e-06,
+      "loss": 0.07,
+      "step": 20400
+    },
+    {
+      "epoch": 77.90076335877862,
+      "grad_norm": 0.5507999658584595,
+      "learning_rate": 2.5435607695957153e-06,
+      "loss": 0.0727,
+      "step": 20410
+    },
+    {
+      "epoch": 77.93893129770993,
+      "grad_norm": 0.36570632457733154,
+      "learning_rate": 2.5387617039546585e-06,
+      "loss": 0.0741,
+      "step": 20420
+    },
+    {
+      "epoch": 77.97709923664122,
+      "grad_norm": 0.589455246925354,
+      "learning_rate": 2.5339656289502105e-06,
+      "loss": 0.0753,
+      "step": 20430
+    },
+    {
+      "epoch": 78.01526717557252,
+      "grad_norm": 0.40846219658851624,
+      "learning_rate": 2.5291725504100563e-06,
+      "loss": 0.0695,
+      "step": 20440
+    },
+    {
+      "epoch": 78.05343511450381,
+      "grad_norm": 0.28900301456451416,
+      "learning_rate": 2.524382474158234e-06,
+      "loss": 0.0691,
+      "step": 20450
+    },
+    {
+      "epoch": 78.09160305343511,
+      "grad_norm": 0.33728376030921936,
+      "learning_rate": 2.5195954060151433e-06,
+      "loss": 0.0755,
+      "step": 20460
+    },
+    {
+      "epoch": 78.12977099236642,
+      "grad_norm": 0.25033625960350037,
+      "learning_rate": 2.5148113517975216e-06,
+      "loss": 0.0724,
+      "step": 20470
+    },
+    {
+      "epoch": 78.16793893129771,
+      "grad_norm": 0.8060714602470398,
+      "learning_rate": 2.510030317318445e-06,
+      "loss": 0.0751,
+      "step": 20480
+    },
+    {
+      "epoch": 78.20610687022901,
+      "grad_norm": 0.5355228781700134,
+      "learning_rate": 2.50525230838732e-06,
+      "loss": 0.0735,
+      "step": 20490
+    },
+    {
+      "epoch": 78.2442748091603,
+      "grad_norm": 0.28783127665519714,
+      "learning_rate": 2.5004773308098814e-06,
+      "loss": 0.0664,
+      "step": 20500
+    },
+    {
+      "epoch": 78.2824427480916,
+      "grad_norm": 0.32949912548065186,
+      "learning_rate": 2.4957053903881736e-06,
+      "loss": 0.0654,
+      "step": 20510
+    },
+    {
+      "epoch": 78.3206106870229,
+      "grad_norm": 0.40475648641586304,
+      "learning_rate": 2.4909364929205575e-06,
+      "loss": 0.0726,
+      "step": 20520
+    },
+    {
+      "epoch": 78.3587786259542,
+      "grad_norm": 0.34688708186149597,
+      "learning_rate": 2.4861706442016923e-06,
+      "loss": 0.0671,
+      "step": 20530
+    },
+    {
+      "epoch": 78.3969465648855,
+      "grad_norm": 0.38922348618507385,
+      "learning_rate": 2.481407850022533e-06,
+      "loss": 0.076,
+      "step": 20540
+    },
+    {
+      "epoch": 78.43511450381679,
+      "grad_norm": 0.25327450037002563,
+      "learning_rate": 2.4766481161703216e-06,
+      "loss": 0.0729,
+      "step": 20550
+    },
+    {
+      "epoch": 78.47328244274809,
+      "grad_norm": 0.36593905091285706,
+      "learning_rate": 2.4718914484285876e-06,
+      "loss": 0.0795,
+      "step": 20560
+    },
+    {
+      "epoch": 78.5114503816794,
+      "grad_norm": 0.34967660903930664,
+      "learning_rate": 2.467137852577129e-06,
+      "loss": 0.0787,
+      "step": 20570
+    },
+    {
+      "epoch": 78.54961832061069,
+      "grad_norm": 0.272087424993515,
+      "learning_rate": 2.4623873343920123e-06,
+      "loss": 0.0722,
+      "step": 20580
+    },
+    {
+      "epoch": 78.58778625954199,
+      "grad_norm": 0.2819080352783203,
+      "learning_rate": 2.4576398996455657e-06,
+      "loss": 0.0631,
+      "step": 20590
+    },
+    {
+      "epoch": 78.62595419847328,
+      "grad_norm": 0.8356397151947021,
+      "learning_rate": 2.4528955541063683e-06,
+      "loss": 0.0724,
+      "step": 20600
+    },
+    {
+      "epoch": 78.66412213740458,
+      "grad_norm": 0.8203865885734558,
+      "learning_rate": 2.448154303539251e-06,
+      "loss": 0.0673,
+      "step": 20610
+    },
+    {
+      "epoch": 78.70229007633588,
+      "grad_norm": 1.2636505365371704,
+      "learning_rate": 2.4434161537052776e-06,
+      "loss": 0.0779,
+      "step": 20620
+    },
+    {
+      "epoch": 78.74045801526718,
+      "grad_norm": 0.42922544479370117,
+      "learning_rate": 2.4386811103617474e-06,
+      "loss": 0.0714,
+      "step": 20630
+    },
+    {
+      "epoch": 78.77862595419847,
+      "grad_norm": 0.4266103208065033,
+      "learning_rate": 2.4339491792621833e-06,
+      "loss": 0.0698,
+      "step": 20640
+    },
+    {
+      "epoch": 78.81679389312977,
+      "grad_norm": 0.2615494132041931,
+      "learning_rate": 2.4292203661563313e-06,
+      "loss": 0.0747,
+      "step": 20650
+    },
+    {
+      "epoch": 78.85496183206106,
+      "grad_norm": 0.2261299043893814,
+      "learning_rate": 2.424494676790141e-06,
+      "loss": 0.0723,
+      "step": 20660
+    },
+    {
+      "epoch": 78.89312977099236,
+      "grad_norm": 0.2378481775522232,
+      "learning_rate": 2.419772116905775e-06,
+      "loss": 0.0674,
+      "step": 20670
+    },
+    {
+      "epoch": 78.93129770992367,
+      "grad_norm": 0.27929824590682983,
+      "learning_rate": 2.4150526922415855e-06,
+      "loss": 0.0772,
+      "step": 20680
+    },
+    {
+      "epoch": 78.96946564885496,
+      "grad_norm": 0.3526342511177063,
+      "learning_rate": 2.41033640853212e-06,
+      "loss": 0.0782,
+      "step": 20690
+    },
+    {
+      "epoch": 79.00763358778626,
+      "grad_norm": 0.3187890350818634,
+      "learning_rate": 2.405623271508108e-06,
+      "loss": 0.0697,
+      "step": 20700
+    },
+    {
+      "epoch": 79.04580152671755,
+      "grad_norm": 0.38696810603141785,
+      "learning_rate": 2.4009132868964525e-06,
+      "loss": 0.0689,
+      "step": 20710
+    },
+    {
+      "epoch": 79.08396946564885,
+      "grad_norm": 0.2635570466518402,
+      "learning_rate": 2.3962064604202327e-06,
+      "loss": 0.0675,
+      "step": 20720
+    },
+    {
+      "epoch": 79.12213740458016,
+      "grad_norm": 0.36084693670272827,
+      "learning_rate": 2.391502797798686e-06,
+      "loss": 0.0697,
+      "step": 20730
+    },
+    {
+      "epoch": 79.16030534351145,
+      "grad_norm": 0.38881757855415344,
+      "learning_rate": 2.386802304747205e-06,
+      "loss": 0.0666,
+      "step": 20740
+    },
+    {
+      "epoch": 79.19847328244275,
+      "grad_norm": 0.4907824397087097,
+      "learning_rate": 2.382104986977332e-06,
+      "loss": 0.0668,
+      "step": 20750
+    },
+    {
+      "epoch": 79.23664122137404,
+      "grad_norm": 0.2782943844795227,
+      "learning_rate": 2.3774108501967493e-06,
+      "loss": 0.0716,
+      "step": 20760
+    },
+    {
+      "epoch": 79.27480916030534,
+      "grad_norm": 0.5779755711555481,
+      "learning_rate": 2.37271990010928e-06,
+      "loss": 0.073,
+      "step": 20770
+    },
+    {
+      "epoch": 79.31297709923665,
+      "grad_norm": 0.3922247886657715,
+      "learning_rate": 2.3680321424148678e-06,
+      "loss": 0.07,
+      "step": 20780
+    },
+    {
+      "epoch": 79.35114503816794,
+      "grad_norm": 0.3388228416442871,
+      "learning_rate": 2.363347582809579e-06,
+      "loss": 0.0672,
+      "step": 20790
+    },
+    {
+      "epoch": 79.38931297709924,
+      "grad_norm": 0.20889022946357727,
+      "learning_rate": 2.358666226985599e-06,
+      "loss": 0.0699,
+      "step": 20800
+    },
+    {
+      "epoch": 79.42748091603053,
+      "grad_norm": 0.5814685225486755,
+      "learning_rate": 2.3539880806312134e-06,
+      "loss": 0.0714,
+      "step": 20810
+    },
+    {
+      "epoch": 79.46564885496183,
+      "grad_norm": 0.2144891619682312,
+      "learning_rate": 2.349313149430814e-06,
+      "loss": 0.0667,
+      "step": 20820
+    },
+    {
+      "epoch": 79.50381679389314,
+      "grad_norm": 0.6816794276237488,
+      "learning_rate": 2.3446414390648815e-06,
+      "loss": 0.0745,
+      "step": 20830
+    },
+    {
+      "epoch": 79.54198473282443,
+      "grad_norm": 0.32124826312065125,
+      "learning_rate": 2.3399729552099844e-06,
+      "loss": 0.0716,
+      "step": 20840
+    },
+    {
+      "epoch": 79.58015267175573,
+      "grad_norm": 0.22228087484836578,
+      "learning_rate": 2.335307703538771e-06,
+      "loss": 0.0693,
+      "step": 20850
+    },
+    {
+      "epoch": 79.61832061068702,
+      "grad_norm": 0.29664310812950134,
+      "learning_rate": 2.330645689719962e-06,
+      "loss": 0.0702,
+      "step": 20860
+    },
+    {
+      "epoch": 79.65648854961832,
+      "grad_norm": 0.35011011362075806,
+      "learning_rate": 2.3259869194183415e-06,
+      "loss": 0.0662,
+      "step": 20870
+    },
+    {
+      "epoch": 79.69465648854961,
+      "grad_norm": 0.5317111015319824,
+      "learning_rate": 2.321331398294759e-06,
+      "loss": 0.0675,
+      "step": 20880
+    },
+    {
+      "epoch": 79.73282442748092,
+      "grad_norm": 0.7479346990585327,
+      "learning_rate": 2.3166791320061095e-06,
+      "loss": 0.0699,
+      "step": 20890
+    },
+    {
+      "epoch": 79.77099236641222,
+      "grad_norm": 0.25332704186439514,
+      "learning_rate": 2.312030126205335e-06,
+      "loss": 0.0709,
+      "step": 20900
+    },
+    {
+      "epoch": 79.80916030534351,
+      "grad_norm": 0.5418967604637146,
+      "learning_rate": 2.3073843865414163e-06,
+      "loss": 0.0714,
+      "step": 20910
+    },
+    {
+      "epoch": 79.8473282442748,
+      "grad_norm": 0.5316504240036011,
+      "learning_rate": 2.302741918659363e-06,
+      "loss": 0.0735,
+      "step": 20920
+    },
+    {
+      "epoch": 79.8854961832061,
+      "grad_norm": 0.31350257992744446,
+      "learning_rate": 2.2981027282002155e-06,
+      "loss": 0.0714,
+      "step": 20930
+    },
+    {
+      "epoch": 79.92366412213741,
+      "grad_norm": 0.47436007857322693,
+      "learning_rate": 2.2934668208010235e-06,
+      "loss": 0.071,
+      "step": 20940
+    },
+    {
+      "epoch": 79.9618320610687,
+      "grad_norm": 0.28028592467308044,
+      "learning_rate": 2.2888342020948556e-06,
+      "loss": 0.0701,
+      "step": 20950
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 0.25115305185317993,
+      "learning_rate": 2.2842048777107783e-06,
+      "loss": 0.0673,
+      "step": 20960
+    },
+    {
+      "epoch": 80.0381679389313,
+      "grad_norm": 0.22422641515731812,
+      "learning_rate": 2.2795788532738555e-06,
+      "loss": 0.0737,
+      "step": 20970
+    },
+    {
+      "epoch": 80.07633587786259,
+      "grad_norm": 0.31794339418411255,
+      "learning_rate": 2.274956134405147e-06,
+      "loss": 0.0674,
+      "step": 20980
+    },
+    {
+      "epoch": 80.1145038167939,
+      "grad_norm": 0.38259294629096985,
+      "learning_rate": 2.2703367267216896e-06,
+      "loss": 0.0634,
+      "step": 20990
+    },
+    {
+      "epoch": 80.1526717557252,
+      "grad_norm": 0.22496770322322845,
+      "learning_rate": 2.2657206358365e-06,
+      "loss": 0.0683,
+      "step": 21000
+    },
+    {
+      "epoch": 80.19083969465649,
+      "grad_norm": 0.24763353168964386,
+      "learning_rate": 2.261107867358563e-06,
+      "loss": 0.0732,
+      "step": 21010
+    },
+    {
+      "epoch": 80.22900763358778,
+      "grad_norm": 0.31791627407073975,
+      "learning_rate": 2.2564984268928264e-06,
+      "loss": 0.0807,
+      "step": 21020
+    },
+    {
+      "epoch": 80.26717557251908,
+      "grad_norm": 0.2604450583457947,
+      "learning_rate": 2.251892320040198e-06,
+      "loss": 0.069,
+      "step": 21030
+    },
+    {
+      "epoch": 80.30534351145039,
+      "grad_norm": 0.21588589251041412,
+      "learning_rate": 2.2472895523975315e-06,
+      "loss": 0.0731,
+      "step": 21040
+    },
+    {
+      "epoch": 80.34351145038168,
+      "grad_norm": 0.8700875043869019,
+      "learning_rate": 2.2426901295576215e-06,
+      "loss": 0.0801,
+      "step": 21050
+    },
+    {
+      "epoch": 80.38167938931298,
+      "grad_norm": 0.2903733551502228,
+      "learning_rate": 2.2380940571092013e-06,
+      "loss": 0.0656,
+      "step": 21060
+    },
+    {
+      "epoch": 80.41984732824427,
+      "grad_norm": 0.41912519931793213,
+      "learning_rate": 2.2335013406369303e-06,
+      "loss": 0.0693,
+      "step": 21070
+    },
+    {
+      "epoch": 80.45801526717557,
+      "grad_norm": 0.5005356669425964,
+      "learning_rate": 2.228911985721397e-06,
+      "loss": 0.0676,
+      "step": 21080
+    },
+    {
+      "epoch": 80.49618320610686,
+      "grad_norm": 0.39985841512680054,
+      "learning_rate": 2.224325997939095e-06,
+      "loss": 0.0703,
+      "step": 21090
+    },
+    {
+      "epoch": 80.53435114503817,
+      "grad_norm": 0.3365218937397003,
+      "learning_rate": 2.2197433828624372e-06,
+      "loss": 0.0669,
+      "step": 21100
+    },
+    {
+      "epoch": 80.57251908396947,
+      "grad_norm": 0.2456941455602646,
+      "learning_rate": 2.2151641460597295e-06,
+      "loss": 0.0694,
+      "step": 21110
+    },
+    {
+      "epoch": 80.61068702290076,
+      "grad_norm": 0.3186664879322052,
+      "learning_rate": 2.210588293095177e-06,
+      "loss": 0.0626,
+      "step": 21120
+    },
+    {
+      "epoch": 80.64885496183206,
+      "grad_norm": 0.3170889616012573,
+      "learning_rate": 2.2060158295288716e-06,
+      "loss": 0.0628,
+      "step": 21130
+    },
+    {
+      "epoch": 80.68702290076335,
+      "grad_norm": 0.49212417006492615,
+      "learning_rate": 2.2014467609167904e-06,
+      "loss": 0.07,
+      "step": 21140
+    },
+    {
+      "epoch": 80.72519083969466,
+      "grad_norm": 0.35683688521385193,
+      "learning_rate": 2.196881092810781e-06,
+      "loss": 0.0707,
+      "step": 21150
+    },
+    {
+      "epoch": 80.76335877862596,
+      "grad_norm": 0.20278224349021912,
+      "learning_rate": 2.192318830758561e-06,
+      "loss": 0.0693,
+      "step": 21160
+    },
+    {
+      "epoch": 80.80152671755725,
+      "grad_norm": 0.27485206723213196,
+      "learning_rate": 2.187759980303708e-06,
+      "loss": 0.0648,
+      "step": 21170
+    },
+    {
+      "epoch": 80.83969465648855,
+      "grad_norm": 0.3982328474521637,
+      "learning_rate": 2.1832045469856544e-06,
+      "loss": 0.0679,
+      "step": 21180
+    },
+    {
+      "epoch": 80.87786259541984,
+      "grad_norm": 0.7986206412315369,
+      "learning_rate": 2.178652536339684e-06,
+      "loss": 0.073,
+      "step": 21190
+    },
+    {
+      "epoch": 80.91603053435115,
+      "grad_norm": 0.45346391201019287,
+      "learning_rate": 2.1741039538969184e-06,
+      "loss": 0.0801,
+      "step": 21200
+    },
+    {
+      "epoch": 80.95419847328245,
+      "grad_norm": 0.33075547218322754,
+      "learning_rate": 2.169558805184313e-06,
+      "loss": 0.0913,
+      "step": 21210
+    },
+    {
+      "epoch": 80.99236641221374,
+      "grad_norm": 0.20161128044128418,
+      "learning_rate": 2.165017095724651e-06,
+      "loss": 0.0724,
+      "step": 21220
+    },
+    {
+      "epoch": 81.03053435114504,
+      "grad_norm": 0.4609990417957306,
+      "learning_rate": 2.1604788310365404e-06,
+      "loss": 0.0798,
+      "step": 21230
+    },
+    {
+      "epoch": 81.06870229007633,
+      "grad_norm": 0.9437888860702515,
+      "learning_rate": 2.155944016634398e-06,
+      "loss": 0.0825,
+      "step": 21240
+    },
+    {
+      "epoch": 81.10687022900764,
+      "grad_norm": 0.3145671486854553,
+      "learning_rate": 2.1514126580284538e-06,
+      "loss": 0.0653,
+      "step": 21250
+    },
+    {
+      "epoch": 81.14503816793894,
+      "grad_norm": 0.25964006781578064,
+      "learning_rate": 2.1468847607247344e-06,
+      "loss": 0.0665,
+      "step": 21260
+    },
+    {
+      "epoch": 81.18320610687023,
+      "grad_norm": 0.41988322138786316,
+      "learning_rate": 2.1423603302250625e-06,
+      "loss": 0.0732,
+      "step": 21270
+    },
+    {
+      "epoch": 81.22137404580153,
+      "grad_norm": 0.2630063593387604,
+      "learning_rate": 2.137839372027047e-06,
+      "loss": 0.073,
+      "step": 21280
+    },
+    {
+      "epoch": 81.25954198473282,
+      "grad_norm": 0.42665886878967285,
+      "learning_rate": 2.133321891624076e-06,
+      "loss": 0.0777,
+      "step": 21290
+    },
+    {
+      "epoch": 81.29770992366412,
+      "grad_norm": 0.6088788509368896,
+      "learning_rate": 2.1288078945053194e-06,
+      "loss": 0.0715,
+      "step": 21300
+    },
+    {
+      "epoch": 81.33587786259542,
+      "grad_norm": 0.7648810744285583,
+      "learning_rate": 2.1242973861557064e-06,
+      "loss": 0.0714,
+      "step": 21310
+    },
+    {
+      "epoch": 81.37404580152672,
+      "grad_norm": 0.4655439555644989,
+      "learning_rate": 2.1197903720559303e-06,
+      "loss": 0.0734,
+      "step": 21320
+    },
+    {
+      "epoch": 81.41221374045801,
+      "grad_norm": 0.8302370309829712,
+      "learning_rate": 2.1152868576824383e-06,
+      "loss": 0.0769,
+      "step": 21330
+    },
+    {
+      "epoch": 81.45038167938931,
+      "grad_norm": 0.7731935977935791,
+      "learning_rate": 2.110786848507423e-06,
+      "loss": 0.0725,
+      "step": 21340
+    },
+    {
+      "epoch": 81.4885496183206,
+      "grad_norm": 0.8815107345581055,
+      "learning_rate": 2.1062903499988235e-06,
+      "loss": 0.0682,
+      "step": 21350
+    },
+    {
+      "epoch": 81.52671755725191,
+      "grad_norm": 0.4499039649963379,
+      "learning_rate": 2.101797367620308e-06,
+      "loss": 0.0738,
+      "step": 21360
+    },
+    {
+      "epoch": 81.56488549618321,
+      "grad_norm": 0.3854046165943146,
+      "learning_rate": 2.0973079068312713e-06,
+      "loss": 0.072,
+      "step": 21370
+    },
+    {
+      "epoch": 81.6030534351145,
+      "grad_norm": 0.33728718757629395,
+      "learning_rate": 2.0928219730868358e-06,
+      "loss": 0.0666,
+      "step": 21380
+    },
+    {
+      "epoch": 81.6412213740458,
+      "grad_norm": 0.24927888810634613,
+      "learning_rate": 2.0883395718378304e-06,
+      "loss": 0.0624,
+      "step": 21390
+    },
+    {
+      "epoch": 81.6793893129771,
+      "grad_norm": 0.45682278275489807,
+      "learning_rate": 2.083860708530798e-06,
+      "loss": 0.0992,
+      "step": 21400
+    },
+    {
+      "epoch": 81.7175572519084,
+      "grad_norm": 0.29621076583862305,
+      "learning_rate": 2.0793853886079794e-06,
+      "loss": 0.0766,
+      "step": 21410
+    },
+    {
+      "epoch": 81.7557251908397,
+      "grad_norm": 0.34035375714302063,
+      "learning_rate": 2.074913617507309e-06,
+      "loss": 0.0694,
+      "step": 21420
+    },
+    {
+      "epoch": 81.79389312977099,
+      "grad_norm": 0.23349541425704956,
+      "learning_rate": 2.0704454006624116e-06,
+      "loss": 0.0709,
+      "step": 21430
+    },
+    {
+      "epoch": 81.83206106870229,
+      "grad_norm": 0.2720065414905548,
+      "learning_rate": 2.0659807435025907e-06,
+      "loss": 0.0791,
+      "step": 21440
+    },
+    {
+      "epoch": 81.87022900763358,
+      "grad_norm": 0.24106143414974213,
+      "learning_rate": 2.061519651452825e-06,
+      "loss": 0.0716,
+      "step": 21450
+    },
+    {
+      "epoch": 81.90839694656489,
+      "grad_norm": 0.5342702865600586,
+      "learning_rate": 2.0570621299337656e-06,
+      "loss": 0.073,
+      "step": 21460
+    },
+    {
+      "epoch": 81.94656488549619,
+      "grad_norm": 0.3628579080104828,
+      "learning_rate": 2.0526081843617183e-06,
+      "loss": 0.0784,
+      "step": 21470
+    },
+    {
+      "epoch": 81.98473282442748,
+      "grad_norm": 0.35211580991744995,
+      "learning_rate": 2.0481578201486484e-06,
+      "loss": 0.0673,
+      "step": 21480
+    },
+    {
+      "epoch": 82.02290076335878,
+      "grad_norm": 0.44048014283180237,
+      "learning_rate": 2.043711042702168e-06,
+      "loss": 0.0757,
+      "step": 21490
+    },
+    {
+      "epoch": 82.06106870229007,
+      "grad_norm": 0.5018337965011597,
+      "learning_rate": 2.039267857425528e-06,
+      "loss": 0.0699,
+      "step": 21500
+    },
+    {
+      "epoch": 82.09923664122137,
+      "grad_norm": 0.4347798824310303,
+      "learning_rate": 2.034828269717622e-06,
+      "loss": 0.0703,
+      "step": 21510
+    },
+    {
+      "epoch": 82.13740458015268,
+      "grad_norm": 0.2908450663089752,
+      "learning_rate": 2.030392284972964e-06,
+      "loss": 0.071,
+      "step": 21520
+    },
+    {
+      "epoch": 82.17557251908397,
+      "grad_norm": 0.39640676975250244,
+      "learning_rate": 2.0259599085816973e-06,
+      "loss": 0.0735,
+      "step": 21530
+    },
+    {
+      "epoch": 82.21374045801527,
+      "grad_norm": 0.8347475528717041,
+      "learning_rate": 2.0215311459295757e-06,
+      "loss": 0.0798,
+      "step": 21540
+    },
+    {
+      "epoch": 82.25190839694656,
+      "grad_norm": 0.3542044758796692,
+      "learning_rate": 2.0171060023979603e-06,
+      "loss": 0.0669,
+      "step": 21550
+    },
+    {
+      "epoch": 82.29007633587786,
+      "grad_norm": 0.5587815046310425,
+      "learning_rate": 2.012684483363823e-06,
+      "loss": 0.0728,
+      "step": 21560
+    },
+    {
+      "epoch": 82.32824427480917,
+      "grad_norm": 0.2797441780567169,
+      "learning_rate": 2.0082665941997236e-06,
+      "loss": 0.0699,
+      "step": 21570
+    },
+    {
+      "epoch": 82.36641221374046,
+      "grad_norm": 0.5870032906532288,
+      "learning_rate": 2.0038523402738147e-06,
+      "loss": 0.0684,
+      "step": 21580
+    },
+    {
+      "epoch": 82.40458015267176,
+      "grad_norm": 0.608632504940033,
+      "learning_rate": 1.99944172694983e-06,
+      "loss": 0.0767,
+      "step": 21590
+    },
+    {
+      "epoch": 82.44274809160305,
+      "grad_norm": 0.28920474648475647,
+      "learning_rate": 1.99503475958708e-06,
+      "loss": 0.0731,
+      "step": 21600
+    },
+    {
+      "epoch": 82.48091603053435,
+      "grad_norm": 0.22390879690647125,
+      "learning_rate": 1.9906314435404484e-06,
+      "loss": 0.0676,
+      "step": 21610
+    },
+    {
+      "epoch": 82.51908396946565,
+      "grad_norm": 0.3375746011734009,
+      "learning_rate": 1.986231784160378e-06,
+      "loss": 0.0724,
+      "step": 21620
+    },
+    {
+      "epoch": 82.55725190839695,
+      "grad_norm": 0.37056073546409607,
+      "learning_rate": 1.9818357867928697e-06,
+      "loss": 0.0741,
+      "step": 21630
+    },
+    {
+      "epoch": 82.59541984732824,
+      "grad_norm": 0.25039762258529663,
+      "learning_rate": 1.9774434567794744e-06,
+      "loss": 0.0709,
+      "step": 21640
+    },
+    {
+      "epoch": 82.63358778625954,
+      "grad_norm": 0.33184704184532166,
+      "learning_rate": 1.973054799457286e-06,
+      "loss": 0.0685,
+      "step": 21650
+    },
+    {
+      "epoch": 82.67175572519083,
+      "grad_norm": 0.4607788920402527,
+      "learning_rate": 1.9686698201589395e-06,
+      "loss": 0.0739,
+      "step": 21660
+    },
+    {
+      "epoch": 82.70992366412214,
+      "grad_norm": 0.7115771770477295,
+      "learning_rate": 1.9642885242125962e-06,
+      "loss": 0.0819,
+      "step": 21670
+    },
+    {
+      "epoch": 82.74809160305344,
+      "grad_norm": 0.3059196472167969,
+      "learning_rate": 1.9599109169419467e-06,
+      "loss": 0.065,
+      "step": 21680
+    },
+    {
+      "epoch": 82.78625954198473,
+      "grad_norm": 0.2690870761871338,
+      "learning_rate": 1.9555370036661946e-06,
+      "loss": 0.069,
+      "step": 21690
+    },
+    {
+      "epoch": 82.82442748091603,
+      "grad_norm": 0.8284600973129272,
+      "learning_rate": 1.9511667897000577e-06,
+      "loss": 0.0672,
+      "step": 21700
+    },
+    {
+      "epoch": 82.86259541984732,
+      "grad_norm": 0.5146779417991638,
+      "learning_rate": 1.946800280353755e-06,
+      "loss": 0.063,
+      "step": 21710
+    },
+    {
+      "epoch": 82.90076335877862,
+      "grad_norm": 0.27559417486190796,
+      "learning_rate": 1.9424374809330117e-06,
+      "loss": 0.0717,
+      "step": 21720
+    },
+    {
+      "epoch": 82.93893129770993,
+      "grad_norm": 0.3968946933746338,
+      "learning_rate": 1.938078396739038e-06,
+      "loss": 0.0662,
+      "step": 21730
+    },
+    {
+      "epoch": 82.97709923664122,
+      "grad_norm": 0.3083120286464691,
+      "learning_rate": 1.9337230330685332e-06,
+      "loss": 0.0686,
+      "step": 21740
+    },
+    {
+      "epoch": 83.01526717557252,
+      "grad_norm": 0.46116316318511963,
+      "learning_rate": 1.929371395213674e-06,
+      "loss": 0.0809,
+      "step": 21750
+    },
+    {
+      "epoch": 83.05343511450381,
+      "grad_norm": 0.3158531188964844,
+      "learning_rate": 1.9250234884621093e-06,
+      "loss": 0.0725,
+      "step": 21760
+    },
+    {
+      "epoch": 83.09160305343511,
+      "grad_norm": 0.5462889671325684,
+      "learning_rate": 1.9206793180969593e-06,
+      "loss": 0.0671,
+      "step": 21770
+    },
+    {
+      "epoch": 83.12977099236642,
+      "grad_norm": 0.4078521430492401,
+      "learning_rate": 1.916338889396798e-06,
+      "loss": 0.0725,
+      "step": 21780
+    },
+    {
+      "epoch": 83.16793893129771,
+      "grad_norm": 0.46893805265426636,
+      "learning_rate": 1.9120022076356577e-06,
+      "loss": 0.0741,
+      "step": 21790
+    },
+    {
+      "epoch": 83.20610687022901,
+      "grad_norm": 0.8775691986083984,
+      "learning_rate": 1.9076692780830115e-06,
+      "loss": 0.0706,
+      "step": 21800
+    },
+    {
+      "epoch": 83.2442748091603,
+      "grad_norm": 0.24274039268493652,
+      "learning_rate": 1.903340106003782e-06,
+      "loss": 0.0711,
+      "step": 21810
+    },
+    {
+      "epoch": 83.2824427480916,
+      "grad_norm": 0.5412287712097168,
+      "learning_rate": 1.8990146966583183e-06,
+      "loss": 0.0716,
+      "step": 21820
+    },
+    {
+      "epoch": 83.3206106870229,
+      "grad_norm": 0.26575616002082825,
+      "learning_rate": 1.8946930553024034e-06,
+      "loss": 0.0693,
+      "step": 21830
+    },
+    {
+      "epoch": 83.3587786259542,
+      "grad_norm": 0.39014455676078796,
+      "learning_rate": 1.8903751871872377e-06,
+      "loss": 0.0669,
+      "step": 21840
+    },
+    {
+      "epoch": 83.3969465648855,
+      "grad_norm": 0.2963654398918152,
+      "learning_rate": 1.8860610975594384e-06,
+      "loss": 0.0783,
+      "step": 21850
+    },
+    {
+      "epoch": 83.43511450381679,
+      "grad_norm": 0.2706157863140106,
+      "learning_rate": 1.8817507916610307e-06,
+      "loss": 0.0688,
+      "step": 21860
+    },
+    {
+      "epoch": 83.47328244274809,
+      "grad_norm": 0.369393914937973,
+      "learning_rate": 1.8774442747294407e-06,
+      "loss": 0.0666,
+      "step": 21870
+    },
+    {
+      "epoch": 83.5114503816794,
+      "grad_norm": 0.44239068031311035,
+      "learning_rate": 1.8731415519974967e-06,
+      "loss": 0.0754,
+      "step": 21880
+    },
+    {
+      "epoch": 83.54961832061069,
+      "grad_norm": 0.29093948006629944,
+      "learning_rate": 1.8688426286934102e-06,
+      "loss": 0.0718,
+      "step": 21890
+    },
+    {
+      "epoch": 83.58778625954199,
+      "grad_norm": 0.45363718271255493,
+      "learning_rate": 1.864547510040779e-06,
+      "loss": 0.0676,
+      "step": 21900
+    },
+    {
+      "epoch": 83.62595419847328,
+      "grad_norm": 0.427654892206192,
+      "learning_rate": 1.8602562012585768e-06,
+      "loss": 0.0667,
+      "step": 21910
+    },
+    {
+      "epoch": 83.66412213740458,
+      "grad_norm": 0.2690734267234802,
+      "learning_rate": 1.8559687075611466e-06,
+      "loss": 0.0648,
+      "step": 21920
+    },
+    {
+      "epoch": 83.70229007633588,
+      "grad_norm": 0.26419100165367126,
+      "learning_rate": 1.8516850341582015e-06,
+      "loss": 0.0647,
+      "step": 21930
+    },
+    {
+      "epoch": 83.74045801526718,
+      "grad_norm": 0.4640803337097168,
+      "learning_rate": 1.847405186254807e-06,
+      "loss": 0.0662,
+      "step": 21940
+    },
+    {
+      "epoch": 83.77862595419847,
+      "grad_norm": 0.3314109742641449,
+      "learning_rate": 1.8431291690513791e-06,
+      "loss": 0.0684,
+      "step": 21950
+    },
+    {
+      "epoch": 83.81679389312977,
+      "grad_norm": 1.6556986570358276,
+      "learning_rate": 1.8388569877436863e-06,
+      "loss": 0.0736,
+      "step": 21960
+    },
+    {
+      "epoch": 83.85496183206106,
+      "grad_norm": 0.2407653033733368,
+      "learning_rate": 1.834588647522828e-06,
+      "loss": 0.0648,
+      "step": 21970
+    },
+    {
+      "epoch": 83.89312977099236,
+      "grad_norm": 0.2633666694164276,
+      "learning_rate": 1.8303241535752437e-06,
+      "loss": 0.0793,
+      "step": 21980
+    },
+    {
+      "epoch": 83.93129770992367,
+      "grad_norm": 0.3340645730495453,
+      "learning_rate": 1.8260635110826936e-06,
+      "loss": 0.0684,
+      "step": 21990
+    },
+    {
+      "epoch": 83.96946564885496,
+      "grad_norm": 0.5088317394256592,
+      "learning_rate": 1.82180672522226e-06,
+      "loss": 0.0668,
+      "step": 22000
+    },
+    {
+      "epoch": 84.00763358778626,
+      "grad_norm": 0.4002598822116852,
+      "learning_rate": 1.817553801166339e-06,
+      "loss": 0.0733,
+      "step": 22010
+    },
+    {
+      "epoch": 84.04580152671755,
+      "grad_norm": 0.34704163670539856,
+      "learning_rate": 1.8133047440826335e-06,
+      "loss": 0.0655,
+      "step": 22020
+    },
+    {
+      "epoch": 84.08396946564885,
+      "grad_norm": 0.27369430661201477,
+      "learning_rate": 1.8090595591341509e-06,
+      "loss": 0.0687,
+      "step": 22030
+    },
+    {
+      "epoch": 84.12213740458016,
+      "grad_norm": 0.6098417043685913,
+      "learning_rate": 1.8048182514791901e-06,
+      "loss": 0.0667,
+      "step": 22040
+    },
+    {
+      "epoch": 84.16030534351145,
+      "grad_norm": 0.3603808581829071,
+      "learning_rate": 1.8005808262713399e-06,
+      "loss": 0.0636,
+      "step": 22050
+    },
+    {
+      "epoch": 84.19847328244275,
+      "grad_norm": 0.5261178016662598,
+      "learning_rate": 1.7963472886594713e-06,
+      "loss": 0.0674,
+      "step": 22060
+    },
+    {
+      "epoch": 84.23664122137404,
+      "grad_norm": 0.27495747804641724,
+      "learning_rate": 1.7921176437877302e-06,
+      "loss": 0.0663,
+      "step": 22070
+    },
+    {
+      "epoch": 84.27480916030534,
+      "grad_norm": 0.5118312835693359,
+      "learning_rate": 1.7878918967955366e-06,
+      "loss": 0.0722,
+      "step": 22080
+    },
+    {
+      "epoch": 84.31297709923665,
+      "grad_norm": 0.2527947723865509,
+      "learning_rate": 1.7836700528175693e-06,
+      "loss": 0.0681,
+      "step": 22090
+    },
+    {
+      "epoch": 84.35114503816794,
+      "grad_norm": 0.25550681352615356,
+      "learning_rate": 1.7794521169837693e-06,
+      "loss": 0.0694,
+      "step": 22100
+    },
+    {
+      "epoch": 84.38931297709924,
+      "grad_norm": 0.6813188195228577,
+      "learning_rate": 1.7752380944193248e-06,
+      "loss": 0.0719,
+      "step": 22110
+    },
+    {
+      "epoch": 84.42748091603053,
+      "grad_norm": 0.6162485480308533,
+      "learning_rate": 1.771027990244671e-06,
+      "loss": 0.0768,
+      "step": 22120
+    },
+    {
+      "epoch": 84.46564885496183,
+      "grad_norm": 0.2738633453845978,
+      "learning_rate": 1.7668218095754797e-06,
+      "loss": 0.0703,
+      "step": 22130
+    },
+    {
+      "epoch": 84.50381679389314,
+      "grad_norm": 0.9221097230911255,
+      "learning_rate": 1.7626195575226595e-06,
+      "loss": 0.0679,
+      "step": 22140
+    },
+    {
+      "epoch": 84.54198473282443,
+      "grad_norm": 0.2497481107711792,
+      "learning_rate": 1.7584212391923428e-06,
+      "loss": 0.0721,
+      "step": 22150
+    },
+    {
+      "epoch": 84.58015267175573,
+      "grad_norm": 0.36537232995033264,
+      "learning_rate": 1.7542268596858813e-06,
+      "loss": 0.0772,
+      "step": 22160
+    },
+    {
+      "epoch": 84.61832061068702,
+      "grad_norm": 0.2652266025543213,
+      "learning_rate": 1.7500364240998412e-06,
+      "loss": 0.0693,
+      "step": 22170
+    },
+    {
+      "epoch": 84.65648854961832,
+      "grad_norm": 0.3559548258781433,
+      "learning_rate": 1.7458499375259957e-06,
+      "loss": 0.0703,
+      "step": 22180
+    },
+    {
+      "epoch": 84.69465648854961,
+      "grad_norm": 0.40493687987327576,
+      "learning_rate": 1.7416674050513243e-06,
+      "loss": 0.0613,
+      "step": 22190
+    },
+    {
+      "epoch": 84.73282442748092,
+      "grad_norm": 0.38015782833099365,
+      "learning_rate": 1.7374888317579968e-06,
+      "loss": 0.0736,
+      "step": 22200
+    },
+    {
+      "epoch": 84.77099236641222,
+      "grad_norm": 0.19811968505382538,
+      "learning_rate": 1.7333142227233728e-06,
+      "loss": 0.0659,
+      "step": 22210
+    },
+    {
+      "epoch": 84.80916030534351,
+      "grad_norm": 0.5039312839508057,
+      "learning_rate": 1.7291435830199954e-06,
+      "loss": 0.0726,
+      "step": 22220
+    },
+    {
+      "epoch": 84.8473282442748,
+      "grad_norm": 0.36674124002456665,
+      "learning_rate": 1.7249769177155879e-06,
+      "loss": 0.0657,
+      "step": 22230
+    },
+    {
+      "epoch": 84.8854961832061,
+      "grad_norm": 0.7190355062484741,
+      "learning_rate": 1.720814231873038e-06,
+      "loss": 0.0783,
+      "step": 22240
+    },
+    {
+      "epoch": 84.92366412213741,
+      "grad_norm": 0.3150536119937897,
+      "learning_rate": 1.716655530550405e-06,
+      "loss": 0.0755,
+      "step": 22250
+    },
+    {
+      "epoch": 84.9618320610687,
+      "grad_norm": 0.29341939091682434,
+      "learning_rate": 1.7125008188009018e-06,
+      "loss": 0.0667,
+      "step": 22260
+    },
+    {
+      "epoch": 85.0,
+      "grad_norm": 0.3069048225879669,
+      "learning_rate": 1.7083501016728944e-06,
+      "loss": 0.0674,
+      "step": 22270
+    },
+    {
+      "epoch": 85.0381679389313,
+      "grad_norm": 0.3248658776283264,
+      "learning_rate": 1.704203384209896e-06,
+      "loss": 0.0649,
+      "step": 22280
+    },
+    {
+      "epoch": 85.07633587786259,
+      "grad_norm": 0.2633165121078491,
+      "learning_rate": 1.700060671450557e-06,
+      "loss": 0.0705,
+      "step": 22290
+    },
+    {
+      "epoch": 85.1145038167939,
+      "grad_norm": 0.2645620107650757,
+      "learning_rate": 1.6959219684286681e-06,
+      "loss": 0.062,
+      "step": 22300
+    },
+    {
+      "epoch": 85.1526717557252,
+      "grad_norm": 0.25467944145202637,
+      "learning_rate": 1.6917872801731417e-06,
+      "loss": 0.0726,
+      "step": 22310
+    },
+    {
+      "epoch": 85.19083969465649,
+      "grad_norm": 0.4649989902973175,
+      "learning_rate": 1.6876566117080139e-06,
+      "loss": 0.0737,
+      "step": 22320
+    },
+    {
+      "epoch": 85.22900763358778,
+      "grad_norm": 0.4027472138404846,
+      "learning_rate": 1.683529968052437e-06,
+      "loss": 0.069,
+      "step": 22330
+    },
+    {
+      "epoch": 85.26717557251908,
+      "grad_norm": 0.35890597105026245,
+      "learning_rate": 1.679407354220669e-06,
+      "loss": 0.0709,
+      "step": 22340
+    },
+    {
+      "epoch": 85.30534351145039,
+      "grad_norm": 0.24897009134292603,
+      "learning_rate": 1.6752887752220792e-06,
+      "loss": 0.0682,
+      "step": 22350
+    },
+    {
+      "epoch": 85.34351145038168,
+      "grad_norm": 0.31344956159591675,
+      "learning_rate": 1.6711742360611277e-06,
+      "loss": 0.0736,
+      "step": 22360
+    },
+    {
+      "epoch": 85.38167938931298,
+      "grad_norm": 0.633032500743866,
+      "learning_rate": 1.6670637417373652e-06,
+      "loss": 0.0769,
+      "step": 22370
+    },
+    {
+      "epoch": 85.41984732824427,
+      "grad_norm": 0.41895341873168945,
+      "learning_rate": 1.6629572972454333e-06,
+      "loss": 0.0666,
+      "step": 22380
+    },
+    {
+      "epoch": 85.45801526717557,
+      "grad_norm": 0.25024113059043884,
+      "learning_rate": 1.6588549075750466e-06,
+      "loss": 0.075,
+      "step": 22390
+    },
+    {
+      "epoch": 85.49618320610686,
+      "grad_norm": 0.3272351622581482,
+      "learning_rate": 1.6547565777109975e-06,
+      "loss": 0.0718,
+      "step": 22400
+    },
+    {
+      "epoch": 85.53435114503817,
+      "grad_norm": 0.6027217507362366,
+      "learning_rate": 1.6506623126331427e-06,
+      "loss": 0.0696,
+      "step": 22410
+    },
+    {
+      "epoch": 85.57251908396947,
+      "grad_norm": 0.4890408515930176,
+      "learning_rate": 1.6465721173164e-06,
+      "loss": 0.0708,
+      "step": 22420
+    },
+    {
+      "epoch": 85.61068702290076,
+      "grad_norm": 0.39123833179473877,
+      "learning_rate": 1.6424859967307427e-06,
+      "loss": 0.0687,
+      "step": 22430
+    },
+    {
+      "epoch": 85.64885496183206,
+      "grad_norm": 0.26333490014076233,
+      "learning_rate": 1.6384039558411902e-06,
+      "loss": 0.0725,
+      "step": 22440
+    },
+    {
+      "epoch": 85.68702290076335,
+      "grad_norm": 0.32792478799819946,
+      "learning_rate": 1.6343259996078109e-06,
+      "loss": 0.0816,
+      "step": 22450
+    },
+    {
+      "epoch": 85.72519083969466,
+      "grad_norm": 0.32284873723983765,
+      "learning_rate": 1.6302521329857046e-06,
+      "loss": 0.0686,
+      "step": 22460
+    },
+    {
+      "epoch": 85.76335877862596,
+      "grad_norm": 0.35827475786209106,
+      "learning_rate": 1.6261823609250027e-06,
+      "loss": 0.0702,
+      "step": 22470
+    },
+    {
+      "epoch": 85.80152671755725,
+      "grad_norm": 0.28327593207359314,
+      "learning_rate": 1.6221166883708629e-06,
+      "loss": 0.0657,
+      "step": 22480
+    },
+    {
+      "epoch": 85.83969465648855,
+      "grad_norm": 0.2737085223197937,
+      "learning_rate": 1.6180551202634603e-06,
+      "loss": 0.0697,
+      "step": 22490
+    },
+    {
+      "epoch": 85.87786259541984,
+      "grad_norm": 0.29437166452407837,
+      "learning_rate": 1.613997661537981e-06,
+      "loss": 0.0692,
+      "step": 22500
+    },
+    {
+      "epoch": 85.91603053435115,
+      "grad_norm": 0.6566985249519348,
+      "learning_rate": 1.6099443171246243e-06,
+      "loss": 0.0664,
+      "step": 22510
+    },
+    {
+      "epoch": 85.95419847328245,
+      "grad_norm": 0.3802522122859955,
+      "learning_rate": 1.6058950919485823e-06,
+      "loss": 0.0635,
+      "step": 22520
+    },
+    {
+      "epoch": 85.99236641221374,
+      "grad_norm": 0.44286301732063293,
+      "learning_rate": 1.6018499909300478e-06,
+      "loss": 0.0697,
+      "step": 22530
+    },
+    {
+      "epoch": 86.03053435114504,
+      "grad_norm": 0.2453600913286209,
+      "learning_rate": 1.5978090189841988e-06,
+      "loss": 0.0671,
+      "step": 22540
+    },
+    {
+      "epoch": 86.06870229007633,
+      "grad_norm": 0.2875606417655945,
+      "learning_rate": 1.5937721810211958e-06,
+      "loss": 0.0676,
+      "step": 22550
+    },
+    {
+      "epoch": 86.10687022900764,
+      "grad_norm": 0.7595479488372803,
+      "learning_rate": 1.5897394819461815e-06,
+      "loss": 0.078,
+      "step": 22560
+    },
+    {
+      "epoch": 86.14503816793894,
+      "grad_norm": 0.2808679938316345,
+      "learning_rate": 1.5857109266592624e-06,
+      "loss": 0.0699,
+      "step": 22570
+    },
+    {
+      "epoch": 86.18320610687023,
+      "grad_norm": 0.27129414677619934,
+      "learning_rate": 1.5816865200555142e-06,
+      "loss": 0.0662,
+      "step": 22580
+    },
+    {
+      "epoch": 86.22137404580153,
+      "grad_norm": 0.27305206656455994,
+      "learning_rate": 1.5776662670249704e-06,
+      "loss": 0.0666,
+      "step": 22590
+    },
+    {
+      "epoch": 86.25954198473282,
+      "grad_norm": 0.4132048487663269,
+      "learning_rate": 1.573650172452615e-06,
+      "loss": 0.0711,
+      "step": 22600
+    },
+    {
+      "epoch": 86.29770992366412,
+      "grad_norm": 0.39085114002227783,
+      "learning_rate": 1.5696382412183853e-06,
+      "loss": 0.0721,
+      "step": 22610
+    },
+    {
+      "epoch": 86.33587786259542,
+      "grad_norm": 0.570540726184845,
+      "learning_rate": 1.5656304781971549e-06,
+      "loss": 0.0726,
+      "step": 22620
+    },
+    {
+      "epoch": 86.37404580152672,
+      "grad_norm": 0.452827513217926,
+      "learning_rate": 1.5616268882587331e-06,
+      "loss": 0.0698,
+      "step": 22630
+    },
+    {
+      "epoch": 86.41221374045801,
+      "grad_norm": 0.2821395993232727,
+      "learning_rate": 1.5576274762678594e-06,
+      "loss": 0.0672,
+      "step": 22640
+    },
+    {
+      "epoch": 86.45038167938931,
+      "grad_norm": 0.30585944652557373,
+      "learning_rate": 1.5536322470841953e-06,
+      "loss": 0.0714,
+      "step": 22650
+    },
+    {
+      "epoch": 86.4885496183206,
+      "grad_norm": 1.8794187307357788,
+      "learning_rate": 1.5496412055623238e-06,
+      "loss": 0.0727,
+      "step": 22660
+    },
+    {
+      "epoch": 86.52671755725191,
+      "grad_norm": 0.3682367205619812,
+      "learning_rate": 1.5456543565517334e-06,
+      "loss": 0.0684,
+      "step": 22670
+    },
+    {
+      "epoch": 86.56488549618321,
+      "grad_norm": 0.6367046236991882,
+      "learning_rate": 1.541671704896825e-06,
+      "loss": 0.0734,
+      "step": 22680
+    },
+    {
+      "epoch": 86.6030534351145,
+      "grad_norm": 0.2984490692615509,
+      "learning_rate": 1.537693255436894e-06,
+      "loss": 0.0801,
+      "step": 22690
+    },
+    {
+      "epoch": 86.6412213740458,
+      "grad_norm": 0.26280051469802856,
+      "learning_rate": 1.533719013006132e-06,
+      "loss": 0.0734,
+      "step": 22700
+    },
+    {
+      "epoch": 86.6793893129771,
+      "grad_norm": 0.6878067255020142,
+      "learning_rate": 1.5297489824336165e-06,
+      "loss": 0.0751,
+      "step": 22710
+    },
+    {
+      "epoch": 86.7175572519084,
+      "grad_norm": 0.42184221744537354,
+      "learning_rate": 1.5257831685433121e-06,
+      "loss": 0.0679,
+      "step": 22720
+    },
+    {
+      "epoch": 86.7557251908397,
+      "grad_norm": 0.33254826068878174,
+      "learning_rate": 1.521821576154055e-06,
+      "loss": 0.0673,
+      "step": 22730
+    },
+    {
+      "epoch": 86.79389312977099,
+      "grad_norm": 0.20484282076358795,
+      "learning_rate": 1.5178642100795543e-06,
+      "loss": 0.0659,
+      "step": 22740
+    },
+    {
+      "epoch": 86.83206106870229,
+      "grad_norm": 0.32377350330352783,
+      "learning_rate": 1.5139110751283819e-06,
+      "loss": 0.0704,
+      "step": 22750
+    },
+    {
+      "epoch": 86.87022900763358,
+      "grad_norm": 0.3272833228111267,
+      "learning_rate": 1.5099621761039684e-06,
+      "loss": 0.0702,
+      "step": 22760
+    },
+    {
+      "epoch": 86.90839694656489,
+      "grad_norm": 0.5103157162666321,
+      "learning_rate": 1.5060175178046017e-06,
+      "loss": 0.0779,
+      "step": 22770
+    },
+    {
+      "epoch": 86.94656488549619,
+      "grad_norm": 0.2767445743083954,
+      "learning_rate": 1.5020771050234118e-06,
+      "loss": 0.065,
+      "step": 22780
+    },
+    {
+      "epoch": 86.98473282442748,
+      "grad_norm": 0.6275362968444824,
+      "learning_rate": 1.4981409425483716e-06,
+      "loss": 0.065,
+      "step": 22790
+    },
+    {
+      "epoch": 87.02290076335878,
+      "grad_norm": 0.2717098593711853,
+      "learning_rate": 1.4942090351622884e-06,
+      "loss": 0.0772,
+      "step": 22800
+    },
+    {
+      "epoch": 87.06106870229007,
+      "grad_norm": 0.5612622499465942,
+      "learning_rate": 1.490281387642804e-06,
+      "loss": 0.0681,
+      "step": 22810
+    },
+    {
+      "epoch": 87.09923664122137,
+      "grad_norm": 0.27303266525268555,
+      "learning_rate": 1.4863580047623767e-06,
+      "loss": 0.068,
+      "step": 22820
+    },
+    {
+      "epoch": 87.13740458015268,
+      "grad_norm": 0.3088218867778778,
+      "learning_rate": 1.4824388912882897e-06,
+      "loss": 0.076,
+      "step": 22830
+    },
+    {
+      "epoch": 87.17557251908397,
+      "grad_norm": 0.22109341621398926,
+      "learning_rate": 1.4785240519826345e-06,
+      "loss": 0.065,
+      "step": 22840
+    },
+    {
+      "epoch": 87.21374045801527,
+      "grad_norm": 0.3760296404361725,
+      "learning_rate": 1.4746134916023096e-06,
+      "loss": 0.0653,
+      "step": 22850
+    },
+    {
+      "epoch": 87.25190839694656,
+      "grad_norm": 0.4311428368091583,
+      "learning_rate": 1.4707072148990142e-06,
+      "loss": 0.0673,
+      "step": 22860
+    },
+    {
+      "epoch": 87.29007633587786,
+      "grad_norm": 0.9990677237510681,
+      "learning_rate": 1.4668052266192423e-06,
+      "loss": 0.0692,
+      "step": 22870
+    },
+    {
+      "epoch": 87.32824427480917,
+      "grad_norm": 0.2663988769054413,
+      "learning_rate": 1.4629075315042795e-06,
+      "loss": 0.0654,
+      "step": 22880
+    },
+    {
+      "epoch": 87.36641221374046,
+      "grad_norm": 0.2906090021133423,
+      "learning_rate": 1.4590141342901926e-06,
+      "loss": 0.0667,
+      "step": 22890
+    },
+    {
+      "epoch": 87.40458015267176,
+      "grad_norm": 0.5465324521064758,
+      "learning_rate": 1.4551250397078253e-06,
+      "loss": 0.0716,
+      "step": 22900
+    },
+    {
+      "epoch": 87.44274809160305,
+      "grad_norm": 0.29929447174072266,
+      "learning_rate": 1.4512402524827945e-06,
+      "loss": 0.0755,
+      "step": 22910
+    },
+    {
+      "epoch": 87.48091603053435,
+      "grad_norm": 0.3241725564002991,
+      "learning_rate": 1.447359777335482e-06,
+      "loss": 0.0635,
+      "step": 22920
+    },
+    {
+      "epoch": 87.51908396946565,
+      "grad_norm": 0.28157156705856323,
+      "learning_rate": 1.4434836189810337e-06,
+      "loss": 0.0651,
+      "step": 22930
+    },
+    {
+      "epoch": 87.55725190839695,
+      "grad_norm": 0.32432863116264343,
+      "learning_rate": 1.4396117821293454e-06,
+      "loss": 0.0682,
+      "step": 22940
+    },
+    {
+      "epoch": 87.59541984732824,
+      "grad_norm": 0.39860352873802185,
+      "learning_rate": 1.4357442714850634e-06,
+      "loss": 0.0703,
+      "step": 22950
+    },
+    {
+      "epoch": 87.63358778625954,
+      "grad_norm": 0.4769950211048126,
+      "learning_rate": 1.43188109174758e-06,
+      "loss": 0.0657,
+      "step": 22960
+    },
+    {
+      "epoch": 87.67175572519083,
+      "grad_norm": 0.42243120074272156,
+      "learning_rate": 1.4280222476110206e-06,
+      "loss": 0.0674,
+      "step": 22970
+    },
+    {
+      "epoch": 87.70992366412214,
+      "grad_norm": 0.6192082166671753,
+      "learning_rate": 1.424167743764247e-06,
+      "loss": 0.0724,
+      "step": 22980
+    },
+    {
+      "epoch": 87.74809160305344,
+      "grad_norm": 0.6476142406463623,
+      "learning_rate": 1.420317584890844e-06,
+      "loss": 0.0671,
+      "step": 22990
+    },
+    {
+      "epoch": 87.78625954198473,
+      "grad_norm": 0.6006608605384827,
+      "learning_rate": 1.4164717756691176e-06,
+      "loss": 0.0727,
+      "step": 23000
+    },
+    {
+      "epoch": 87.82442748091603,
+      "grad_norm": 0.5018774271011353,
+      "learning_rate": 1.4126303207720882e-06,
+      "loss": 0.0644,
+      "step": 23010
+    },
+    {
+      "epoch": 87.86259541984732,
+      "grad_norm": 0.19781306385993958,
+      "learning_rate": 1.408793224867484e-06,
+      "loss": 0.0683,
+      "step": 23020
+    },
+    {
+      "epoch": 87.90076335877862,
+      "grad_norm": 0.3770202100276947,
+      "learning_rate": 1.4049604926177423e-06,
+      "loss": 0.0748,
+      "step": 23030
+    },
+    {
+      "epoch": 87.93893129770993,
+      "grad_norm": 0.23254378139972687,
+      "learning_rate": 1.4011321286799918e-06,
+      "loss": 0.071,
+      "step": 23040
+    },
+    {
+      "epoch": 87.97709923664122,
+      "grad_norm": 0.2871849238872528,
+      "learning_rate": 1.3973081377060565e-06,
+      "loss": 0.0685,
+      "step": 23050
+    },
+    {
+      "epoch": 88.01526717557252,
+      "grad_norm": 0.34911373257637024,
+      "learning_rate": 1.393488524342445e-06,
+      "loss": 0.0669,
+      "step": 23060
+    },
+    {
+      "epoch": 88.05343511450381,
+      "grad_norm": 0.23986873030662537,
+      "learning_rate": 1.3896732932303485e-06,
+      "loss": 0.0698,
+      "step": 23070
+    },
+    {
+      "epoch": 88.09160305343511,
+      "grad_norm": 0.26337021589279175,
+      "learning_rate": 1.3858624490056304e-06,
+      "loss": 0.067,
+      "step": 23080
+    },
+    {
+      "epoch": 88.12977099236642,
+      "grad_norm": 0.4079955220222473,
+      "learning_rate": 1.38205599629883e-06,
+      "loss": 0.0705,
+      "step": 23090
+    },
+    {
+      "epoch": 88.16793893129771,
+      "grad_norm": 0.5138763785362244,
+      "learning_rate": 1.378253939735142e-06,
+      "loss": 0.0691,
+      "step": 23100
+    },
+    {
+      "epoch": 88.20610687022901,
+      "grad_norm": 0.4184453785419464,
+      "learning_rate": 1.3744562839344267e-06,
+      "loss": 0.0688,
+      "step": 23110
+    },
+    {
+      "epoch": 88.2442748091603,
+      "grad_norm": 0.24194218218326569,
+      "learning_rate": 1.3706630335111932e-06,
+      "loss": 0.0689,
+      "step": 23120
+    },
+    {
+      "epoch": 88.2824427480916,
+      "grad_norm": 0.48007258772850037,
+      "learning_rate": 1.3668741930745966e-06,
+      "loss": 0.0766,
+      "step": 23130
+    },
+    {
+      "epoch": 88.3206106870229,
+      "grad_norm": 0.2578524351119995,
+      "learning_rate": 1.3630897672284382e-06,
+      "loss": 0.0634,
+      "step": 23140
+    },
+    {
+      "epoch": 88.3587786259542,
+      "grad_norm": 0.40297532081604004,
+      "learning_rate": 1.3593097605711508e-06,
+      "loss": 0.0656,
+      "step": 23150
+    },
+    {
+      "epoch": 88.3969465648855,
+      "grad_norm": 0.31493544578552246,
+      "learning_rate": 1.3555341776957992e-06,
+      "loss": 0.0729,
+      "step": 23160
+    },
+    {
+      "epoch": 88.43511450381679,
+      "grad_norm": 0.2946300208568573,
+      "learning_rate": 1.3517630231900724e-06,
+      "loss": 0.0722,
+      "step": 23170
+    },
+    {
+      "epoch": 88.47328244274809,
+      "grad_norm": 0.23656029999256134,
+      "learning_rate": 1.3479963016362768e-06,
+      "loss": 0.0721,
+      "step": 23180
+    },
+    {
+      "epoch": 88.5114503816794,
+      "grad_norm": 0.3117351233959198,
+      "learning_rate": 1.3442340176113378e-06,
+      "loss": 0.0699,
+      "step": 23190
+    },
+    {
+      "epoch": 88.54961832061069,
+      "grad_norm": 0.4488414227962494,
+      "learning_rate": 1.340476175686784e-06,
+      "loss": 0.0683,
+      "step": 23200
+    },
+    {
+      "epoch": 88.58778625954199,
+      "grad_norm": 0.3813345730304718,
+      "learning_rate": 1.336722780428747e-06,
+      "loss": 0.0699,
+      "step": 23210
+    },
+    {
+      "epoch": 88.62595419847328,
+      "grad_norm": 0.2527889311313629,
+      "learning_rate": 1.3329738363979561e-06,
+      "loss": 0.0717,
+      "step": 23220
+    },
+    {
+      "epoch": 88.66412213740458,
+      "grad_norm": 0.5992899537086487,
+      "learning_rate": 1.329229348149731e-06,
+      "loss": 0.07,
+      "step": 23230
+    },
+    {
+      "epoch": 88.70229007633588,
+      "grad_norm": 1.4383306503295898,
+      "learning_rate": 1.3254893202339798e-06,
+      "loss": 0.0725,
+      "step": 23240
+    },
+    {
+      "epoch": 88.74045801526718,
+      "grad_norm": 0.48335134983062744,
+      "learning_rate": 1.3217537571951872e-06,
+      "loss": 0.0736,
+      "step": 23250
+    },
+    {
+      "epoch": 88.77862595419847,
+      "grad_norm": 0.23111307621002197,
+      "learning_rate": 1.3180226635724169e-06,
+      "loss": 0.0851,
+      "step": 23260
+    },
+    {
+      "epoch": 88.81679389312977,
+      "grad_norm": 2.051677942276001,
+      "learning_rate": 1.314296043899298e-06,
+      "loss": 0.0734,
+      "step": 23270
+    },
+    {
+      "epoch": 88.85496183206106,
+      "grad_norm": 0.24527736008167267,
+      "learning_rate": 1.3105739027040248e-06,
+      "loss": 0.0708,
+      "step": 23280
+    },
+    {
+      "epoch": 88.89312977099236,
+      "grad_norm": 0.3417549729347229,
+      "learning_rate": 1.306856244509349e-06,
+      "loss": 0.0741,
+      "step": 23290
+    },
+    {
+      "epoch": 88.93129770992367,
+      "grad_norm": 0.42347151041030884,
+      "learning_rate": 1.3031430738325778e-06,
+      "loss": 0.0738,
+      "step": 23300
+    },
+    {
+      "epoch": 88.96946564885496,
+      "grad_norm": 0.3796163499355316,
+      "learning_rate": 1.299434395185563e-06,
+      "loss": 0.0675,
+      "step": 23310
+    },
+    {
+      "epoch": 89.00763358778626,
+      "grad_norm": 0.5870466828346252,
+      "learning_rate": 1.295730213074699e-06,
+      "loss": 0.0689,
+      "step": 23320
+    },
+    {
+      "epoch": 89.04580152671755,
+      "grad_norm": 0.3642314076423645,
+      "learning_rate": 1.2920305320009153e-06,
+      "loss": 0.064,
+      "step": 23330
+    },
+    {
+      "epoch": 89.08396946564885,
+      "grad_norm": 0.23387549817562103,
+      "learning_rate": 1.2883353564596729e-06,
+      "loss": 0.0675,
+      "step": 23340
+    },
+    {
+      "epoch": 89.12213740458016,
+      "grad_norm": 0.29250991344451904,
+      "learning_rate": 1.2846446909409605e-06,
+      "loss": 0.0627,
+      "step": 23350
+    },
+    {
+      "epoch": 89.16030534351145,
+      "grad_norm": 0.3102540075778961,
+      "learning_rate": 1.280958539929284e-06,
+      "loss": 0.0674,
+      "step": 23360
+    },
+    {
+      "epoch": 89.19847328244275,
+      "grad_norm": 0.2419544756412506,
+      "learning_rate": 1.2772769079036639e-06,
+      "loss": 0.0644,
+      "step": 23370
+    },
+    {
+      "epoch": 89.23664122137404,
+      "grad_norm": 0.26334354281425476,
+      "learning_rate": 1.2735997993376299e-06,
+      "loss": 0.0717,
+      "step": 23380
+    },
+    {
+      "epoch": 89.27480916030534,
+      "grad_norm": 0.30459722876548767,
+      "learning_rate": 1.2699272186992168e-06,
+      "loss": 0.0671,
+      "step": 23390
+    },
+    {
+      "epoch": 89.31297709923665,
+      "grad_norm": 0.49876540899276733,
+      "learning_rate": 1.2662591704509548e-06,
+      "loss": 0.068,
+      "step": 23400
+    },
+    {
+      "epoch": 89.35114503816794,
+      "grad_norm": 0.8938238620758057,
+      "learning_rate": 1.2625956590498712e-06,
+      "loss": 0.0736,
+      "step": 23410
+    },
+    {
+      "epoch": 89.38931297709924,
+      "grad_norm": 0.3588089942932129,
+      "learning_rate": 1.2589366889474758e-06,
+      "loss": 0.0731,
+      "step": 23420
+    },
+    {
+      "epoch": 89.42748091603053,
+      "grad_norm": 0.24369336664676666,
+      "learning_rate": 1.2552822645897623e-06,
+      "loss": 0.0688,
+      "step": 23430
+    },
+    {
+      "epoch": 89.46564885496183,
+      "grad_norm": 0.9177067279815674,
+      "learning_rate": 1.2516323904172001e-06,
+      "loss": 0.0654,
+      "step": 23440
+    },
+    {
+      "epoch": 89.50381679389314,
+      "grad_norm": 0.2567913830280304,
+      "learning_rate": 1.2479870708647324e-06,
+      "loss": 0.0669,
+      "step": 23450
+    },
+    {
+      "epoch": 89.54198473282443,
+      "grad_norm": 1.152028203010559,
+      "learning_rate": 1.2443463103617658e-06,
+      "loss": 0.066,
+      "step": 23460
+    },
+    {
+      "epoch": 89.58015267175573,
+      "grad_norm": 0.28860822319984436,
+      "learning_rate": 1.240710113332167e-06,
+      "loss": 0.068,
+      "step": 23470
+    },
+    {
+      "epoch": 89.61832061068702,
+      "grad_norm": 0.17750433087348938,
+      "learning_rate": 1.2370784841942584e-06,
+      "loss": 0.0676,
+      "step": 23480
+    },
+    {
+      "epoch": 89.65648854961832,
+      "grad_norm": 0.28479427099227905,
+      "learning_rate": 1.2334514273608117e-06,
+      "loss": 0.0668,
+      "step": 23490
+    },
+    {
+      "epoch": 89.69465648854961,
+      "grad_norm": 0.2883140444755554,
+      "learning_rate": 1.2298289472390417e-06,
+      "loss": 0.0837,
+      "step": 23500
+    },
+    {
+      "epoch": 89.73282442748092,
+      "grad_norm": 0.22831976413726807,
+      "learning_rate": 1.226211048230606e-06,
+      "loss": 0.073,
+      "step": 23510
+    },
+    {
+      "epoch": 89.77099236641222,
+      "grad_norm": 0.20338012278079987,
+      "learning_rate": 1.222597734731592e-06,
+      "loss": 0.0617,
+      "step": 23520
+    },
+    {
+      "epoch": 89.80916030534351,
+      "grad_norm": 0.4108172655105591,
+      "learning_rate": 1.2189890111325149e-06,
+      "loss": 0.0682,
+      "step": 23530
+    },
+    {
+      "epoch": 89.8473282442748,
+      "grad_norm": 0.3931494951248169,
+      "learning_rate": 1.2153848818183161e-06,
+      "loss": 0.0656,
+      "step": 23540
+    },
+    {
+      "epoch": 89.8854961832061,
+      "grad_norm": 0.5028300285339355,
+      "learning_rate": 1.2117853511683509e-06,
+      "loss": 0.0717,
+      "step": 23550
+    },
+    {
+      "epoch": 89.92366412213741,
+      "grad_norm": 0.5321047902107239,
+      "learning_rate": 1.2081904235563908e-06,
+      "loss": 0.0713,
+      "step": 23560
+    },
+    {
+      "epoch": 89.9618320610687,
+      "grad_norm": 0.40350303053855896,
+      "learning_rate": 1.20460010335061e-06,
+      "loss": 0.0666,
+      "step": 23570
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 0.23981277644634247,
+      "learning_rate": 1.2010143949135866e-06,
+      "loss": 0.0726,
+      "step": 23580
+    },
+    {
+      "epoch": 90.0381679389313,
+      "grad_norm": 0.3623555600643158,
+      "learning_rate": 1.1974333026022939e-06,
+      "loss": 0.0692,
+      "step": 23590
+    },
+    {
+      "epoch": 90.07633587786259,
+      "grad_norm": 0.23631367087364197,
+      "learning_rate": 1.1938568307680965e-06,
+      "loss": 0.0698,
+      "step": 23600
+    },
+    {
+      "epoch": 90.1145038167939,
+      "grad_norm": 0.4967484176158905,
+      "learning_rate": 1.1902849837567466e-06,
+      "loss": 0.0695,
+      "step": 23610
+    },
+    {
+      "epoch": 90.1526717557252,
+      "grad_norm": 0.24320781230926514,
+      "learning_rate": 1.186717765908374e-06,
+      "loss": 0.0705,
+      "step": 23620
+    },
+    {
+      "epoch": 90.19083969465649,
+      "grad_norm": 0.7703472375869751,
+      "learning_rate": 1.1831551815574848e-06,
+      "loss": 0.0642,
+      "step": 23630
+    },
+    {
+      "epoch": 90.22900763358778,
+      "grad_norm": 0.24756668508052826,
+      "learning_rate": 1.1795972350329554e-06,
+      "loss": 0.0712,
+      "step": 23640
+    },
+    {
+      "epoch": 90.26717557251908,
+      "grad_norm": 0.45252394676208496,
+      "learning_rate": 1.176043930658024e-06,
+      "loss": 0.0674,
+      "step": 23650
+    },
+    {
+      "epoch": 90.30534351145039,
+      "grad_norm": 0.47805851697921753,
+      "learning_rate": 1.1724952727502942e-06,
+      "loss": 0.0754,
+      "step": 23660
+    },
+    {
+      "epoch": 90.34351145038168,
+      "grad_norm": 0.6508886814117432,
+      "learning_rate": 1.1689512656217178e-06,
+      "loss": 0.0732,
+      "step": 23670
+    },
+    {
+      "epoch": 90.38167938931298,
+      "grad_norm": 0.37030789256095886,
+      "learning_rate": 1.1654119135785964e-06,
+      "loss": 0.072,
+      "step": 23680
+    },
+    {
+      "epoch": 90.41984732824427,
+      "grad_norm": 0.43612009286880493,
+      "learning_rate": 1.1618772209215795e-06,
+      "loss": 0.0714,
+      "step": 23690
+    },
+    {
+      "epoch": 90.45801526717557,
+      "grad_norm": 0.3412275016307831,
+      "learning_rate": 1.1583471919456506e-06,
+      "loss": 0.0713,
+      "step": 23700
+    },
+    {
+      "epoch": 90.49618320610686,
+      "grad_norm": 0.24611587822437286,
+      "learning_rate": 1.1548218309401267e-06,
+      "loss": 0.075,
+      "step": 23710
+    },
+    {
+      "epoch": 90.53435114503817,
+      "grad_norm": 0.4314908981323242,
+      "learning_rate": 1.1513011421886554e-06,
+      "loss": 0.0646,
+      "step": 23720
+    },
+    {
+      "epoch": 90.57251908396947,
+      "grad_norm": 0.38820695877075195,
+      "learning_rate": 1.1477851299692056e-06,
+      "loss": 0.0712,
+      "step": 23730
+    },
+    {
+      "epoch": 90.61068702290076,
+      "grad_norm": 0.28595876693725586,
+      "learning_rate": 1.1442737985540631e-06,
+      "loss": 0.0646,
+      "step": 23740
+    },
+    {
+      "epoch": 90.64885496183206,
+      "grad_norm": 0.6347960233688354,
+      "learning_rate": 1.1407671522098262e-06,
+      "loss": 0.0855,
+      "step": 23750
+    },
+    {
+      "epoch": 90.68702290076335,
+      "grad_norm": 0.19028426706790924,
+      "learning_rate": 1.1372651951974001e-06,
+      "loss": 0.0702,
+      "step": 23760
+    },
+    {
+      "epoch": 90.72519083969466,
+      "grad_norm": 0.49181249737739563,
+      "learning_rate": 1.1337679317719952e-06,
+      "loss": 0.0758,
+      "step": 23770
+    },
+    {
+      "epoch": 90.76335877862596,
+      "grad_norm": 0.3415592312812805,
+      "learning_rate": 1.130275366183115e-06,
+      "loss": 0.0721,
+      "step": 23780
+    },
+    {
+      "epoch": 90.80152671755725,
+      "grad_norm": 0.3656644821166992,
+      "learning_rate": 1.1267875026745562e-06,
+      "loss": 0.0716,
+      "step": 23790
+    },
+    {
+      "epoch": 90.83969465648855,
+      "grad_norm": 0.2646436393260956,
+      "learning_rate": 1.1233043454844017e-06,
+      "loss": 0.067,
+      "step": 23800
+    },
+    {
+      "epoch": 90.87786259541984,
+      "grad_norm": 0.7572306394577026,
+      "learning_rate": 1.1198258988450145e-06,
+      "loss": 0.0698,
+      "step": 23810
+    },
+    {
+      "epoch": 90.91603053435115,
+      "grad_norm": 0.42010462284088135,
+      "learning_rate": 1.116352166983037e-06,
+      "loss": 0.0675,
+      "step": 23820
+    },
+    {
+      "epoch": 90.95419847328245,
+      "grad_norm": 0.26916852593421936,
+      "learning_rate": 1.112883154119382e-06,
+      "loss": 0.0651,
+      "step": 23830
+    },
+    {
+      "epoch": 90.99236641221374,
+      "grad_norm": 0.30210235714912415,
+      "learning_rate": 1.1094188644692255e-06,
+      "loss": 0.0734,
+      "step": 23840
+    },
+    {
+      "epoch": 91.03053435114504,
+      "grad_norm": 0.41560983657836914,
+      "learning_rate": 1.1059593022420067e-06,
+      "loss": 0.063,
+      "step": 23850
+    },
+    {
+      "epoch": 91.06870229007633,
+      "grad_norm": 0.37110573053359985,
+      "learning_rate": 1.1025044716414185e-06,
+      "loss": 0.0748,
+      "step": 23860
+    },
+    {
+      "epoch": 91.10687022900764,
+      "grad_norm": 0.5967647433280945,
+      "learning_rate": 1.0990543768654084e-06,
+      "loss": 0.0675,
+      "step": 23870
+    },
+    {
+      "epoch": 91.14503816793894,
+      "grad_norm": 0.2824907600879669,
+      "learning_rate": 1.0956090221061655e-06,
+      "loss": 0.0612,
+      "step": 23880
+    },
+    {
+      "epoch": 91.18320610687023,
+      "grad_norm": 0.536435604095459,
+      "learning_rate": 1.0921684115501208e-06,
+      "loss": 0.06,
+      "step": 23890
+    },
+    {
+      "epoch": 91.22137404580153,
+      "grad_norm": 0.25722911953926086,
+      "learning_rate": 1.0887325493779405e-06,
+      "loss": 0.0671,
+      "step": 23900
+    },
+    {
+      "epoch": 91.25954198473282,
+      "grad_norm": 0.28526943922042847,
+      "learning_rate": 1.0853014397645205e-06,
+      "loss": 0.0684,
+      "step": 23910
+    },
+    {
+      "epoch": 91.29770992366412,
+      "grad_norm": 0.28562769293785095,
+      "learning_rate": 1.0818750868789828e-06,
+      "loss": 0.0637,
+      "step": 23920
+    },
+    {
+      "epoch": 91.33587786259542,
+      "grad_norm": 0.3431946039199829,
+      "learning_rate": 1.0784534948846704e-06,
+      "loss": 0.0651,
+      "step": 23930
+    },
+    {
+      "epoch": 91.37404580152672,
+      "grad_norm": 1.2695426940917969,
+      "learning_rate": 1.0750366679391393e-06,
+      "loss": 0.0718,
+      "step": 23940
+    },
+    {
+      "epoch": 91.41221374045801,
+      "grad_norm": 0.27434664964675903,
+      "learning_rate": 1.0716246101941558e-06,
+      "loss": 0.0692,
+      "step": 23950
+    },
+    {
+      "epoch": 91.45038167938931,
+      "grad_norm": 0.30581653118133545,
+      "learning_rate": 1.0682173257956935e-06,
+      "loss": 0.0685,
+      "step": 23960
+    },
+    {
+      "epoch": 91.4885496183206,
+      "grad_norm": 0.2575470209121704,
+      "learning_rate": 1.064814818883922e-06,
+      "loss": 0.0675,
+      "step": 23970
+    },
+    {
+      "epoch": 91.52671755725191,
+      "grad_norm": 0.31820449233055115,
+      "learning_rate": 1.06141709359321e-06,
+      "loss": 0.0752,
+      "step": 23980
+    },
+    {
+      "epoch": 91.56488549618321,
+      "grad_norm": 0.5742216110229492,
+      "learning_rate": 1.0580241540521142e-06,
+      "loss": 0.0847,
+      "step": 23990
+    },
+    {
+      "epoch": 91.6030534351145,
+      "grad_norm": 0.545124351978302,
+      "learning_rate": 1.0546360043833747e-06,
+      "loss": 0.0692,
+      "step": 24000
+    },
+    {
+      "epoch": 91.6412213740458,
+      "grad_norm": 0.27798354625701904,
+      "learning_rate": 1.0512526487039138e-06,
+      "loss": 0.071,
+      "step": 24010
+    },
+    {
+      "epoch": 91.6793893129771,
+      "grad_norm": 0.6156341433525085,
+      "learning_rate": 1.0478740911248259e-06,
+      "loss": 0.0694,
+      "step": 24020
+    },
+    {
+      "epoch": 91.7175572519084,
+      "grad_norm": 0.318088173866272,
+      "learning_rate": 1.0445003357513794e-06,
+      "loss": 0.0695,
+      "step": 24030
+    },
+    {
+      "epoch": 91.7557251908397,
+      "grad_norm": 0.28530198335647583,
+      "learning_rate": 1.0411313866830042e-06,
+      "loss": 0.0656,
+      "step": 24040
+    },
+    {
+      "epoch": 91.79389312977099,
+      "grad_norm": 0.717268168926239,
+      "learning_rate": 1.0377672480132917e-06,
+      "loss": 0.0689,
+      "step": 24050
+    },
+    {
+      "epoch": 91.83206106870229,
+      "grad_norm": 0.2844495475292206,
+      "learning_rate": 1.0344079238299865e-06,
+      "loss": 0.0672,
+      "step": 24060
+    },
+    {
+      "epoch": 91.87022900763358,
+      "grad_norm": 0.8445849418640137,
+      "learning_rate": 1.0310534182149835e-06,
+      "loss": 0.074,
+      "step": 24070
+    },
+    {
+      "epoch": 91.90839694656489,
+      "grad_norm": 0.6028487682342529,
+      "learning_rate": 1.0277037352443258e-06,
+      "loss": 0.0714,
+      "step": 24080
+    },
+    {
+      "epoch": 91.94656488549619,
+      "grad_norm": 0.29483121633529663,
+      "learning_rate": 1.0243588789881931e-06,
+      "loss": 0.0693,
+      "step": 24090
+    },
+    {
+      "epoch": 91.98473282442748,
+      "grad_norm": 0.2845158576965332,
+      "learning_rate": 1.0210188535108995e-06,
+      "loss": 0.0692,
+      "step": 24100
+    },
+    {
+      "epoch": 92.02290076335878,
+      "grad_norm": 0.28095564246177673,
+      "learning_rate": 1.0176836628708937e-06,
+      "loss": 0.0705,
+      "step": 24110
+    },
+    {
+      "epoch": 92.06106870229007,
+      "grad_norm": 0.6553061008453369,
+      "learning_rate": 1.0143533111207455e-06,
+      "loss": 0.0724,
+      "step": 24120
+    },
+    {
+      "epoch": 92.09923664122137,
+      "grad_norm": 0.22242268919944763,
+      "learning_rate": 1.0110278023071445e-06,
+      "loss": 0.0688,
+      "step": 24130
+    },
+    {
+      "epoch": 92.13740458015268,
+      "grad_norm": 0.4113728106021881,
+      "learning_rate": 1.007707140470901e-06,
+      "loss": 0.0726,
+      "step": 24140
+    },
+    {
+      "epoch": 92.17557251908397,
+      "grad_norm": 0.22345533967018127,
+      "learning_rate": 1.00439132964693e-06,
+      "loss": 0.0668,
+      "step": 24150
+    },
+    {
+      "epoch": 92.21374045801527,
+      "grad_norm": 0.7509258389472961,
+      "learning_rate": 1.001080373864255e-06,
+      "loss": 0.0783,
+      "step": 24160
+    },
+    {
+      "epoch": 92.25190839694656,
+      "grad_norm": 0.3895869255065918,
+      "learning_rate": 9.977742771459992e-07,
+      "loss": 0.0698,
+      "step": 24170
+    },
+    {
+      "epoch": 92.29007633587786,
+      "grad_norm": 0.360806941986084,
+      "learning_rate": 9.944730435093803e-07,
+      "loss": 0.0658,
+      "step": 24180
+    },
+    {
+      "epoch": 92.32824427480917,
+      "grad_norm": 0.6112274527549744,
+      "learning_rate": 9.911766769657116e-07,
+      "loss": 0.0724,
+      "step": 24190
+    },
+    {
+      "epoch": 92.36641221374046,
+      "grad_norm": 0.5387519598007202,
+      "learning_rate": 9.878851815203883e-07,
+      "loss": 0.0705,
+      "step": 24200
+    },
+    {
+      "epoch": 92.40458015267176,
+      "grad_norm": 0.9864296913146973,
+      "learning_rate": 9.845985611728864e-07,
+      "loss": 0.0681,
+      "step": 24210
+    },
+    {
+      "epoch": 92.44274809160305,
+      "grad_norm": 0.337456613779068,
+      "learning_rate": 9.813168199167604e-07,
+      "loss": 0.0694,
+      "step": 24220
+    },
+    {
+      "epoch": 92.48091603053435,
+      "grad_norm": 0.3321627974510193,
+      "learning_rate": 9.78039961739634e-07,
+      "loss": 0.0657,
+      "step": 24230
+    },
+    {
+      "epoch": 92.51908396946565,
+      "grad_norm": 0.5132209062576294,
+      "learning_rate": 9.747679906232016e-07,
+      "loss": 0.0679,
+      "step": 24240
+    },
+    {
+      "epoch": 92.55725190839695,
+      "grad_norm": 0.503035843372345,
+      "learning_rate": 9.71500910543214e-07,
+      "loss": 0.0673,
+      "step": 24250
+    },
+    {
+      "epoch": 92.59541984732824,
+      "grad_norm": 0.21332845091819763,
+      "learning_rate": 9.682387254694835e-07,
+      "loss": 0.073,
+      "step": 24260
+    },
+    {
+      "epoch": 92.63358778625954,
+      "grad_norm": 0.32724636793136597,
+      "learning_rate": 9.649814393658725e-07,
+      "loss": 0.0662,
+      "step": 24270
+    },
+    {
+      "epoch": 92.67175572519083,
+      "grad_norm": 0.607822597026825,
+      "learning_rate": 9.61729056190288e-07,
+      "loss": 0.0681,
+      "step": 24280
+    },
+    {
+      "epoch": 92.70992366412214,
+      "grad_norm": 0.3636825978755951,
+      "learning_rate": 9.584815798946862e-07,
+      "loss": 0.072,
+      "step": 24290
+    },
+    {
+      "epoch": 92.74809160305344,
+      "grad_norm": 0.22937163710594177,
+      "learning_rate": 9.552390144250552e-07,
+      "loss": 0.0665,
+      "step": 24300
+    },
+    {
+      "epoch": 92.78625954198473,
+      "grad_norm": 0.2677508294582367,
+      "learning_rate": 9.520013637214176e-07,
+      "loss": 0.0675,
+      "step": 24310
+    },
+    {
+      "epoch": 92.82442748091603,
+      "grad_norm": 0.2615829408168793,
+      "learning_rate": 9.487686317178241e-07,
+      "loss": 0.0653,
+      "step": 24320
+    },
+    {
+      "epoch": 92.86259541984732,
+      "grad_norm": 0.2584886848926544,
+      "learning_rate": 9.455408223423496e-07,
+      "loss": 0.0643,
+      "step": 24330
+    },
+    {
+      "epoch": 92.90076335877862,
+      "grad_norm": 0.24199534952640533,
+      "learning_rate": 9.423179395170845e-07,
+      "loss": 0.0702,
+      "step": 24340
+    },
+    {
+      "epoch": 92.93893129770993,
+      "grad_norm": 0.3466721475124359,
+      "learning_rate": 9.390999871581391e-07,
+      "loss": 0.0767,
+      "step": 24350
+    },
+    {
+      "epoch": 92.97709923664122,
+      "grad_norm": 0.5451226830482483,
+      "learning_rate": 9.358869691756273e-07,
+      "loss": 0.0705,
+      "step": 24360
+    },
+    {
+      "epoch": 93.01526717557252,
+      "grad_norm": 0.5285539627075195,
+      "learning_rate": 9.326788894736688e-07,
+      "loss": 0.0692,
+      "step": 24370
+    },
+    {
+      "epoch": 93.05343511450381,
+      "grad_norm": 0.5585796236991882,
+      "learning_rate": 9.294757519503811e-07,
+      "loss": 0.0677,
+      "step": 24380
+    },
+    {
+      "epoch": 93.09160305343511,
+      "grad_norm": 0.31735959649086,
+      "learning_rate": 9.262775604978819e-07,
+      "loss": 0.073,
+      "step": 24390
+    },
+    {
+      "epoch": 93.12977099236642,
+      "grad_norm": 0.2296765148639679,
+      "learning_rate": 9.230843190022726e-07,
+      "loss": 0.0698,
+      "step": 24400
+    },
+    {
+      "epoch": 93.16793893129771,
+      "grad_norm": 0.5008897185325623,
+      "learning_rate": 9.198960313436444e-07,
+      "loss": 0.0874,
+      "step": 24410
+    },
+    {
+      "epoch": 93.20610687022901,
+      "grad_norm": 0.4472108483314514,
+      "learning_rate": 9.16712701396067e-07,
+      "loss": 0.0697,
+      "step": 24420
+    },
+    {
+      "epoch": 93.2442748091603,
+      "grad_norm": 0.2684642970561981,
+      "learning_rate": 9.135343330275864e-07,
+      "loss": 0.0654,
+      "step": 24430
+    },
+    {
+      "epoch": 93.2824427480916,
+      "grad_norm": 0.4187399744987488,
+      "learning_rate": 9.103609301002181e-07,
+      "loss": 0.0692,
+      "step": 24440
+    },
+    {
+      "epoch": 93.3206106870229,
+      "grad_norm": 0.6167700886726379,
+      "learning_rate": 9.071924964699491e-07,
+      "loss": 0.0724,
+      "step": 24450
+    },
+    {
+      "epoch": 93.3587786259542,
+      "grad_norm": 0.34823620319366455,
+      "learning_rate": 9.040290359867232e-07,
+      "loss": 0.0674,
+      "step": 24460
+    },
+    {
+      "epoch": 93.3969465648855,
+      "grad_norm": 0.27436694502830505,
+      "learning_rate": 9.008705524944439e-07,
+      "loss": 0.0692,
+      "step": 24470
+    },
+    {
+      "epoch": 93.43511450381679,
+      "grad_norm": 0.2943238914012909,
+      "learning_rate": 8.977170498309651e-07,
+      "loss": 0.0635,
+      "step": 24480
+    },
+    {
+      "epoch": 93.47328244274809,
+      "grad_norm": 0.6278515458106995,
+      "learning_rate": 8.945685318280917e-07,
+      "loss": 0.079,
+      "step": 24490
+    },
+    {
+      "epoch": 93.5114503816794,
+      "grad_norm": 0.23879125714302063,
+      "learning_rate": 8.914250023115672e-07,
+      "loss": 0.0696,
+      "step": 24500
+    },
+    {
+      "epoch": 93.54961832061069,
+      "grad_norm": 0.35707348585128784,
+      "learning_rate": 8.882864651010798e-07,
+      "loss": 0.0733,
+      "step": 24510
+    },
+    {
+      "epoch": 93.58778625954199,
+      "grad_norm": 0.3247864246368408,
+      "learning_rate": 8.851529240102464e-07,
+      "loss": 0.0663,
+      "step": 24520
+    },
+    {
+      "epoch": 93.62595419847328,
+      "grad_norm": 0.4654683768749237,
+      "learning_rate": 8.820243828466135e-07,
+      "loss": 0.072,
+      "step": 24530
+    },
+    {
+      "epoch": 93.66412213740458,
+      "grad_norm": 0.504030704498291,
+      "learning_rate": 8.789008454116566e-07,
+      "loss": 0.0712,
+      "step": 24540
+    },
+    {
+      "epoch": 93.70229007633588,
+      "grad_norm": 0.6368444561958313,
+      "learning_rate": 8.757823155007655e-07,
+      "loss": 0.0685,
+      "step": 24550
+    },
+    {
+      "epoch": 93.74045801526718,
+      "grad_norm": 0.28386011719703674,
+      "learning_rate": 8.72668796903251e-07,
+      "loss": 0.0714,
+      "step": 24560
+    },
+    {
+      "epoch": 93.77862595419847,
+      "grad_norm": 0.3366830348968506,
+      "learning_rate": 8.6956029340233e-07,
+      "loss": 0.0704,
+      "step": 24570
+    },
+    {
+      "epoch": 93.81679389312977,
+      "grad_norm": 0.391422837972641,
+      "learning_rate": 8.664568087751274e-07,
+      "loss": 0.069,
+      "step": 24580
+    },
+    {
+      "epoch": 93.85496183206106,
+      "grad_norm": 0.41178223490715027,
+      "learning_rate": 8.633583467926698e-07,
+      "loss": 0.0687,
+      "step": 24590
+    },
+    {
+      "epoch": 93.89312977099236,
+      "grad_norm": 0.23154325783252716,
+      "learning_rate": 8.602649112198796e-07,
+      "loss": 0.0677,
+      "step": 24600
+    },
+    {
+      "epoch": 93.93129770992367,
+      "grad_norm": 0.3986658751964569,
+      "learning_rate": 8.571765058155745e-07,
+      "loss": 0.0698,
+      "step": 24610
+    },
+    {
+      "epoch": 93.96946564885496,
+      "grad_norm": 0.3332964777946472,
+      "learning_rate": 8.540931343324582e-07,
+      "loss": 0.0714,
+      "step": 24620
+    },
+    {
+      "epoch": 94.00763358778626,
+      "grad_norm": 0.2473592758178711,
+      "learning_rate": 8.510148005171171e-07,
+      "loss": 0.0684,
+      "step": 24630
+    },
+    {
+      "epoch": 94.04580152671755,
+      "grad_norm": 0.5109586715698242,
+      "learning_rate": 8.479415081100167e-07,
+      "loss": 0.0721,
+      "step": 24640
+    },
+    {
+      "epoch": 94.08396946564885,
+      "grad_norm": 0.2871462404727936,
+      "learning_rate": 8.448732608454968e-07,
+      "loss": 0.065,
+      "step": 24650
+    },
+    {
+      "epoch": 94.12213740458016,
+      "grad_norm": 0.27044838666915894,
+      "learning_rate": 8.418100624517688e-07,
+      "loss": 0.0678,
+      "step": 24660
+    },
+    {
+      "epoch": 94.16030534351145,
+      "grad_norm": 0.2179548740386963,
+      "learning_rate": 8.387519166509062e-07,
+      "loss": 0.0654,
+      "step": 24670
+    },
+    {
+      "epoch": 94.19847328244275,
+      "grad_norm": 0.2881402373313904,
+      "learning_rate": 8.356988271588445e-07,
+      "loss": 0.0738,
+      "step": 24680
+    },
+    {
+      "epoch": 94.23664122137404,
+      "grad_norm": 0.23480482399463654,
+      "learning_rate": 8.326507976853765e-07,
+      "loss": 0.067,
+      "step": 24690
+    },
+    {
+      "epoch": 94.27480916030534,
+      "grad_norm": 0.5004714131355286,
+      "learning_rate": 8.296078319341444e-07,
+      "loss": 0.0707,
+      "step": 24700
+    },
+    {
+      "epoch": 94.31297709923665,
+      "grad_norm": 0.41305282711982727,
+      "learning_rate": 8.265699336026384e-07,
+      "loss": 0.0795,
+      "step": 24710
+    },
+    {
+      "epoch": 94.35114503816794,
+      "grad_norm": 0.7068077325820923,
+      "learning_rate": 8.235371063821923e-07,
+      "loss": 0.0639,
+      "step": 24720
+    },
+    {
+      "epoch": 94.38931297709924,
+      "grad_norm": 0.3450511693954468,
+      "learning_rate": 8.205093539579768e-07,
+      "loss": 0.0691,
+      "step": 24730
+    },
+    {
+      "epoch": 94.42748091603053,
+      "grad_norm": 0.3504796624183655,
+      "learning_rate": 8.174866800089964e-07,
+      "loss": 0.0697,
+      "step": 24740
+    },
+    {
+      "epoch": 94.46564885496183,
+      "grad_norm": 0.9486528635025024,
+      "learning_rate": 8.144690882080853e-07,
+      "loss": 0.0706,
+      "step": 24750
+    },
+    {
+      "epoch": 94.50381679389314,
+      "grad_norm": 0.4473707973957062,
+      "learning_rate": 8.114565822219006e-07,
+      "loss": 0.0683,
+      "step": 24760
+    },
+    {
+      "epoch": 94.54198473282443,
+      "grad_norm": 0.2520901560783386,
+      "learning_rate": 8.084491657109233e-07,
+      "loss": 0.0697,
+      "step": 24770
+    },
+    {
+      "epoch": 94.58015267175573,
+      "grad_norm": 0.31963804364204407,
+      "learning_rate": 8.054468423294476e-07,
+      "loss": 0.072,
+      "step": 24780
+    },
+    {
+      "epoch": 94.61832061068702,
+      "grad_norm": 0.6297641396522522,
+      "learning_rate": 8.024496157255784e-07,
+      "loss": 0.0716,
+      "step": 24790
+    },
+    {
+      "epoch": 94.65648854961832,
+      "grad_norm": 0.4397490322589874,
+      "learning_rate": 7.994574895412294e-07,
+      "loss": 0.0687,
+      "step": 24800
+    },
+    {
+      "epoch": 94.69465648854961,
+      "grad_norm": 0.3805553615093231,
+      "learning_rate": 7.964704674121149e-07,
+      "loss": 0.0654,
+      "step": 24810
+    },
+    {
+      "epoch": 94.73282442748092,
+      "grad_norm": 0.3672832250595093,
+      "learning_rate": 7.934885529677505e-07,
+      "loss": 0.068,
+      "step": 24820
+    },
+    {
+      "epoch": 94.77099236641222,
+      "grad_norm": 0.2700408399105072,
+      "learning_rate": 7.905117498314413e-07,
+      "loss": 0.0652,
+      "step": 24830
+    },
+    {
+      "epoch": 94.80916030534351,
+      "grad_norm": 0.20384560525417328,
+      "learning_rate": 7.875400616202861e-07,
+      "loss": 0.074,
+      "step": 24840
+    },
+    {
+      "epoch": 94.8473282442748,
+      "grad_norm": 0.27636826038360596,
+      "learning_rate": 7.845734919451647e-07,
+      "loss": 0.0737,
+      "step": 24850
+    },
+    {
+      "epoch": 94.8854961832061,
+      "grad_norm": 0.9613207578659058,
+      "learning_rate": 7.816120444107384e-07,
+      "loss": 0.0823,
+      "step": 24860
+    },
+    {
+      "epoch": 94.92366412213741,
+      "grad_norm": 0.3635960817337036,
+      "learning_rate": 7.786557226154473e-07,
+      "loss": 0.0747,
+      "step": 24870
+    },
+    {
+      "epoch": 94.9618320610687,
+      "grad_norm": 0.276128888130188,
+      "learning_rate": 7.757045301514998e-07,
+      "loss": 0.0715,
+      "step": 24880
+    },
+    {
+      "epoch": 95.0,
+      "grad_norm": 0.4147992432117462,
+      "learning_rate": 7.727584706048735e-07,
+      "loss": 0.0695,
+      "step": 24890
+    },
+    {
+      "epoch": 95.0381679389313,
+      "grad_norm": 0.3812028467655182,
+      "learning_rate": 7.698175475553076e-07,
+      "loss": 0.0689,
+      "step": 24900
+    },
+    {
+      "epoch": 95.07633587786259,
+      "grad_norm": 0.3648328483104706,
+      "learning_rate": 7.668817645763021e-07,
+      "loss": 0.073,
+      "step": 24910
+    },
+    {
+      "epoch": 95.1145038167939,
+      "grad_norm": 0.4475592076778412,
+      "learning_rate": 7.63951125235109e-07,
+      "loss": 0.0769,
+      "step": 24920
+    },
+    {
+      "epoch": 95.1526717557252,
+      "grad_norm": 0.3380667269229889,
+      "learning_rate": 7.610256330927323e-07,
+      "loss": 0.0676,
+      "step": 24930
+    },
+    {
+      "epoch": 95.19083969465649,
+      "grad_norm": 0.8378775119781494,
+      "learning_rate": 7.581052917039211e-07,
+      "loss": 0.0733,
+      "step": 24940
+    },
+    {
+      "epoch": 95.22900763358778,
+      "grad_norm": 0.21355819702148438,
+      "learning_rate": 7.551901046171645e-07,
+      "loss": 0.0696,
+      "step": 24950
+    },
+    {
+      "epoch": 95.26717557251908,
+      "grad_norm": 0.41850680112838745,
+      "learning_rate": 7.522800753746895e-07,
+      "loss": 0.0662,
+      "step": 24960
+    },
+    {
+      "epoch": 95.30534351145039,
+      "grad_norm": 0.3715193271636963,
+      "learning_rate": 7.493752075124577e-07,
+      "loss": 0.0705,
+      "step": 24970
+    },
+    {
+      "epoch": 95.34351145038168,
+      "grad_norm": 0.6367314457893372,
+      "learning_rate": 7.464755045601557e-07,
+      "loss": 0.0645,
+      "step": 24980
+    },
+    {
+      "epoch": 95.38167938931298,
+      "grad_norm": 0.2814936935901642,
+      "learning_rate": 7.435809700411972e-07,
+      "loss": 0.0774,
+      "step": 24990
+    },
+    {
+      "epoch": 95.41984732824427,
+      "grad_norm": 0.22556699812412262,
+      "learning_rate": 7.40691607472715e-07,
+      "loss": 0.067,
+      "step": 25000
+    },
+    {
+      "epoch": 95.45801526717557,
+      "grad_norm": 0.8220465183258057,
+      "learning_rate": 7.378074203655561e-07,
+      "loss": 0.0731,
+      "step": 25010
+    },
+    {
+      "epoch": 95.49618320610686,
+      "grad_norm": 0.2574046552181244,
+      "learning_rate": 7.349284122242783e-07,
+      "loss": 0.0736,
+      "step": 25020
+    },
+    {
+      "epoch": 95.53435114503817,
+      "grad_norm": 0.5456118583679199,
+      "learning_rate": 7.320545865471513e-07,
+      "loss": 0.0725,
+      "step": 25030
+    },
+    {
+      "epoch": 95.57251908396947,
+      "grad_norm": 0.6138007044792175,
+      "learning_rate": 7.291859468261426e-07,
+      "loss": 0.0662,
+      "step": 25040
+    },
+    {
+      "epoch": 95.61068702290076,
+      "grad_norm": 0.7047519683837891,
+      "learning_rate": 7.263224965469195e-07,
+      "loss": 0.0729,
+      "step": 25050
+    },
+    {
+      "epoch": 95.64885496183206,
+      "grad_norm": 0.20071052014827728,
+      "learning_rate": 7.234642391888446e-07,
+      "loss": 0.0697,
+      "step": 25060
+    },
+    {
+      "epoch": 95.68702290076335,
+      "grad_norm": 0.43756791949272156,
+      "learning_rate": 7.206111782249698e-07,
+      "loss": 0.0681,
+      "step": 25070
+    },
+    {
+      "epoch": 95.72519083969466,
+      "grad_norm": 0.4008914530277252,
+      "learning_rate": 7.177633171220339e-07,
+      "loss": 0.0742,
+      "step": 25080
+    },
+    {
+      "epoch": 95.76335877862596,
+      "grad_norm": 0.9968006014823914,
+      "learning_rate": 7.149206593404562e-07,
+      "loss": 0.0692,
+      "step": 25090
+    },
+    {
+      "epoch": 95.80152671755725,
+      "grad_norm": 0.2601996958255768,
+      "learning_rate": 7.120832083343337e-07,
+      "loss": 0.0672,
+      "step": 25100
+    },
+    {
+      "epoch": 95.83969465648855,
+      "grad_norm": 0.2922241687774658,
+      "learning_rate": 7.092509675514369e-07,
+      "loss": 0.0655,
+      "step": 25110
+    },
+    {
+      "epoch": 95.87786259541984,
+      "grad_norm": 0.40154197812080383,
+      "learning_rate": 7.064239404332063e-07,
+      "loss": 0.0678,
+      "step": 25120
+    },
+    {
+      "epoch": 95.91603053435115,
+      "grad_norm": 0.8301547169685364,
+      "learning_rate": 7.03602130414745e-07,
+      "loss": 0.0677,
+      "step": 25130
+    },
+    {
+      "epoch": 95.95419847328245,
+      "grad_norm": 0.28282907605171204,
+      "learning_rate": 7.007855409248199e-07,
+      "loss": 0.066,
+      "step": 25140
+    },
+    {
+      "epoch": 95.99236641221374,
+      "grad_norm": 0.278046578168869,
+      "learning_rate": 6.979741753858521e-07,
+      "loss": 0.072,
+      "step": 25150
+    },
+    {
+      "epoch": 96.03053435114504,
+      "grad_norm": 0.22601468861103058,
+      "learning_rate": 6.951680372139158e-07,
+      "loss": 0.0701,
+      "step": 25160
+    },
+    {
+      "epoch": 96.06870229007633,
+      "grad_norm": 0.3387024700641632,
+      "learning_rate": 6.923671298187335e-07,
+      "loss": 0.0757,
+      "step": 25170
+    },
+    {
+      "epoch": 96.10687022900764,
+      "grad_norm": 0.26798492670059204,
+      "learning_rate": 6.895714566036705e-07,
+      "loss": 0.0624,
+      "step": 25180
+    },
+    {
+      "epoch": 96.14503816793894,
+      "grad_norm": 0.2048361450433731,
+      "learning_rate": 6.86781020965736e-07,
+      "loss": 0.0675,
+      "step": 25190
+    },
+    {
+      "epoch": 96.18320610687023,
+      "grad_norm": 0.4887497127056122,
+      "learning_rate": 6.839958262955709e-07,
+      "loss": 0.0733,
+      "step": 25200
+    },
+    {
+      "epoch": 96.22137404580153,
+      "grad_norm": 0.37875017523765564,
+      "learning_rate": 6.812158759774489e-07,
+      "loss": 0.065,
+      "step": 25210
+    },
+    {
+      "epoch": 96.25954198473282,
+      "grad_norm": 0.3595539629459381,
+      "learning_rate": 6.784411733892732e-07,
+      "loss": 0.0682,
+      "step": 25220
+    },
+    {
+      "epoch": 96.29770992366412,
+      "grad_norm": 0.44329530000686646,
+      "learning_rate": 6.756717219025666e-07,
+      "loss": 0.0657,
+      "step": 25230
+    },
+    {
+      "epoch": 96.33587786259542,
+      "grad_norm": 0.3791472613811493,
+      "learning_rate": 6.729075248824762e-07,
+      "loss": 0.0641,
+      "step": 25240
+    },
+    {
+      "epoch": 96.37404580152672,
+      "grad_norm": 0.5714471936225891,
+      "learning_rate": 6.701485856877615e-07,
+      "loss": 0.0714,
+      "step": 25250
+    },
+    {
+      "epoch": 96.41221374045801,
+      "grad_norm": 0.5484432578086853,
+      "learning_rate": 6.673949076707925e-07,
+      "loss": 0.0691,
+      "step": 25260
+    },
+    {
+      "epoch": 96.45038167938931,
+      "grad_norm": 0.32033249735832214,
+      "learning_rate": 6.646464941775499e-07,
+      "loss": 0.0655,
+      "step": 25270
+    },
+    {
+      "epoch": 96.4885496183206,
+      "grad_norm": 0.4129583537578583,
+      "learning_rate": 6.619033485476129e-07,
+      "loss": 0.0723,
+      "step": 25280
+    },
+    {
+      "epoch": 96.52671755725191,
+      "grad_norm": 0.216600239276886,
+      "learning_rate": 6.591654741141639e-07,
+      "loss": 0.0672,
+      "step": 25290
+    },
+    {
+      "epoch": 96.56488549618321,
+      "grad_norm": 0.29276999831199646,
+      "learning_rate": 6.564328742039782e-07,
+      "loss": 0.0637,
+      "step": 25300
+    },
+    {
+      "epoch": 96.6030534351145,
+      "grad_norm": 0.20918267965316772,
+      "learning_rate": 6.537055521374219e-07,
+      "loss": 0.0667,
+      "step": 25310
+    },
+    {
+      "epoch": 96.6412213740458,
+      "grad_norm": 0.5599342584609985,
+      "learning_rate": 6.509835112284485e-07,
+      "loss": 0.0699,
+      "step": 25320
+    },
+    {
+      "epoch": 96.6793893129771,
+      "grad_norm": 0.253470242023468,
+      "learning_rate": 6.482667547845944e-07,
+      "loss": 0.069,
+      "step": 25330
+    },
+    {
+      "epoch": 96.7175572519084,
+      "grad_norm": 0.2381214201450348,
+      "learning_rate": 6.455552861069736e-07,
+      "loss": 0.0646,
+      "step": 25340
+    },
+    {
+      "epoch": 96.7557251908397,
+      "grad_norm": 0.3462165296077728,
+      "learning_rate": 6.428491084902788e-07,
+      "loss": 0.0656,
+      "step": 25350
+    },
+    {
+      "epoch": 96.79389312977099,
+      "grad_norm": 0.2496246099472046,
+      "learning_rate": 6.401482252227697e-07,
+      "loss": 0.0689,
+      "step": 25360
+    },
+    {
+      "epoch": 96.83206106870229,
+      "grad_norm": 0.4224330484867096,
+      "learning_rate": 6.37452639586274e-07,
+      "loss": 0.0723,
+      "step": 25370
+    },
+    {
+      "epoch": 96.87022900763358,
+      "grad_norm": 0.22836147248744965,
+      "learning_rate": 6.347623548561827e-07,
+      "loss": 0.0675,
+      "step": 25380
+    },
+    {
+      "epoch": 96.90839694656489,
+      "grad_norm": 0.33980873227119446,
+      "learning_rate": 6.320773743014441e-07,
+      "loss": 0.0696,
+      "step": 25390
+    },
+    {
+      "epoch": 96.94656488549619,
+      "grad_norm": 0.5132980942726135,
+      "learning_rate": 6.293977011845648e-07,
+      "loss": 0.0704,
+      "step": 25400
+    },
+    {
+      "epoch": 96.98473282442748,
+      "grad_norm": 0.2276553362607956,
+      "learning_rate": 6.267233387615984e-07,
+      "loss": 0.068,
+      "step": 25410
+    },
+    {
+      "epoch": 97.02290076335878,
+      "grad_norm": 0.3203546106815338,
+      "learning_rate": 6.2405429028215e-07,
+      "loss": 0.077,
+      "step": 25420
+    },
+    {
+      "epoch": 97.06106870229007,
+      "grad_norm": 0.5133867859840393,
+      "learning_rate": 6.213905589893632e-07,
+      "loss": 0.0696,
+      "step": 25430
+    },
+    {
+      "epoch": 97.09923664122137,
+      "grad_norm": 0.810599684715271,
+      "learning_rate": 6.187321481199221e-07,
+      "loss": 0.07,
+      "step": 25440
+    },
+    {
+      "epoch": 97.13740458015268,
+      "grad_norm": 0.36244019865989685,
+      "learning_rate": 6.16079060904049e-07,
+      "loss": 0.0705,
+      "step": 25450
+    },
+    {
+      "epoch": 97.17557251908397,
+      "grad_norm": 0.383052259683609,
+      "learning_rate": 6.134313005654929e-07,
+      "loss": 0.0666,
+      "step": 25460
+    },
+    {
+      "epoch": 97.21374045801527,
+      "grad_norm": 0.5795229077339172,
+      "learning_rate": 6.107888703215337e-07,
+      "loss": 0.0699,
+      "step": 25470
+    },
+    {
+      "epoch": 97.25190839694656,
+      "grad_norm": 0.37245771288871765,
+      "learning_rate": 6.081517733829723e-07,
+      "loss": 0.0668,
+      "step": 25480
+    },
+    {
+      "epoch": 97.29007633587786,
+      "grad_norm": 0.47016429901123047,
+      "learning_rate": 6.055200129541294e-07,
+      "loss": 0.0689,
+      "step": 25490
+    },
+    {
+      "epoch": 97.32824427480917,
+      "grad_norm": 0.21799707412719727,
+      "learning_rate": 6.028935922328444e-07,
+      "loss": 0.0626,
+      "step": 25500
+    },
+    {
+      "epoch": 97.36641221374046,
+      "grad_norm": 0.2931235432624817,
+      "learning_rate": 6.002725144104649e-07,
+      "loss": 0.0733,
+      "step": 25510
+    },
+    {
+      "epoch": 97.40458015267176,
+      "grad_norm": 0.5701701641082764,
+      "learning_rate": 5.976567826718476e-07,
+      "loss": 0.0735,
+      "step": 25520
+    },
+    {
+      "epoch": 97.44274809160305,
+      "grad_norm": 0.5489487648010254,
+      "learning_rate": 5.950464001953532e-07,
+      "loss": 0.0786,
+      "step": 25530
+    },
+    {
+      "epoch": 97.48091603053435,
+      "grad_norm": 0.4438866376876831,
+      "learning_rate": 5.924413701528415e-07,
+      "loss": 0.0679,
+      "step": 25540
+    },
+    {
+      "epoch": 97.51908396946565,
+      "grad_norm": 0.6267343163490295,
+      "learning_rate": 5.898416957096704e-07,
+      "loss": 0.0724,
+      "step": 25550
+    },
+    {
+      "epoch": 97.55725190839695,
+      "grad_norm": 0.3363583981990814,
+      "learning_rate": 5.872473800246914e-07,
+      "loss": 0.0681,
+      "step": 25560
+    },
+    {
+      "epoch": 97.59541984732824,
+      "grad_norm": 0.5211305022239685,
+      "learning_rate": 5.846584262502403e-07,
+      "loss": 0.0794,
+      "step": 25570
+    },
+    {
+      "epoch": 97.63358778625954,
+      "grad_norm": 0.32496604323387146,
+      "learning_rate": 5.820748375321411e-07,
+      "loss": 0.0661,
+      "step": 25580
+    },
+    {
+      "epoch": 97.67175572519083,
+      "grad_norm": 0.29008033871650696,
+      "learning_rate": 5.794966170096977e-07,
+      "loss": 0.0665,
+      "step": 25590
+    },
+    {
+      "epoch": 97.70992366412214,
+      "grad_norm": 0.4152953624725342,
+      "learning_rate": 5.769237678156897e-07,
+      "loss": 0.0687,
+      "step": 25600
+    },
+    {
+      "epoch": 97.74809160305344,
+      "grad_norm": 0.33877304196357727,
+      "learning_rate": 5.743562930763735e-07,
+      "loss": 0.067,
+      "step": 25610
+    },
+    {
+      "epoch": 97.78625954198473,
+      "grad_norm": 0.2797228693962097,
+      "learning_rate": 5.717941959114726e-07,
+      "loss": 0.076,
+      "step": 25620
+    },
+    {
+      "epoch": 97.82442748091603,
+      "grad_norm": 0.3312159776687622,
+      "learning_rate": 5.69237479434176e-07,
+      "loss": 0.0684,
+      "step": 25630
+    },
+    {
+      "epoch": 97.86259541984732,
+      "grad_norm": 0.36744144558906555,
+      "learning_rate": 5.666861467511353e-07,
+      "loss": 0.0686,
+      "step": 25640
+    },
+    {
+      "epoch": 97.90076335877862,
+      "grad_norm": 0.40221524238586426,
+      "learning_rate": 5.641402009624591e-07,
+      "loss": 0.0697,
+      "step": 25650
+    },
+    {
+      "epoch": 97.93893129770993,
+      "grad_norm": 0.8071361184120178,
+      "learning_rate": 5.615996451617145e-07,
+      "loss": 0.0737,
+      "step": 25660
+    },
+    {
+      "epoch": 97.97709923664122,
+      "grad_norm": 0.2889866232872009,
+      "learning_rate": 5.590644824359148e-07,
+      "loss": 0.0682,
+      "step": 25670
+    },
+    {
+      "epoch": 98.01526717557252,
+      "grad_norm": 0.27220702171325684,
+      "learning_rate": 5.56534715865521e-07,
+      "loss": 0.066,
+      "step": 25680
+    },
+    {
+      "epoch": 98.05343511450381,
+      "grad_norm": 0.2493731528520584,
+      "learning_rate": 5.540103485244397e-07,
+      "loss": 0.062,
+      "step": 25690
+    },
+    {
+      "epoch": 98.09160305343511,
+      "grad_norm": 0.33816832304000854,
+      "learning_rate": 5.514913834800134e-07,
+      "loss": 0.0676,
+      "step": 25700
+    },
+    {
+      "epoch": 98.12977099236642,
+      "grad_norm": 0.7667673230171204,
+      "learning_rate": 5.489778237930238e-07,
+      "loss": 0.0755,
+      "step": 25710
+    },
+    {
+      "epoch": 98.16793893129771,
+      "grad_norm": 0.35118091106414795,
+      "learning_rate": 5.464696725176827e-07,
+      "loss": 0.0683,
+      "step": 25720
+    },
+    {
+      "epoch": 98.20610687022901,
+      "grad_norm": 0.28815001249313354,
+      "learning_rate": 5.439669327016295e-07,
+      "loss": 0.0702,
+      "step": 25730
+    },
+    {
+      "epoch": 98.2442748091603,
+      "grad_norm": 0.5837283134460449,
+      "learning_rate": 5.414696073859299e-07,
+      "loss": 0.0691,
+      "step": 25740
+    },
+    {
+      "epoch": 98.2824427480916,
+      "grad_norm": 0.37461745738983154,
+      "learning_rate": 5.389776996050695e-07,
+      "loss": 0.0708,
+      "step": 25750
+    },
+    {
+      "epoch": 98.3206106870229,
+      "grad_norm": 0.23216131329536438,
+      "learning_rate": 5.364912123869493e-07,
+      "loss": 0.0671,
+      "step": 25760
+    },
+    {
+      "epoch": 98.3587786259542,
+      "grad_norm": 0.3143027424812317,
+      "learning_rate": 5.340101487528887e-07,
+      "loss": 0.0693,
+      "step": 25770
+    },
+    {
+      "epoch": 98.3969465648855,
+      "grad_norm": 0.35901618003845215,
+      "learning_rate": 5.315345117176129e-07,
+      "loss": 0.0713,
+      "step": 25780
+    },
+    {
+      "epoch": 98.43511450381679,
+      "grad_norm": 0.22512131929397583,
+      "learning_rate": 5.290643042892541e-07,
+      "loss": 0.0714,
+      "step": 25790
+    },
+    {
+      "epoch": 98.47328244274809,
+      "grad_norm": 0.2000698745250702,
+      "learning_rate": 5.265995294693472e-07,
+      "loss": 0.0681,
+      "step": 25800
+    },
+    {
+      "epoch": 98.5114503816794,
+      "grad_norm": 1.0218034982681274,
+      "learning_rate": 5.241401902528253e-07,
+      "loss": 0.0682,
+      "step": 25810
+    },
+    {
+      "epoch": 98.54961832061069,
+      "grad_norm": 0.3218908905982971,
+      "learning_rate": 5.21686289628019e-07,
+      "loss": 0.066,
+      "step": 25820
+    },
+    {
+      "epoch": 98.58778625954199,
+      "grad_norm": 0.4360622465610504,
+      "learning_rate": 5.192378305766472e-07,
+      "loss": 0.0684,
+      "step": 25830
+    },
+    {
+      "epoch": 98.62595419847328,
+      "grad_norm": 0.2918654680252075,
+      "learning_rate": 5.167948160738206e-07,
+      "loss": 0.0634,
+      "step": 25840
+    },
+    {
+      "epoch": 98.66412213740458,
+      "grad_norm": 0.5152965188026428,
+      "learning_rate": 5.143572490880311e-07,
+      "loss": 0.0678,
+      "step": 25850
+    },
+    {
+      "epoch": 98.70229007633588,
+      "grad_norm": 1.0082851648330688,
+      "learning_rate": 5.119251325811513e-07,
+      "loss": 0.066,
+      "step": 25860
+    },
+    {
+      "epoch": 98.74045801526718,
+      "grad_norm": 0.28817683458328247,
+      "learning_rate": 5.094984695084348e-07,
+      "loss": 0.0629,
+      "step": 25870
+    },
+    {
+      "epoch": 98.77862595419847,
+      "grad_norm": 0.5672718286514282,
+      "learning_rate": 5.070772628185039e-07,
+      "loss": 0.0688,
+      "step": 25880
+    },
+    {
+      "epoch": 98.81679389312977,
+      "grad_norm": 0.7285252213478088,
+      "learning_rate": 5.046615154533535e-07,
+      "loss": 0.0681,
+      "step": 25890
+    },
+    {
+      "epoch": 98.85496183206106,
+      "grad_norm": 0.4292791783809662,
+      "learning_rate": 5.022512303483451e-07,
+      "loss": 0.0658,
+      "step": 25900
+    },
+    {
+      "epoch": 98.89312977099236,
+      "grad_norm": 0.20751267671585083,
+      "learning_rate": 4.998464104322015e-07,
+      "loss": 0.0703,
+      "step": 25910
+    },
+    {
+      "epoch": 98.93129770992367,
+      "grad_norm": 0.2615070939064026,
+      "learning_rate": 4.974470586270047e-07,
+      "loss": 0.0651,
+      "step": 25920
+    },
+    {
+      "epoch": 98.96946564885496,
+      "grad_norm": 0.2248445302248001,
+      "learning_rate": 4.950531778481965e-07,
+      "loss": 0.0672,
+      "step": 25930
+    },
+    {
+      "epoch": 99.00763358778626,
+      "grad_norm": 0.4531436264514923,
+      "learning_rate": 4.926647710045652e-07,
+      "loss": 0.0625,
+      "step": 25940
+    },
+    {
+      "epoch": 99.04580152671755,
+      "grad_norm": 0.4915529787540436,
+      "learning_rate": 4.902818409982513e-07,
+      "loss": 0.066,
+      "step": 25950
+    },
+    {
+      "epoch": 99.08396946564885,
+      "grad_norm": 0.28831353783607483,
+      "learning_rate": 4.879043907247383e-07,
+      "loss": 0.0672,
+      "step": 25960
+    },
+    {
+      "epoch": 99.12213740458016,
+      "grad_norm": 0.2686160206794739,
+      "learning_rate": 4.855324230728542e-07,
+      "loss": 0.0696,
+      "step": 25970
+    },
+    {
+      "epoch": 99.16030534351145,
+      "grad_norm": 0.3270827829837799,
+      "learning_rate": 4.831659409247619e-07,
+      "loss": 0.0645,
+      "step": 25980
+    },
+    {
+      "epoch": 99.19847328244275,
+      "grad_norm": 0.25841906666755676,
+      "learning_rate": 4.808049471559617e-07,
+      "loss": 0.0696,
+      "step": 25990
+    },
+    {
+      "epoch": 99.23664122137404,
+      "grad_norm": 0.4139740467071533,
+      "learning_rate": 4.784494446352833e-07,
+      "loss": 0.0721,
+      "step": 26000
+    },
+    {
+      "epoch": 99.27480916030534,
+      "grad_norm": 0.6838223338127136,
+      "learning_rate": 4.7609943622488333e-07,
+      "loss": 0.0711,
+      "step": 26010
+    },
+    {
+      "epoch": 99.31297709923665,
+      "grad_norm": 0.5432331562042236,
+      "learning_rate": 4.737549247802442e-07,
+      "loss": 0.0803,
+      "step": 26020
+    },
+    {
+      "epoch": 99.35114503816794,
+      "grad_norm": 0.7913231253623962,
+      "learning_rate": 4.714159131501689e-07,
+      "loss": 0.0696,
+      "step": 26030
+    },
+    {
+      "epoch": 99.38931297709924,
+      "grad_norm": 0.5764281749725342,
+      "learning_rate": 4.690824041767766e-07,
+      "loss": 0.0753,
+      "step": 26040
+    },
+    {
+      "epoch": 99.42748091603053,
+      "grad_norm": 0.793904185295105,
+      "learning_rate": 4.66754400695501e-07,
+      "loss": 0.0798,
+      "step": 26050
+    },
+    {
+      "epoch": 99.46564885496183,
+      "grad_norm": 0.38731345534324646,
+      "learning_rate": 4.6443190553508597e-07,
+      "loss": 0.0695,
+      "step": 26060
+    },
+    {
+      "epoch": 99.50381679389314,
+      "grad_norm": 0.3625410199165344,
+      "learning_rate": 4.6211492151758076e-07,
+      "loss": 0.0681,
+      "step": 26070
+    },
+    {
+      "epoch": 99.54198473282443,
+      "grad_norm": 0.47990864515304565,
+      "learning_rate": 4.598034514583416e-07,
+      "loss": 0.0696,
+      "step": 26080
+    },
+    {
+      "epoch": 99.58015267175573,
+      "grad_norm": 0.2346879243850708,
+      "learning_rate": 4.574974981660213e-07,
+      "loss": 0.0675,
+      "step": 26090
+    },
+    {
+      "epoch": 99.61832061068702,
+      "grad_norm": 0.29999473690986633,
+      "learning_rate": 4.5519706444257073e-07,
+      "loss": 0.0679,
+      "step": 26100
+    },
+    {
+      "epoch": 99.65648854961832,
+      "grad_norm": 0.516541063785553,
+      "learning_rate": 4.529021530832328e-07,
+      "loss": 0.0648,
+      "step": 26110
+    },
+    {
+      "epoch": 99.69465648854961,
+      "grad_norm": 0.32835254073143005,
+      "learning_rate": 4.506127668765431e-07,
+      "loss": 0.0666,
+      "step": 26120
+    },
+    {
+      "epoch": 99.73282442748092,
+      "grad_norm": 0.3316701054573059,
+      "learning_rate": 4.483289086043197e-07,
+      "loss": 0.0686,
+      "step": 26130
+    },
+    {
+      "epoch": 99.77099236641222,
+      "grad_norm": 0.27133581042289734,
+      "learning_rate": 4.460505810416682e-07,
+      "loss": 0.0713,
+      "step": 26140
+    },
+    {
+      "epoch": 99.80916030534351,
+      "grad_norm": 0.6455039978027344,
+      "learning_rate": 4.437777869569698e-07,
+      "loss": 0.0716,
+      "step": 26150
+    },
+    {
+      "epoch": 99.8473282442748,
+      "grad_norm": 0.3414127826690674,
+      "learning_rate": 4.415105291118843e-07,
+      "loss": 0.0654,
+      "step": 26160
+    },
+    {
+      "epoch": 99.8854961832061,
+      "grad_norm": 0.28300490975379944,
+      "learning_rate": 4.392488102613435e-07,
+      "loss": 0.0684,
+      "step": 26170
+    },
+    {
+      "epoch": 99.92366412213741,
+      "grad_norm": 0.3191199004650116,
+      "learning_rate": 4.3699263315354735e-07,
+      "loss": 0.0706,
+      "step": 26180
+    },
+    {
+      "epoch": 99.9618320610687,
+      "grad_norm": 0.25766992568969727,
+      "learning_rate": 4.3474200052996685e-07,
+      "loss": 0.0659,
+      "step": 26190
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 0.3972707986831665,
+      "learning_rate": 4.324969151253317e-07,
+      "loss": 0.0705,
+      "step": 26200
+    },
+    {
+      "epoch": 100.0381679389313,
+      "grad_norm": 0.21596112847328186,
+      "learning_rate": 4.302573796676313e-07,
+      "loss": 0.0675,
+      "step": 26210
+    },
+    {
+      "epoch": 100.07633587786259,
+      "grad_norm": 0.3397051692008972,
+      "learning_rate": 4.280233968781139e-07,
+      "loss": 0.0658,
+      "step": 26220
+    },
+    {
+      "epoch": 100.1145038167939,
+      "grad_norm": 0.32072919607162476,
+      "learning_rate": 4.257949694712771e-07,
+      "loss": 0.0702,
+      "step": 26230
+    },
+    {
+      "epoch": 100.1526717557252,
+      "grad_norm": 0.2912977337837219,
+      "learning_rate": 4.235721001548726e-07,
+      "loss": 0.0697,
+      "step": 26240
+    },
+    {
+      "epoch": 100.19083969465649,
+      "grad_norm": 0.3459882140159607,
+      "learning_rate": 4.21354791629896e-07,
+      "loss": 0.069,
+      "step": 26250
+    },
+    {
+      "epoch": 100.22900763358778,
+      "grad_norm": 0.19958975911140442,
+      "learning_rate": 4.1914304659058437e-07,
+      "loss": 0.0626,
+      "step": 26260
+    },
+    {
+      "epoch": 100.26717557251908,
+      "grad_norm": 0.35708945989608765,
+      "learning_rate": 4.1693686772441846e-07,
+      "loss": 0.0687,
+      "step": 26270
+    },
+    {
+      "epoch": 100.30534351145039,
+      "grad_norm": 0.7189772129058838,
+      "learning_rate": 4.1473625771211224e-07,
+      "loss": 0.0748,
+      "step": 26280
+    },
+    {
+      "epoch": 100.34351145038168,
+      "grad_norm": 0.21925044059753418,
+      "learning_rate": 4.125412192276157e-07,
+      "loss": 0.0681,
+      "step": 26290
+    },
+    {
+      "epoch": 100.38167938931298,
+      "grad_norm": 0.3142671585083008,
+      "learning_rate": 4.1035175493810696e-07,
+      "loss": 0.0725,
+      "step": 26300
+    },
+    {
+      "epoch": 100.41984732824427,
+      "grad_norm": 0.5476031303405762,
+      "learning_rate": 4.081678675039913e-07,
+      "loss": 0.0667,
+      "step": 26310
+    },
+    {
+      "epoch": 100.45801526717557,
+      "grad_norm": 0.2075609415769577,
+      "learning_rate": 4.059895595788987e-07,
+      "loss": 0.0669,
+      "step": 26320
+    },
+    {
+      "epoch": 100.49618320610686,
+      "grad_norm": 0.3402668237686157,
+      "learning_rate": 4.038168338096776e-07,
+      "loss": 0.0715,
+      "step": 26330
+    },
+    {
+      "epoch": 100.53435114503817,
+      "grad_norm": 0.2339000403881073,
+      "learning_rate": 4.016496928363944e-07,
+      "loss": 0.0724,
+      "step": 26340
+    },
+    {
+      "epoch": 100.57251908396947,
+      "grad_norm": 0.435660183429718,
+      "learning_rate": 3.994881392923317e-07,
+      "loss": 0.072,
+      "step": 26350
+    },
+    {
+      "epoch": 100.61068702290076,
+      "grad_norm": 0.24794331192970276,
+      "learning_rate": 3.973321758039794e-07,
+      "loss": 0.0633,
+      "step": 26360
+    },
+    {
+      "epoch": 100.64885496183206,
+      "grad_norm": 0.2800830602645874,
+      "learning_rate": 3.9518180499103665e-07,
+      "loss": 0.0644,
+      "step": 26370
+    },
+    {
+      "epoch": 100.68702290076335,
+      "grad_norm": 0.25040721893310547,
+      "learning_rate": 3.930370294664071e-07,
+      "loss": 0.0698,
+      "step": 26380
+    },
+    {
+      "epoch": 100.72519083969466,
+      "grad_norm": 0.2757255733013153,
+      "learning_rate": 3.9089785183619386e-07,
+      "loss": 0.0644,
+      "step": 26390
+    },
+    {
+      "epoch": 100.76335877862596,
+      "grad_norm": 0.6161481738090515,
+      "learning_rate": 3.8876427469970167e-07,
+      "loss": 0.0693,
+      "step": 26400
+    },
+    {
+      "epoch": 100.80152671755725,
+      "grad_norm": 0.24093779921531677,
+      "learning_rate": 3.866363006494256e-07,
+      "loss": 0.0653,
+      "step": 26410
+    },
+    {
+      "epoch": 100.83969465648855,
+      "grad_norm": 0.27081120014190674,
+      "learning_rate": 3.845139322710573e-07,
+      "loss": 0.0662,
+      "step": 26420
+    },
+    {
+      "epoch": 100.87786259541984,
+      "grad_norm": 0.2640981376171112,
+      "learning_rate": 3.823971721434727e-07,
+      "loss": 0.0657,
+      "step": 26430
+    },
+    {
+      "epoch": 100.91603053435115,
+      "grad_norm": 1.1142594814300537,
+      "learning_rate": 3.8028602283873504e-07,
+      "loss": 0.0662,
+      "step": 26440
+    },
+    {
+      "epoch": 100.95419847328245,
+      "grad_norm": 0.3841104209423065,
+      "learning_rate": 3.781804869220912e-07,
+      "loss": 0.07,
+      "step": 26450
+    },
+    {
+      "epoch": 100.99236641221374,
+      "grad_norm": 0.23676466941833496,
+      "learning_rate": 3.760805669519646e-07,
+      "loss": 0.0676,
+      "step": 26460
+    },
+    {
+      "epoch": 101.03053435114504,
+      "grad_norm": 0.7228853702545166,
+      "learning_rate": 3.7398626547995585e-07,
+      "loss": 0.0881,
+      "step": 26470
+    },
+    {
+      "epoch": 101.06870229007633,
+      "grad_norm": 0.665624737739563,
+      "learning_rate": 3.7189758505083994e-07,
+      "loss": 0.0639,
+      "step": 26480
+    },
+    {
+      "epoch": 101.10687022900764,
+      "grad_norm": 0.30101853609085083,
+      "learning_rate": 3.6981452820255836e-07,
+      "loss": 0.0621,
+      "step": 26490
+    },
+    {
+      "epoch": 101.14503816793894,
+      "grad_norm": 0.36952438950538635,
+      "learning_rate": 3.6773709746622365e-07,
+      "loss": 0.0672,
+      "step": 26500
+    },
+    {
+      "epoch": 101.18320610687023,
+      "grad_norm": 0.528901219367981,
+      "learning_rate": 3.6566529536611e-07,
+      "loss": 0.07,
+      "step": 26510
+    },
+    {
+      "epoch": 101.22137404580153,
+      "grad_norm": 0.3763343393802643,
+      "learning_rate": 3.635991244196513e-07,
+      "loss": 0.0788,
+      "step": 26520
+    },
+    {
+      "epoch": 101.25954198473282,
+      "grad_norm": 0.33685335516929626,
+      "learning_rate": 3.615385871374405e-07,
+      "loss": 0.0696,
+      "step": 26530
+    },
+    {
+      "epoch": 101.29770992366412,
+      "grad_norm": 0.2437128871679306,
+      "learning_rate": 3.594836860232237e-07,
+      "loss": 0.0682,
+      "step": 26540
+    },
+    {
+      "epoch": 101.33587786259542,
+      "grad_norm": 0.2727994918823242,
+      "learning_rate": 3.574344235739019e-07,
+      "loss": 0.0698,
+      "step": 26550
+    },
+    {
+      "epoch": 101.37404580152672,
+      "grad_norm": 0.6007474064826965,
+      "learning_rate": 3.553908022795194e-07,
+      "loss": 0.0701,
+      "step": 26560
+    },
+    {
+      "epoch": 101.41221374045801,
+      "grad_norm": 0.23269033432006836,
+      "learning_rate": 3.5335282462327093e-07,
+      "loss": 0.0656,
+      "step": 26570
+    },
+    {
+      "epoch": 101.45038167938931,
+      "grad_norm": 0.3645854592323303,
+      "learning_rate": 3.513204930814912e-07,
+      "loss": 0.068,
+      "step": 26580
+    },
+    {
+      "epoch": 101.4885496183206,
+      "grad_norm": 0.19710974395275116,
+      "learning_rate": 3.4929381012365424e-07,
+      "loss": 0.0682,
+      "step": 26590
+    },
+    {
+      "epoch": 101.52671755725191,
+      "grad_norm": 0.5920630097389221,
+      "learning_rate": 3.472727782123697e-07,
+      "loss": 0.0692,
+      "step": 26600
+    },
+    {
+      "epoch": 101.56488549618321,
+      "grad_norm": 0.7211366891860962,
+      "learning_rate": 3.452573998033842e-07,
+      "loss": 0.0689,
+      "step": 26610
+    },
+    {
+      "epoch": 101.6030534351145,
+      "grad_norm": 0.2851048409938812,
+      "learning_rate": 3.432476773455712e-07,
+      "loss": 0.0753,
+      "step": 26620
+    },
+    {
+      "epoch": 101.6412213740458,
+      "grad_norm": 0.38267195224761963,
+      "learning_rate": 3.412436132809338e-07,
+      "loss": 0.068,
+      "step": 26630
+    },
+    {
+      "epoch": 101.6793893129771,
+      "grad_norm": 0.5825901031494141,
+      "learning_rate": 3.392452100445975e-07,
+      "loss": 0.0682,
+      "step": 26640
+    },
+    {
+      "epoch": 101.7175572519084,
+      "grad_norm": 0.37534067034721375,
+      "learning_rate": 3.3725247006481167e-07,
+      "loss": 0.0746,
+      "step": 26650
+    },
+    {
+      "epoch": 101.7557251908397,
+      "grad_norm": 0.3646638095378876,
+      "learning_rate": 3.3526539576294315e-07,
+      "loss": 0.0671,
+      "step": 26660
+    },
+    {
+      "epoch": 101.79389312977099,
+      "grad_norm": 0.24642281234264374,
+      "learning_rate": 3.332839895534745e-07,
+      "loss": 0.0706,
+      "step": 26670
+    },
+    {
+      "epoch": 101.83206106870229,
+      "grad_norm": 0.45950332283973694,
+      "learning_rate": 3.3130825384400156e-07,
+      "loss": 0.0662,
+      "step": 26680
+    },
+    {
+      "epoch": 101.87022900763358,
+      "grad_norm": 0.3571898639202118,
+      "learning_rate": 3.293381910352278e-07,
+      "loss": 0.0646,
+      "step": 26690
+    },
+    {
+      "epoch": 101.90839694656489,
+      "grad_norm": 0.2520214021205902,
+      "learning_rate": 3.273738035209678e-07,
+      "loss": 0.0577,
+      "step": 26700
+    },
+    {
+      "epoch": 101.94656488549619,
+      "grad_norm": 0.28303033113479614,
+      "learning_rate": 3.254150936881356e-07,
+      "loss": 0.0689,
+      "step": 26710
+    },
+    {
+      "epoch": 101.98473282442748,
+      "grad_norm": 0.3684123754501343,
+      "learning_rate": 3.234620639167496e-07,
+      "loss": 0.0655,
+      "step": 26720
+    },
+    {
+      "epoch": 102.02290076335878,
+      "grad_norm": 0.5102269053459167,
+      "learning_rate": 3.2151471657992485e-07,
+      "loss": 0.0746,
+      "step": 26730
+    },
+    {
+      "epoch": 102.06106870229007,
+      "grad_norm": 0.24563086032867432,
+      "learning_rate": 3.1957305404387187e-07,
+      "loss": 0.0726,
+      "step": 26740
+    },
+    {
+      "epoch": 102.09923664122137,
+      "grad_norm": 0.37572887539863586,
+      "learning_rate": 3.176370786678934e-07,
+      "loss": 0.0636,
+      "step": 26750
+    },
+    {
+      "epoch": 102.13740458015268,
+      "grad_norm": 0.2450253665447235,
+      "learning_rate": 3.1570679280438163e-07,
+      "loss": 0.0726,
+      "step": 26760
+    },
+    {
+      "epoch": 102.17557251908397,
+      "grad_norm": 0.48279571533203125,
+      "learning_rate": 3.1378219879881764e-07,
+      "loss": 0.0705,
+      "step": 26770
+    },
+    {
+      "epoch": 102.21374045801527,
+      "grad_norm": 0.39897432923316956,
+      "learning_rate": 3.1186329898976296e-07,
+      "loss": 0.0731,
+      "step": 26780
+    },
+    {
+      "epoch": 102.25190839694656,
+      "grad_norm": 0.23803192377090454,
+      "learning_rate": 3.0995009570886305e-07,
+      "loss": 0.0779,
+      "step": 26790
+    },
+    {
+      "epoch": 102.29007633587786,
+      "grad_norm": 0.6093907356262207,
+      "learning_rate": 3.0804259128083936e-07,
+      "loss": 0.0667,
+      "step": 26800
+    },
+    {
+      "epoch": 102.32824427480917,
+      "grad_norm": 0.275802880525589,
+      "learning_rate": 3.0614078802348903e-07,
+      "loss": 0.0642,
+      "step": 26810
+    },
+    {
+      "epoch": 102.36641221374046,
+      "grad_norm": 0.2756710648536682,
+      "learning_rate": 3.042446882476846e-07,
+      "loss": 0.0668,
+      "step": 26820
+    },
+    {
+      "epoch": 102.40458015267176,
+      "grad_norm": 0.2288149744272232,
+      "learning_rate": 3.023542942573643e-07,
+      "loss": 0.0716,
+      "step": 26830
+    },
+    {
+      "epoch": 102.44274809160305,
+      "grad_norm": 0.2968932092189789,
+      "learning_rate": 3.004696083495351e-07,
+      "loss": 0.0671,
+      "step": 26840
+    },
+    {
+      "epoch": 102.48091603053435,
+      "grad_norm": 0.27723342180252075,
+      "learning_rate": 2.98590632814269e-07,
+      "loss": 0.0673,
+      "step": 26850
+    },
+    {
+      "epoch": 102.51908396946565,
+      "grad_norm": 0.36572501063346863,
+      "learning_rate": 2.96717369934697e-07,
+      "loss": 0.0679,
+      "step": 26860
+    },
+    {
+      "epoch": 102.55725190839695,
+      "grad_norm": 0.19056031107902527,
+      "learning_rate": 2.948498219870122e-07,
+      "loss": 0.067,
+      "step": 26870
+    },
+    {
+      "epoch": 102.59541984732824,
+      "grad_norm": 0.220974862575531,
+      "learning_rate": 2.929879912404604e-07,
+      "loss": 0.067,
+      "step": 26880
+    },
+    {
+      "epoch": 102.63358778625954,
+      "grad_norm": 0.2869727909564972,
+      "learning_rate": 2.911318799573415e-07,
+      "loss": 0.0699,
+      "step": 26890
+    },
+    {
+      "epoch": 102.67175572519083,
+      "grad_norm": 1.0001291036605835,
+      "learning_rate": 2.8928149039300525e-07,
+      "loss": 0.0713,
+      "step": 26900
+    },
+    {
+      "epoch": 102.70992366412214,
+      "grad_norm": 0.3044949173927307,
+      "learning_rate": 2.8743682479584977e-07,
+      "loss": 0.0816,
+      "step": 26910
+    },
+    {
+      "epoch": 102.74809160305344,
+      "grad_norm": 0.2789320945739746,
+      "learning_rate": 2.8559788540731826e-07,
+      "loss": 0.0637,
+      "step": 26920
+    },
+    {
+      "epoch": 102.78625954198473,
+      "grad_norm": 0.7257505655288696,
+      "learning_rate": 2.837646744618949e-07,
+      "loss": 0.0823,
+      "step": 26930
+    },
+    {
+      "epoch": 102.82442748091603,
+      "grad_norm": 0.3718293309211731,
+      "learning_rate": 2.8193719418710405e-07,
+      "loss": 0.0703,
+      "step": 26940
+    },
+    {
+      "epoch": 102.86259541984732,
+      "grad_norm": 0.32060879468917847,
+      "learning_rate": 2.8011544680350667e-07,
+      "loss": 0.0709,
+      "step": 26950
+    },
+    {
+      "epoch": 102.90076335877862,
+      "grad_norm": 0.6443219780921936,
+      "learning_rate": 2.7829943452469753e-07,
+      "loss": 0.0755,
+      "step": 26960
+    },
+    {
+      "epoch": 102.93893129770993,
+      "grad_norm": 0.5253495573997498,
+      "learning_rate": 2.764891595573022e-07,
+      "loss": 0.0703,
+      "step": 26970
+    },
+    {
+      "epoch": 102.97709923664122,
+      "grad_norm": 0.3999308943748474,
+      "learning_rate": 2.746846241009765e-07,
+      "loss": 0.0673,
+      "step": 26980
+    },
+    {
+      "epoch": 103.01526717557252,
+      "grad_norm": 0.2527059018611908,
+      "learning_rate": 2.7288583034839944e-07,
+      "loss": 0.0687,
+      "step": 26990
+    },
+    {
+      "epoch": 103.05343511450381,
+      "grad_norm": 0.296678751707077,
+      "learning_rate": 2.7109278048527756e-07,
+      "loss": 0.0654,
+      "step": 27000
+    },
+    {
+      "epoch": 103.09160305343511,
+      "grad_norm": 0.7945778965950012,
+      "learning_rate": 2.6930547669033415e-07,
+      "loss": 0.0737,
+      "step": 27010
+    },
+    {
+      "epoch": 103.12977099236642,
+      "grad_norm": 0.32029277086257935,
+      "learning_rate": 2.675239211353109e-07,
+      "loss": 0.0693,
+      "step": 27020
+    },
+    {
+      "epoch": 103.16793893129771,
+      "grad_norm": 0.2556557357311249,
+      "learning_rate": 2.6574811598496787e-07,
+      "loss": 0.0755,
+      "step": 27030
+    },
+    {
+      "epoch": 103.20610687022901,
+      "grad_norm": 0.23679769039154053,
+      "learning_rate": 2.6397806339707457e-07,
+      "loss": 0.0756,
+      "step": 27040
+    },
+    {
+      "epoch": 103.2442748091603,
+      "grad_norm": 0.4383762776851654,
+      "learning_rate": 2.622137655224122e-07,
+      "loss": 0.0758,
+      "step": 27050
+    },
+    {
+      "epoch": 103.2824427480916,
+      "grad_norm": 0.5516987442970276,
+      "learning_rate": 2.6045522450476814e-07,
+      "loss": 0.0728,
+      "step": 27060
+    },
+    {
+      "epoch": 103.3206106870229,
+      "grad_norm": 0.4582926630973816,
+      "learning_rate": 2.587024424809359e-07,
+      "loss": 0.0779,
+      "step": 27070
+    },
+    {
+      "epoch": 103.3587786259542,
+      "grad_norm": 0.5267468094825745,
+      "learning_rate": 2.5695542158071187e-07,
+      "loss": 0.0677,
+      "step": 27080
+    },
+    {
+      "epoch": 103.3969465648855,
+      "grad_norm": 0.3101442754268646,
+      "learning_rate": 2.5521416392689066e-07,
+      "loss": 0.0666,
+      "step": 27090
+    },
+    {
+      "epoch": 103.43511450381679,
+      "grad_norm": 0.22319281101226807,
+      "learning_rate": 2.5347867163526387e-07,
+      "loss": 0.0672,
+      "step": 27100
+    },
+    {
+      "epoch": 103.47328244274809,
+      "grad_norm": 0.2605898678302765,
+      "learning_rate": 2.517489468146189e-07,
+      "loss": 0.0673,
+      "step": 27110
+    },
+    {
+      "epoch": 103.5114503816794,
+      "grad_norm": 0.2470434010028839,
+      "learning_rate": 2.500249915667341e-07,
+      "loss": 0.0683,
+      "step": 27120
+    },
+    {
+      "epoch": 103.54961832061069,
+      "grad_norm": 0.18632711470127106,
+      "learning_rate": 2.4830680798637817e-07,
+      "loss": 0.0679,
+      "step": 27130
+    },
+    {
+      "epoch": 103.58778625954199,
+      "grad_norm": 0.2968946695327759,
+      "learning_rate": 2.4659439816130557e-07,
+      "loss": 0.0667,
+      "step": 27140
+    },
+    {
+      "epoch": 103.62595419847328,
+      "grad_norm": 0.3109892010688782,
+      "learning_rate": 2.448877641722569e-07,
+      "loss": 0.0656,
+      "step": 27150
+    },
+    {
+      "epoch": 103.66412213740458,
+      "grad_norm": 0.6580647826194763,
+      "learning_rate": 2.431869080929522e-07,
+      "loss": 0.0691,
+      "step": 27160
+    },
+    {
+      "epoch": 103.70229007633588,
+      "grad_norm": 0.44333863258361816,
+      "learning_rate": 2.414918319900922e-07,
+      "loss": 0.0796,
+      "step": 27170
+    },
+    {
+      "epoch": 103.74045801526718,
+      "grad_norm": 0.201785147190094,
+      "learning_rate": 2.398025379233543e-07,
+      "loss": 0.0642,
+      "step": 27180
+    },
+    {
+      "epoch": 103.77862595419847,
+      "grad_norm": 0.3678027391433716,
+      "learning_rate": 2.381190279453899e-07,
+      "loss": 0.0824,
+      "step": 27190
+    },
+    {
+      "epoch": 103.81679389312977,
+      "grad_norm": 0.3141831159591675,
+      "learning_rate": 2.364413041018232e-07,
+      "loss": 0.0694,
+      "step": 27200
+    },
+    {
+      "epoch": 103.85496183206106,
+      "grad_norm": 0.35555902123451233,
+      "learning_rate": 2.3476936843124633e-07,
+      "loss": 0.0665,
+      "step": 27210
+    },
+    {
+      "epoch": 103.89312977099236,
+      "grad_norm": 0.2217165231704712,
+      "learning_rate": 2.3310322296521859e-07,
+      "loss": 0.0652,
+      "step": 27220
+    },
+    {
+      "epoch": 103.93129770992367,
+      "grad_norm": 0.36410123109817505,
+      "learning_rate": 2.314428697282628e-07,
+      "loss": 0.0719,
+      "step": 27230
+    },
+    {
+      "epoch": 103.96946564885496,
+      "grad_norm": 0.7411630749702454,
+      "learning_rate": 2.2978831073786735e-07,
+      "loss": 0.0803,
+      "step": 27240
+    },
+    {
+      "epoch": 104.00763358778626,
+      "grad_norm": 0.2194436937570572,
+      "learning_rate": 2.2813954800447512e-07,
+      "loss": 0.0631,
+      "step": 27250
+    },
+    {
+      "epoch": 104.04580152671755,
+      "grad_norm": 0.3712310194969177,
+      "learning_rate": 2.2649658353148974e-07,
+      "loss": 0.0655,
+      "step": 27260
+    },
+    {
+      "epoch": 104.08396946564885,
+      "grad_norm": 0.2021886706352234,
+      "learning_rate": 2.2485941931526646e-07,
+      "loss": 0.0663,
+      "step": 27270
+    },
+    {
+      "epoch": 104.12213740458016,
+      "grad_norm": 0.5154406428337097,
+      "learning_rate": 2.232280573451151e-07,
+      "loss": 0.0712,
+      "step": 27280
+    },
+    {
+      "epoch": 104.16030534351145,
+      "grad_norm": 0.28259921073913574,
+      "learning_rate": 2.216024996032945e-07,
+      "loss": 0.0654,
+      "step": 27290
+    },
+    {
+      "epoch": 104.19847328244275,
+      "grad_norm": 0.23785048723220825,
+      "learning_rate": 2.1998274806501074e-07,
+      "loss": 0.0725,
+      "step": 27300
+    },
+    {
+      "epoch": 104.23664122137404,
+      "grad_norm": 0.3378708064556122,
+      "learning_rate": 2.1836880469841391e-07,
+      "loss": 0.071,
+      "step": 27310
+    },
+    {
+      "epoch": 104.27480916030534,
+      "grad_norm": 0.4880586266517639,
+      "learning_rate": 2.1676067146459812e-07,
+      "loss": 0.072,
+      "step": 27320
+    },
+    {
+      "epoch": 104.31297709923665,
+      "grad_norm": 0.4210817217826843,
+      "learning_rate": 2.151583503175958e-07,
+      "loss": 0.0669,
+      "step": 27330
+    },
+    {
+      "epoch": 104.35114503816794,
+      "grad_norm": 0.5876713991165161,
+      "learning_rate": 2.1356184320437955e-07,
+      "loss": 0.0688,
+      "step": 27340
+    },
+    {
+      "epoch": 104.38931297709924,
+      "grad_norm": 0.3950154483318329,
+      "learning_rate": 2.1197115206485542e-07,
+      "loss": 0.0646,
+      "step": 27350
+    },
+    {
+      "epoch": 104.42748091603053,
+      "grad_norm": 0.5265299081802368,
+      "learning_rate": 2.103862788318628e-07,
+      "loss": 0.0676,
+      "step": 27360
+    },
+    {
+      "epoch": 104.46564885496183,
+      "grad_norm": 0.4005798399448395,
+      "learning_rate": 2.0880722543117293e-07,
+      "loss": 0.0629,
+      "step": 27370
+    },
+    {
+      "epoch": 104.50381679389314,
+      "grad_norm": 0.8654969930648804,
+      "learning_rate": 2.0723399378148434e-07,
+      "loss": 0.0687,
+      "step": 27380
+    },
+    {
+      "epoch": 104.54198473282443,
+      "grad_norm": 0.2732725739479065,
+      "learning_rate": 2.0566658579442067e-07,
+      "loss": 0.0669,
+      "step": 27390
+    },
+    {
+      "epoch": 104.58015267175573,
+      "grad_norm": 0.5799111723899841,
+      "learning_rate": 2.0410500337453176e-07,
+      "loss": 0.0705,
+      "step": 27400
+    },
+    {
+      "epoch": 104.61832061068702,
+      "grad_norm": 0.28128835558891296,
+      "learning_rate": 2.0254924841928647e-07,
+      "loss": 0.0654,
+      "step": 27410
+    },
+    {
+      "epoch": 104.65648854961832,
+      "grad_norm": 0.41618332266807556,
+      "learning_rate": 2.0099932281907542e-07,
+      "loss": 0.0699,
+      "step": 27420
+    },
+    {
+      "epoch": 104.69465648854961,
+      "grad_norm": 0.5150026679039001,
+      "learning_rate": 1.9945522845720323e-07,
+      "loss": 0.0704,
+      "step": 27430
+    },
+    {
+      "epoch": 104.73282442748092,
+      "grad_norm": 0.2852458953857422,
+      "learning_rate": 1.9791696720988963e-07,
+      "loss": 0.0631,
+      "step": 27440
+    },
+    {
+      "epoch": 104.77099236641222,
+      "grad_norm": 0.2500865161418915,
+      "learning_rate": 1.9638454094626836e-07,
+      "loss": 0.0695,
+      "step": 27450
+    },
+    {
+      "epoch": 104.80916030534351,
+      "grad_norm": 0.3002980947494507,
+      "learning_rate": 1.9485795152838105e-07,
+      "loss": 0.0703,
+      "step": 27460
+    },
+    {
+      "epoch": 104.8473282442748,
+      "grad_norm": 0.22461900115013123,
+      "learning_rate": 1.933372008111778e-07,
+      "loss": 0.0646,
+      "step": 27470
+    },
+    {
+      "epoch": 104.8854961832061,
+      "grad_norm": 0.2733502686023712,
+      "learning_rate": 1.9182229064251433e-07,
+      "loss": 0.0649,
+      "step": 27480
+    },
+    {
+      "epoch": 104.92366412213741,
+      "grad_norm": 0.35162800550460815,
+      "learning_rate": 1.9031322286314878e-07,
+      "loss": 0.0745,
+      "step": 27490
+    },
+    {
+      "epoch": 104.9618320610687,
+      "grad_norm": 0.22003084421157837,
+      "learning_rate": 1.8880999930674216e-07,
+      "loss": 0.0635,
+      "step": 27500
+    },
+    {
+      "epoch": 105.0,
+      "grad_norm": 0.39051511883735657,
+      "learning_rate": 1.8731262179985166e-07,
+      "loss": 0.0752,
+      "step": 27510
+    },
+    {
+      "epoch": 105.0381679389313,
+      "grad_norm": 0.4942927658557892,
+      "learning_rate": 1.8582109216193245e-07,
+      "loss": 0.0925,
+      "step": 27520
+    },
+    {
+      "epoch": 105.07633587786259,
+      "grad_norm": 0.26188600063323975,
+      "learning_rate": 1.8433541220533368e-07,
+      "loss": 0.0646,
+      "step": 27530
+    },
+    {
+      "epoch": 105.1145038167939,
+      "grad_norm": 0.28856149315834045,
+      "learning_rate": 1.8285558373529578e-07,
+      "loss": 0.0676,
+      "step": 27540
+    },
+    {
+      "epoch": 105.1526717557252,
+      "grad_norm": 0.24751274287700653,
+      "learning_rate": 1.8138160854995145e-07,
+      "loss": 0.0711,
+      "step": 27550
+    },
+    {
+      "epoch": 105.19083969465649,
+      "grad_norm": 0.3480677902698517,
+      "learning_rate": 1.7991348844031864e-07,
+      "loss": 0.0659,
+      "step": 27560
+    },
+    {
+      "epoch": 105.22900763358778,
+      "grad_norm": 0.42214107513427734,
+      "learning_rate": 1.78451225190302e-07,
+      "loss": 0.0706,
+      "step": 27570
+    },
+    {
+      "epoch": 105.26717557251908,
+      "grad_norm": 0.47172361612319946,
+      "learning_rate": 1.7699482057668916e-07,
+      "loss": 0.0669,
+      "step": 27580
+    },
+    {
+      "epoch": 105.30534351145039,
+      "grad_norm": 0.6258100867271423,
+      "learning_rate": 1.7554427636914894e-07,
+      "loss": 0.0689,
+      "step": 27590
+    },
+    {
+      "epoch": 105.34351145038168,
+      "grad_norm": 0.21334940195083618,
+      "learning_rate": 1.7409959433022873e-07,
+      "loss": 0.0653,
+      "step": 27600
+    },
+    {
+      "epoch": 105.38167938931298,
+      "grad_norm": 0.39631593227386475,
+      "learning_rate": 1.726607762153548e-07,
+      "loss": 0.0843,
+      "step": 27610
+    },
+    {
+      "epoch": 105.41984732824427,
+      "grad_norm": 0.49448487162590027,
+      "learning_rate": 1.7122782377282598e-07,
+      "loss": 0.0667,
+      "step": 27620
+    },
+    {
+      "epoch": 105.45801526717557,
+      "grad_norm": 0.26243850588798523,
+      "learning_rate": 1.6980073874381497e-07,
+      "loss": 0.0646,
+      "step": 27630
+    },
+    {
+      "epoch": 105.49618320610686,
+      "grad_norm": 0.3036261200904846,
+      "learning_rate": 1.6837952286236415e-07,
+      "loss": 0.0757,
+      "step": 27640
+    },
+    {
+      "epoch": 105.53435114503817,
+      "grad_norm": 0.37669137120246887,
+      "learning_rate": 1.6696417785538487e-07,
+      "loss": 0.0729,
+      "step": 27650
+    },
+    {
+      "epoch": 105.57251908396947,
+      "grad_norm": 0.2965763807296753,
+      "learning_rate": 1.6555470544265539e-07,
+      "loss": 0.0724,
+      "step": 27660
+    },
+    {
+      "epoch": 105.61068702290076,
+      "grad_norm": 0.4320843815803528,
+      "learning_rate": 1.6415110733681737e-07,
+      "loss": 0.0712,
+      "step": 27670
+    },
+    {
+      "epoch": 105.64885496183206,
+      "grad_norm": 0.2053402066230774,
+      "learning_rate": 1.6275338524337437e-07,
+      "loss": 0.0683,
+      "step": 27680
+    },
+    {
+      "epoch": 105.68702290076335,
+      "grad_norm": 0.27618706226348877,
+      "learning_rate": 1.6136154086069054e-07,
+      "loss": 0.064,
+      "step": 27690
+    },
+    {
+      "epoch": 105.72519083969466,
+      "grad_norm": 0.5700408220291138,
+      "learning_rate": 1.599755758799887e-07,
+      "loss": 0.0694,
+      "step": 27700
+    },
+    {
+      "epoch": 105.76335877862596,
+      "grad_norm": 0.31304672360420227,
+      "learning_rate": 1.5859549198534607e-07,
+      "loss": 0.0695,
+      "step": 27710
+    },
+    {
+      "epoch": 105.80152671755725,
+      "grad_norm": 0.23126287758350372,
+      "learning_rate": 1.572212908536963e-07,
+      "loss": 0.0689,
+      "step": 27720
+    },
+    {
+      "epoch": 105.83969465648855,
+      "grad_norm": 0.3003278374671936,
+      "learning_rate": 1.5585297415482203e-07,
+      "loss": 0.0724,
+      "step": 27730
+    },
+    {
+      "epoch": 105.87786259541984,
+      "grad_norm": 0.4534234404563904,
+      "learning_rate": 1.5449054355135718e-07,
+      "loss": 0.0669,
+      "step": 27740
+    },
+    {
+      "epoch": 105.91603053435115,
+      "grad_norm": 0.5015247464179993,
+      "learning_rate": 1.5313400069878416e-07,
+      "loss": 0.0691,
+      "step": 27750
+    },
+    {
+      "epoch": 105.95419847328245,
+      "grad_norm": 0.2227436900138855,
+      "learning_rate": 1.5178334724542887e-07,
+      "loss": 0.0653,
+      "step": 27760
+    },
+    {
+      "epoch": 105.99236641221374,
+      "grad_norm": 0.9543538093566895,
+      "learning_rate": 1.504385848324641e-07,
+      "loss": 0.0665,
+      "step": 27770
+    },
+    {
+      "epoch": 106.03053435114504,
+      "grad_norm": 0.7465227246284485,
+      "learning_rate": 1.4909971509390332e-07,
+      "loss": 0.072,
+      "step": 27780
+    },
+    {
+      "epoch": 106.06870229007633,
+      "grad_norm": 0.23705200850963593,
+      "learning_rate": 1.4776673965659793e-07,
+      "loss": 0.0608,
+      "step": 27790
+    },
+    {
+      "epoch": 106.10687022900764,
+      "grad_norm": 0.29993176460266113,
+      "learning_rate": 1.4643966014023958e-07,
+      "loss": 0.0686,
+      "step": 27800
+    },
+    {
+      "epoch": 106.14503816793894,
+      "grad_norm": 0.24530579149723053,
+      "learning_rate": 1.4511847815735503e-07,
+      "loss": 0.0666,
+      "step": 27810
+    },
+    {
+      "epoch": 106.18320610687023,
+      "grad_norm": 0.24449963867664337,
+      "learning_rate": 1.438031953133051e-07,
+      "loss": 0.0694,
+      "step": 27820
+    },
+    {
+      "epoch": 106.22137404580153,
+      "grad_norm": 0.3024006485939026,
+      "learning_rate": 1.4249381320628197e-07,
+      "loss": 0.0774,
+      "step": 27830
+    },
+    {
+      "epoch": 106.25954198473282,
+      "grad_norm": 0.3465965986251831,
+      "learning_rate": 1.4119033342730904e-07,
+      "loss": 0.0644,
+      "step": 27840
+    },
+    {
+      "epoch": 106.29770992366412,
+      "grad_norm": 0.2610500156879425,
+      "learning_rate": 1.3989275756023658e-07,
+      "loss": 0.0673,
+      "step": 27850
+    },
+    {
+      "epoch": 106.33587786259542,
+      "grad_norm": 0.34050482511520386,
+      "learning_rate": 1.386010871817417e-07,
+      "loss": 0.0695,
+      "step": 27860
+    },
+    {
+      "epoch": 106.37404580152672,
+      "grad_norm": 0.3278212249279022,
+      "learning_rate": 1.3731532386132617e-07,
+      "loss": 0.0736,
+      "step": 27870
+    },
+    {
+      "epoch": 106.41221374045801,
+      "grad_norm": 0.27031412720680237,
+      "learning_rate": 1.360354691613136e-07,
+      "loss": 0.0695,
+      "step": 27880
+    },
+    {
+      "epoch": 106.45038167938931,
+      "grad_norm": 0.2940131723880768,
+      "learning_rate": 1.3476152463684778e-07,
+      "loss": 0.0666,
+      "step": 27890
+    },
+    {
+      "epoch": 106.4885496183206,
+      "grad_norm": 0.25805696845054626,
+      "learning_rate": 1.3349349183589155e-07,
+      "loss": 0.0729,
+      "step": 27900
+    },
+    {
+      "epoch": 106.52671755725191,
+      "grad_norm": 1.4068719148635864,
+      "learning_rate": 1.3223137229922356e-07,
+      "loss": 0.0731,
+      "step": 27910
+    },
+    {
+      "epoch": 106.56488549618321,
+      "grad_norm": 0.6141703724861145,
+      "learning_rate": 1.3097516756043982e-07,
+      "loss": 0.0695,
+      "step": 27920
+    },
+    {
+      "epoch": 106.6030534351145,
+      "grad_norm": 0.2119428515434265,
+      "learning_rate": 1.29724879145946e-07,
+      "loss": 0.0727,
+      "step": 27930
+    },
+    {
+      "epoch": 106.6412213740458,
+      "grad_norm": 0.24076199531555176,
+      "learning_rate": 1.284805085749613e-07,
+      "loss": 0.0788,
+      "step": 27940
+    },
+    {
+      "epoch": 106.6793893129771,
+      "grad_norm": 0.4202733337879181,
+      "learning_rate": 1.2724205735951288e-07,
+      "loss": 0.0714,
+      "step": 27950
+    },
+    {
+      "epoch": 106.7175572519084,
+      "grad_norm": 0.2482680380344391,
+      "learning_rate": 1.2600952700443592e-07,
+      "loss": 0.0708,
+      "step": 27960
+    },
+    {
+      "epoch": 106.7557251908397,
+      "grad_norm": 0.65274977684021,
+      "learning_rate": 1.247829190073707e-07,
+      "loss": 0.0727,
+      "step": 27970
+    },
+    {
+      "epoch": 106.79389312977099,
+      "grad_norm": 0.23994283378124237,
+      "learning_rate": 1.2356223485876174e-07,
+      "loss": 0.072,
+      "step": 27980
+    },
+    {
+      "epoch": 106.83206106870229,
+      "grad_norm": 0.2923242747783661,
+      "learning_rate": 1.2234747604185527e-07,
+      "loss": 0.0726,
+      "step": 27990
+    },
+    {
+      "epoch": 106.87022900763358,
+      "grad_norm": 0.2507231533527374,
+      "learning_rate": 1.2113864403269837e-07,
+      "loss": 0.0723,
+      "step": 28000
+    },
+    {
+      "epoch": 106.90839694656489,
+      "grad_norm": 0.6731197237968445,
+      "learning_rate": 1.1993574030013554e-07,
+      "loss": 0.0674,
+      "step": 28010
+    },
+    {
+      "epoch": 106.94656488549619,
+      "grad_norm": 0.17780010402202606,
+      "learning_rate": 1.1873876630580816e-07,
+      "loss": 0.0758,
+      "step": 28020
+    },
+    {
+      "epoch": 106.98473282442748,
+      "grad_norm": 0.7125159502029419,
+      "learning_rate": 1.1754772350415278e-07,
+      "loss": 0.0748,
+      "step": 28030
+    },
+    {
+      "epoch": 107.02290076335878,
+      "grad_norm": 0.355917364358902,
+      "learning_rate": 1.163626133423984e-07,
+      "loss": 0.0743,
+      "step": 28040
+    },
+    {
+      "epoch": 107.06106870229007,
+      "grad_norm": 0.4013403058052063,
+      "learning_rate": 1.151834372605659e-07,
+      "loss": 0.0645,
+      "step": 28050
+    },
+    {
+      "epoch": 107.09923664122137,
+      "grad_norm": 0.6267454624176025,
+      "learning_rate": 1.1401019669146474e-07,
+      "loss": 0.0681,
+      "step": 28060
+    },
+    {
+      "epoch": 107.13740458015268,
+      "grad_norm": 0.3011610209941864,
+      "learning_rate": 1.128428930606934e-07,
+      "loss": 0.0671,
+      "step": 28070
+    },
+    {
+      "epoch": 107.17557251908397,
+      "grad_norm": 0.5167744159698486,
+      "learning_rate": 1.1168152778663621e-07,
+      "loss": 0.0678,
+      "step": 28080
+    },
+    {
+      "epoch": 107.21374045801527,
+      "grad_norm": 0.2572309970855713,
+      "learning_rate": 1.10526102280461e-07,
+      "loss": 0.0666,
+      "step": 28090
+    },
+    {
+      "epoch": 107.25190839694656,
+      "grad_norm": 0.2860301434993744,
+      "learning_rate": 1.0937661794611865e-07,
+      "loss": 0.0628,
+      "step": 28100
+    },
+    {
+      "epoch": 107.29007633587786,
+      "grad_norm": 0.7779892086982727,
+      "learning_rate": 1.0823307618034129e-07,
+      "loss": 0.0683,
+      "step": 28110
+    },
+    {
+      "epoch": 107.32824427480917,
+      "grad_norm": 0.26767176389694214,
+      "learning_rate": 1.0709547837263967e-07,
+      "loss": 0.062,
+      "step": 28120
+    },
+    {
+      "epoch": 107.36641221374046,
+      "grad_norm": 0.38694190979003906,
+      "learning_rate": 1.0596382590530252e-07,
+      "loss": 0.0651,
+      "step": 28130
+    },
+    {
+      "epoch": 107.40458015267176,
+      "grad_norm": 0.9800488948822021,
+      "learning_rate": 1.0483812015339435e-07,
+      "loss": 0.0708,
+      "step": 28140
+    },
+    {
+      "epoch": 107.44274809160305,
+      "grad_norm": 0.5889298319816589,
+      "learning_rate": 1.0371836248475431e-07,
+      "loss": 0.0698,
+      "step": 28150
+    },
+    {
+      "epoch": 107.48091603053435,
+      "grad_norm": 0.2541733384132385,
+      "learning_rate": 1.0260455425999239e-07,
+      "loss": 0.0704,
+      "step": 28160
+    },
+    {
+      "epoch": 107.51908396946565,
+      "grad_norm": 0.26119011640548706,
+      "learning_rate": 1.0149669683249209e-07,
+      "loss": 0.0698,
+      "step": 28170
+    },
+    {
+      "epoch": 107.55725190839695,
+      "grad_norm": 0.42216214537620544,
+      "learning_rate": 1.0039479154840271e-07,
+      "loss": 0.0663,
+      "step": 28180
+    },
+    {
+      "epoch": 107.59541984732824,
+      "grad_norm": 0.7032060623168945,
+      "learning_rate": 9.92988397466449e-08,
+      "loss": 0.0772,
+      "step": 28190
+    },
+    {
+      "epoch": 107.63358778625954,
+      "grad_norm": 0.26121023297309875,
+      "learning_rate": 9.820884275890286e-08,
+      "loss": 0.0682,
+      "step": 28200
+    },
+    {
+      "epoch": 107.67175572519083,
+      "grad_norm": 0.7753889560699463,
+      "learning_rate": 9.712480190962548e-08,
+      "loss": 0.0703,
+      "step": 28210
+    },
+    {
+      "epoch": 107.70992366412214,
+      "grad_norm": 0.48209503293037415,
+      "learning_rate": 9.604671851602466e-08,
+      "loss": 0.0697,
+      "step": 28220
+    },
+    {
+      "epoch": 107.74809160305344,
+      "grad_norm": 0.3045661449432373,
+      "learning_rate": 9.497459388807306e-08,
+      "loss": 0.0659,
+      "step": 28230
+    },
+    {
+      "epoch": 107.78625954198473,
+      "grad_norm": 0.4404143691062927,
+      "learning_rate": 9.390842932850364e-08,
+      "loss": 0.0724,
+      "step": 28240
+    },
+    {
+      "epoch": 107.82442748091603,
+      "grad_norm": 0.34857988357543945,
+      "learning_rate": 9.284822613280731e-08,
+      "loss": 0.0706,
+      "step": 28250
+    },
+    {
+      "epoch": 107.86259541984732,
+      "grad_norm": 0.654318630695343,
+      "learning_rate": 9.179398558923024e-08,
+      "loss": 0.0696,
+      "step": 28260
+    },
+    {
+      "epoch": 107.90076335877862,
+      "grad_norm": 0.6860091686248779,
+      "learning_rate": 9.074570897877388e-08,
+      "loss": 0.0774,
+      "step": 28270
+    },
+    {
+      "epoch": 107.93893129770993,
+      "grad_norm": 0.36173543334007263,
+      "learning_rate": 8.970339757519375e-08,
+      "loss": 0.0646,
+      "step": 28280
+    },
+    {
+      "epoch": 107.97709923664122,
+      "grad_norm": 0.37401559948921204,
+      "learning_rate": 8.86670526449962e-08,
+      "loss": 0.0652,
+      "step": 28290
+    },
+    {
+      "epoch": 108.01526717557252,
+      "grad_norm": 0.4267309904098511,
+      "learning_rate": 8.763667544743837e-08,
+      "loss": 0.0682,
+      "step": 28300
+    },
+    {
+      "epoch": 108.05343511450381,
+      "grad_norm": 0.9610293507575989,
+      "learning_rate": 8.661226723452542e-08,
+      "loss": 0.0796,
+      "step": 28310
+    },
+    {
+      "epoch": 108.09160305343511,
+      "grad_norm": 0.8470723032951355,
+      "learning_rate": 8.559382925101e-08,
+      "loss": 0.0682,
+      "step": 28320
+    },
+    {
+      "epoch": 108.12977099236642,
+      "grad_norm": 0.5021095871925354,
+      "learning_rate": 8.458136273438943e-08,
+      "loss": 0.0686,
+      "step": 28330
+    },
+    {
+      "epoch": 108.16793893129771,
+      "grad_norm": 0.4912349581718445,
+      "learning_rate": 8.357486891490796e-08,
+      "loss": 0.064,
+      "step": 28340
+    },
+    {
+      "epoch": 108.20610687022901,
+      "grad_norm": 0.2633496820926666,
+      "learning_rate": 8.257434901554895e-08,
+      "loss": 0.0684,
+      "step": 28350
+    },
+    {
+      "epoch": 108.2442748091603,
+      "grad_norm": 0.25884994864463806,
+      "learning_rate": 8.157980425203938e-08,
+      "loss": 0.0704,
+      "step": 28360
+    },
+    {
+      "epoch": 108.2824427480916,
+      "grad_norm": 0.5608441233634949,
+      "learning_rate": 8.059123583284368e-08,
+      "loss": 0.0731,
+      "step": 28370
+    },
+    {
+      "epoch": 108.3206106870229,
+      "grad_norm": 0.32129549980163574,
+      "learning_rate": 7.960864495916653e-08,
+      "loss": 0.0691,
+      "step": 28380
+    },
+    {
+      "epoch": 108.3587786259542,
+      "grad_norm": 0.367439866065979,
+      "learning_rate": 7.863203282494846e-08,
+      "loss": 0.0764,
+      "step": 28390
+    },
+    {
+      "epoch": 108.3969465648855,
+      "grad_norm": 0.5697895884513855,
+      "learning_rate": 7.766140061686522e-08,
+      "loss": 0.0698,
+      "step": 28400
+    },
+    {
+      "epoch": 108.43511450381679,
+      "grad_norm": 0.5481321215629578,
+      "learning_rate": 7.669674951432615e-08,
+      "loss": 0.067,
+      "step": 28410
+    },
+    {
+      "epoch": 108.47328244274809,
+      "grad_norm": 0.2596859931945801,
+      "learning_rate": 7.573808068947363e-08,
+      "loss": 0.0645,
+      "step": 28420
+    },
+    {
+      "epoch": 108.5114503816794,
+      "grad_norm": 0.3332163989543915,
+      "learning_rate": 7.478539530718087e-08,
+      "loss": 0.0775,
+      "step": 28430
+    },
+    {
+      "epoch": 108.54961832061069,
+      "grad_norm": 0.3313477635383606,
+      "learning_rate": 7.383869452504966e-08,
+      "loss": 0.0666,
+      "step": 28440
+    },
+    {
+      "epoch": 108.58778625954199,
+      "grad_norm": 0.28801289200782776,
+      "learning_rate": 7.289797949341204e-08,
+      "loss": 0.0711,
+      "step": 28450
+    },
+    {
+      "epoch": 108.62595419847328,
+      "grad_norm": 0.4016149640083313,
+      "learning_rate": 7.196325135532422e-08,
+      "loss": 0.0649,
+      "step": 28460
+    },
+    {
+      "epoch": 108.66412213740458,
+      "grad_norm": 0.27881985902786255,
+      "learning_rate": 7.10345112465699e-08,
+      "loss": 0.0619,
+      "step": 28470
+    },
+    {
+      "epoch": 108.70229007633588,
+      "grad_norm": 0.5351571440696716,
+      "learning_rate": 7.011176029565525e-08,
+      "loss": 0.0687,
+      "step": 28480
+    },
+    {
+      "epoch": 108.74045801526718,
+      "grad_norm": 0.34915560483932495,
+      "learning_rate": 6.919499962381004e-08,
+      "loss": 0.0717,
+      "step": 28490
+    },
+    {
+      "epoch": 108.77862595419847,
+      "grad_norm": 0.3323579430580139,
+      "learning_rate": 6.828423034498488e-08,
+      "loss": 0.0642,
+      "step": 28500
+    },
+    {
+      "epoch": 108.81679389312977,
+      "grad_norm": 0.23722094297409058,
+      "learning_rate": 6.737945356585008e-08,
+      "loss": 0.067,
+      "step": 28510
+    },
+    {
+      "epoch": 108.85496183206106,
+      "grad_norm": 0.4655064344406128,
+      "learning_rate": 6.648067038579508e-08,
+      "loss": 0.0644,
+      "step": 28520
+    },
+    {
+      "epoch": 108.89312977099236,
+      "grad_norm": 0.27879801392555237,
+      "learning_rate": 6.558788189692578e-08,
+      "loss": 0.0705,
+      "step": 28530
+    },
+    {
+      "epoch": 108.93129770992367,
+      "grad_norm": 0.3153517246246338,
+      "learning_rate": 6.470108918406493e-08,
+      "loss": 0.0733,
+      "step": 28540
+    },
+    {
+      "epoch": 108.96946564885496,
+      "grad_norm": 0.42960867285728455,
+      "learning_rate": 6.382029332474893e-08,
+      "loss": 0.0667,
+      "step": 28550
+    },
+    {
+      "epoch": 109.00763358778626,
+      "grad_norm": 0.28105250000953674,
+      "learning_rate": 6.294549538922778e-08,
+      "loss": 0.0622,
+      "step": 28560
+    },
+    {
+      "epoch": 109.04580152671755,
+      "grad_norm": 0.597682535648346,
+      "learning_rate": 6.207669644046344e-08,
+      "loss": 0.0681,
+      "step": 28570
+    },
+    {
+      "epoch": 109.08396946564885,
+      "grad_norm": 0.31500354409217834,
+      "learning_rate": 6.121389753412866e-08,
+      "loss": 0.0698,
+      "step": 28580
+    },
+    {
+      "epoch": 109.12213740458016,
+      "grad_norm": 0.2600068151950836,
+      "learning_rate": 6.035709971860592e-08,
+      "loss": 0.0747,
+      "step": 28590
+    },
+    {
+      "epoch": 109.16030534351145,
+      "grad_norm": 0.2531625032424927,
+      "learning_rate": 5.95063040349847e-08,
+      "loss": 0.0688,
+      "step": 28600
+    },
+    {
+      "epoch": 109.19847328244275,
+      "grad_norm": 0.42508846521377563,
+      "learning_rate": 5.86615115170619e-08,
+      "loss": 0.0695,
+      "step": 28610
+    },
+    {
+      "epoch": 109.23664122137404,
+      "grad_norm": 0.38278618454933167,
+      "learning_rate": 5.782272319134086e-08,
+      "loss": 0.0705,
+      "step": 28620
+    },
+    {
+      "epoch": 109.27480916030534,
+      "grad_norm": 0.2914545238018036,
+      "learning_rate": 5.698994007702796e-08,
+      "loss": 0.0983,
+      "step": 28630
+    },
+    {
+      "epoch": 109.31297709923665,
+      "grad_norm": 0.2625536620616913,
+      "learning_rate": 5.616316318603321e-08,
+      "loss": 0.0712,
+      "step": 28640
+    },
+    {
+      "epoch": 109.35114503816794,
+      "grad_norm": 0.21834562718868256,
+      "learning_rate": 5.5342393522968e-08,
+      "loss": 0.0681,
+      "step": 28650
+    },
+    {
+      "epoch": 109.38931297709924,
+      "grad_norm": 0.305205374956131,
+      "learning_rate": 5.452763208514622e-08,
+      "loss": 0.0677,
+      "step": 28660
+    },
+    {
+      "epoch": 109.42748091603053,
+      "grad_norm": 0.4056416153907776,
+      "learning_rate": 5.371887986257873e-08,
+      "loss": 0.0676,
+      "step": 28670
+    },
+    {
+      "epoch": 109.46564885496183,
+      "grad_norm": 0.2562243640422821,
+      "learning_rate": 5.291613783797611e-08,
+      "loss": 0.067,
+      "step": 28680
+    },
+    {
+      "epoch": 109.50381679389314,
+      "grad_norm": 0.4858242869377136,
+      "learning_rate": 5.2119406986745336e-08,
+      "loss": 0.0686,
+      "step": 28690
+    },
+    {
+      "epoch": 109.54198473282443,
+      "grad_norm": 0.8462430238723755,
+      "learning_rate": 5.132868827698978e-08,
+      "loss": 0.0723,
+      "step": 28700
+    },
+    {
+      "epoch": 109.58015267175573,
+      "grad_norm": 0.46360403299331665,
+      "learning_rate": 5.0543982669507554e-08,
+      "loss": 0.0666,
+      "step": 28710
+    },
+    {
+      "epoch": 109.61832061068702,
+      "grad_norm": 0.2263144552707672,
+      "learning_rate": 4.976529111778872e-08,
+      "loss": 0.0672,
+      "step": 28720
+    },
+    {
+      "epoch": 109.65648854961832,
+      "grad_norm": 0.5241805911064148,
+      "learning_rate": 4.8992614568018624e-08,
+      "loss": 0.0671,
+      "step": 28730
+    },
+    {
+      "epoch": 109.69465648854961,
+      "grad_norm": 0.3242310881614685,
+      "learning_rate": 4.822595395907126e-08,
+      "loss": 0.0679,
+      "step": 28740
+    },
+    {
+      "epoch": 109.73282442748092,
+      "grad_norm": 0.300382137298584,
+      "learning_rate": 4.7465310222510885e-08,
+      "loss": 0.0718,
+      "step": 28750
+    },
+    {
+      "epoch": 109.77099236641222,
+      "grad_norm": 0.25149857997894287,
+      "learning_rate": 4.671068428259318e-08,
+      "loss": 0.0707,
+      "step": 28760
+    },
+    {
+      "epoch": 109.80916030534351,
+      "grad_norm": 0.49300718307495117,
+      "learning_rate": 4.596207705625799e-08,
+      "loss": 0.0614,
+      "step": 28770
+    },
+    {
+      "epoch": 109.8473282442748,
+      "grad_norm": 0.2667870819568634,
+      "learning_rate": 4.521948945313492e-08,
+      "loss": 0.0733,
+      "step": 28780
+    },
+    {
+      "epoch": 109.8854961832061,
+      "grad_norm": 0.32633841037750244,
+      "learning_rate": 4.4482922375537196e-08,
+      "loss": 0.073,
+      "step": 28790
+    },
+    {
+      "epoch": 109.92366412213741,
+      "grad_norm": 0.23223261535167694,
+      "learning_rate": 4.375237671846333e-08,
+      "loss": 0.0684,
+      "step": 28800
+    },
+    {
+      "epoch": 109.9618320610687,
+      "grad_norm": 0.2435031235218048,
+      "learning_rate": 4.302785336959547e-08,
+      "loss": 0.0648,
+      "step": 28810
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 0.273689866065979,
+      "learning_rate": 4.230935320929774e-08,
+      "loss": 0.0732,
+      "step": 28820
+    },
+    {
+      "epoch": 110.0381679389313,
+      "grad_norm": 0.290106862783432,
+      "learning_rate": 4.159687711061566e-08,
+      "loss": 0.0728,
+      "step": 28830
+    },
+    {
+      "epoch": 110.07633587786259,
+      "grad_norm": 0.3028109669685364,
+      "learning_rate": 4.089042593927506e-08,
+      "loss": 0.0642,
+      "step": 28840
+    },
+    {
+      "epoch": 110.1145038167939,
+      "grad_norm": 0.2728022634983063,
+      "learning_rate": 4.019000055367983e-08,
+      "loss": 0.0614,
+      "step": 28850
+    },
+    {
+      "epoch": 110.1526717557252,
+      "grad_norm": 0.3827283978462219,
+      "learning_rate": 3.949560180491363e-08,
+      "loss": 0.0665,
+      "step": 28860
+    },
+    {
+      "epoch": 110.19083969465649,
+      "grad_norm": 0.29793211817741394,
+      "learning_rate": 3.880723053673652e-08,
+      "loss": 0.0652,
+      "step": 28870
+    },
+    {
+      "epoch": 110.22900763358778,
+      "grad_norm": 0.3043680191040039,
+      "learning_rate": 3.812488758558386e-08,
+      "loss": 0.065,
+      "step": 28880
+    },
+    {
+      "epoch": 110.26717557251908,
+      "grad_norm": 0.47717395424842834,
+      "learning_rate": 3.744857378056743e-08,
+      "loss": 0.0696,
+      "step": 28890
+    },
+    {
+      "epoch": 110.30534351145039,
+      "grad_norm": 0.47816434502601624,
+      "learning_rate": 3.677828994347155e-08,
+      "loss": 0.0655,
+      "step": 28900
+    },
+    {
+      "epoch": 110.34351145038168,
+      "grad_norm": 0.4901565611362457,
+      "learning_rate": 3.61140368887547e-08,
+      "loss": 0.074,
+      "step": 28910
+    },
+    {
+      "epoch": 110.38167938931298,
+      "grad_norm": 0.22826752066612244,
+      "learning_rate": 3.5455815423546815e-08,
+      "loss": 0.0712,
+      "step": 28920
+    },
+    {
+      "epoch": 110.41984732824427,
+      "grad_norm": 0.44958510994911194,
+      "learning_rate": 3.480362634764922e-08,
+      "loss": 0.0687,
+      "step": 28930
+    },
+    {
+      "epoch": 110.45801526717557,
+      "grad_norm": 0.21928606927394867,
+      "learning_rate": 3.4157470453533015e-08,
+      "loss": 0.0683,
+      "step": 28940
+    },
+    {
+      "epoch": 110.49618320610686,
+      "grad_norm": 0.28822770714759827,
+      "learning_rate": 3.3517348526339034e-08,
+      "loss": 0.0678,
+      "step": 28950
+    },
+    {
+      "epoch": 110.53435114503817,
+      "grad_norm": 0.44351062178611755,
+      "learning_rate": 3.288326134387454e-08,
+      "loss": 0.0647,
+      "step": 28960
+    },
+    {
+      "epoch": 110.57251908396947,
+      "grad_norm": 0.7302173972129822,
+      "learning_rate": 3.225520967661655e-08,
+      "loss": 0.0683,
+      "step": 28970
+    },
+    {
+      "epoch": 110.61068702290076,
+      "grad_norm": 0.4752961993217468,
+      "learning_rate": 3.163319428770628e-08,
+      "loss": 0.0713,
+      "step": 28980
+    },
+    {
+      "epoch": 110.64885496183206,
+      "grad_norm": 0.35356375575065613,
+      "learning_rate": 3.1017215932951375e-08,
+      "loss": 0.0694,
+      "step": 28990
+    },
+    {
+      "epoch": 110.68702290076335,
+      "grad_norm": 0.2346401959657669,
+      "learning_rate": 3.040727536082366e-08,
+      "loss": 0.0752,
+      "step": 29000
+    },
+    {
+      "epoch": 110.72519083969466,
+      "grad_norm": 0.3200841546058655,
+      "learning_rate": 2.980337331245864e-08,
+      "loss": 0.0686,
+      "step": 29010
+    },
+    {
+      "epoch": 110.76335877862596,
+      "grad_norm": 0.26796045899391174,
+      "learning_rate": 2.9205510521653214e-08,
+      "loss": 0.0625,
+      "step": 29020
+    },
+    {
+      "epoch": 110.80152671755725,
+      "grad_norm": 0.297757089138031,
+      "learning_rate": 2.8613687714868497e-08,
+      "loss": 0.0635,
+      "step": 29030
+    },
+    {
+      "epoch": 110.83969465648855,
+      "grad_norm": 0.256091833114624,
+      "learning_rate": 2.8027905611223704e-08,
+      "loss": 0.0645,
+      "step": 29040
+    },
+    {
+      "epoch": 110.87786259541984,
+      "grad_norm": 0.27508917450904846,
+      "learning_rate": 2.7448164922500577e-08,
+      "loss": 0.0867,
+      "step": 29050
+    },
+    {
+      "epoch": 110.91603053435115,
+      "grad_norm": 0.4821559488773346,
+      "learning_rate": 2.687446635313784e-08,
+      "loss": 0.0717,
+      "step": 29060
+    },
+    {
+      "epoch": 110.95419847328245,
+      "grad_norm": 0.26554882526397705,
+      "learning_rate": 2.6306810600233435e-08,
+      "loss": 0.0655,
+      "step": 29070
+    },
+    {
+      "epoch": 110.99236641221374,
+      "grad_norm": 0.19737420976161957,
+      "learning_rate": 2.5745198353542834e-08,
+      "loss": 0.0663,
+      "step": 29080
+    },
+    {
+      "epoch": 111.03053435114504,
+      "grad_norm": 0.258440762758255,
+      "learning_rate": 2.518963029547794e-08,
+      "loss": 0.0601,
+      "step": 29090
+    },
+    {
+      "epoch": 111.06870229007633,
+      "grad_norm": 0.25001877546310425,
+      "learning_rate": 2.464010710110598e-08,
+      "loss": 0.0745,
+      "step": 29100
+    },
+    {
+      "epoch": 111.10687022900764,
+      "grad_norm": 0.2975136935710907,
+      "learning_rate": 2.4096629438150055e-08,
+      "loss": 0.0669,
+      "step": 29110
+    },
+    {
+      "epoch": 111.14503816793894,
+      "grad_norm": 0.2732112407684326,
+      "learning_rate": 2.3559197966985802e-08,
+      "loss": 0.0692,
+      "step": 29120
+    },
+    {
+      "epoch": 111.18320610687023,
+      "grad_norm": 0.3679260313510895,
+      "learning_rate": 2.302781334064419e-08,
+      "loss": 0.063,
+      "step": 29130
+    },
+    {
+      "epoch": 111.22137404580153,
+      "grad_norm": 0.2609327435493469,
+      "learning_rate": 2.2502476204807056e-08,
+      "loss": 0.0694,
+      "step": 29140
+    },
+    {
+      "epoch": 111.25954198473282,
+      "grad_norm": 0.436776727437973,
+      "learning_rate": 2.1983187197808788e-08,
+      "loss": 0.0758,
+      "step": 29150
+    },
+    {
+      "epoch": 111.29770992366412,
+      "grad_norm": 0.6425755023956299,
+      "learning_rate": 2.1469946950634645e-08,
+      "loss": 0.0716,
+      "step": 29160
+    },
+    {
+      "epoch": 111.33587786259542,
+      "grad_norm": 0.22093476355075836,
+      "learning_rate": 2.0962756086919112e-08,
+      "loss": 0.0765,
+      "step": 29170
+    },
+    {
+      "epoch": 111.37404580152672,
+      "grad_norm": 0.5801058411598206,
+      "learning_rate": 2.046161522294754e-08,
+      "loss": 0.0782,
+      "step": 29180
+    },
+    {
+      "epoch": 111.41221374045801,
+      "grad_norm": 0.5053842663764954,
+      "learning_rate": 1.9966524967653943e-08,
+      "loss": 0.0657,
+      "step": 29190
+    },
+    {
+      "epoch": 111.45038167938931,
+      "grad_norm": 0.31466546654701233,
+      "learning_rate": 1.9477485922618222e-08,
+      "loss": 0.0752,
+      "step": 29200
+    },
+    {
+      "epoch": 111.4885496183206,
+      "grad_norm": 0.3375924825668335,
+      "learning_rate": 1.899449868207004e-08,
+      "loss": 0.0683,
+      "step": 29210
+    },
+    {
+      "epoch": 111.52671755725191,
+      "grad_norm": 0.49920454621315,
+      "learning_rate": 1.8517563832884387e-08,
+      "loss": 0.0734,
+      "step": 29220
+    },
+    {
+      "epoch": 111.56488549618321,
+      "grad_norm": 0.6840377449989319,
+      "learning_rate": 1.8046681954581035e-08,
+      "loss": 0.0684,
+      "step": 29230
+    },
+    {
+      "epoch": 111.6030534351145,
+      "grad_norm": 0.37621980905532837,
+      "learning_rate": 1.7581853619327295e-08,
+      "loss": 0.0656,
+      "step": 29240
+    },
+    {
+      "epoch": 111.6412213740458,
+      "grad_norm": 0.23118160665035248,
+      "learning_rate": 1.7123079391932474e-08,
+      "loss": 0.0704,
+      "step": 29250
+    },
+    {
+      "epoch": 111.6793893129771,
+      "grad_norm": 0.3853679299354553,
+      "learning_rate": 1.6670359829850657e-08,
+      "loss": 0.07,
+      "step": 29260
+    },
+    {
+      "epoch": 111.7175572519084,
+      "grad_norm": 0.20712091028690338,
+      "learning_rate": 1.6223695483179035e-08,
+      "loss": 0.0667,
+      "step": 29270
+    },
+    {
+      "epoch": 111.7557251908397,
+      "grad_norm": 0.45825672149658203,
+      "learning_rate": 1.5783086894656796e-08,
+      "loss": 0.0658,
+      "step": 29280
+    },
+    {
+      "epoch": 111.79389312977099,
+      "grad_norm": 0.25062382221221924,
+      "learning_rate": 1.5348534599665122e-08,
+      "loss": 0.0656,
+      "step": 29290
+    },
+    {
+      "epoch": 111.83206106870229,
+      "grad_norm": 0.7445496320724487,
+      "learning_rate": 1.4920039126225528e-08,
+      "loss": 0.0756,
+      "step": 29300
+    },
+    {
+      "epoch": 111.87022900763358,
+      "grad_norm": 0.4825686812400818,
+      "learning_rate": 1.449760099500097e-08,
+      "loss": 0.0684,
+      "step": 29310
+    },
+    {
+      "epoch": 111.90839694656489,
+      "grad_norm": 0.2946581244468689,
+      "learning_rate": 1.4081220719293076e-08,
+      "loss": 0.0647,
+      "step": 29320
+    },
+    {
+      "epoch": 111.94656488549619,
+      "grad_norm": 0.2547703981399536,
+      "learning_rate": 1.3670898805043798e-08,
+      "loss": 0.0762,
+      "step": 29330
+    },
+    {
+      "epoch": 111.98473282442748,
+      "grad_norm": 0.44975385069847107,
+      "learning_rate": 1.326663575083209e-08,
+      "loss": 0.0698,
+      "step": 29340
+    },
+    {
+      "epoch": 112.02290076335878,
+      "grad_norm": 0.2827472686767578,
+      "learning_rate": 1.286843204787669e-08,
+      "loss": 0.0679,
+      "step": 29350
+    },
+    {
+      "epoch": 112.06106870229007,
+      "grad_norm": 0.4250585734844208,
+      "learning_rate": 1.2476288180032214e-08,
+      "loss": 0.0638,
+      "step": 29360
+    },
+    {
+      "epoch": 112.09923664122137,
+      "grad_norm": 0.3234623670578003,
+      "learning_rate": 1.2090204623790292e-08,
+      "loss": 0.0611,
+      "step": 29370
+    },
+    {
+      "epoch": 112.13740458015268,
+      "grad_norm": 0.20758773386478424,
+      "learning_rate": 1.1710181848278435e-08,
+      "loss": 0.0717,
+      "step": 29380
+    },
+    {
+      "epoch": 112.17557251908397,
+      "grad_norm": 0.476921945810318,
+      "learning_rate": 1.133622031526116e-08,
+      "loss": 0.0667,
+      "step": 29390
+    },
+    {
+      "epoch": 112.21374045801527,
+      "grad_norm": 0.809885561466217,
+      "learning_rate": 1.0968320479136097e-08,
+      "loss": 0.0722,
+      "step": 29400
+    },
+    {
+      "epoch": 112.25190839694656,
+      "grad_norm": 0.218302920460701,
+      "learning_rate": 1.0606482786936768e-08,
+      "loss": 0.0659,
+      "step": 29410
+    },
+    {
+      "epoch": 112.29007633587786,
+      "grad_norm": 0.9261871576309204,
+      "learning_rate": 1.0250707678329808e-08,
+      "loss": 0.0777,
+      "step": 29420
+    },
+    {
+      "epoch": 112.32824427480917,
+      "grad_norm": 0.3251490890979767,
+      "learning_rate": 9.900995585615525e-09,
+      "loss": 0.0663,
+      "step": 29430
+    },
+    {
+      "epoch": 112.36641221374046,
+      "grad_norm": 0.4006260931491852,
+      "learning_rate": 9.55734693372734e-09,
+      "loss": 0.067,
+      "step": 29440
+    },
+    {
+      "epoch": 112.40458015267176,
+      "grad_norm": 0.3395634889602661,
+      "learning_rate": 9.219762140231237e-09,
+      "loss": 0.0695,
+      "step": 29450
+    },
+    {
+      "epoch": 112.44274809160305,
+      "grad_norm": 0.5408310294151306,
+      "learning_rate": 8.888241615322979e-09,
+      "loss": 0.0658,
+      "step": 29460
+    },
+    {
+      "epoch": 112.48091603053435,
+      "grad_norm": 0.22297032177448273,
+      "learning_rate": 8.562785761833114e-09,
+      "loss": 0.0761,
+      "step": 29470
+    },
+    {
+      "epoch": 112.51908396946565,
+      "grad_norm": 0.1803329437971115,
+      "learning_rate": 8.243394975219753e-09,
+      "loss": 0.0647,
+      "step": 29480
+    },
+    {
+      "epoch": 112.55725190839695,
+      "grad_norm": 0.4298841059207916,
+      "learning_rate": 7.930069643573568e-09,
+      "loss": 0.07,
+      "step": 29490
+    },
+    {
+      "epoch": 112.59541984732824,
+      "grad_norm": 0.27020174264907837,
+      "learning_rate": 7.622810147614457e-09,
+      "loss": 0.0712,
+      "step": 29500
+    },
+    {
+      "epoch": 112.63358778625954,
+      "grad_norm": 0.2004048377275467,
+      "learning_rate": 7.321616860690995e-09,
+      "loss": 0.0665,
+      "step": 29510
+    },
+    {
+      "epoch": 112.67175572519083,
+      "grad_norm": 0.25727108120918274,
+      "learning_rate": 7.026490148782095e-09,
+      "loss": 0.0726,
+      "step": 29520
+    },
+    {
+      "epoch": 112.70992366412214,
+      "grad_norm": 0.3128455877304077,
+      "learning_rate": 6.737430370494236e-09,
+      "loss": 0.0684,
+      "step": 29530
+    },
+    {
+      "epoch": 112.74809160305344,
+      "grad_norm": 0.27916234731674194,
+      "learning_rate": 6.4544378770625695e-09,
+      "loss": 0.0669,
+      "step": 29540
+    },
+    {
+      "epoch": 112.78625954198473,
+      "grad_norm": 0.26003262400627136,
+      "learning_rate": 6.177513012349257e-09,
+      "loss": 0.0651,
+      "step": 29550
+    },
+    {
+      "epoch": 112.82442748091603,
+      "grad_norm": 0.6421084403991699,
+      "learning_rate": 5.906656112844578e-09,
+      "loss": 0.0683,
+      "step": 29560
+    },
+    {
+      "epoch": 112.86259541984732,
+      "grad_norm": 0.2419942170381546,
+      "learning_rate": 5.641867507664156e-09,
+      "loss": 0.066,
+      "step": 29570
+    },
+    {
+      "epoch": 112.90076335877862,
+      "grad_norm": 0.4693322479724884,
+      "learning_rate": 5.383147518552845e-09,
+      "loss": 0.0672,
+      "step": 29580
+    },
+    {
+      "epoch": 112.93893129770993,
+      "grad_norm": 0.29389244318008423,
+      "learning_rate": 5.13049645987862e-09,
+      "loss": 0.0692,
+      "step": 29590
+    },
+    {
+      "epoch": 112.97709923664122,
+      "grad_norm": 0.5387037396430969,
+      "learning_rate": 4.883914638636467e-09,
+      "loss": 0.074,
+      "step": 29600
+    },
+    {
+      "epoch": 113.01526717557252,
+      "grad_norm": 0.2849062383174896,
+      "learning_rate": 4.6434023544467135e-09,
+      "loss": 0.0698,
+      "step": 29610
+    },
+    {
+      "epoch": 113.05343511450381,
+      "grad_norm": 0.2868102192878723,
+      "learning_rate": 4.408959899554477e-09,
+      "loss": 0.0694,
+      "step": 29620
+    },
+    {
+      "epoch": 113.09160305343511,
+      "grad_norm": 0.21665853261947632,
+      "learning_rate": 4.180587558829663e-09,
+      "loss": 0.0681,
+      "step": 29630
+    },
+    {
+      "epoch": 113.12977099236642,
+      "grad_norm": 0.2538401484489441,
+      "learning_rate": 3.9582856097658554e-09,
+      "loss": 0.0667,
+      "step": 29640
+    },
+    {
+      "epoch": 113.16793893129771,
+      "grad_norm": 0.4532374143600464,
+      "learning_rate": 3.74205432248087e-09,
+      "loss": 0.0852,
+      "step": 29650
+    },
+    {
+      "epoch": 113.20610687022901,
+      "grad_norm": 0.3058018386363983,
+      "learning_rate": 3.531893959716204e-09,
+      "loss": 0.0707,
+      "step": 29660
+    },
+    {
+      "epoch": 113.2442748091603,
+      "grad_norm": 0.21909840404987335,
+      "learning_rate": 3.327804776837029e-09,
+      "loss": 0.0932,
+      "step": 29670
+    },
+    {
+      "epoch": 113.2824427480916,
+      "grad_norm": 0.49850842356681824,
+      "learning_rate": 3.1297870218299776e-09,
+      "loss": 0.0721,
+      "step": 29680
+    },
+    {
+      "epoch": 113.3206106870229,
+      "grad_norm": 0.4520060420036316,
+      "learning_rate": 2.9378409353059133e-09,
+      "loss": 0.0695,
+      "step": 29690
+    },
+    {
+      "epoch": 113.3587786259542,
+      "grad_norm": 0.24441687762737274,
+      "learning_rate": 2.7519667504971593e-09,
+      "loss": 0.078,
+      "step": 29700
+    },
+    {
+      "epoch": 113.3969465648855,
+      "grad_norm": 0.4024795889854431,
+      "learning_rate": 2.572164693258605e-09,
+      "loss": 0.0649,
+      "step": 29710
+    },
+    {
+      "epoch": 113.43511450381679,
+      "grad_norm": 0.2969071567058563,
+      "learning_rate": 2.3984349820665997e-09,
+      "loss": 0.0697,
+      "step": 29720
+    },
+    {
+      "epoch": 113.47328244274809,
+      "grad_norm": 0.24398191273212433,
+      "learning_rate": 2.2307778280189484e-09,
+      "loss": 0.0691,
+      "step": 29730
+    },
+    {
+      "epoch": 113.5114503816794,
+      "grad_norm": 0.5313243269920349,
+      "learning_rate": 2.06919343483547e-09,
+      "loss": 0.0686,
+      "step": 29740
+    },
+    {
+      "epoch": 113.54961832061069,
+      "grad_norm": 0.33933746814727783,
+      "learning_rate": 1.9136819988557765e-09,
+      "loss": 0.0683,
+      "step": 29750
+    },
+    {
+      "epoch": 113.58778625954199,
+      "grad_norm": 0.39412805438041687,
+      "learning_rate": 1.764243709041491e-09,
+      "loss": 0.0755,
+      "step": 29760
+    },
+    {
+      "epoch": 113.62595419847328,
+      "grad_norm": 0.3232634961605072,
+      "learning_rate": 1.6208787469734755e-09,
+      "loss": 0.0709,
+      "step": 29770
+    },
+    {
+      "epoch": 113.66412213740458,
+      "grad_norm": 0.34813275933265686,
+      "learning_rate": 1.4835872868546041e-09,
+      "loss": 0.0706,
+      "step": 29780
+    },
+    {
+      "epoch": 113.70229007633588,
+      "grad_norm": 0.5944547057151794,
+      "learning_rate": 1.3523694955064338e-09,
+      "loss": 0.0737,
+      "step": 29790
+    },
+    {
+      "epoch": 113.74045801526718,
+      "grad_norm": 0.5624353289604187,
+      "learning_rate": 1.2272255323708682e-09,
+      "loss": 0.0642,
+      "step": 29800
+    },
+    {
+      "epoch": 113.77862595419847,
+      "grad_norm": 0.7441560626029968,
+      "learning_rate": 1.1081555495096042e-09,
+      "loss": 0.0699,
+      "step": 29810
+    },
+    {
+      "epoch": 113.81679389312977,
+      "grad_norm": 0.8023766875267029,
+      "learning_rate": 9.951596916041307e-10,
+      "loss": 0.0705,
+      "step": 29820
+    },
+    {
+      "epoch": 113.85496183206106,
+      "grad_norm": 0.3070337176322937,
+      "learning_rate": 8.88238095955174e-10,
+      "loss": 0.0678,
+      "step": 29830
+    },
+    {
+      "epoch": 113.89312977099236,
+      "grad_norm": 0.3391276001930237,
+      "learning_rate": 7.873908924821427e-10,
+      "loss": 0.0659,
+      "step": 29840
+    },
+    {
+      "epoch": 113.93129770992367,
+      "grad_norm": 0.2922404110431671,
+      "learning_rate": 6.926182037242379e-10,
+      "loss": 0.0701,
+      "step": 29850
+    },
+    {
+      "epoch": 113.96946564885496,
+      "grad_norm": 0.34531670808792114,
+      "learning_rate": 6.039201448387877e-10,
+      "loss": 0.0731,
+      "step": 29860
+    },
+    {
+      "epoch": 114.00763358778626,
+      "grad_norm": 0.4919748902320862,
+      "learning_rate": 5.212968236029126e-10,
+      "loss": 0.0744,
+      "step": 29870
+    },
+    {
+      "epoch": 114.04580152671755,
+      "grad_norm": 0.35671672224998474,
+      "learning_rate": 4.447483404118602e-10,
+      "loss": 0.07,
+      "step": 29880
+    },
+    {
+      "epoch": 114.08396946564885,
+      "grad_norm": 0.25368714332580566,
+      "learning_rate": 3.7427478827845033e-10,
+      "loss": 0.0624,
+      "step": 29890
+    },
+    {
+      "epoch": 114.12213740458016,
+      "grad_norm": 1.127943992614746,
+      "learning_rate": 3.0987625283473986e-10,
+      "loss": 0.09,
+      "step": 29900
+    },
+    {
+      "epoch": 114.16030534351145,
+      "grad_norm": 0.26309409737586975,
+      "learning_rate": 2.5155281233202324e-10,
+      "loss": 0.066,
+      "step": 29910
+    },
+    {
+      "epoch": 114.19847328244275,
+      "grad_norm": 0.26694461703300476,
+      "learning_rate": 1.9930453763750135e-10,
+      "loss": 0.0771,
+      "step": 29920
+    },
+    {
+      "epoch": 114.23664122137404,
+      "grad_norm": 0.3160324692726135,
+      "learning_rate": 1.531314922387228e-10,
+      "loss": 0.0669,
+      "step": 29930
+    },
+    {
+      "epoch": 114.27480916030534,
+      "grad_norm": 0.24074247479438782,
+      "learning_rate": 1.1303373224025305e-10,
+      "loss": 0.0718,
+      "step": 29940
+    },
+    {
+      "epoch": 114.31297709923665,
+      "grad_norm": 0.47037428617477417,
+      "learning_rate": 7.901130636367437e-11,
+      "loss": 0.0722,
+      "step": 29950
+    },
+    {
+      "epoch": 114.35114503816794,
+      "grad_norm": 0.5677265524864197,
+      "learning_rate": 5.1064255950361574e-11,
+      "loss": 0.0716,
+      "step": 29960
+    },
+    {
+      "epoch": 114.38931297709924,
+      "grad_norm": 0.3310537338256836,
+      "learning_rate": 2.9192614958706286e-11,
+      "loss": 0.0715,
+      "step": 29970
+    },
+    {
+      "epoch": 114.42748091603053,
+      "grad_norm": 0.5864256024360657,
+      "learning_rate": 1.3396409964117062e-11,
+      "loss": 0.072,
+      "step": 29980
+    },
+    {
+      "epoch": 114.46564885496183,
+      "grad_norm": 0.6437695026397705,
+      "learning_rate": 3.6756601606846574e-12,
+      "loss": 0.0669,
+      "step": 29990
+    },
+    {
+      "epoch": 114.50381679389314,
+      "grad_norm": 0.734368085861206,
+      "learning_rate": 3.037736062694308e-14,
+      "loss": 0.0751,
+      "step": 30000
+    },
+    {
+      "epoch": 114.50381679389314,
+      "step": 30000,
+      "total_flos": 0.0,
+      "train_loss": 0.0959479827662309,
+      "train_runtime": 13839.331,
+      "train_samples_per_second": 17.342,
+      "train_steps_per_second": 2.168
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 30000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 115,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}