diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -2,162412 +2,100812 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.7692320189761337,
+  "epoch": 2.3395233221231173,
   "eval_steps": 500,
-  "global_step": 232000,
+  "global_step": 144000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.00016246689736966092,
-      "grad_norm": 3.21875,
+      "grad_norm": 18.5,
       "learning_rate": 5e-05,
-      "loss": 1.0066,
+      "loss": 4.2222,
       "step": 10
     },
     {
       "epoch": 0.00032493379473932184,
-      "grad_norm": 2.015625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.7951,
+      "loss": 2.8702,
       "step": 20
     },
     {
       "epoch": 0.0004874006921089828,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.8157,
+      "loss": 2.8067,
       "step": 30
     },
     {
       "epoch": 0.0006498675894786437,
-      "grad_norm": 1.9765625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.7483,
+      "loss": 2.7599,
       "step": 40
     },
     {
       "epoch": 0.0008123344868483047,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.7824,
+      "loss": 2.7352,
       "step": 50
     },
     {
       "epoch": 0.0009748013842179656,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.7569,
+      "loss": 2.7011,
       "step": 60
     },
     {
       "epoch": 0.0011372682815876266,
-      "grad_norm": 2.140625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.7243,
+      "loss": 2.6803,
       "step": 70
     },
     {
       "epoch": 0.0012997351789572874,
-      "grad_norm": 2.328125,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.7144,
+      "loss": 2.6885,
       "step": 80
     },
     {
       "epoch": 0.0014622020763269484,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.7287,
+      "loss": 2.6393,
       "step": 90
     },
     {
       "epoch": 0.0016246689736966094,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.7012,
+      "loss": 2.623,
       "step": 100
     },
     {
       "epoch": 0.0017871358710662704,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.7384,
+      "loss": 2.5805,
       "step": 110
     },
     {
       "epoch": 0.0019496027684359311,
-      "grad_norm": 1.671875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.6988,
+      "loss": 2.5268,
       "step": 120
     },
     {
       "epoch": 0.002112069665805592,
-      "grad_norm": 2.4375,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.7101,
+      "loss": 2.5629,
       "step": 130
     },
     {
       "epoch": 0.002274536563175253,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.7127,
+      "loss": 2.516,
       "step": 140
     },
     {
       "epoch": 0.002437003460544914,
-      "grad_norm": 2.484375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.6875,
+      "loss": 2.5065,
       "step": 150
     },
     {
       "epoch": 0.0025994703579145747,
-      "grad_norm": 2.5,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.6538,
+      "loss": 2.4767,
       "step": 160
     },
     {
       "epoch": 0.002761937255284236,
-      "grad_norm": 2.3125,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.6603,
+      "loss": 2.365,
       "step": 170
     },
     {
       "epoch": 0.0029244041526538967,
-      "grad_norm": 2.796875,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.6824,
+      "loss": 2.3641,
       "step": 180
     },
     {
       "epoch": 0.0030868710500235575,
-      "grad_norm": 2.65625,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.6743,
+      "loss": 2.3778,
       "step": 190
     },
     {
       "epoch": 0.0032493379473932187,
-      "grad_norm": 2.90625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.6485,
+      "loss": 2.3157,
       "step": 200
     },
     {
       "epoch": 0.0034118048447628795,
-      "grad_norm": 2.40625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.6561,
+      "loss": 2.351,
       "step": 210
     },
     {
       "epoch": 0.0035742717421325407,
-      "grad_norm": 3.625,
+      "grad_norm": 12.0625,
       "learning_rate": 5e-05,
-      "loss": 0.6336,
+      "loss": 2.3476,
       "step": 220
     },
     {
       "epoch": 0.0037367386395022015,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.6448,
+      "loss": 2.3315,
       "step": 230
     },
     {
       "epoch": 0.0038992055368718623,
-      "grad_norm": 3.5,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.6522,
+      "loss": 2.2686,
       "step": 240
     },
     {
       "epoch": 0.0040616724342415235,
-      "grad_norm": 2.546875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.6449,
+      "loss": 2.2057,
       "step": 250
     },
     {
       "epoch": 0.004224139331611184,
-      "grad_norm": 3.890625,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.6624,
+      "loss": 2.255,
       "step": 260
     },
     {
       "epoch": 0.004386606228980845,
-      "grad_norm": 3.875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.6278,
+      "loss": 2.2444,
       "step": 270
     },
     {
       "epoch": 0.004549073126350506,
-      "grad_norm": 3.875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.6308,
+      "loss": 2.2084,
       "step": 280
     },
     {
       "epoch": 0.004711540023720167,
-      "grad_norm": 2.75,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.6149,
+      "loss": 2.1988,
       "step": 290
     },
     {
       "epoch": 0.004874006921089828,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.6097,
+      "loss": 2.1301,
       "step": 300
     },
     {
       "epoch": 0.005036473818459489,
-      "grad_norm": 3.53125,
+      "grad_norm": 20.125,
       "learning_rate": 5e-05,
-      "loss": 0.632,
+      "loss": 2.2097,
       "step": 310
     },
     {
       "epoch": 0.005198940715829149,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.6255,
+      "loss": 2.2341,
       "step": 320
     },
     {
       "epoch": 0.005361407613198811,
-      "grad_norm": 3.109375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.5955,
+      "loss": 2.2017,
       "step": 330
     },
     {
       "epoch": 0.005523874510568472,
-      "grad_norm": 3.625,
+      "grad_norm": 12.25,
       "learning_rate": 5e-05,
-      "loss": 0.6413,
+      "loss": 2.1777,
       "step": 340
     },
     {
       "epoch": 0.005686341407938132,
-      "grad_norm": 2.640625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.6127,
+      "loss": 2.1847,
       "step": 350
     },
     {
       "epoch": 0.005848808305307793,
-      "grad_norm": 4.40625,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.6269,
+      "loss": 2.1179,
       "step": 360
     },
     {
       "epoch": 0.006011275202677455,
-      "grad_norm": 3.875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.6175,
+      "loss": 2.16,
       "step": 370
     },
     {
       "epoch": 0.006173742100047115,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.6031,
+      "loss": 2.1363,
       "step": 380
     },
     {
       "epoch": 0.006336208997416776,
-      "grad_norm": 4.125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.6269,
+      "loss": 2.0848,
       "step": 390
     },
     {
       "epoch": 0.006498675894786437,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5863,
+      "loss": 2.1271,
       "step": 400
     },
     {
       "epoch": 0.006661142792156098,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5944,
+      "loss": 2.1607,
       "step": 410
     },
     {
       "epoch": 0.006823609689525759,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.5719,
+      "loss": 2.1711,
       "step": 420
     },
     {
       "epoch": 0.00698607658689542,
-      "grad_norm": 2.640625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.6048,
+      "loss": 2.1329,
       "step": 430
     },
     {
       "epoch": 0.007148543484265081,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.6376,
+      "loss": 2.1782,
       "step": 440
     },
     {
       "epoch": 0.007311010381634742,
-      "grad_norm": 2.765625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.5809,
+      "loss": 2.155,
       "step": 450
     },
     {
       "epoch": 0.007473477279004403,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5967,
+      "loss": 2.1267,
       "step": 460
     },
     {
       "epoch": 0.007635944176374064,
-      "grad_norm": 3.015625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.5654,
+      "loss": 2.062,
       "step": 470
     },
     {
       "epoch": 0.0077984110737437246,
-      "grad_norm": 4.0625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.6064,
+      "loss": 2.1209,
       "step": 480
     },
     {
       "epoch": 0.007960877971113386,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5931,
+      "loss": 2.0781,
       "step": 490
     },
     {
       "epoch": 0.008123344868483047,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5953,
+      "loss": 2.1085,
       "step": 500
     },
     {
       "epoch": 0.008285811765852708,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.5465,
+      "loss": 2.0701,
       "step": 510
     },
     {
       "epoch": 0.008448278663222368,
-      "grad_norm": 3.5,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.5826,
+      "loss": 2.1291,
       "step": 520
     },
     {
       "epoch": 0.008610745560592029,
-      "grad_norm": 3.390625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.587,
+      "loss": 2.1417,
       "step": 530
     },
     {
       "epoch": 0.00877321245796169,
-      "grad_norm": 2.828125,
+      "grad_norm": 12.5,
       "learning_rate": 5e-05,
-      "loss": 0.5662,
+      "loss": 2.1319,
       "step": 540
     },
     {
       "epoch": 0.008935679355331351,
-      "grad_norm": 2.59375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5789,
+      "loss": 2.1132,
       "step": 550
     },
     {
       "epoch": 0.009098146252701013,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5627,
+      "loss": 2.0785,
       "step": 560
     },
     {
       "epoch": 0.009260613150070674,
-      "grad_norm": 3.109375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5633,
+      "loss": 2.0269,
       "step": 570
     },
     {
       "epoch": 0.009423080047440333,
-      "grad_norm": 3.140625,
+      "grad_norm": 15.875,
       "learning_rate": 5e-05,
-      "loss": 0.5736,
+      "loss": 2.0257,
       "step": 580
     },
     {
       "epoch": 0.009585546944809994,
-      "grad_norm": 4.0,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.5617,
+      "loss": 2.0357,
       "step": 590
     },
     {
       "epoch": 0.009748013842179656,
-      "grad_norm": 3.546875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5861,
+      "loss": 2.0786,
       "step": 600
     },
     {
       "epoch": 0.009910480739549317,
-      "grad_norm": 2.390625,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5692,
+      "loss": 2.0636,
       "step": 610
     },
     {
       "epoch": 0.010072947636918978,
-      "grad_norm": 2.984375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.5865,
+      "loss": 2.0761,
       "step": 620
     },
     {
       "epoch": 0.01023541453428864,
-      "grad_norm": 4.9375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5693,
+      "loss": 2.0936,
       "step": 630
     },
     {
       "epoch": 0.010397881431658299,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5736,
+      "loss": 2.0929,
       "step": 640
     },
     {
       "epoch": 0.01056034832902796,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.5493,
+      "loss": 2.0383,
       "step": 650
     },
     {
       "epoch": 0.010722815226397621,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5237,
+      "loss": 2.0966,
       "step": 660
     },
     {
       "epoch": 0.010885282123767282,
-      "grad_norm": 4.46875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5831,
+      "loss": 2.0044,
       "step": 670
     },
     {
       "epoch": 0.011047749021136944,
-      "grad_norm": 3.90625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5964,
+      "loss": 1.9833,
       "step": 680
     },
     {
       "epoch": 0.011210215918506605,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5678,
+      "loss": 2.0724,
       "step": 690
     },
     {
       "epoch": 0.011372682815876264,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.5772,
+      "loss": 2.0792,
       "step": 700
     },
     {
       "epoch": 0.011535149713245926,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.5437,
+      "loss": 2.0345,
       "step": 710
     },
     {
       "epoch": 0.011697616610615587,
-      "grad_norm": 2.71875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.551,
+      "loss": 2.0374,
       "step": 720
     },
     {
       "epoch": 0.011860083507985248,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5805,
+      "loss": 2.0347,
       "step": 730
     },
     {
       "epoch": 0.01202255040535491,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5655,
+      "loss": 2.0389,
       "step": 740
     },
     {
       "epoch": 0.01218501730272457,
-      "grad_norm": 3.96875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.5644,
+      "loss": 2.0409,
       "step": 750
     },
     {
       "epoch": 0.01234748420009423,
-      "grad_norm": 4.375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5548,
+      "loss": 1.9977,
       "step": 760
     },
     {
       "epoch": 0.012509951097463891,
-      "grad_norm": 4.0,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5766,
+      "loss": 2.0257,
       "step": 770
     },
     {
       "epoch": 0.012672417994833552,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5525,
+      "loss": 2.0504,
       "step": 780
     },
     {
       "epoch": 0.012834884892203214,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.568,
+      "loss": 2.0268,
       "step": 790
     },
     {
       "epoch": 0.012997351789572875,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.5306,
+      "loss": 1.99,
       "step": 800
     },
     {
       "epoch": 0.013159818686942536,
-      "grad_norm": 2.59375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.553,
+      "loss": 2.0369,
       "step": 810
     },
     {
       "epoch": 0.013322285584312196,
-      "grad_norm": 4.0,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5774,
+      "loss": 2.0161,
       "step": 820
     },
     {
       "epoch": 0.013484752481681857,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5627,
+      "loss": 1.9958,
       "step": 830
     },
     {
       "epoch": 0.013647219379051518,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5863,
+      "loss": 2.0108,
       "step": 840
     },
     {
       "epoch": 0.01380968627642118,
-      "grad_norm": 4.28125,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5595,
+      "loss": 2.005,
       "step": 850
     },
     {
       "epoch": 0.01397215317379084,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.577,
+      "loss": 2.0494,
       "step": 860
     },
     {
       "epoch": 0.014134620071160502,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5762,
+      "loss": 2.0173,
       "step": 870
     },
     {
       "epoch": 0.014297086968530163,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5727,
+      "loss": 2.009,
       "step": 880
     },
     {
       "epoch": 0.014459553865899822,
-      "grad_norm": 2.71875,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.5575,
+      "loss": 2.0045,
       "step": 890
     },
     {
       "epoch": 0.014622020763269484,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.5525,
+      "loss": 1.999,
       "step": 900
     },
     {
       "epoch": 0.014784487660639145,
-      "grad_norm": 3.828125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5563,
+      "loss": 2.0119,
       "step": 910
     },
     {
       "epoch": 0.014946954558008806,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.5376,
+      "loss": 1.9934,
       "step": 920
     },
     {
       "epoch": 0.015109421455378467,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5451,
+      "loss": 2.0727,
       "step": 930
     },
     {
       "epoch": 0.015271888352748128,
-      "grad_norm": 3.921875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5302,
+      "loss": 1.9875,
       "step": 940
     },
     {
       "epoch": 0.015434355250117788,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.5226,
+      "loss": 1.9839,
       "step": 950
     },
     {
       "epoch": 0.015596822147487449,
-      "grad_norm": 4.5,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5468,
+      "loss": 1.9896,
       "step": 960
     },
     {
       "epoch": 0.01575928904485711,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5577,
+      "loss": 1.9408,
       "step": 970
     },
     {
       "epoch": 0.01592175594222677,
-      "grad_norm": 2.46875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.546,
+      "loss": 1.9486,
       "step": 980
     },
     {
       "epoch": 0.016084222839596433,
-      "grad_norm": 3.25,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5501,
+      "loss": 1.9862,
       "step": 990
     },
     {
       "epoch": 0.016246689736966094,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.5513,
+      "loss": 2.0006,
       "step": 1000
     },
     {
       "epoch": 0.016409156634335755,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5597,
+      "loss": 1.9865,
       "step": 1010
     },
     {
       "epoch": 0.016571623531705416,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5574,
+      "loss": 1.9855,
       "step": 1020
     },
     {
       "epoch": 0.016734090429075078,
-      "grad_norm": 4.65625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.5803,
+      "loss": 2.0077,
       "step": 1030
     },
     {
       "epoch": 0.016896557326444735,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.5938,
+      "loss": 1.9773,
       "step": 1040
     },
     {
       "epoch": 0.017059024223814397,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5554,
+      "loss": 1.9923,
       "step": 1050
     },
     {
       "epoch": 0.017221491121184058,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5378,
+      "loss": 1.9769,
       "step": 1060
     },
     {
       "epoch": 0.01738395801855372,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5429,
+      "loss": 1.9912,
       "step": 1070
     },
     {
       "epoch": 0.01754642491592338,
-      "grad_norm": 2.84375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.5355,
+      "loss": 1.9507,
       "step": 1080
     },
     {
       "epoch": 0.01770889181329304,
-      "grad_norm": 3.4375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.5686,
+      "loss": 1.9716,
       "step": 1090
     },
     {
       "epoch": 0.017871358710662703,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5438,
+      "loss": 1.9511,
       "step": 1100
     },
     {
       "epoch": 0.018033825608032364,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5228,
+      "loss": 1.9659,
       "step": 1110
     },
     {
       "epoch": 0.018196292505402025,
-      "grad_norm": 2.625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.5712,
+      "loss": 1.9807,
       "step": 1120
     },
     {
       "epoch": 0.018358759402771686,
-      "grad_norm": 2.65625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5662,
+      "loss": 1.953,
       "step": 1130
     },
     {
       "epoch": 0.018521226300141348,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.5599,
+      "loss": 1.9439,
       "step": 1140
     },
     {
       "epoch": 0.01868369319751101,
-      "grad_norm": 3.75,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5511,
+      "loss": 1.9661,
       "step": 1150
     },
     {
       "epoch": 0.018846160094880667,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.5468,
+      "loss": 1.9723,
       "step": 1160
     },
     {
       "epoch": 0.019008626992250328,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.545,
+      "loss": 1.9882,
       "step": 1170
     },
     {
       "epoch": 0.01917109388961999,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.557,
+      "loss": 1.9708,
       "step": 1180
     },
     {
       "epoch": 0.01933356078698965,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5456,
+      "loss": 1.9717,
       "step": 1190
     },
     {
       "epoch": 0.01949602768435931,
-      "grad_norm": 2.875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.5305,
+      "loss": 1.9375,
       "step": 1200
     },
     {
       "epoch": 0.019658494581728973,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5195,
+      "loss": 1.921,
       "step": 1210
     },
     {
       "epoch": 0.019820961479098634,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5442,
+      "loss": 1.9272,
       "step": 1220
     },
     {
       "epoch": 0.019983428376468295,
-      "grad_norm": 2.53125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5324,
+      "loss": 1.919,
       "step": 1230
     },
     {
       "epoch": 0.020145895273837956,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.5525,
+      "loss": 1.9992,
       "step": 1240
     },
     {
       "epoch": 0.020308362171207617,
-      "grad_norm": 3.03125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5451,
+      "loss": 1.8881,
       "step": 1250
     },
     {
       "epoch": 0.02047082906857728,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5489,
+      "loss": 1.9595,
       "step": 1260
     },
     {
       "epoch": 0.02063329596594694,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5415,
+      "loss": 1.9441,
       "step": 1270
     },
     {
       "epoch": 0.020795762863316598,
-      "grad_norm": 4.125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.5343,
+      "loss": 1.9065,
       "step": 1280
     },
     {
       "epoch": 0.02095822976068626,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5382,
+      "loss": 1.9703,
       "step": 1290
     },
     {
       "epoch": 0.02112069665805592,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.542,
+      "loss": 1.9717,
       "step": 1300
     },
     {
       "epoch": 0.02128316355542558,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.5437,
+      "loss": 1.9648,
       "step": 1310
     },
     {
       "epoch": 0.021445630452795243,
-      "grad_norm": 3.484375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.5462,
+      "loss": 1.9126,
       "step": 1320
     },
     {
       "epoch": 0.021608097350164904,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.5419,
+      "loss": 1.9794,
       "step": 1330
     },
     {
       "epoch": 0.021770564247534565,
-      "grad_norm": 3.765625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5692,
+      "loss": 1.9236,
       "step": 1340
     },
     {
       "epoch": 0.021933031144904226,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.5275,
+      "loss": 1.9317,
       "step": 1350
     },
     {
       "epoch": 0.022095498042273887,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.5104,
+      "loss": 1.9245,
       "step": 1360
     },
     {
       "epoch": 0.02225796493964355,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.5338,
+      "loss": 1.9526,
       "step": 1370
     },
     {
       "epoch": 0.02242043183701321,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5563,
+      "loss": 2.0052,
       "step": 1380
     },
     {
       "epoch": 0.02258289873438287,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.5508,
+      "loss": 1.8721,
       "step": 1390
     },
     {
       "epoch": 0.02274536563175253,
-      "grad_norm": 3.15625,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5457,
+      "loss": 1.9513,
       "step": 1400
     },
     {
       "epoch": 0.02290783252912219,
-      "grad_norm": 3.21875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.5231,
+      "loss": 1.9132,
       "step": 1410
     },
     {
       "epoch": 0.02307029942649185,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.5393,
+      "loss": 1.8867,
       "step": 1420
     },
     {
       "epoch": 0.023232766323861512,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.5501,
+      "loss": 1.9067,
       "step": 1430
     },
     {
       "epoch": 0.023395233221231174,
-      "grad_norm": 2.59375,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5355,
+      "loss": 1.8866,
       "step": 1440
     },
     {
       "epoch": 0.023557700118600835,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.5275,
+      "loss": 1.9581,
       "step": 1450
     },
     {
       "epoch": 0.023720167015970496,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5258,
+      "loss": 1.9323,
       "step": 1460
     },
     {
       "epoch": 0.023882633913340157,
-      "grad_norm": 4.125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.5355,
+      "loss": 1.8765,
       "step": 1470
     },
     {
       "epoch": 0.02404510081070982,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5267,
+      "loss": 2.0147,
       "step": 1480
     },
     {
       "epoch": 0.02420756770807948,
-      "grad_norm": 4.1875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.5283,
+      "loss": 1.9294,
       "step": 1490
     },
     {
       "epoch": 0.02437003460544914,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.532,
+      "loss": 1.8393,
       "step": 1500
     },
     {
       "epoch": 0.024532501502818802,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.539,
+      "loss": 1.8806,
       "step": 1510
     },
     {
       "epoch": 0.02469496840018846,
-      "grad_norm": 2.78125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.541,
+      "loss": 1.9108,
       "step": 1520
     },
     {
       "epoch": 0.02485743529755812,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.5426,
+      "loss": 1.8955,
       "step": 1530
     },
     {
       "epoch": 0.025019902194927782,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.5198,
+      "loss": 1.92,
       "step": 1540
     },
     {
       "epoch": 0.025182369092297444,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5339,
+      "loss": 1.9345,
       "step": 1550
     },
     {
       "epoch": 0.025344835989667105,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5301,
+      "loss": 1.9404,
       "step": 1560
     },
     {
       "epoch": 0.025507302887036766,
-      "grad_norm": 3.125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.5376,
+      "loss": 1.9687,
       "step": 1570
     },
     {
       "epoch": 0.025669769784406427,
-      "grad_norm": 2.90625,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.5279,
+      "loss": 1.916,
       "step": 1580
     },
     {
       "epoch": 0.02583223668177609,
-      "grad_norm": 3.734375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5138,
+      "loss": 1.9022,
       "step": 1590
     },
     {
       "epoch": 0.02599470357914575,
-      "grad_norm": 3.390625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5472,
+      "loss": 1.9425,
       "step": 1600
     },
     {
       "epoch": 0.02615717047651541,
-      "grad_norm": 3.375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5422,
+      "loss": 1.8871,
       "step": 1610
     },
     {
       "epoch": 0.026319637373885072,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5423,
+      "loss": 1.9168,
       "step": 1620
     },
     {
       "epoch": 0.026482104271254733,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5065,
+      "loss": 1.7908,
       "step": 1630
     },
     {
       "epoch": 0.02664457116862439,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.5288,
+      "loss": 1.9419,
       "step": 1640
     },
     {
       "epoch": 0.026807038065994052,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5147,
+      "loss": 1.8631,
       "step": 1650
     },
     {
       "epoch": 0.026969504963363714,
-      "grad_norm": 3.375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5483,
+      "loss": 1.8995,
       "step": 1660
     },
     {
       "epoch": 0.027131971860733375,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5282,
+      "loss": 1.9424,
       "step": 1670
     },
     {
       "epoch": 0.027294438758103036,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5255,
+      "loss": 1.8994,
       "step": 1680
     },
     {
       "epoch": 0.027456905655472697,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.512,
+      "loss": 1.8724,
       "step": 1690
     },
     {
       "epoch": 0.02761937255284236,
-      "grad_norm": 2.71875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5236,
+      "loss": 1.8762,
       "step": 1700
     },
     {
       "epoch": 0.02778183945021202,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.523,
+      "loss": 1.9012,
       "step": 1710
     },
     {
       "epoch": 0.02794430634758168,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5637,
+      "loss": 1.9093,
       "step": 1720
     },
     {
       "epoch": 0.028106773244951342,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.5265,
+      "loss": 1.8886,
       "step": 1730
     },
     {
       "epoch": 0.028269240142321003,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.5145,
+      "loss": 1.9103,
       "step": 1740
     },
     {
       "epoch": 0.028431707039690664,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5002,
+      "loss": 1.8413,
       "step": 1750
     },
     {
       "epoch": 0.028594173937060326,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.5098,
+      "loss": 1.8148,
       "step": 1760
     },
     {
       "epoch": 0.028756640834429983,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.5103,
+      "loss": 1.9225,
       "step": 1770
     },
     {
       "epoch": 0.028919107731799645,
-      "grad_norm": 4.03125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5369,
+      "loss": 1.8839,
       "step": 1780
     },
     {
       "epoch": 0.029081574629169306,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.511,
+      "loss": 1.8992,
       "step": 1790
     },
     {
       "epoch": 0.029244041526538967,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.5324,
+      "loss": 1.893,
       "step": 1800
     },
     {
       "epoch": 0.02940650842390863,
-      "grad_norm": 4.59375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.5391,
+      "loss": 1.8234,
       "step": 1810
     },
     {
       "epoch": 0.02956897532127829,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5024,
+      "loss": 1.9001,
       "step": 1820
     },
     {
       "epoch": 0.02973144221864795,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.5525,
+      "loss": 1.8568,
       "step": 1830
     },
     {
       "epoch": 0.029893909116017612,
-      "grad_norm": 3.25,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5406,
+      "loss": 1.8893,
       "step": 1840
     },
     {
       "epoch": 0.030056376013387273,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5327,
+      "loss": 1.8858,
       "step": 1850
     },
     {
       "epoch": 0.030218842910756934,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.5312,
+      "loss": 1.8771,
       "step": 1860
     },
     {
       "epoch": 0.030381309808126596,
-      "grad_norm": 4.75,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.5136,
+      "loss": 1.8805,
       "step": 1870
     },
     {
       "epoch": 0.030543776705496257,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5105,
+      "loss": 1.898,
       "step": 1880
     },
     {
       "epoch": 0.030706243602865915,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.538,
+      "loss": 1.7919,
       "step": 1890
     },
     {
       "epoch": 0.030868710500235576,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5115,
+      "loss": 1.8997,
       "step": 1900
     },
     {
       "epoch": 0.031031177397605237,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.525,
+      "loss": 1.9184,
       "step": 1910
     },
     {
       "epoch": 0.031193644294974898,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.5534,
+      "loss": 1.8856,
       "step": 1920
     },
     {
       "epoch": 0.03135611119234456,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.5133,
+      "loss": 1.8914,
       "step": 1930
     },
     {
       "epoch": 0.03151857808971422,
-      "grad_norm": 3.9375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.5114,
+      "loss": 1.8983,
       "step": 1940
     },
     {
       "epoch": 0.03168104498708388,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.5433,
+      "loss": 1.8814,
       "step": 1950
     },
     {
       "epoch": 0.03184351188445354,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.5465,
+      "loss": 1.8934,
       "step": 1960
     },
     {
       "epoch": 0.0320059787818232,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4845,
+      "loss": 1.8018,
       "step": 1970
     },
     {
       "epoch": 0.032168445679192866,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.5311,
+      "loss": 1.8521,
       "step": 1980
     },
     {
       "epoch": 0.03233091257656252,
-      "grad_norm": 2.125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.5444,
+      "loss": 1.8517,
       "step": 1990
     },
     {
       "epoch": 0.03249337947393219,
-      "grad_norm": 3.25,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5321,
+      "loss": 1.925,
       "step": 2000
     },
     {
       "epoch": 0.032655846371301846,
-      "grad_norm": 3.0,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5373,
+      "loss": 1.8674,
       "step": 2010
     },
     {
       "epoch": 0.03281831326867151,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.5273,
+      "loss": 1.867,
       "step": 2020
     },
     {
       "epoch": 0.03298078016604117,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4987,
+      "loss": 1.8817,
       "step": 2030
     },
     {
       "epoch": 0.03314324706341083,
-      "grad_norm": 2.75,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.5216,
+      "loss": 1.8527,
       "step": 2040
     },
     {
       "epoch": 0.03330571396078049,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5193,
+      "loss": 1.8443,
       "step": 2050
     },
     {
       "epoch": 0.033468180858150155,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5318,
+      "loss": 1.8478,
       "step": 2060
     },
     {
       "epoch": 0.03363064775551981,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5166,
+      "loss": 1.8793,
       "step": 2070
     },
     {
       "epoch": 0.03379311465288947,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5334,
+      "loss": 1.8445,
       "step": 2080
     },
     {
       "epoch": 0.033955581550259135,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.5001,
+      "loss": 1.8189,
       "step": 2090
     },
     {
       "epoch": 0.03411804844762879,
-      "grad_norm": 2.546875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5295,
+      "loss": 1.8637,
       "step": 2100
     },
     {
       "epoch": 0.03428051534499846,
-      "grad_norm": 2.5,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.51,
+      "loss": 1.85,
       "step": 2110
     },
     {
       "epoch": 0.034442982242368116,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5093,
+      "loss": 1.9001,
       "step": 2120
     },
     {
       "epoch": 0.03460544913973778,
-      "grad_norm": 3.0,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.5209,
+      "loss": 1.8521,
       "step": 2130
     },
     {
       "epoch": 0.03476791603710744,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.5111,
+      "loss": 1.8267,
       "step": 2140
     },
     {
       "epoch": 0.0349303829344771,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.5215,
+      "loss": 1.8491,
       "step": 2150
     },
     {
       "epoch": 0.03509284983184676,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4999,
+      "loss": 1.8285,
       "step": 2160
     },
     {
       "epoch": 0.035255316729216425,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5264,
+      "loss": 1.8566,
       "step": 2170
     },
     {
       "epoch": 0.03541778362658608,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.5292,
+      "loss": 1.8169,
       "step": 2180
     },
     {
       "epoch": 0.03558025052395574,
-      "grad_norm": 3.390625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5275,
+      "loss": 1.8938,
       "step": 2190
     },
     {
       "epoch": 0.035742717421325405,
-      "grad_norm": 2.953125,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.5188,
+      "loss": 1.8532,
       "step": 2200
     },
     {
       "epoch": 0.03590518431869506,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.5215,
+      "loss": 1.863,
       "step": 2210
     },
     {
       "epoch": 0.03606765121606473,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.5314,
+      "loss": 1.88,
       "step": 2220
     },
     {
       "epoch": 0.036230118113434386,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.5341,
+      "loss": 1.8143,
       "step": 2230
     },
     {
       "epoch": 0.03639258501080405,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.5086,
+      "loss": 1.8619,
       "step": 2240
     },
     {
       "epoch": 0.03655505190817371,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5368,
+      "loss": 1.8283,
       "step": 2250
     },
     {
       "epoch": 0.03671751880554337,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5302,
+      "loss": 1.7903,
       "step": 2260
     },
     {
       "epoch": 0.03687998570291303,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5288,
+      "loss": 1.8053,
       "step": 2270
     },
     {
       "epoch": 0.037042452600282695,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.553,
+      "loss": 1.7992,
       "step": 2280
     },
     {
       "epoch": 0.03720491949765235,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.5065,
+      "loss": 1.8087,
       "step": 2290
     },
     {
       "epoch": 0.03736738639502202,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4968,
+      "loss": 1.8324,
       "step": 2300
     },
     {
       "epoch": 0.037529853292391675,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.5422,
+      "loss": 1.8718,
       "step": 2310
     },
     {
       "epoch": 0.03769232018976133,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.5015,
+      "loss": 1.8082,
       "step": 2320
     },
     {
       "epoch": 0.037854787087131,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.5278,
+      "loss": 1.8131,
       "step": 2330
     },
     {
       "epoch": 0.038017253984500655,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.518,
+      "loss": 1.8354,
       "step": 2340
     },
     {
       "epoch": 0.03817972088187032,
-      "grad_norm": 2.625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.515,
+      "loss": 1.7845,
       "step": 2350
     },
     {
       "epoch": 0.03834218777923998,
-      "grad_norm": 3.125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5209,
+      "loss": 1.804,
       "step": 2360
     },
     {
       "epoch": 0.03850465467660964,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4882,
+      "loss": 1.8226,
       "step": 2370
     },
     {
       "epoch": 0.0386671215739793,
-      "grad_norm": 2.5,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5491,
+      "loss": 1.8022,
       "step": 2380
     },
     {
       "epoch": 0.038829588471348965,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.51,
+      "loss": 1.8297,
       "step": 2390
     },
     {
       "epoch": 0.03899205536871862,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5149,
+      "loss": 1.7524,
       "step": 2400
     },
     {
       "epoch": 0.03915452226608829,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.5118,
+      "loss": 1.848,
       "step": 2410
     },
     {
       "epoch": 0.039316989163457945,
-      "grad_norm": 4.03125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5224,
+      "loss": 1.8467,
       "step": 2420
     },
     {
       "epoch": 0.0394794560608276,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.5045,
+      "loss": 1.866,
       "step": 2430
     },
     {
       "epoch": 0.03964192295819727,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4879,
+      "loss": 1.8111,
       "step": 2440
     },
     {
       "epoch": 0.039804389855566925,
-      "grad_norm": 3.828125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5017,
+      "loss": 1.8193,
       "step": 2450
     },
     {
       "epoch": 0.03996685675293659,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5041,
+      "loss": 1.8222,
       "step": 2460
     },
     {
       "epoch": 0.04012932365030625,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4858,
+      "loss": 1.8226,
       "step": 2470
     },
     {
       "epoch": 0.04029179054767591,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4953,
+      "loss": 1.8185,
       "step": 2480
     },
     {
       "epoch": 0.04045425744504557,
-      "grad_norm": 3.0,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4973,
+      "loss": 1.8827,
       "step": 2490
     },
     {
       "epoch": 0.040616724342415235,
-      "grad_norm": 3.5,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.498,
+      "loss": 1.8436,
       "step": 2500
     },
     {
       "epoch": 0.04077919123978489,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5288,
+      "loss": 1.8364,
       "step": 2510
     },
     {
       "epoch": 0.04094165813715456,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.509,
+      "loss": 1.8858,
       "step": 2520
     },
     {
       "epoch": 0.041104125034524215,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.5065,
+      "loss": 1.8513,
       "step": 2530
     },
     {
       "epoch": 0.04126659193189388,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.5228,
+      "loss": 1.8579,
       "step": 2540
     },
     {
       "epoch": 0.04142905882926354,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5034,
+      "loss": 1.8191,
       "step": 2550
     },
     {
       "epoch": 0.041591525726633195,
-      "grad_norm": 2.46875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.533,
+      "loss": 1.7919,
       "step": 2560
     },
     {
       "epoch": 0.04175399262400286,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.5366,
+      "loss": 1.8537,
       "step": 2570
     },
     {
       "epoch": 0.04191645952137252,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.5051,
+      "loss": 1.8356,
       "step": 2580
     },
     {
       "epoch": 0.04207892641874218,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.5078,
+      "loss": 1.7649,
       "step": 2590
     },
     {
       "epoch": 0.04224139331611184,
-      "grad_norm": 2.84375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5057,
+      "loss": 1.795,
       "step": 2600
     },
     {
       "epoch": 0.042403860213481505,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5208,
+      "loss": 1.8169,
       "step": 2610
     },
     {
       "epoch": 0.04256632711085116,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5167,
+      "loss": 1.8144,
       "step": 2620
     },
     {
       "epoch": 0.04272879400822083,
-      "grad_norm": 4.4375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.5204,
+      "loss": 1.8678,
       "step": 2630
     },
     {
       "epoch": 0.042891260905590485,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.5305,
+      "loss": 1.8315,
       "step": 2640
     },
     {
       "epoch": 0.04305372780296015,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5019,
+      "loss": 1.8467,
       "step": 2650
     },
     {
       "epoch": 0.04321619470032981,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4978,
+      "loss": 1.8167,
       "step": 2660
     },
     {
       "epoch": 0.04337866159769947,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5065,
+      "loss": 1.7885,
       "step": 2670
     },
     {
       "epoch": 0.04354112849506913,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.5178,
+      "loss": 1.852,
       "step": 2680
     },
     {
       "epoch": 0.04370359539243879,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4997,
+      "loss": 1.7799,
       "step": 2690
     },
     {
       "epoch": 0.04386606228980845,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.5105,
+      "loss": 1.901,
       "step": 2700
     },
     {
       "epoch": 0.04402852918717811,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.5006,
+      "loss": 1.8487,
       "step": 2710
     },
     {
       "epoch": 0.044190996084547775,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4963,
+      "loss": 1.8063,
       "step": 2720
     },
     {
       "epoch": 0.04435346298191743,
-      "grad_norm": 3.765625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4908,
+      "loss": 1.8423,
       "step": 2730
     },
     {
       "epoch": 0.0445159298792871,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.5338,
+      "loss": 1.8448,
       "step": 2740
     },
     {
       "epoch": 0.044678396776656755,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.5176,
+      "loss": 1.8164,
       "step": 2750
     },
     {
       "epoch": 0.04484086367402642,
-      "grad_norm": 2.921875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.5261,
+      "loss": 1.8369,
       "step": 2760
     },
     {
       "epoch": 0.04500333057139608,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.5021,
+      "loss": 1.7832,
       "step": 2770
     },
     {
       "epoch": 0.04516579746876574,
-      "grad_norm": 2.84375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4945,
+      "loss": 1.8083,
       "step": 2780
     },
     {
       "epoch": 0.0453282643661354,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.5166,
+      "loss": 1.8192,
       "step": 2790
     },
     {
       "epoch": 0.04549073126350506,
-      "grad_norm": 3.0,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5067,
+      "loss": 1.875,
       "step": 2800
     },
     {
       "epoch": 0.04565319816087472,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4977,
+      "loss": 1.7782,
       "step": 2810
     },
     {
       "epoch": 0.04581566505824438,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4854,
+      "loss": 1.808,
       "step": 2820
     },
     {
       "epoch": 0.045978131955614045,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5296,
+      "loss": 1.8154,
       "step": 2830
     },
     {
       "epoch": 0.0461405988529837,
-      "grad_norm": 4.1875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4979,
+      "loss": 1.816,
       "step": 2840
     },
     {
       "epoch": 0.04630306575035337,
-      "grad_norm": 3.828125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4771,
+      "loss": 1.8379,
       "step": 2850
     },
     {
       "epoch": 0.046465532647723025,
-      "grad_norm": 2.4375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.4846,
+      "loss": 1.7865,
       "step": 2860
     },
     {
       "epoch": 0.04662799954509269,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4965,
+      "loss": 1.7962,
       "step": 2870
     },
     {
       "epoch": 0.04679046644246235,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4903,
+      "loss": 1.7959,
       "step": 2880
     },
     {
       "epoch": 0.04695293333983201,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4837,
+      "loss": 1.7987,
       "step": 2890
     },
     {
       "epoch": 0.04711540023720167,
-      "grad_norm": 3.0,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.5015,
+      "loss": 1.8463,
       "step": 2900
     },
     {
       "epoch": 0.047277867134571334,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4985,
+      "loss": 1.7798,
       "step": 2910
     },
     {
       "epoch": 0.04744033403194099,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4776,
+      "loss": 1.7573,
       "step": 2920
     },
     {
       "epoch": 0.04760280092931065,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4722,
+      "loss": 1.7726,
       "step": 2930
     },
     {
       "epoch": 0.047765267826680315,
-      "grad_norm": 2.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4873,
+      "loss": 1.7923,
       "step": 2940
     },
     {
       "epoch": 0.04792773472404997,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4965,
+      "loss": 1.7849,
       "step": 2950
     },
     {
       "epoch": 0.04809020162141964,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.5322,
+      "loss": 1.8086,
       "step": 2960
     },
     {
       "epoch": 0.048252668518789295,
-      "grad_norm": 4.625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4881,
+      "loss": 1.7831,
       "step": 2970
     },
     {
       "epoch": 0.04841513541615896,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.5168,
+      "loss": 1.8096,
       "step": 2980
     },
     {
       "epoch": 0.04857760231352862,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4691,
+      "loss": 1.8366,
       "step": 2990
     },
     {
       "epoch": 0.04874006921089828,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4932,
+      "loss": 1.7399,
       "step": 3000
     },
     {
       "epoch": 0.04890253610826794,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.518,
+      "loss": 1.8117,
       "step": 3010
     },
     {
       "epoch": 0.049065003005637604,
-      "grad_norm": 3.265625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.5022,
+      "loss": 1.7757,
       "step": 3020
     },
     {
       "epoch": 0.04922746990300726,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4976,
+      "loss": 1.7725,
       "step": 3030
     },
     {
       "epoch": 0.04938993680037692,
-      "grad_norm": 2.5,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4941,
+      "loss": 1.8023,
       "step": 3040
     },
     {
       "epoch": 0.049552403697746585,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4912,
+      "loss": 1.8026,
       "step": 3050
     },
     {
       "epoch": 0.04971487059511624,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4942,
+      "loss": 1.7869,
       "step": 3060
     },
     {
       "epoch": 0.04987733749248591,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.472,
+      "loss": 1.7411,
       "step": 3070
     },
     {
       "epoch": 0.050039804389855565,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4971,
+      "loss": 1.7994,
       "step": 3080
     },
     {
       "epoch": 0.05020227128722523,
-      "grad_norm": 2.421875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.4767,
+      "loss": 1.8092,
       "step": 3090
     },
     {
       "epoch": 0.05036473818459489,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5082,
+      "loss": 1.7706,
       "step": 3100
     },
     {
       "epoch": 0.05052720508196455,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4739,
+      "loss": 1.7814,
       "step": 3110
     },
     {
       "epoch": 0.05068967197933421,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4639,
+      "loss": 1.811,
       "step": 3120
     },
     {
       "epoch": 0.050852138876703874,
-      "grad_norm": 3.9375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4888,
+      "loss": 1.7808,
       "step": 3130
     },
     {
       "epoch": 0.05101460577407353,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.5012,
+      "loss": 1.8141,
       "step": 3140
     },
     {
       "epoch": 0.0511770726714432,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4734,
+      "loss": 1.7944,
       "step": 3150
     },
     {
       "epoch": 0.051339539568812854,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.5001,
+      "loss": 1.8139,
       "step": 3160
     },
     {
       "epoch": 0.05150200646618251,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4744,
+      "loss": 1.7823,
       "step": 3170
     },
     {
       "epoch": 0.05166447336355218,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5186,
+      "loss": 1.8137,
       "step": 3180
     },
     {
       "epoch": 0.051826940260921835,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4938,
+      "loss": 1.7791,
       "step": 3190
     },
     {
       "epoch": 0.0519894071582915,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.5049,
+      "loss": 1.7663,
       "step": 3200
     },
     {
       "epoch": 0.05215187405566116,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.489,
+      "loss": 1.7568,
       "step": 3210
     },
     {
       "epoch": 0.05231434095303082,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4643,
+      "loss": 1.7536,
       "step": 3220
     },
     {
       "epoch": 0.05247680785040048,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4785,
+      "loss": 1.7857,
       "step": 3230
     },
     {
       "epoch": 0.052639274747770144,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4926,
+      "loss": 1.7958,
       "step": 3240
     },
     {
       "epoch": 0.0528017416451398,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4976,
+      "loss": 1.7384,
       "step": 3250
     },
     {
       "epoch": 0.05296420854250947,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.5111,
+      "loss": 1.7904,
       "step": 3260
     },
     {
       "epoch": 0.053126675439879124,
-      "grad_norm": 2.375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.473,
+      "loss": 1.8061,
       "step": 3270
     },
     {
       "epoch": 0.05328914233724878,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.5039,
+      "loss": 1.7625,
       "step": 3280
     },
     {
       "epoch": 0.05345160923461845,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.5058,
+      "loss": 1.769,
       "step": 3290
     },
     {
       "epoch": 0.053614076131988105,
-      "grad_norm": 3.6875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.5071,
+      "loss": 1.7668,
       "step": 3300
     },
     {
       "epoch": 0.05377654302935777,
-      "grad_norm": 4.0625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.498,
+      "loss": 1.797,
       "step": 3310
     },
     {
       "epoch": 0.05393900992672743,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4807,
+      "loss": 1.7675,
       "step": 3320
     },
     {
       "epoch": 0.05410147682409709,
-      "grad_norm": 4.375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4688,
+      "loss": 1.7085,
       "step": 3330
     },
     {
       "epoch": 0.05426394372146675,
-      "grad_norm": 4.1875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.5027,
+      "loss": 1.7297,
       "step": 3340
     },
     {
       "epoch": 0.054426410618836414,
-      "grad_norm": 4.53125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4901,
+      "loss": 1.7487,
       "step": 3350
     },
     {
       "epoch": 0.05458887751620607,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.5154,
+      "loss": 1.7584,
       "step": 3360
     },
     {
       "epoch": 0.05475134441357574,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4845,
+      "loss": 1.7941,
       "step": 3370
     },
     {
       "epoch": 0.054913811310945394,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.512,
+      "loss": 1.7868,
       "step": 3380
     },
     {
       "epoch": 0.05507627820831506,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4777,
+      "loss": 1.7995,
       "step": 3390
     },
     {
       "epoch": 0.05523874510568472,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4831,
+      "loss": 1.7553,
       "step": 3400
     },
     {
       "epoch": 0.055401212003054374,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5117,
+      "loss": 1.7704,
       "step": 3410
     },
     {
       "epoch": 0.05556367890042404,
-      "grad_norm": 2.75,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5205,
+      "loss": 1.6645,
       "step": 3420
     },
     {
       "epoch": 0.0557261457977937,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.493,
+      "loss": 1.7738,
       "step": 3430
     },
     {
       "epoch": 0.05588861269516336,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4864,
+      "loss": 1.7887,
       "step": 3440
     },
     {
       "epoch": 0.05605107959253302,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.467,
+      "loss": 1.8069,
       "step": 3450
     },
     {
       "epoch": 0.056213546489902684,
-      "grad_norm": 3.5,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4858,
+      "loss": 1.7526,
       "step": 3460
     },
     {
       "epoch": 0.05637601338727234,
-      "grad_norm": 2.75,
+      "grad_norm": 12.0,
       "learning_rate": 5e-05,
-      "loss": 0.4749,
+      "loss": 1.7321,
       "step": 3470
     },
     {
       "epoch": 0.056538480284642006,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4813,
+      "loss": 1.8251,
       "step": 3480
     },
     {
       "epoch": 0.056700947182011664,
-      "grad_norm": 3.0,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4793,
+      "loss": 1.765,
       "step": 3490
     },
     {
       "epoch": 0.05686341407938133,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4677,
+      "loss": 1.7775,
       "step": 3500
     },
     {
       "epoch": 0.05702588097675099,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4817,
+      "loss": 1.7914,
       "step": 3510
     },
     {
       "epoch": 0.05718834787412065,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.5034,
+      "loss": 1.7471,
       "step": 3520
     },
     {
       "epoch": 0.05735081477149031,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4948,
+      "loss": 1.7496,
       "step": 3530
     },
     {
       "epoch": 0.05751328166885997,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4722,
+      "loss": 1.8152,
       "step": 3540
     },
     {
       "epoch": 0.05767574856622963,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4888,
+      "loss": 1.7157,
       "step": 3550
     },
     {
       "epoch": 0.05783821546359929,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4681,
+      "loss": 1.7592,
       "step": 3560
     },
     {
       "epoch": 0.058000682360968954,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.516,
+      "loss": 1.7713,
       "step": 3570
     },
     {
       "epoch": 0.05816314925833861,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.469,
+      "loss": 1.731,
       "step": 3580
     },
     {
       "epoch": 0.058325616155708276,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4875,
+      "loss": 1.7449,
       "step": 3590
     },
     {
       "epoch": 0.058488083053077934,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4961,
+      "loss": 1.7929,
       "step": 3600
     },
     {
       "epoch": 0.0586505499504476,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4919,
+      "loss": 1.7485,
       "step": 3610
     },
     {
       "epoch": 0.05881301684781726,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4837,
+      "loss": 1.7617,
       "step": 3620
     },
     {
       "epoch": 0.05897548374518692,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4791,
+      "loss": 1.7604,
       "step": 3630
     },
     {
       "epoch": 0.05913795064255658,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4961,
+      "loss": 1.7353,
       "step": 3640
     },
     {
       "epoch": 0.05930041753992624,
-      "grad_norm": 2.5,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.5136,
+      "loss": 1.7612,
       "step": 3650
     },
     {
       "epoch": 0.0594628844372959,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4638,
+      "loss": 1.7614,
       "step": 3660
     },
     {
       "epoch": 0.05962535133466556,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.5141,
+      "loss": 1.789,
       "step": 3670
     },
     {
       "epoch": 0.059787818232035224,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.5075,
+      "loss": 1.7414,
       "step": 3680
     },
     {
       "epoch": 0.05995028512940488,
-      "grad_norm": 2.375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4808,
+      "loss": 1.74,
       "step": 3690
     },
     {
       "epoch": 0.060112752026774546,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4629,
+      "loss": 1.7172,
       "step": 3700
     },
     {
       "epoch": 0.060275218924144204,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4952,
+      "loss": 1.7425,
       "step": 3710
     },
     {
       "epoch": 0.06043768582151387,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4879,
+      "loss": 1.7717,
       "step": 3720
     },
     {
       "epoch": 0.060600152718883527,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4918,
+      "loss": 1.7767,
       "step": 3730
     },
     {
       "epoch": 0.06076261961625319,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4737,
+      "loss": 1.7432,
       "step": 3740
     },
     {
       "epoch": 0.06092508651362285,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4811,
+      "loss": 1.7933,
       "step": 3750
     },
     {
       "epoch": 0.061087553410992514,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4905,
+      "loss": 1.7348,
       "step": 3760
     },
     {
       "epoch": 0.06125002030836217,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.493,
+      "loss": 1.7278,
       "step": 3770
     },
     {
       "epoch": 0.06141248720573183,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4766,
+      "loss": 1.7339,
       "step": 3780
     },
     {
       "epoch": 0.061574954103101494,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.5237,
+      "loss": 1.7996,
       "step": 3790
     },
     {
       "epoch": 0.06173742100047115,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.491,
+      "loss": 1.7264,
       "step": 3800
     },
     {
       "epoch": 0.061899887897840816,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4988,
+      "loss": 1.7524,
       "step": 3810
     },
     {
       "epoch": 0.062062354795210474,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4985,
+      "loss": 1.7165,
       "step": 3820
     },
     {
       "epoch": 0.06222482169258014,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5009,
+      "loss": 1.765,
       "step": 3830
     },
     {
       "epoch": 0.062387288589949796,
-      "grad_norm": 4.5,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4895,
+      "loss": 1.7042,
       "step": 3840
     },
     {
       "epoch": 0.06254975548731946,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4585,
+      "loss": 1.6955,
       "step": 3850
     },
     {
       "epoch": 0.06271222238468913,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4894,
+      "loss": 1.7294,
       "step": 3860
     },
     {
       "epoch": 0.06287468928205878,
-      "grad_norm": 2.3125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4677,
+      "loss": 1.7702,
       "step": 3870
     },
     {
       "epoch": 0.06303715617942844,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4934,
+      "loss": 1.7128,
       "step": 3880
     },
     {
       "epoch": 0.0631996230767981,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.5007,
+      "loss": 1.7306,
       "step": 3890
     },
     {
       "epoch": 0.06336208997416776,
-      "grad_norm": 3.921875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.4863,
+      "loss": 1.7252,
       "step": 3900
     },
     {
       "epoch": 0.06352455687153742,
-      "grad_norm": 2.296875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4919,
+      "loss": 1.7412,
       "step": 3910
     },
     {
       "epoch": 0.06368702376890709,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4865,
+      "loss": 1.7243,
       "step": 3920
     },
     {
       "epoch": 0.06384949066627675,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4964,
+      "loss": 1.7559,
       "step": 3930
     },
     {
       "epoch": 0.0640119575636464,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4826,
+      "loss": 1.6993,
       "step": 3940
     },
     {
       "epoch": 0.06417442446101607,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4945,
+      "loss": 1.6901,
       "step": 3950
     },
     {
       "epoch": 0.06433689135838573,
-      "grad_norm": 2.625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4763,
+      "loss": 1.7384,
       "step": 3960
     },
     {
       "epoch": 0.0644993582557554,
-      "grad_norm": 2.5,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4829,
+      "loss": 1.7489,
       "step": 3970
     },
     {
       "epoch": 0.06466182515312505,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4685,
+      "loss": 1.7043,
       "step": 3980
     },
     {
       "epoch": 0.06482429205049471,
-      "grad_norm": 3.0,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4745,
+      "loss": 1.7535,
       "step": 3990
     },
     {
       "epoch": 0.06498675894786438,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4486,
+      "loss": 1.7467,
       "step": 4000
     },
     {
       "epoch": 0.06514922584523403,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4891,
+      "loss": 1.671,
       "step": 4010
     },
     {
       "epoch": 0.06531169274260369,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.458,
+      "loss": 1.7374,
       "step": 4020
     },
     {
       "epoch": 0.06547415963997336,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4824,
+      "loss": 1.7623,
       "step": 4030
     },
     {
       "epoch": 0.06563662653734302,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4745,
+      "loss": 1.7513,
       "step": 4040
     },
     {
       "epoch": 0.06579909343471267,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.4755,
+      "loss": 1.7179,
       "step": 4050
     },
     {
       "epoch": 0.06596156033208234,
-      "grad_norm": 3.640625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5001,
+      "loss": 1.7864,
       "step": 4060
     },
     {
       "epoch": 0.066124027229452,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4829,
+      "loss": 1.7167,
       "step": 4070
     },
     {
       "epoch": 0.06628649412682167,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.5134,
+      "loss": 1.7385,
       "step": 4080
     },
     {
       "epoch": 0.06644896102419132,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.46,
+      "loss": 1.7749,
       "step": 4090
     },
     {
       "epoch": 0.06661142792156098,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4961,
+      "loss": 1.7072,
       "step": 4100
     },
     {
       "epoch": 0.06677389481893065,
-      "grad_norm": 2.546875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4592,
+      "loss": 1.6869,
       "step": 4110
     },
     {
       "epoch": 0.06693636171630031,
-      "grad_norm": 3.265625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4749,
+      "loss": 1.7569,
       "step": 4120
     },
     {
       "epoch": 0.06709882861366996,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4637,
+      "loss": 1.7291,
       "step": 4130
     },
     {
       "epoch": 0.06726129551103963,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.5083,
+      "loss": 1.752,
       "step": 4140
     },
     {
       "epoch": 0.06742376240840929,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4828,
+      "loss": 1.7284,
       "step": 4150
     },
     {
       "epoch": 0.06758622930577894,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4717,
+      "loss": 1.7071,
       "step": 4160
     },
     {
       "epoch": 0.0677486962031486,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4883,
+      "loss": 1.7479,
       "step": 4170
     },
     {
       "epoch": 0.06791116310051827,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4957,
+      "loss": 1.7407,
       "step": 4180
     },
     {
       "epoch": 0.06807362999788794,
-      "grad_norm": 2.375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4974,
+      "loss": 1.6655,
       "step": 4190
     },
     {
       "epoch": 0.06823609689525759,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4996,
+      "loss": 1.6884,
       "step": 4200
     },
     {
       "epoch": 0.06839856379262725,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.5128,
+      "loss": 1.6985,
       "step": 4210
     },
     {
       "epoch": 0.06856103068999692,
-      "grad_norm": 4.46875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4725,
+      "loss": 1.7051,
       "step": 4220
     },
     {
       "epoch": 0.06872349758736658,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4706,
+      "loss": 1.7475,
       "step": 4230
     },
     {
       "epoch": 0.06888596448473623,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4802,
+      "loss": 1.7579,
       "step": 4240
     },
     {
       "epoch": 0.0690484313821059,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.472,
+      "loss": 1.7052,
       "step": 4250
     },
     {
       "epoch": 0.06921089827947556,
-      "grad_norm": 2.625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4932,
+      "loss": 1.7756,
       "step": 4260
     },
     {
       "epoch": 0.06937336517684521,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4809,
+      "loss": 1.6722,
       "step": 4270
     },
     {
       "epoch": 0.06953583207421488,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4997,
+      "loss": 1.723,
       "step": 4280
     },
     {
       "epoch": 0.06969829897158454,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.475,
+      "loss": 1.7427,
       "step": 4290
     },
     {
       "epoch": 0.0698607658689542,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4668,
+      "loss": 1.7217,
       "step": 4300
     },
     {
       "epoch": 0.07002323276632386,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4662,
+      "loss": 1.7298,
       "step": 4310
     },
     {
       "epoch": 0.07018569966369352,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4755,
+      "loss": 1.7156,
       "step": 4320
     },
     {
       "epoch": 0.07034816656106319,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.5064,
+      "loss": 1.7343,
       "step": 4330
     },
     {
       "epoch": 0.07051063345843285,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.512,
+      "loss": 1.7113,
       "step": 4340
     },
     {
       "epoch": 0.0706731003558025,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4711,
+      "loss": 1.7286,
       "step": 4350
     },
     {
       "epoch": 0.07083556725317217,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4752,
+      "loss": 1.7242,
       "step": 4360
     },
     {
       "epoch": 0.07099803415054183,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4862,
+      "loss": 1.6946,
       "step": 4370
     },
     {
       "epoch": 0.07116050104791148,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.5003,
+      "loss": 1.7219,
       "step": 4380
     },
     {
       "epoch": 0.07132296794528115,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4698,
+      "loss": 1.7203,
       "step": 4390
     },
     {
       "epoch": 0.07148543484265081,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4826,
+      "loss": 1.7343,
       "step": 4400
     },
     {
       "epoch": 0.07164790174002048,
-      "grad_norm": 2.625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4643,
+      "loss": 1.7308,
       "step": 4410
     },
     {
       "epoch": 0.07181036863739013,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.464,
+      "loss": 1.7486,
       "step": 4420
     },
     {
       "epoch": 0.07197283553475979,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4521,
+      "loss": 1.7168,
       "step": 4430
     },
     {
       "epoch": 0.07213530243212946,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4392,
+      "loss": 1.6936,
       "step": 4440
     },
     {
       "epoch": 0.07229776932949912,
-      "grad_norm": 3.03125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.495,
+      "loss": 1.706,
       "step": 4450
     },
     {
       "epoch": 0.07246023622686877,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.5058,
+      "loss": 1.6701,
       "step": 4460
     },
     {
       "epoch": 0.07262270312423844,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4853,
+      "loss": 1.698,
       "step": 4470
     },
     {
       "epoch": 0.0727851700216081,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4648,
+      "loss": 1.7345,
       "step": 4480
     },
     {
       "epoch": 0.07294763691897777,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4701,
+      "loss": 1.7381,
       "step": 4490
     },
     {
       "epoch": 0.07311010381634742,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4763,
+      "loss": 1.7132,
       "step": 4500
     },
     {
       "epoch": 0.07327257071371708,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4923,
+      "loss": 1.6657,
       "step": 4510
     },
     {
       "epoch": 0.07343503761108675,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4918,
+      "loss": 1.691,
       "step": 4520
     },
     {
       "epoch": 0.0735975045084564,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4754,
+      "loss": 1.6947,
       "step": 4530
     },
     {
       "epoch": 0.07375997140582606,
-      "grad_norm": 4.0,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4866,
+      "loss": 1.674,
       "step": 4540
     },
     {
       "epoch": 0.07392243830319573,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4624,
+      "loss": 1.7421,
       "step": 4550
     },
     {
       "epoch": 0.07408490520056539,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4674,
+      "loss": 1.6998,
       "step": 4560
     },
     {
       "epoch": 0.07424737209793504,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4913,
+      "loss": 1.6972,
       "step": 4570
     },
     {
       "epoch": 0.0744098389953047,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4716,
+      "loss": 1.7226,
       "step": 4580
     },
     {
       "epoch": 0.07457230589267437,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4645,
+      "loss": 1.7648,
       "step": 4590
     },
     {
       "epoch": 0.07473477279004404,
-      "grad_norm": 3.5,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4579,
+      "loss": 1.734,
       "step": 4600
     },
     {
       "epoch": 0.07489723968741369,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4937,
+      "loss": 1.7609,
       "step": 4610
     },
     {
       "epoch": 0.07505970658478335,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4753,
+      "loss": 1.6962,
       "step": 4620
     },
     {
       "epoch": 0.07522217348215302,
-      "grad_norm": 3.484375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4724,
+      "loss": 1.7308,
       "step": 4630
     },
     {
       "epoch": 0.07538464037952267,
-      "grad_norm": 2.5,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4636,
+      "loss": 1.716,
       "step": 4640
     },
     {
       "epoch": 0.07554710727689233,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4739,
+      "loss": 1.7319,
       "step": 4650
     },
     {
       "epoch": 0.075709574174262,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4932,
+      "loss": 1.6636,
       "step": 4660
     },
     {
       "epoch": 0.07587204107163166,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4904,
+      "loss": 1.6922,
       "step": 4670
     },
     {
       "epoch": 0.07603450796900131,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4701,
+      "loss": 1.6471,
       "step": 4680
     },
     {
       "epoch": 0.07619697486637098,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4896,
+      "loss": 1.7414,
       "step": 4690
     },
     {
       "epoch": 0.07635944176374064,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4789,
+      "loss": 1.6967,
       "step": 4700
     },
     {
       "epoch": 0.0765219086611103,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4605,
+      "loss": 1.7051,
       "step": 4710
     },
     {
       "epoch": 0.07668437555847996,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4687,
+      "loss": 1.643,
       "step": 4720
     },
     {
       "epoch": 0.07684684245584962,
-      "grad_norm": 4.78125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4726,
+      "loss": 1.6833,
       "step": 4730
     },
     {
       "epoch": 0.07700930935321929,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4582,
+      "loss": 1.6724,
       "step": 4740
     },
     {
       "epoch": 0.07717177625058894,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4733,
+      "loss": 1.751,
       "step": 4750
     },
     {
       "epoch": 0.0773342431479586,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4788,
+      "loss": 1.6551,
       "step": 4760
     },
     {
       "epoch": 0.07749671004532827,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4731,
+      "loss": 1.6991,
       "step": 4770
     },
     {
       "epoch": 0.07765917694269793,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.4755,
+      "loss": 1.7378,
       "step": 4780
     },
     {
       "epoch": 0.07782164384006758,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4684,
+      "loss": 1.7074,
       "step": 4790
     },
     {
       "epoch": 0.07798411073743725,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4472,
+      "loss": 1.6778,
       "step": 4800
     },
     {
       "epoch": 0.07814657763480691,
-      "grad_norm": 1.8984375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4662,
+      "loss": 1.7151,
       "step": 4810
     },
     {
       "epoch": 0.07830904453217657,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.5012,
+      "loss": 1.7688,
       "step": 4820
     },
     {
       "epoch": 0.07847151142954623,
-      "grad_norm": 4.125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4602,
+      "loss": 1.6581,
       "step": 4830
     },
     {
       "epoch": 0.07863397832691589,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.4904,
+      "loss": 1.6661,
       "step": 4840
     },
     {
       "epoch": 0.07879644522428556,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4793,
+      "loss": 1.6261,
       "step": 4850
     },
     {
       "epoch": 0.0789589121216552,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4775,
+      "loss": 1.7372,
       "step": 4860
     },
     {
       "epoch": 0.07912137901902487,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4998,
+      "loss": 1.7341,
       "step": 4870
     },
     {
       "epoch": 0.07928384591639454,
-      "grad_norm": 3.4375,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.4642,
+      "loss": 1.6765,
       "step": 4880
     },
     {
       "epoch": 0.0794463128137642,
-      "grad_norm": 2.875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4954,
+      "loss": 1.7103,
       "step": 4890
     },
     {
       "epoch": 0.07960877971113385,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4672,
+      "loss": 1.6938,
       "step": 4900
     },
     {
       "epoch": 0.07977124660850352,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4805,
+      "loss": 1.7391,
       "step": 4910
     },
     {
       "epoch": 0.07993371350587318,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4986,
+      "loss": 1.6558,
       "step": 4920
     },
     {
       "epoch": 0.08009618040324284,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4932,
+      "loss": 1.693,
       "step": 4930
     },
     {
       "epoch": 0.0802586473006125,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4576,
+      "loss": 1.6365,
       "step": 4940
     },
     {
       "epoch": 0.08042111419798216,
-      "grad_norm": 2.75,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4681,
+      "loss": 1.7376,
       "step": 4950
     },
     {
       "epoch": 0.08058358109535183,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4944,
+      "loss": 1.6967,
       "step": 4960
     },
     {
       "epoch": 0.08074604799272149,
-      "grad_norm": 11.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4821,
+      "loss": 1.6898,
       "step": 4970
     },
     {
       "epoch": 0.08090851489009114,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4545,
+      "loss": 1.6645,
       "step": 4980
     },
     {
       "epoch": 0.0810709817874608,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4443,
+      "loss": 1.6844,
       "step": 4990
     },
     {
       "epoch": 0.08123344868483047,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4664,
+      "loss": 1.7255,
       "step": 5000
     },
     {
       "epoch": 0.08139591558220012,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4532,
+      "loss": 1.7216,
       "step": 5010
     },
     {
       "epoch": 0.08155838247956979,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4554,
+      "loss": 1.6713,
       "step": 5020
     },
     {
       "epoch": 0.08172084937693945,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4712,
+      "loss": 1.7102,
       "step": 5030
     },
     {
       "epoch": 0.08188331627430911,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4719,
+      "loss": 1.7929,
       "step": 5040
     },
     {
       "epoch": 0.08204578317167877,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4756,
+      "loss": 1.6801,
       "step": 5050
     },
     {
       "epoch": 0.08220825006904843,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4772,
+      "loss": 1.6788,
       "step": 5060
     },
     {
       "epoch": 0.0823707169664181,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4592,
+      "loss": 1.6463,
       "step": 5070
     },
     {
       "epoch": 0.08253318386378776,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4841,
+      "loss": 1.7143,
       "step": 5080
     },
     {
       "epoch": 0.08269565076115741,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4748,
+      "loss": 1.6636,
       "step": 5090
     },
     {
       "epoch": 0.08285811765852708,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4638,
+      "loss": 1.664,
       "step": 5100
     },
     {
       "epoch": 0.08302058455589674,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4874,
+      "loss": 1.6763,
       "step": 5110
     },
     {
       "epoch": 0.08318305145326639,
-      "grad_norm": 3.921875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4794,
+      "loss": 1.6872,
       "step": 5120
     },
     {
       "epoch": 0.08334551835063606,
-      "grad_norm": 2.875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.472,
+      "loss": 1.6827,
       "step": 5130
     },
     {
       "epoch": 0.08350798524800572,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.488,
+      "loss": 1.76,
       "step": 5140
     },
     {
       "epoch": 0.08367045214537538,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4729,
+      "loss": 1.7283,
       "step": 5150
     },
     {
       "epoch": 0.08383291904274504,
-      "grad_norm": 2.921875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4693,
+      "loss": 1.7079,
       "step": 5160
     },
     {
       "epoch": 0.0839953859401147,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4706,
+      "loss": 1.7071,
       "step": 5170
     },
     {
       "epoch": 0.08415785283748436,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4816,
+      "loss": 1.6826,
       "step": 5180
     },
     {
       "epoch": 0.08432031973485403,
-      "grad_norm": 2.875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4471,
+      "loss": 1.7167,
       "step": 5190
     },
     {
       "epoch": 0.08448278663222368,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4552,
+      "loss": 1.7117,
       "step": 5200
     },
     {
       "epoch": 0.08464525352959335,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4801,
+      "loss": 1.7016,
       "step": 5210
     },
     {
       "epoch": 0.08480772042696301,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4781,
+      "loss": 1.7176,
       "step": 5220
     },
     {
       "epoch": 0.08497018732433266,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4915,
+      "loss": 1.69,
       "step": 5230
     },
     {
       "epoch": 0.08513265422170233,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4776,
+      "loss": 1.6877,
       "step": 5240
     },
     {
       "epoch": 0.08529512111907199,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4833,
+      "loss": 1.6295,
       "step": 5250
     },
     {
       "epoch": 0.08545758801644165,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4692,
+      "loss": 1.6136,
       "step": 5260
     },
     {
       "epoch": 0.0856200549138113,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4764,
+      "loss": 1.6583,
       "step": 5270
     },
     {
       "epoch": 0.08578252181118097,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4628,
+      "loss": 1.6506,
       "step": 5280
     },
     {
       "epoch": 0.08594498870855063,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4571,
+      "loss": 1.6355,
       "step": 5290
     },
     {
       "epoch": 0.0861074556059203,
-      "grad_norm": 2.3125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.4704,
+      "loss": 1.6329,
       "step": 5300
     },
     {
       "epoch": 0.08626992250328995,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4621,
+      "loss": 1.6814,
       "step": 5310
     },
     {
       "epoch": 0.08643238940065961,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4695,
+      "loss": 1.6988,
       "step": 5320
     },
     {
       "epoch": 0.08659485629802928,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4871,
+      "loss": 1.6607,
       "step": 5330
     },
     {
       "epoch": 0.08675732319539894,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4728,
+      "loss": 1.6993,
       "step": 5340
     },
     {
       "epoch": 0.0869197900927686,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4749,
+      "loss": 1.6847,
       "step": 5350
     },
     {
       "epoch": 0.08708225699013826,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4846,
+      "loss": 1.7041,
       "step": 5360
     },
     {
       "epoch": 0.08724472388750792,
-      "grad_norm": 2.921875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.4771,
+      "loss": 1.711,
       "step": 5370
     },
     {
       "epoch": 0.08740719078487758,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4801,
+      "loss": 1.7096,
       "step": 5380
     },
     {
       "epoch": 0.08756965768224724,
-      "grad_norm": 2.375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4587,
+      "loss": 1.647,
       "step": 5390
     },
     {
       "epoch": 0.0877321245796169,
-      "grad_norm": 2.640625,
+      "grad_norm": 12.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4607,
+      "loss": 1.7082,
       "step": 5400
     },
     {
       "epoch": 0.08789459147698657,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4828,
+      "loss": 1.6458,
       "step": 5410
     },
     {
       "epoch": 0.08805705837435622,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4671,
+      "loss": 1.6996,
       "step": 5420
     },
     {
       "epoch": 0.08821952527172588,
-      "grad_norm": 3.5,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.49,
+      "loss": 1.6664,
       "step": 5430
     },
     {
       "epoch": 0.08838199216909555,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4544,
+      "loss": 1.6938,
       "step": 5440
     },
     {
       "epoch": 0.08854445906646521,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4858,
+      "loss": 1.7531,
       "step": 5450
     },
     {
       "epoch": 0.08870692596383487,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4871,
+      "loss": 1.6965,
       "step": 5460
     },
     {
       "epoch": 0.08886939286120453,
-      "grad_norm": 2.875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4794,
+      "loss": 1.6955,
       "step": 5470
     },
     {
       "epoch": 0.0890318597585742,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4639,
+      "loss": 1.7175,
       "step": 5480
     },
     {
       "epoch": 0.08919432665594385,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4817,
+      "loss": 1.6401,
       "step": 5490
     },
     {
       "epoch": 0.08935679355331351,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.5099,
+      "loss": 1.6833,
       "step": 5500
     },
     {
       "epoch": 0.08951926045068317,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4854,
+      "loss": 1.6775,
       "step": 5510
     },
     {
       "epoch": 0.08968172734805284,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4668,
+      "loss": 1.6871,
       "step": 5520
     },
     {
       "epoch": 0.08984419424542249,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.448,
+      "loss": 1.6723,
       "step": 5530
     },
     {
       "epoch": 0.09000666114279215,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4714,
+      "loss": 1.6767,
       "step": 5540
     },
     {
       "epoch": 0.09016912804016182,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4691,
+      "loss": 1.6479,
       "step": 5550
     },
     {
       "epoch": 0.09033159493753148,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4689,
+      "loss": 1.635,
       "step": 5560
     },
     {
       "epoch": 0.09049406183490113,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4593,
+      "loss": 1.6953,
       "step": 5570
     },
     {
       "epoch": 0.0906565287322708,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4705,
+      "loss": 1.6522,
       "step": 5580
     },
     {
       "epoch": 0.09081899562964046,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4736,
+      "loss": 1.6489,
       "step": 5590
     },
     {
       "epoch": 0.09098146252701012,
-      "grad_norm": 3.703125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.4632,
+      "loss": 1.6811,
       "step": 5600
     },
     {
       "epoch": 0.09114392942437978,
-      "grad_norm": 3.5,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4604,
+      "loss": 1.6664,
       "step": 5610
     },
     {
       "epoch": 0.09130639632174944,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4714,
+      "loss": 1.6505,
       "step": 5620
     },
     {
       "epoch": 0.09146886321911911,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4614,
+      "loss": 1.6651,
       "step": 5630
     },
     {
       "epoch": 0.09163133011648876,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4754,
+      "loss": 1.6132,
       "step": 5640
     },
     {
       "epoch": 0.09179379701385842,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4688,
+      "loss": 1.6922,
       "step": 5650
     },
     {
       "epoch": 0.09195626391122809,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4993,
+      "loss": 1.7268,
       "step": 5660
     },
     {
       "epoch": 0.09211873080859775,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4571,
+      "loss": 1.6479,
       "step": 5670
     },
     {
       "epoch": 0.0922811977059674,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4575,
+      "loss": 1.7005,
       "step": 5680
     },
     {
       "epoch": 0.09244366460333707,
-      "grad_norm": 2.5,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4654,
+      "loss": 1.6803,
       "step": 5690
     },
     {
       "epoch": 0.09260613150070673,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4559,
+      "loss": 1.6432,
       "step": 5700
     },
     {
       "epoch": 0.09276859839807639,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4653,
+      "loss": 1.6572,
       "step": 5710
     },
     {
       "epoch": 0.09293106529544605,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4494,
+      "loss": 1.703,
       "step": 5720
     },
     {
       "epoch": 0.09309353219281571,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4629,
+      "loss": 1.7156,
       "step": 5730
     },
     {
       "epoch": 0.09325599909018538,
-      "grad_norm": 2.25,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4741,
+      "loss": 1.679,
       "step": 5740
     },
     {
       "epoch": 0.09341846598755503,
-      "grad_norm": 2.8125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.4804,
+      "loss": 1.6309,
       "step": 5750
     },
     {
       "epoch": 0.0935809328849247,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.4607,
+      "loss": 1.6312,
       "step": 5760
     },
     {
       "epoch": 0.09374339978229436,
-      "grad_norm": 2.078125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4701,
+      "loss": 1.6522,
       "step": 5770
     },
     {
       "epoch": 0.09390586667966402,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.469,
+      "loss": 1.6437,
       "step": 5780
     },
     {
       "epoch": 0.09406833357703367,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.463,
+      "loss": 1.6619,
       "step": 5790
     },
     {
       "epoch": 0.09423080047440334,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.465,
+      "loss": 1.6559,
       "step": 5800
     },
     {
       "epoch": 0.094393267371773,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4607,
+      "loss": 1.6617,
       "step": 5810
     },
     {
       "epoch": 0.09455573426914267,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4791,
+      "loss": 1.6538,
       "step": 5820
     },
     {
       "epoch": 0.09471820116651232,
-      "grad_norm": 2.140625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.4633,
+      "loss": 1.6383,
       "step": 5830
     },
     {
       "epoch": 0.09488066806388198,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4867,
+      "loss": 1.645,
       "step": 5840
     },
     {
       "epoch": 0.09504313496125165,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4785,
+      "loss": 1.6432,
       "step": 5850
     },
     {
       "epoch": 0.0952056018586213,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.478,
+      "loss": 1.6259,
       "step": 5860
     },
     {
       "epoch": 0.09536806875599096,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4617,
+      "loss": 1.6839,
       "step": 5870
     },
     {
       "epoch": 0.09553053565336063,
-      "grad_norm": 2.75,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4467,
+      "loss": 1.6937,
       "step": 5880
     },
     {
       "epoch": 0.0956930025507303,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4742,
+      "loss": 1.7218,
       "step": 5890
     },
     {
       "epoch": 0.09585546944809994,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.7718,
       "step": 5900
     },
     {
       "epoch": 0.09601793634546961,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4837,
+      "loss": 1.64,
       "step": 5910
     },
     {
       "epoch": 0.09618040324283927,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4988,
+      "loss": 1.7288,
       "step": 5920
     },
     {
       "epoch": 0.09634287014020894,
-      "grad_norm": 3.125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4718,
+      "loss": 1.6864,
       "step": 5930
     },
     {
       "epoch": 0.09650533703757859,
-      "grad_norm": 5.0625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4596,
+      "loss": 1.6519,
       "step": 5940
     },
     {
       "epoch": 0.09666780393494825,
-      "grad_norm": 1.859375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4923,
+      "loss": 1.7087,
       "step": 5950
     },
     {
       "epoch": 0.09683027083231792,
-      "grad_norm": 2.375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4367,
+      "loss": 1.6692,
       "step": 5960
     },
     {
       "epoch": 0.09699273772968757,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4412,
+      "loss": 1.6315,
       "step": 5970
     },
     {
       "epoch": 0.09715520462705723,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4782,
+      "loss": 1.6439,
       "step": 5980
     },
     {
       "epoch": 0.0973176715244269,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4596,
+      "loss": 1.5991,
       "step": 5990
     },
     {
       "epoch": 0.09748013842179656,
-      "grad_norm": 4.03125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4845,
+      "loss": 1.6364,
       "step": 6000
     },
     {
       "epoch": 0.09764260531916621,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4824,
+      "loss": 1.6908,
       "step": 6010
     },
     {
       "epoch": 0.09780507221653588,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.433,
+      "loss": 1.6644,
       "step": 6020
     },
     {
       "epoch": 0.09796753911390554,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4849,
+      "loss": 1.6636,
       "step": 6030
     },
     {
       "epoch": 0.09813000601127521,
-      "grad_norm": 3.34375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4525,
+      "loss": 1.667,
       "step": 6040
     },
     {
       "epoch": 0.09829247290864486,
-      "grad_norm": 3.78125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4748,
+      "loss": 1.6443,
       "step": 6050
     },
     {
       "epoch": 0.09845493980601452,
-      "grad_norm": 2.875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4651,
+      "loss": 1.6882,
       "step": 6060
     },
     {
       "epoch": 0.09861740670338419,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4614,
+      "loss": 1.6558,
       "step": 6070
     },
     {
       "epoch": 0.09877987360075384,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4661,
+      "loss": 1.6661,
       "step": 6080
     },
     {
       "epoch": 0.0989423404981235,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4519,
+      "loss": 1.694,
       "step": 6090
     },
     {
       "epoch": 0.09910480739549317,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.468,
+      "loss": 1.675,
       "step": 6100
     },
     {
       "epoch": 0.09926727429286283,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4528,
+      "loss": 1.5784,
       "step": 6110
     },
     {
       "epoch": 0.09942974119023248,
-      "grad_norm": 3.25,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.474,
+      "loss": 1.6885,
       "step": 6120
     },
     {
       "epoch": 0.09959220808760215,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4757,
+      "loss": 1.6745,
       "step": 6130
     },
     {
       "epoch": 0.09975467498497181,
-      "grad_norm": 2.625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4502,
+      "loss": 1.601,
       "step": 6140
     },
     {
       "epoch": 0.09991714188234148,
-      "grad_norm": 3.75,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4869,
+      "loss": 1.6743,
       "step": 6150
     },
     {
       "epoch": 0.10007960877971113,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4667,
+      "loss": 1.668,
       "step": 6160
     },
     {
       "epoch": 0.1002420756770808,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4695,
+      "loss": 1.6609,
       "step": 6170
     },
     {
       "epoch": 0.10040454257445046,
-      "grad_norm": 2.75,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4564,
+      "loss": 1.6374,
       "step": 6180
     },
     {
       "epoch": 0.10056700947182012,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.459,
+      "loss": 1.6634,
       "step": 6190
     },
     {
       "epoch": 0.10072947636918977,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4418,
+      "loss": 1.6548,
       "step": 6200
     },
     {
       "epoch": 0.10089194326655944,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4708,
+      "loss": 1.6496,
       "step": 6210
     },
     {
       "epoch": 0.1010544101639291,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.4742,
+      "loss": 1.6276,
       "step": 6220
     },
     {
       "epoch": 0.10121687706129875,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4609,
+      "loss": 1.6048,
       "step": 6230
     },
     {
       "epoch": 0.10137934395866842,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4668,
+      "loss": 1.6315,
       "step": 6240
     },
     {
       "epoch": 0.10154181085603808,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4647,
+      "loss": 1.6367,
       "step": 6250
     },
     {
       "epoch": 0.10170427775340775,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4548,
+      "loss": 1.6422,
       "step": 6260
     },
     {
       "epoch": 0.1018667446507774,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4433,
+      "loss": 1.653,
       "step": 6270
     },
     {
       "epoch": 0.10202921154814706,
-      "grad_norm": 7.25,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.6703,
       "step": 6280
     },
     {
       "epoch": 0.10219167844551673,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4498,
+      "loss": 1.6573,
       "step": 6290
     },
     {
       "epoch": 0.1023541453428864,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4228,
+      "loss": 1.6545,
       "step": 6300
     },
     {
       "epoch": 0.10251661224025604,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.473,
+      "loss": 1.6078,
       "step": 6310
     },
     {
       "epoch": 0.10267907913762571,
-      "grad_norm": 2.75,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4663,
+      "loss": 1.6859,
       "step": 6320
     },
     {
       "epoch": 0.10284154603499537,
-      "grad_norm": 5.53125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4604,
+      "loss": 1.6812,
       "step": 6330
     },
     {
       "epoch": 0.10300401293236502,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4547,
+      "loss": 1.6223,
       "step": 6340
     },
     {
       "epoch": 0.10316647982973469,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4367,
+      "loss": 1.6682,
       "step": 6350
     },
     {
       "epoch": 0.10332894672710435,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4659,
+      "loss": 1.6749,
       "step": 6360
     },
     {
       "epoch": 0.10349141362447402,
-      "grad_norm": 2.125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4823,
+      "loss": 1.6498,
       "step": 6370
     },
     {
       "epoch": 0.10365388052184367,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4648,
+      "loss": 1.6153,
       "step": 6380
     },
     {
       "epoch": 0.10381634741921333,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4655,
+      "loss": 1.7107,
       "step": 6390
     },
     {
       "epoch": 0.103978814316583,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4435,
+      "loss": 1.6532,
       "step": 6400
     },
     {
       "epoch": 0.10414128121395266,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4658,
+      "loss": 1.6814,
       "step": 6410
     },
     {
       "epoch": 0.10430374811132231,
-      "grad_norm": 2.375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4685,
+      "loss": 1.6432,
       "step": 6420
     },
     {
       "epoch": 0.10446621500869198,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.47,
+      "loss": 1.676,
       "step": 6430
     },
     {
       "epoch": 0.10462868190606164,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4695,
+      "loss": 1.6776,
       "step": 6440
     },
     {
       "epoch": 0.1047911488034313,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4699,
+      "loss": 1.6491,
       "step": 6450
     },
     {
       "epoch": 0.10495361570080096,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4615,
+      "loss": 1.6353,
       "step": 6460
     },
     {
       "epoch": 0.10511608259817062,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4645,
+      "loss": 1.6654,
       "step": 6470
     },
     {
       "epoch": 0.10527854949554029,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.6889,
       "step": 6480
     },
     {
       "epoch": 0.10544101639290994,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4616,
+      "loss": 1.6688,
       "step": 6490
     },
     {
       "epoch": 0.1056034832902796,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4529,
+      "loss": 1.6294,
       "step": 6500
     },
     {
       "epoch": 0.10576595018764927,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4558,
+      "loss": 1.6195,
       "step": 6510
     },
     {
       "epoch": 0.10592841708501893,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4844,
+      "loss": 1.6101,
       "step": 6520
     },
     {
       "epoch": 0.10609088398238858,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4496,
+      "loss": 1.658,
       "step": 6530
     },
     {
       "epoch": 0.10625335087975825,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4475,
+      "loss": 1.5964,
       "step": 6540
     },
     {
       "epoch": 0.10641581777712791,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4464,
+      "loss": 1.6305,
       "step": 6550
     },
     {
       "epoch": 0.10657828467449756,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4819,
+      "loss": 1.6119,
       "step": 6560
     },
     {
       "epoch": 0.10674075157186723,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4743,
+      "loss": 1.6456,
       "step": 6570
     },
     {
       "epoch": 0.1069032184692369,
-      "grad_norm": 3.796875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4643,
+      "loss": 1.6666,
       "step": 6580
     },
     {
       "epoch": 0.10706568536660656,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4481,
+      "loss": 1.6322,
       "step": 6590
     },
     {
       "epoch": 0.10722815226397621,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4462,
+      "loss": 1.6621,
       "step": 6600
     },
     {
       "epoch": 0.10739061916134587,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4505,
+      "loss": 1.6101,
       "step": 6610
     },
     {
       "epoch": 0.10755308605871554,
-      "grad_norm": 5.0625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.447,
+      "loss": 1.7102,
       "step": 6620
     },
     {
       "epoch": 0.1077155529560852,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4621,
+      "loss": 1.6233,
       "step": 6630
     },
     {
       "epoch": 0.10787801985345485,
-      "grad_norm": 2.75,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4662,
+      "loss": 1.6387,
       "step": 6640
     },
     {
       "epoch": 0.10804048675082452,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4381,
+      "loss": 1.6307,
       "step": 6650
     },
     {
       "epoch": 0.10820295364819418,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4851,
+      "loss": 1.6427,
       "step": 6660
     },
     {
       "epoch": 0.10836542054556385,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4515,
+      "loss": 1.6534,
       "step": 6670
     },
     {
       "epoch": 0.1085278874429335,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4629,
+      "loss": 1.6039,
       "step": 6680
     },
     {
       "epoch": 0.10869035434030316,
-      "grad_norm": 2.5,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4602,
+      "loss": 1.6367,
       "step": 6690
     },
     {
       "epoch": 0.10885282123767283,
-      "grad_norm": 3.453125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4524,
+      "loss": 1.6193,
       "step": 6700
     },
     {
       "epoch": 0.10901528813504248,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4426,
+      "loss": 1.5779,
       "step": 6710
     },
     {
       "epoch": 0.10917775503241214,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4815,
+      "loss": 1.6559,
       "step": 6720
     },
     {
       "epoch": 0.10934022192978181,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4635,
+      "loss": 1.6551,
       "step": 6730
     },
     {
       "epoch": 0.10950268882715147,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4459,
+      "loss": 1.6301,
       "step": 6740
     },
     {
       "epoch": 0.10966515572452112,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.445,
+      "loss": 1.6401,
       "step": 6750
     },
     {
       "epoch": 0.10982762262189079,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4485,
+      "loss": 1.6571,
       "step": 6760
     },
     {
       "epoch": 0.10999008951926045,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4867,
+      "loss": 1.665,
       "step": 6770
     },
     {
       "epoch": 0.11015255641663012,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.451,
+      "loss": 1.6588,
       "step": 6780
     },
     {
       "epoch": 0.11031502331399977,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4692,
+      "loss": 1.6141,
       "step": 6790
     },
     {
       "epoch": 0.11047749021136943,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4778,
+      "loss": 1.6609,
       "step": 6800
     },
     {
       "epoch": 0.1106399571087391,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4657,
+      "loss": 1.6571,
       "step": 6810
     },
     {
       "epoch": 0.11080242400610875,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4901,
+      "loss": 1.6277,
       "step": 6820
     },
     {
       "epoch": 0.11096489090347841,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4496,
+      "loss": 1.6298,
       "step": 6830
     },
     {
       "epoch": 0.11112735780084808,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4664,
+      "loss": 1.6635,
       "step": 6840
     },
     {
       "epoch": 0.11128982469821774,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.474,
+      "loss": 1.6749,
       "step": 6850
     },
     {
       "epoch": 0.1114522915955874,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4431,
+      "loss": 1.6745,
       "step": 6860
     },
     {
       "epoch": 0.11161475849295706,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4743,
+      "loss": 1.6459,
       "step": 6870
     },
     {
       "epoch": 0.11177722539032672,
-      "grad_norm": 2.515625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4716,
+      "loss": 1.669,
       "step": 6880
     },
     {
       "epoch": 0.11193969228769639,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4618,
+      "loss": 1.6553,
       "step": 6890
     },
     {
       "epoch": 0.11210215918506604,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4684,
+      "loss": 1.6332,
       "step": 6900
     },
     {
       "epoch": 0.1122646260824357,
-      "grad_norm": 3.0,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4736,
+      "loss": 1.6513,
       "step": 6910
     },
     {
       "epoch": 0.11242709297980537,
-      "grad_norm": 2.125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.457,
+      "loss": 1.5883,
       "step": 6920
     },
     {
       "epoch": 0.11258955987717502,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4581,
+      "loss": 1.6433,
       "step": 6930
     },
     {
       "epoch": 0.11275202677454468,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4525,
+      "loss": 1.6429,
       "step": 6940
     },
     {
       "epoch": 0.11291449367191435,
-      "grad_norm": 2.5,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4617,
+      "loss": 1.6453,
       "step": 6950
     },
     {
       "epoch": 0.11307696056928401,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.468,
+      "loss": 1.6804,
       "step": 6960
     },
     {
       "epoch": 0.11323942746665366,
-      "grad_norm": 3.84375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4807,
+      "loss": 1.7259,
       "step": 6970
     },
     {
       "epoch": 0.11340189436402333,
-      "grad_norm": 2.71875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.4837,
+      "loss": 1.6883,
       "step": 6980
     },
     {
       "epoch": 0.113564361261393,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4686,
+      "loss": 1.672,
       "step": 6990
     },
     {
       "epoch": 0.11372682815876266,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4558,
+      "loss": 1.6463,
       "step": 7000
     },
     {
       "epoch": 0.11388929505613231,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4476,
+      "loss": 1.6396,
       "step": 7010
     },
     {
       "epoch": 0.11405176195350197,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4609,
+      "loss": 1.5739,
       "step": 7020
     },
     {
       "epoch": 0.11421422885087164,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4581,
+      "loss": 1.6775,
       "step": 7030
     },
     {
       "epoch": 0.1143766957482413,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4544,
+      "loss": 1.6989,
       "step": 7040
     },
     {
       "epoch": 0.11453916264561095,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4615,
+      "loss": 1.6554,
       "step": 7050
     },
     {
       "epoch": 0.11470162954298062,
-      "grad_norm": 3.0,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4537,
+      "loss": 1.6037,
       "step": 7060
     },
     {
       "epoch": 0.11486409644035028,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4616,
+      "loss": 1.6147,
       "step": 7070
     },
     {
       "epoch": 0.11502656333771993,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4544,
+      "loss": 1.6405,
       "step": 7080
     },
     {
       "epoch": 0.1151890302350896,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4499,
+      "loss": 1.6321,
       "step": 7090
     },
     {
       "epoch": 0.11535149713245926,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.6421,
       "step": 7100
     },
     {
       "epoch": 0.11551396402982893,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4422,
+      "loss": 1.6467,
       "step": 7110
     },
     {
       "epoch": 0.11567643092719858,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4434,
+      "loss": 1.6347,
       "step": 7120
     },
     {
       "epoch": 0.11583889782456824,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4395,
+      "loss": 1.6162,
       "step": 7130
     },
     {
       "epoch": 0.11600136472193791,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4622,
+      "loss": 1.6701,
       "step": 7140
     },
     {
       "epoch": 0.11616383161930757,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4702,
+      "loss": 1.6613,
       "step": 7150
     },
     {
       "epoch": 0.11632629851667722,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4334,
+      "loss": 1.6798,
       "step": 7160
     },
     {
       "epoch": 0.11648876541404689,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4677,
+      "loss": 1.6489,
       "step": 7170
     },
     {
       "epoch": 0.11665123231141655,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4531,
+      "loss": 1.6365,
       "step": 7180
     },
     {
       "epoch": 0.1168136992087862,
-      "grad_norm": 2.578125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.4406,
+      "loss": 1.6308,
       "step": 7190
     },
     {
       "epoch": 0.11697616610615587,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4647,
+      "loss": 1.6146,
       "step": 7200
     },
     {
       "epoch": 0.11713863300352553,
-      "grad_norm": 2.5,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4461,
+      "loss": 1.6543,
       "step": 7210
     },
     {
       "epoch": 0.1173010999008952,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4747,
+      "loss": 1.5975,
       "step": 7220
     },
     {
       "epoch": 0.11746356679826485,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4643,
+      "loss": 1.6191,
       "step": 7230
     },
     {
       "epoch": 0.11762603369563451,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4634,
+      "loss": 1.6236,
       "step": 7240
     },
     {
       "epoch": 0.11778850059300418,
-      "grad_norm": 3.125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.438,
+      "loss": 1.6481,
       "step": 7250
     },
     {
       "epoch": 0.11795096749037384,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4781,
+      "loss": 1.6725,
       "step": 7260
     },
     {
       "epoch": 0.1181134343877435,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.6711,
       "step": 7270
     },
     {
       "epoch": 0.11827590128511316,
-      "grad_norm": 3.375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4586,
+      "loss": 1.6001,
       "step": 7280
     },
     {
       "epoch": 0.11843836818248282,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4647,
+      "loss": 1.6563,
       "step": 7290
     },
     {
       "epoch": 0.11860083507985247,
-      "grad_norm": 2.375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4686,
+      "loss": 1.6324,
       "step": 7300
     },
     {
       "epoch": 0.11876330197722214,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4538,
+      "loss": 1.6619,
       "step": 7310
     },
     {
       "epoch": 0.1189257688745918,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4456,
+      "loss": 1.6182,
       "step": 7320
     },
     {
       "epoch": 0.11908823577196147,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4901,
+      "loss": 1.6449,
       "step": 7330
     },
     {
       "epoch": 0.11925070266933112,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4637,
+      "loss": 1.681,
       "step": 7340
     },
     {
       "epoch": 0.11941316956670078,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4449,
+      "loss": 1.6197,
       "step": 7350
     },
     {
       "epoch": 0.11957563646407045,
-      "grad_norm": 3.96875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4565,
+      "loss": 1.6355,
       "step": 7360
     },
     {
       "epoch": 0.11973810336144011,
-      "grad_norm": 2.125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4626,
+      "loss": 1.6572,
       "step": 7370
     },
     {
       "epoch": 0.11990057025880976,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4439,
+      "loss": 1.621,
       "step": 7380
     },
     {
       "epoch": 0.12006303715617943,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.645,
       "step": 7390
     },
     {
       "epoch": 0.12022550405354909,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4573,
+      "loss": 1.6198,
       "step": 7400
     },
     {
       "epoch": 0.12038797095091874,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4606,
+      "loss": 1.6179,
       "step": 7410
     },
     {
       "epoch": 0.12055043784828841,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4671,
+      "loss": 1.6009,
       "step": 7420
     },
     {
       "epoch": 0.12071290474565807,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4605,
+      "loss": 1.6658,
       "step": 7430
     },
     {
       "epoch": 0.12087537164302774,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4651,
+      "loss": 1.6429,
       "step": 7440
     },
     {
       "epoch": 0.12103783854039739,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4716,
+      "loss": 1.6054,
       "step": 7450
     },
     {
       "epoch": 0.12120030543776705,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4595,
+      "loss": 1.6177,
       "step": 7460
     },
     {
       "epoch": 0.12136277233513672,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.469,
+      "loss": 1.6761,
       "step": 7470
     },
     {
       "epoch": 0.12152523923250638,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.461,
+      "loss": 1.6447,
       "step": 7480
     },
     {
       "epoch": 0.12168770612987603,
-      "grad_norm": 1.6875,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.475,
+      "loss": 1.591,
       "step": 7490
     },
     {
       "epoch": 0.1218501730272457,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.4656,
+      "loss": 1.5899,
       "step": 7500
     },
     {
       "epoch": 0.12201263992461536,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4595,
+      "loss": 1.656,
       "step": 7510
     },
     {
       "epoch": 0.12217510682198503,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.6423,
       "step": 7520
     },
     {
       "epoch": 0.12233757371935468,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4464,
+      "loss": 1.6384,
       "step": 7530
     },
     {
       "epoch": 0.12250004061672434,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.449,
+      "loss": 1.6611,
       "step": 7540
     },
     {
       "epoch": 0.12266250751409401,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4624,
+      "loss": 1.659,
       "step": 7550
     },
     {
       "epoch": 0.12282497441146366,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4626,
+      "loss": 1.5474,
       "step": 7560
     },
     {
       "epoch": 0.12298744130883332,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4484,
+      "loss": 1.6056,
       "step": 7570
     },
     {
       "epoch": 0.12314990820620299,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4767,
+      "loss": 1.6352,
       "step": 7580
     },
     {
       "epoch": 0.12331237510357265,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4506,
+      "loss": 1.6377,
       "step": 7590
     },
     {
       "epoch": 0.1234748420009423,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4443,
+      "loss": 1.6772,
       "step": 7600
     },
     {
       "epoch": 0.12363730889831197,
-      "grad_norm": 2.46875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.4621,
+      "loss": 1.6462,
       "step": 7610
     },
     {
       "epoch": 0.12379977579568163,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4523,
+      "loss": 1.5812,
       "step": 7620
     },
     {
       "epoch": 0.1239622426930513,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4826,
+      "loss": 1.6442,
       "step": 7630
     },
     {
       "epoch": 0.12412470959042095,
-      "grad_norm": 3.375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4491,
+      "loss": 1.6166,
       "step": 7640
     },
     {
       "epoch": 0.12428717648779061,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4642,
+      "loss": 1.6263,
       "step": 7650
     },
     {
       "epoch": 0.12444964338516028,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4529,
+      "loss": 1.6388,
       "step": 7660
     },
     {
       "epoch": 0.12461211028252993,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4308,
+      "loss": 1.658,
       "step": 7670
     },
     {
       "epoch": 0.12477457717989959,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4426,
+      "loss": 1.6316,
       "step": 7680
     },
     {
       "epoch": 0.12493704407726926,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4616,
+      "loss": 1.666,
       "step": 7690
     },
     {
       "epoch": 0.12509951097463892,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4688,
+      "loss": 1.6246,
       "step": 7700
     },
     {
       "epoch": 0.12526197787200857,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4383,
+      "loss": 1.6347,
       "step": 7710
     },
     {
       "epoch": 0.12542444476937825,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.471,
+      "loss": 1.5334,
       "step": 7720
     },
     {
       "epoch": 0.1255869116667479,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.462,
+      "loss": 1.6167,
       "step": 7730
     },
     {
       "epoch": 0.12574937856411755,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4439,
+      "loss": 1.5898,
       "step": 7740
     },
     {
       "epoch": 0.12591184546148723,
-      "grad_norm": 3.546875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4465,
+      "loss": 1.6192,
       "step": 7750
     },
     {
       "epoch": 0.12607431235885688,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4683,
+      "loss": 1.566,
       "step": 7760
     },
     {
       "epoch": 0.12623677925622653,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4494,
+      "loss": 1.6434,
       "step": 7770
     },
     {
       "epoch": 0.1263992461535962,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4538,
+      "loss": 1.6484,
       "step": 7780
     },
     {
       "epoch": 0.12656171305096586,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4584,
+      "loss": 1.6932,
       "step": 7790
     },
     {
       "epoch": 0.1267241799483355,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4313,
+      "loss": 1.5846,
       "step": 7800
     },
     {
       "epoch": 0.1268866468457052,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4535,
+      "loss": 1.6435,
       "step": 7810
     },
     {
       "epoch": 0.12704911374307484,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4423,
+      "loss": 1.6504,
       "step": 7820
     },
     {
       "epoch": 0.12721158064044452,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4356,
+      "loss": 1.609,
       "step": 7830
     },
     {
       "epoch": 0.12737404753781417,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4629,
+      "loss": 1.6014,
       "step": 7840
     },
     {
       "epoch": 0.12753651443518382,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4581,
+      "loss": 1.6755,
       "step": 7850
     },
     {
       "epoch": 0.1276989813325535,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.44,
+      "loss": 1.6334,
       "step": 7860
     },
     {
       "epoch": 0.12786144822992315,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4504,
+      "loss": 1.6214,
       "step": 7870
     },
     {
       "epoch": 0.1280239151272928,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4771,
+      "loss": 1.5478,
       "step": 7880
     },
     {
       "epoch": 0.12818638202466248,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4401,
+      "loss": 1.7008,
       "step": 7890
     },
     {
       "epoch": 0.12834884892203213,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4609,
+      "loss": 1.662,
       "step": 7900
     },
     {
       "epoch": 0.12851131581940178,
-      "grad_norm": 3.28125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.4725,
+      "loss": 1.6171,
       "step": 7910
     },
     {
       "epoch": 0.12867378271677146,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4539,
+      "loss": 1.6539,
       "step": 7920
     },
     {
       "epoch": 0.1288362496141411,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4637,
+      "loss": 1.6199,
       "step": 7930
     },
     {
       "epoch": 0.1289987165115108,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4458,
+      "loss": 1.6677,
       "step": 7940
     },
     {
       "epoch": 0.12916118340888044,
-      "grad_norm": 1.9765625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4615,
+      "loss": 1.6555,
       "step": 7950
     },
     {
       "epoch": 0.1293236503062501,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4553,
+      "loss": 1.6117,
       "step": 7960
     },
     {
       "epoch": 0.12948611720361977,
-      "grad_norm": 1.90625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.5896,
       "step": 7970
     },
     {
       "epoch": 0.12964858410098942,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4633,
+      "loss": 1.6367,
       "step": 7980
     },
     {
       "epoch": 0.12981105099835907,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4484,
+      "loss": 1.5863,
       "step": 7990
     },
     {
       "epoch": 0.12997351789572875,
-      "grad_norm": 3.828125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4572,
+      "loss": 1.5887,
       "step": 8000
     },
     {
       "epoch": 0.1301359847930984,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4692,
+      "loss": 1.6279,
       "step": 8010
     },
     {
       "epoch": 0.13029845169046805,
-      "grad_norm": 3.0,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4662,
+      "loss": 1.6003,
       "step": 8020
     },
     {
       "epoch": 0.13046091858783773,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4461,
+      "loss": 1.6237,
       "step": 8030
     },
     {
       "epoch": 0.13062338548520738,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4507,
+      "loss": 1.6026,
       "step": 8040
     },
     {
       "epoch": 0.13078585238257706,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4431,
+      "loss": 1.5598,
       "step": 8050
     },
     {
       "epoch": 0.1309483192799467,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4668,
+      "loss": 1.6642,
       "step": 8060
     },
     {
       "epoch": 0.13111078617731636,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4467,
+      "loss": 1.6175,
       "step": 8070
     },
     {
       "epoch": 0.13127325307468604,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4485,
+      "loss": 1.7081,
       "step": 8080
     },
     {
       "epoch": 0.1314357199720557,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4489,
+      "loss": 1.5968,
       "step": 8090
     },
     {
       "epoch": 0.13159818686942534,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4692,
+      "loss": 1.6173,
       "step": 8100
     },
     {
       "epoch": 0.13176065376679502,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4308,
+      "loss": 1.6099,
       "step": 8110
     },
     {
       "epoch": 0.13192312066416467,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4425,
+      "loss": 1.5923,
       "step": 8120
     },
     {
       "epoch": 0.13208558756153435,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4728,
+      "loss": 1.6103,
       "step": 8130
     },
     {
       "epoch": 0.132248054458904,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4265,
+      "loss": 1.6313,
       "step": 8140
     },
     {
       "epoch": 0.13241052135627365,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4585,
+      "loss": 1.6616,
       "step": 8150
     },
     {
       "epoch": 0.13257298825364333,
-      "grad_norm": 2.75,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4366,
+      "loss": 1.6329,
       "step": 8160
     },
     {
       "epoch": 0.13273545515101298,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4555,
+      "loss": 1.6176,
       "step": 8170
     },
     {
       "epoch": 0.13289792204838263,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.455,
+      "loss": 1.6302,
       "step": 8180
     },
     {
       "epoch": 0.1330603889457523,
-      "grad_norm": 4.625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4686,
+      "loss": 1.5867,
       "step": 8190
     },
     {
       "epoch": 0.13322285584312196,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4572,
+      "loss": 1.619,
       "step": 8200
     },
     {
       "epoch": 0.1333853227404916,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4768,
+      "loss": 1.6175,
       "step": 8210
     },
     {
       "epoch": 0.1335477896378613,
-      "grad_norm": 4.0,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4567,
+      "loss": 1.6183,
       "step": 8220
     },
     {
       "epoch": 0.13371025653523094,
-      "grad_norm": 4.0,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.4512,
+      "loss": 1.5878,
       "step": 8230
     },
     {
       "epoch": 0.13387272343260062,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4425,
+      "loss": 1.5793,
       "step": 8240
     },
     {
       "epoch": 0.13403519032997027,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4483,
+      "loss": 1.6242,
       "step": 8250
     },
     {
       "epoch": 0.13419765722733992,
-      "grad_norm": 2.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4564,
+      "loss": 1.5925,
       "step": 8260
     },
     {
       "epoch": 0.1343601241247096,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.449,
+      "loss": 1.5923,
       "step": 8270
     },
     {
       "epoch": 0.13452259102207925,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4388,
+      "loss": 1.6665,
       "step": 8280
     },
     {
       "epoch": 0.1346850579194489,
-      "grad_norm": 3.75,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.4796,
+      "loss": 1.6416,
       "step": 8290
     },
     {
       "epoch": 0.13484752481681858,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4349,
+      "loss": 1.5829,
       "step": 8300
     },
     {
       "epoch": 0.13500999171418823,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4509,
+      "loss": 1.5888,
       "step": 8310
     },
     {
       "epoch": 0.13517245861155788,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4759,
+      "loss": 1.6284,
       "step": 8320
     },
     {
       "epoch": 0.13533492550892756,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.625,
       "step": 8330
     },
     {
       "epoch": 0.1354973924062972,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4574,
+      "loss": 1.5979,
       "step": 8340
     },
     {
       "epoch": 0.1356598593036669,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4395,
+      "loss": 1.5964,
       "step": 8350
     },
     {
       "epoch": 0.13582232620103654,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4593,
+      "loss": 1.6308,
       "step": 8360
     },
     {
       "epoch": 0.1359847930984062,
-      "grad_norm": 5.34375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4703,
+      "loss": 1.6273,
       "step": 8370
     },
     {
       "epoch": 0.13614725999577587,
-      "grad_norm": 2.375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4463,
+      "loss": 1.5749,
       "step": 8380
     },
     {
       "epoch": 0.13630972689314552,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4398,
+      "loss": 1.5986,
       "step": 8390
     },
     {
       "epoch": 0.13647219379051517,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.441,
+      "loss": 1.6056,
       "step": 8400
     },
     {
       "epoch": 0.13663466068788485,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4617,
+      "loss": 1.5848,
       "step": 8410
     },
     {
       "epoch": 0.1367971275852545,
-      "grad_norm": 2.546875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.4753,
+      "loss": 1.5853,
       "step": 8420
     },
     {
       "epoch": 0.13695959448262415,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.438,
+      "loss": 1.6203,
       "step": 8430
     },
     {
       "epoch": 0.13712206137999383,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4636,
+      "loss": 1.6462,
       "step": 8440
     },
     {
       "epoch": 0.13728452827736348,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4637,
+      "loss": 1.6165,
       "step": 8450
     },
     {
       "epoch": 0.13744699517473316,
-      "grad_norm": 2.25,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4291,
+      "loss": 1.6152,
       "step": 8460
     },
     {
       "epoch": 0.1376094620721028,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4501,
+      "loss": 1.625,
       "step": 8470
     },
     {
       "epoch": 0.13777192896947246,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4369,
+      "loss": 1.6075,
       "step": 8480
     },
     {
       "epoch": 0.13793439586684214,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4621,
+      "loss": 1.6255,
       "step": 8490
     },
     {
       "epoch": 0.1380968627642118,
-      "grad_norm": 1.875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.435,
+      "loss": 1.584,
       "step": 8500
     },
     {
       "epoch": 0.13825932966158144,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4574,
+      "loss": 1.6469,
       "step": 8510
     },
     {
       "epoch": 0.13842179655895112,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4334,
+      "loss": 1.5851,
       "step": 8520
     },
     {
       "epoch": 0.13858426345632077,
-      "grad_norm": 2.25,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.5647,
       "step": 8530
     },
     {
       "epoch": 0.13874673035369042,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.6073,
       "step": 8540
     },
     {
       "epoch": 0.1389091972510601,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4689,
+      "loss": 1.6396,
       "step": 8550
     },
     {
       "epoch": 0.13907166414842975,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4633,
+      "loss": 1.5766,
       "step": 8560
     },
     {
       "epoch": 0.13923413104579943,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4675,
+      "loss": 1.5852,
       "step": 8570
     },
     {
       "epoch": 0.13939659794316908,
-      "grad_norm": 2.234375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.467,
+      "loss": 1.6859,
       "step": 8580
     },
     {
       "epoch": 0.13955906484053873,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4598,
+      "loss": 1.6298,
       "step": 8590
     },
     {
       "epoch": 0.1397215317379084,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4644,
+      "loss": 1.5978,
       "step": 8600
     },
     {
       "epoch": 0.13988399863527806,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4633,
+      "loss": 1.5875,
       "step": 8610
     },
     {
       "epoch": 0.1400464655326477,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4391,
+      "loss": 1.6214,
       "step": 8620
     },
     {
       "epoch": 0.1402089324300174,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4576,
+      "loss": 1.6177,
       "step": 8630
     },
     {
       "epoch": 0.14037139932738704,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4495,
+      "loss": 1.6294,
       "step": 8640
     },
     {
       "epoch": 0.1405338662247567,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4478,
+      "loss": 1.5441,
       "step": 8650
     },
     {
       "epoch": 0.14069633312212637,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4505,
+      "loss": 1.6173,
       "step": 8660
     },
     {
       "epoch": 0.14085880001949602,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4324,
+      "loss": 1.636,
       "step": 8670
     },
     {
       "epoch": 0.1410212669168657,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.459,
+      "loss": 1.5774,
       "step": 8680
     },
     {
       "epoch": 0.14118373381423535,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4519,
+      "loss": 1.6248,
       "step": 8690
     },
     {
       "epoch": 0.141346200711605,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4508,
+      "loss": 1.6181,
       "step": 8700
     },
     {
       "epoch": 0.14150866760897468,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4579,
+      "loss": 1.617,
       "step": 8710
     },
     {
       "epoch": 0.14167113450634433,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4667,
+      "loss": 1.6375,
       "step": 8720
     },
     {
       "epoch": 0.14183360140371398,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4465,
+      "loss": 1.5705,
       "step": 8730
     },
     {
       "epoch": 0.14199606830108366,
-      "grad_norm": 3.265625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4471,
+      "loss": 1.6009,
       "step": 8740
     },
     {
       "epoch": 0.1421585351984533,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4442,
+      "loss": 1.6373,
       "step": 8750
     },
     {
       "epoch": 0.14232100209582296,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.442,
+      "loss": 1.6493,
       "step": 8760
     },
     {
       "epoch": 0.14248346899319264,
-      "grad_norm": 2.578125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.467,
+      "loss": 1.6235,
       "step": 8770
     },
     {
       "epoch": 0.1426459358905623,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4207,
+      "loss": 1.6019,
       "step": 8780
     },
     {
       "epoch": 0.14280840278793197,
-      "grad_norm": 3.5,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.459,
+      "loss": 1.5669,
       "step": 8790
     },
     {
       "epoch": 0.14297086968530162,
-      "grad_norm": 2.0,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.484,
+      "loss": 1.6103,
       "step": 8800
     },
     {
       "epoch": 0.14313333658267127,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.5784,
       "step": 8810
     },
     {
       "epoch": 0.14329580348004095,
-      "grad_norm": 3.296875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.448,
+      "loss": 1.6257,
       "step": 8820
     },
     {
       "epoch": 0.1434582703774106,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.477,
+      "loss": 1.6559,
       "step": 8830
     },
     {
       "epoch": 0.14362073727478025,
-      "grad_norm": 2.0,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4594,
+      "loss": 1.6427,
       "step": 8840
     },
     {
       "epoch": 0.14378320417214993,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.476,
+      "loss": 1.5646,
       "step": 8850
     },
     {
       "epoch": 0.14394567106951958,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4517,
+      "loss": 1.597,
       "step": 8860
     },
     {
       "epoch": 0.14410813796688923,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.437,
+      "loss": 1.636,
       "step": 8870
     },
     {
       "epoch": 0.1442706048642589,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.6362,
       "step": 8880
     },
     {
       "epoch": 0.14443307176162856,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4398,
+      "loss": 1.6256,
       "step": 8890
     },
     {
       "epoch": 0.14459553865899824,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4409,
+      "loss": 1.5474,
       "step": 8900
     },
     {
       "epoch": 0.1447580055563679,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4641,
+      "loss": 1.6104,
       "step": 8910
     },
     {
       "epoch": 0.14492047245373754,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4544,
+      "loss": 1.5915,
       "step": 8920
     },
     {
       "epoch": 0.14508293935110722,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4645,
+      "loss": 1.6802,
       "step": 8930
     },
     {
       "epoch": 0.14524540624847687,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4472,
+      "loss": 1.5791,
       "step": 8940
     },
     {
       "epoch": 0.14540787314584652,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4476,
+      "loss": 1.608,
       "step": 8950
     },
     {
       "epoch": 0.1455703400432162,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4623,
+      "loss": 1.6317,
       "step": 8960
     },
     {
       "epoch": 0.14573280694058585,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4347,
+      "loss": 1.6218,
       "step": 8970
     },
     {
       "epoch": 0.14589527383795553,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4577,
+      "loss": 1.6237,
       "step": 8980
     },
     {
       "epoch": 0.14605774073532518,
-      "grad_norm": 3.265625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4631,
+      "loss": 1.5783,
       "step": 8990
     },
     {
       "epoch": 0.14622020763269483,
-      "grad_norm": 3.5,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4515,
+      "loss": 1.6064,
       "step": 9000
     },
     {
       "epoch": 0.1463826745300645,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4791,
+      "loss": 1.5842,
       "step": 9010
     },
     {
       "epoch": 0.14654514142743416,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4649,
+      "loss": 1.5627,
       "step": 9020
     },
     {
       "epoch": 0.1467076083248038,
-      "grad_norm": 2.375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.6516,
       "step": 9030
     },
     {
       "epoch": 0.1468700752221735,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.4454,
+      "loss": 1.5592,
       "step": 9040
     },
     {
       "epoch": 0.14703254211954314,
-      "grad_norm": 3.5625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4424,
+      "loss": 1.6541,
       "step": 9050
     },
     {
       "epoch": 0.1471950090169128,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4558,
+      "loss": 1.6369,
       "step": 9060
     },
     {
       "epoch": 0.14735747591428247,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4626,
+      "loss": 1.6005,
       "step": 9070
     },
     {
       "epoch": 0.14751994281165212,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4353,
+      "loss": 1.5836,
       "step": 9080
     },
     {
       "epoch": 0.1476824097090218,
-      "grad_norm": 2.734375,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4461,
+      "loss": 1.5006,
       "step": 9090
     },
     {
       "epoch": 0.14784487660639145,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4556,
+      "loss": 1.5858,
       "step": 9100
     },
     {
       "epoch": 0.1480073435037611,
-      "grad_norm": 3.546875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.426,
+      "loss": 1.6098,
       "step": 9110
     },
     {
       "epoch": 0.14816981040113078,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4562,
+      "loss": 1.6035,
       "step": 9120
     },
     {
       "epoch": 0.14833227729850043,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4639,
+      "loss": 1.621,
       "step": 9130
     },
     {
       "epoch": 0.14849474419587008,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4628,
+      "loss": 1.6493,
       "step": 9140
     },
     {
       "epoch": 0.14865721109323976,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4635,
+      "loss": 1.5762,
       "step": 9150
     },
     {
       "epoch": 0.1488196779906094,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.6035,
       "step": 9160
     },
     {
       "epoch": 0.14898214488797906,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4605,
+      "loss": 1.6005,
       "step": 9170
     },
     {
       "epoch": 0.14914461178534874,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4541,
+      "loss": 1.5828,
       "step": 9180
     },
     {
       "epoch": 0.1493070786827184,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4454,
+      "loss": 1.5358,
       "step": 9190
     },
     {
       "epoch": 0.14946954558008807,
-      "grad_norm": 3.0,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4389,
+      "loss": 1.5682,
       "step": 9200
     },
     {
       "epoch": 0.14963201247745772,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4484,
+      "loss": 1.5677,
       "step": 9210
     },
     {
       "epoch": 0.14979447937482737,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4691,
+      "loss": 1.5823,
       "step": 9220
     },
     {
       "epoch": 0.14995694627219705,
-      "grad_norm": 1.9453125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4325,
+      "loss": 1.5835,
       "step": 9230
     },
     {
       "epoch": 0.1501194131695667,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4394,
+      "loss": 1.6034,
       "step": 9240
     },
     {
       "epoch": 0.15028188006693635,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.5916,
       "step": 9250
     },
     {
       "epoch": 0.15044434696430603,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4507,
+      "loss": 1.6199,
       "step": 9260
     },
     {
       "epoch": 0.15060681386167568,
-      "grad_norm": 4.28125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4447,
+      "loss": 1.603,
       "step": 9270
     },
     {
       "epoch": 0.15076928075904533,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4556,
+      "loss": 1.547,
       "step": 9280
     },
     {
       "epoch": 0.150931747656415,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4466,
+      "loss": 1.662,
       "step": 9290
     },
     {
       "epoch": 0.15109421455378466,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4586,
+      "loss": 1.5935,
       "step": 9300
     },
     {
       "epoch": 0.15125668145115434,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4546,
+      "loss": 1.6907,
       "step": 9310
     },
     {
       "epoch": 0.151419148348524,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.451,
+      "loss": 1.6446,
       "step": 9320
     },
     {
       "epoch": 0.15158161524589364,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4717,
+      "loss": 1.5639,
       "step": 9330
     },
     {
       "epoch": 0.15174408214326332,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4435,
+      "loss": 1.5879,
       "step": 9340
     },
     {
       "epoch": 0.15190654904063297,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4563,
+      "loss": 1.6097,
       "step": 9350
     },
     {
       "epoch": 0.15206901593800262,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4611,
+      "loss": 1.6003,
       "step": 9360
     },
     {
       "epoch": 0.1522314828353723,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4599,
+      "loss": 1.569,
       "step": 9370
     },
     {
       "epoch": 0.15239394973274195,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4642,
+      "loss": 1.596,
       "step": 9380
     },
     {
       "epoch": 0.1525564166301116,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4304,
+      "loss": 1.585,
       "step": 9390
     },
     {
       "epoch": 0.15271888352748128,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4561,
+      "loss": 1.6072,
       "step": 9400
     },
     {
       "epoch": 0.15288135042485093,
-      "grad_norm": 2.59375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.45,
+      "loss": 1.5843,
       "step": 9410
     },
     {
       "epoch": 0.1530438173222206,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4138,
+      "loss": 1.5986,
       "step": 9420
     },
     {
       "epoch": 0.15320628421959026,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.6122,
       "step": 9430
     },
     {
       "epoch": 0.1533687511169599,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4284,
+      "loss": 1.5987,
       "step": 9440
     },
     {
       "epoch": 0.1535312180143296,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.6073,
       "step": 9450
     },
     {
       "epoch": 0.15369368491169924,
-      "grad_norm": 3.125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4364,
+      "loss": 1.5877,
       "step": 9460
     },
     {
       "epoch": 0.1538561518090689,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4617,
+      "loss": 1.5871,
       "step": 9470
     },
     {
       "epoch": 0.15401861870643857,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4624,
+      "loss": 1.6517,
       "step": 9480
     },
     {
       "epoch": 0.15418108560380822,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4288,
+      "loss": 1.6201,
       "step": 9490
     },
     {
       "epoch": 0.15434355250117787,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.448,
+      "loss": 1.6022,
       "step": 9500
     },
     {
       "epoch": 0.15450601939854755,
-      "grad_norm": 3.25,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4368,
+      "loss": 1.5707,
       "step": 9510
     },
     {
       "epoch": 0.1546684862959172,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4623,
+      "loss": 1.5512,
       "step": 9520
     },
     {
       "epoch": 0.15483095319328688,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4613,
+      "loss": 1.6235,
       "step": 9530
     },
     {
       "epoch": 0.15499342009065653,
-      "grad_norm": 2.296875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4531,
+      "loss": 1.5745,
       "step": 9540
     },
     {
       "epoch": 0.15515588698802618,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.466,
+      "loss": 1.6228,
       "step": 9550
     },
     {
       "epoch": 0.15531835388539586,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.445,
+      "loss": 1.5852,
       "step": 9560
     },
     {
       "epoch": 0.1554808207827655,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4606,
+      "loss": 1.5838,
       "step": 9570
     },
     {
       "epoch": 0.15564328768013516,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4628,
+      "loss": 1.573,
       "step": 9580
     },
     {
       "epoch": 0.15580575457750484,
-      "grad_norm": 2.75,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.549,
       "step": 9590
     },
     {
       "epoch": 0.1559682214748745,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4352,
+      "loss": 1.6055,
       "step": 9600
     },
     {
       "epoch": 0.15613068837224414,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4318,
+      "loss": 1.6327,
       "step": 9610
     },
     {
       "epoch": 0.15629315526961382,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4384,
+      "loss": 1.5991,
       "step": 9620
     },
     {
       "epoch": 0.15645562216698347,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4411,
+      "loss": 1.6355,
       "step": 9630
     },
     {
       "epoch": 0.15661808906435315,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4341,
+      "loss": 1.5741,
       "step": 9640
     },
     {
       "epoch": 0.1567805559617228,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4556,
+      "loss": 1.5458,
       "step": 9650
     },
     {
       "epoch": 0.15694302285909245,
-      "grad_norm": 2.5,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4407,
+      "loss": 1.5824,
       "step": 9660
     },
     {
       "epoch": 0.15710548975646213,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4443,
+      "loss": 1.5765,
       "step": 9670
     },
     {
       "epoch": 0.15726795665383178,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4658,
+      "loss": 1.5613,
       "step": 9680
     },
     {
       "epoch": 0.15743042355120143,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4269,
+      "loss": 1.6414,
       "step": 9690
     },
     {
       "epoch": 0.1575928904485711,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4667,
+      "loss": 1.585,
       "step": 9700
     },
     {
       "epoch": 0.15775535734594076,
-      "grad_norm": 1.8828125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.5911,
       "step": 9710
     },
     {
       "epoch": 0.1579178242433104,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4604,
+      "loss": 1.5381,
       "step": 9720
     },
     {
       "epoch": 0.1580802911406801,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4767,
+      "loss": 1.619,
       "step": 9730
     },
     {
       "epoch": 0.15824275803804974,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4504,
+      "loss": 1.586,
       "step": 9740
     },
     {
       "epoch": 0.15840522493541942,
-      "grad_norm": 3.875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4489,
+      "loss": 1.5483,
       "step": 9750
     },
     {
       "epoch": 0.15856769183278907,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4471,
+      "loss": 1.5909,
       "step": 9760
     },
     {
       "epoch": 0.15873015873015872,
-      "grad_norm": 1.765625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.6048,
       "step": 9770
     },
     {
       "epoch": 0.1588926256275284,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4625,
+      "loss": 1.6145,
       "step": 9780
     },
     {
       "epoch": 0.15905509252489805,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4639,
+      "loss": 1.6026,
       "step": 9790
     },
     {
       "epoch": 0.1592175594222677,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4387,
+      "loss": 1.5977,
       "step": 9800
     },
     {
       "epoch": 0.15938002631963738,
-      "grad_norm": 4.71875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4567,
+      "loss": 1.5787,
       "step": 9810
     },
     {
       "epoch": 0.15954249321700703,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4505,
+      "loss": 1.5738,
       "step": 9820
     },
     {
       "epoch": 0.1597049601143767,
-      "grad_norm": 2.375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4479,
+      "loss": 1.577,
       "step": 9830
     },
     {
       "epoch": 0.15986742701174636,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4581,
+      "loss": 1.5261,
       "step": 9840
     },
     {
       "epoch": 0.160029893909116,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4371,
+      "loss": 1.5703,
       "step": 9850
     },
     {
       "epoch": 0.1601923608064857,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4617,
+      "loss": 1.6188,
       "step": 9860
     },
     {
       "epoch": 0.16035482770385534,
-      "grad_norm": 3.84375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.441,
+      "loss": 1.6101,
       "step": 9870
     },
     {
       "epoch": 0.160517294601225,
-      "grad_norm": 2.0625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4536,
+      "loss": 1.563,
       "step": 9880
     },
     {
       "epoch": 0.16067976149859467,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.443,
+      "loss": 1.593,
       "step": 9890
     },
     {
       "epoch": 0.16084222839596432,
-      "grad_norm": 2.5,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4409,
+      "loss": 1.5964,
       "step": 9900
     },
     {
       "epoch": 0.16100469529333397,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.557,
       "step": 9910
     },
     {
       "epoch": 0.16116716219070365,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.444,
+      "loss": 1.5854,
       "step": 9920
     },
     {
       "epoch": 0.1613296290880733,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4562,
+      "loss": 1.6027,
       "step": 9930
     },
     {
       "epoch": 0.16149209598544298,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.6116,
       "step": 9940
     },
     {
       "epoch": 0.16165456288281263,
-      "grad_norm": 3.546875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4428,
+      "loss": 1.562,
       "step": 9950
     },
     {
       "epoch": 0.16181702978018228,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4492,
+      "loss": 1.5926,
       "step": 9960
     },
     {
       "epoch": 0.16197949667755196,
-      "grad_norm": 2.296875,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.457,
+      "loss": 1.5641,
       "step": 9970
     },
     {
       "epoch": 0.1621419635749216,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.5539,
       "step": 9980
     },
     {
       "epoch": 0.16230443047229126,
-      "grad_norm": 3.703125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4479,
+      "loss": 1.5665,
       "step": 9990
     },
     {
       "epoch": 0.16246689736966094,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.4461,
+      "loss": 1.5504,
       "step": 10000
     },
     {
       "epoch": 0.1626293642670306,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4579,
+      "loss": 1.5749,
       "step": 10010
     },
     {
       "epoch": 0.16279183116440024,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4497,
+      "loss": 1.5803,
       "step": 10020
     },
     {
       "epoch": 0.16295429806176992,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4722,
+      "loss": 1.57,
       "step": 10030
     },
     {
       "epoch": 0.16311676495913957,
-      "grad_norm": 3.25,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4433,
+      "loss": 1.6002,
       "step": 10040
     },
     {
       "epoch": 0.16327923185650925,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4376,
+      "loss": 1.5467,
       "step": 10050
     },
     {
       "epoch": 0.1634416987538789,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4323,
+      "loss": 1.5372,
       "step": 10060
     },
     {
       "epoch": 0.16360416565124855,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.5835,
       "step": 10070
     },
     {
       "epoch": 0.16376663254861823,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4468,
+      "loss": 1.6117,
       "step": 10080
     },
     {
       "epoch": 0.16392909944598788,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4329,
+      "loss": 1.5634,
       "step": 10090
     },
     {
       "epoch": 0.16409156634335753,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4302,
+      "loss": 1.6142,
       "step": 10100
     },
     {
       "epoch": 0.1642540332407272,
-      "grad_norm": 3.96875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.5547,
       "step": 10110
     },
     {
       "epoch": 0.16441650013809686,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.6484,
       "step": 10120
     },
     {
       "epoch": 0.1645789670354665,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4409,
+      "loss": 1.5782,
       "step": 10130
     },
     {
       "epoch": 0.1647414339328362,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4682,
+      "loss": 1.5781,
       "step": 10140
     },
     {
       "epoch": 0.16490390083020584,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.6042,
       "step": 10150
     },
     {
       "epoch": 0.16506636772757552,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.4495,
+      "loss": 1.5761,
       "step": 10160
     },
     {
       "epoch": 0.16522883462494517,
-      "grad_norm": 5.4375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4418,
+      "loss": 1.5983,
       "step": 10170
     },
     {
       "epoch": 0.16539130152231482,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4469,
+      "loss": 1.565,
       "step": 10180
     },
     {
       "epoch": 0.1655537684196845,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4531,
+      "loss": 1.5751,
       "step": 10190
     },
     {
       "epoch": 0.16571623531705415,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4379,
+      "loss": 1.5732,
       "step": 10200
     },
     {
       "epoch": 0.1658787022144238,
-      "grad_norm": 2.0,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.4472,
+      "loss": 1.6159,
       "step": 10210
     },
     {
       "epoch": 0.16604116911179348,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4477,
+      "loss": 1.5575,
       "step": 10220
     },
     {
       "epoch": 0.16620363600916313,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4469,
+      "loss": 1.5969,
       "step": 10230
     },
     {
       "epoch": 0.16636610290653278,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.6191,
       "step": 10240
     },
     {
       "epoch": 0.16652856980390246,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4415,
+      "loss": 1.559,
       "step": 10250
     },
     {
       "epoch": 0.1666910367012721,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4358,
+      "loss": 1.5855,
       "step": 10260
     },
     {
       "epoch": 0.1668535035986418,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4346,
+      "loss": 1.5365,
       "step": 10270
     },
     {
       "epoch": 0.16701597049601144,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.435,
+      "loss": 1.5925,
       "step": 10280
     },
     {
       "epoch": 0.1671784373933811,
-      "grad_norm": 3.578125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4496,
+      "loss": 1.5985,
       "step": 10290
     },
     {
       "epoch": 0.16734090429075077,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4453,
+      "loss": 1.5692,
       "step": 10300
     },
     {
       "epoch": 0.16750337118812042,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4399,
+      "loss": 1.5466,
       "step": 10310
     },
     {
       "epoch": 0.16766583808549007,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4398,
+      "loss": 1.5692,
       "step": 10320
     },
     {
       "epoch": 0.16782830498285975,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4507,
+      "loss": 1.5606,
       "step": 10330
     },
     {
       "epoch": 0.1679907718802294,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4729,
+      "loss": 1.5907,
       "step": 10340
     },
     {
       "epoch": 0.16815323877759905,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4313,
+      "loss": 1.6089,
       "step": 10350
     },
     {
       "epoch": 0.16831570567496873,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4258,
+      "loss": 1.5851,
       "step": 10360
     },
     {
       "epoch": 0.16847817257233838,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4623,
+      "loss": 1.6002,
       "step": 10370
     },
     {
       "epoch": 0.16864063946970806,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.5415,
       "step": 10380
     },
     {
       "epoch": 0.1688031063670777,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.6051,
       "step": 10390
     },
     {
       "epoch": 0.16896557326444736,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.6037,
       "step": 10400
     },
     {
       "epoch": 0.16912804016181704,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.438,
+      "loss": 1.5999,
       "step": 10410
     },
     {
       "epoch": 0.1692905070591867,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4268,
+      "loss": 1.5848,
       "step": 10420
     },
     {
       "epoch": 0.16945297395655634,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4306,
+      "loss": 1.5892,
       "step": 10430
     },
     {
       "epoch": 0.16961544085392602,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4602,
+      "loss": 1.5964,
       "step": 10440
     },
     {
       "epoch": 0.16977790775129567,
-      "grad_norm": 2.25,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4642,
+      "loss": 1.5452,
       "step": 10450
     },
     {
       "epoch": 0.16994037464866532,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.5432,
       "step": 10460
     },
     {
       "epoch": 0.170102841546035,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4494,
+      "loss": 1.608,
       "step": 10470
     },
     {
       "epoch": 0.17026530844340465,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4674,
+      "loss": 1.597,
       "step": 10480
     },
     {
       "epoch": 0.17042777534077433,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4584,
+      "loss": 1.5356,
       "step": 10490
     },
     {
       "epoch": 0.17059024223814398,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4478,
+      "loss": 1.5562,
       "step": 10500
     },
     {
       "epoch": 0.17075270913551363,
-      "grad_norm": 1.953125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4393,
+      "loss": 1.567,
       "step": 10510
     },
     {
       "epoch": 0.1709151760328833,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4588,
+      "loss": 1.5345,
       "step": 10520
     },
     {
       "epoch": 0.17107764293025296,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.6001,
       "step": 10530
     },
     {
       "epoch": 0.1712401098276226,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4344,
+      "loss": 1.6107,
       "step": 10540
     },
     {
       "epoch": 0.1714025767249923,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.4428,
+      "loss": 1.6237,
       "step": 10550
     },
     {
       "epoch": 0.17156504362236194,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4583,
+      "loss": 1.565,
       "step": 10560
     },
     {
       "epoch": 0.1717275105197316,
-      "grad_norm": 2.015625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4449,
+      "loss": 1.5618,
       "step": 10570
     },
     {
       "epoch": 0.17188997741710127,
-      "grad_norm": 3.25,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.45,
+      "loss": 1.5442,
       "step": 10580
     },
     {
       "epoch": 0.17205244431447092,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4633,
+      "loss": 1.6018,
       "step": 10590
     },
     {
       "epoch": 0.1722149112118406,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4384,
+      "loss": 1.5536,
       "step": 10600
     },
     {
       "epoch": 0.17237737810921025,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4464,
+      "loss": 1.6028,
       "step": 10610
     },
     {
       "epoch": 0.1725398450065799,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4665,
+      "loss": 1.6033,
       "step": 10620
     },
     {
       "epoch": 0.17270231190394958,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4498,
+      "loss": 1.5728,
       "step": 10630
     },
     {
       "epoch": 0.17286477880131923,
-      "grad_norm": 1.9296875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4246,
+      "loss": 1.6193,
       "step": 10640
     },
     {
       "epoch": 0.17302724569868888,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4431,
+      "loss": 1.5399,
       "step": 10650
     },
     {
       "epoch": 0.17318971259605856,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4308,
+      "loss": 1.5142,
       "step": 10660
     },
     {
       "epoch": 0.1733521794934282,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4466,
+      "loss": 1.5768,
       "step": 10670
     },
     {
       "epoch": 0.1735146463907979,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.5364,
       "step": 10680
     },
     {
       "epoch": 0.17367711328816754,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4487,
+      "loss": 1.5748,
       "step": 10690
     },
     {
       "epoch": 0.1738395801855372,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4634,
+      "loss": 1.5158,
       "step": 10700
     },
     {
       "epoch": 0.17400204708290687,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4519,
+      "loss": 1.5743,
       "step": 10710
     },
     {
       "epoch": 0.17416451398027652,
-      "grad_norm": 2.125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4423,
+      "loss": 1.5211,
       "step": 10720
     },
     {
       "epoch": 0.17432698087764617,
-      "grad_norm": 6.3125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4304,
+      "loss": 1.5439,
       "step": 10730
     },
     {
       "epoch": 0.17448944777501585,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.5508,
       "step": 10740
     },
     {
       "epoch": 0.1746519146723855,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4715,
+      "loss": 1.5341,
       "step": 10750
     },
     {
       "epoch": 0.17481438156975515,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4388,
+      "loss": 1.5918,
       "step": 10760
     },
     {
       "epoch": 0.17497684846712483,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4539,
+      "loss": 1.5414,
       "step": 10770
     },
     {
       "epoch": 0.17513931536449448,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4437,
+      "loss": 1.5636,
       "step": 10780
     },
     {
       "epoch": 0.17530178226186416,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.5527,
       "step": 10790
     },
     {
       "epoch": 0.1754642491592338,
-      "grad_norm": 2.375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.5978,
       "step": 10800
     },
     {
       "epoch": 0.17562671605660346,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4296,
+      "loss": 1.5931,
       "step": 10810
     },
     {
       "epoch": 0.17578918295397314,
-      "grad_norm": 1.7578125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4432,
+      "loss": 1.5646,
       "step": 10820
     },
     {
       "epoch": 0.1759516498513428,
-      "grad_norm": 2.71875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4429,
+      "loss": 1.561,
       "step": 10830
     },
     {
       "epoch": 0.17611411674871244,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4418,
+      "loss": 1.5387,
       "step": 10840
     },
     {
       "epoch": 0.17627658364608212,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4475,
+      "loss": 1.6037,
       "step": 10850
     },
     {
       "epoch": 0.17643905054345177,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4432,
+      "loss": 1.5451,
       "step": 10860
     },
     {
       "epoch": 0.17660151744082142,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4448,
+      "loss": 1.5721,
       "step": 10870
     },
     {
       "epoch": 0.1767639843381911,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.5831,
       "step": 10880
     },
     {
       "epoch": 0.17692645123556075,
-      "grad_norm": 2.5,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.454,
+      "loss": 1.5982,
       "step": 10890
     },
     {
       "epoch": 0.17708891813293043,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.5746,
       "step": 10900
     },
     {
       "epoch": 0.17725138503030008,
-      "grad_norm": 2.25,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4477,
+      "loss": 1.5588,
       "step": 10910
     },
     {
       "epoch": 0.17741385192766973,
-      "grad_norm": 2.25,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4325,
+      "loss": 1.6331,
       "step": 10920
     },
     {
       "epoch": 0.1775763188250394,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4381,
+      "loss": 1.5897,
       "step": 10930
     },
     {
       "epoch": 0.17773878572240906,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4365,
+      "loss": 1.607,
       "step": 10940
     },
     {
       "epoch": 0.1779012526197787,
-      "grad_norm": 2.375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.5811,
       "step": 10950
     },
     {
       "epoch": 0.1780637195171484,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4605,
+      "loss": 1.5959,
       "step": 10960
     },
     {
       "epoch": 0.17822618641451804,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4417,
+      "loss": 1.5756,
       "step": 10970
     },
     {
       "epoch": 0.1783886533118877,
-      "grad_norm": 2.0625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4546,
+      "loss": 1.6133,
       "step": 10980
     },
     {
       "epoch": 0.17855112020925737,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4413,
+      "loss": 1.528,
       "step": 10990
     },
     {
       "epoch": 0.17871358710662702,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4407,
+      "loss": 1.55,
       "step": 11000
     },
     {
       "epoch": 0.1788760540039967,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4372,
+      "loss": 1.5719,
       "step": 11010
     },
     {
       "epoch": 0.17903852090136635,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4408,
+      "loss": 1.5798,
       "step": 11020
     },
     {
       "epoch": 0.179200987798736,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.438,
+      "loss": 1.6016,
       "step": 11030
     },
     {
       "epoch": 0.17936345469610568,
-      "grad_norm": 1.921875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4404,
+      "loss": 1.5662,
       "step": 11040
     },
     {
       "epoch": 0.17952592159347533,
-      "grad_norm": 4.8125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4457,
+      "loss": 1.6069,
       "step": 11050
     },
     {
       "epoch": 0.17968838849084498,
-      "grad_norm": 2.625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.5521,
       "step": 11060
     },
     {
       "epoch": 0.17985085538821466,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4499,
+      "loss": 1.5546,
       "step": 11070
     },
     {
       "epoch": 0.1800133222855843,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4367,
+      "loss": 1.5833,
       "step": 11080
     },
     {
       "epoch": 0.18017578918295396,
-      "grad_norm": 3.828125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4587,
+      "loss": 1.563,
       "step": 11090
     },
     {
       "epoch": 0.18033825608032364,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4396,
+      "loss": 1.5777,
       "step": 11100
     },
     {
       "epoch": 0.1805007229776933,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4268,
+      "loss": 1.5865,
       "step": 11110
     },
     {
       "epoch": 0.18066318987506297,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4474,
+      "loss": 1.5607,
       "step": 11120
     },
     {
       "epoch": 0.18082565677243262,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.5661,
       "step": 11130
     },
     {
       "epoch": 0.18098812366980227,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4511,
+      "loss": 1.5156,
       "step": 11140
     },
     {
       "epoch": 0.18115059056717195,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4517,
+      "loss": 1.5795,
       "step": 11150
     },
     {
       "epoch": 0.1813130574645416,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4355,
+      "loss": 1.5632,
       "step": 11160
     },
     {
       "epoch": 0.18147552436191125,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4355,
+      "loss": 1.597,
       "step": 11170
     },
     {
       "epoch": 0.18163799125928093,
-      "grad_norm": 2.375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4316,
+      "loss": 1.6249,
       "step": 11180
     },
     {
       "epoch": 0.18180045815665058,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.5736,
       "step": 11190
     },
     {
       "epoch": 0.18196292505402023,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.5955,
       "step": 11200
     },
     {
       "epoch": 0.1821253919513899,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4441,
+      "loss": 1.5365,
       "step": 11210
     },
     {
       "epoch": 0.18228785884875956,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.5259,
       "step": 11220
     },
     {
       "epoch": 0.18245032574612924,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4493,
+      "loss": 1.5787,
       "step": 11230
     },
     {
       "epoch": 0.1826127926434989,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.4457,
+      "loss": 1.5713,
       "step": 11240
     },
     {
       "epoch": 0.18277525954086854,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4527,
+      "loss": 1.5696,
       "step": 11250
     },
     {
       "epoch": 0.18293772643823822,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4563,
+      "loss": 1.5844,
       "step": 11260
     },
     {
       "epoch": 0.18310019333560787,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4453,
+      "loss": 1.5853,
       "step": 11270
     },
     {
       "epoch": 0.18326266023297752,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.5538,
       "step": 11280
     },
     {
       "epoch": 0.1834251271303472,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4332,
+      "loss": 1.6148,
       "step": 11290
     },
     {
       "epoch": 0.18358759402771685,
-      "grad_norm": 2.375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.438,
+      "loss": 1.5841,
       "step": 11300
     },
     {
       "epoch": 0.1837500609250865,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4295,
+      "loss": 1.5642,
       "step": 11310
     },
     {
       "epoch": 0.18391252782245618,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4722,
+      "loss": 1.5548,
       "step": 11320
     },
     {
       "epoch": 0.18407499471982583,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4432,
+      "loss": 1.5881,
       "step": 11330
     },
     {
       "epoch": 0.1842374616171955,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4464,
+      "loss": 1.586,
       "step": 11340
     },
     {
       "epoch": 0.18439992851456516,
-      "grad_norm": 2.9375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4619,
+      "loss": 1.5335,
       "step": 11350
     },
     {
       "epoch": 0.1845623954119348,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.453,
+      "loss": 1.5759,
       "step": 11360
     },
     {
       "epoch": 0.1847248623093045,
-      "grad_norm": 2.5,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4477,
+      "loss": 1.5715,
       "step": 11370
     },
     {
       "epoch": 0.18488732920667414,
-      "grad_norm": 3.890625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.5577,
       "step": 11380
     },
     {
       "epoch": 0.1850497961040438,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.5534,
       "step": 11390
     },
     {
       "epoch": 0.18521226300141347,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4564,
+      "loss": 1.5708,
       "step": 11400
     },
     {
       "epoch": 0.18537472989878312,
-      "grad_norm": 2.71875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4514,
+      "loss": 1.5579,
       "step": 11410
     },
     {
       "epoch": 0.18553719679615277,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.5826,
       "step": 11420
     },
     {
       "epoch": 0.18569966369352245,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4397,
+      "loss": 1.5965,
       "step": 11430
     },
     {
       "epoch": 0.1858621305908921,
-      "grad_norm": 3.5,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.45,
+      "loss": 1.6008,
       "step": 11440
     },
     {
       "epoch": 0.18602459748826178,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4463,
+      "loss": 1.575,
       "step": 11450
     },
     {
       "epoch": 0.18618706438563143,
-      "grad_norm": 3.125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.5787,
       "step": 11460
     },
     {
       "epoch": 0.18634953128300108,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4329,
+      "loss": 1.5523,
       "step": 11470
     },
     {
       "epoch": 0.18651199818037076,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.5426,
       "step": 11480
     },
     {
       "epoch": 0.1866744650777404,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.433,
+      "loss": 1.5845,
       "step": 11490
     },
     {
       "epoch": 0.18683693197511006,
-      "grad_norm": 2.125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4398,
+      "loss": 1.5756,
       "step": 11500
     },
     {
       "epoch": 0.18699939887247974,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.4499,
+      "loss": 1.616,
       "step": 11510
     },
     {
       "epoch": 0.1871618657698494,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4366,
+      "loss": 1.5199,
       "step": 11520
     },
     {
       "epoch": 0.18732433266721907,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4455,
+      "loss": 1.5681,
       "step": 11530
     },
     {
       "epoch": 0.18748679956458872,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4523,
+      "loss": 1.6279,
       "step": 11540
     },
     {
       "epoch": 0.18764926646195837,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.44,
+      "loss": 1.542,
       "step": 11550
     },
     {
       "epoch": 0.18781173335932805,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5176,
       "step": 11560
     },
     {
       "epoch": 0.1879742002566977,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4208,
+      "loss": 1.5227,
       "step": 11570
     },
     {
       "epoch": 0.18813666715406735,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4163,
+      "loss": 1.5583,
       "step": 11580
     },
     {
       "epoch": 0.18829913405143703,
-      "grad_norm": 3.5,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4395,
+      "loss": 1.5601,
       "step": 11590
     },
     {
       "epoch": 0.18846160094880668,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.6072,
       "step": 11600
     },
     {
       "epoch": 0.18862406784617633,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4462,
+      "loss": 1.5582,
       "step": 11610
     },
     {
       "epoch": 0.188786534743546,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.445,
+      "loss": 1.5386,
       "step": 11620
     },
     {
       "epoch": 0.18894900164091566,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4308,
+      "loss": 1.5205,
       "step": 11630
     },
     {
       "epoch": 0.18911146853828534,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4294,
+      "loss": 1.5279,
       "step": 11640
     },
     {
       "epoch": 0.189273935435655,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.5826,
       "step": 11650
     },
     {
       "epoch": 0.18943640233302464,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.5711,
       "step": 11660
     },
     {
       "epoch": 0.18959886923039432,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4511,
+      "loss": 1.5584,
       "step": 11670
     },
     {
       "epoch": 0.18976133612776397,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.6076,
       "step": 11680
     },
     {
       "epoch": 0.18992380302513362,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4501,
+      "loss": 1.5426,
       "step": 11690
     },
     {
       "epoch": 0.1900862699225033,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4325,
+      "loss": 1.5592,
       "step": 11700
     },
     {
       "epoch": 0.19024873681987295,
-      "grad_norm": 2.75,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4429,
+      "loss": 1.5407,
       "step": 11710
     },
     {
       "epoch": 0.1904112037172426,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4617,
+      "loss": 1.5996,
       "step": 11720
     },
     {
       "epoch": 0.19057367061461228,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.465,
+      "loss": 1.5933,
       "step": 11730
     },
     {
       "epoch": 0.19073613751198193,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4555,
+      "loss": 1.5513,
       "step": 11740
     },
     {
       "epoch": 0.1908986044093516,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4207,
+      "loss": 1.5989,
       "step": 11750
     },
     {
       "epoch": 0.19106107130672126,
-      "grad_norm": 2.21875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.4547,
+      "loss": 1.5471,
       "step": 11760
     },
     {
       "epoch": 0.1912235382040909,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4359,
+      "loss": 1.5721,
       "step": 11770
     },
     {
       "epoch": 0.1913860051014606,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4418,
+      "loss": 1.5747,
       "step": 11780
     },
     {
       "epoch": 0.19154847199883024,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4287,
+      "loss": 1.5992,
       "step": 11790
     },
     {
       "epoch": 0.1917109388961999,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4542,
+      "loss": 1.5995,
       "step": 11800
     },
     {
       "epoch": 0.19187340579356957,
-      "grad_norm": 2.125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4651,
+      "loss": 1.5545,
       "step": 11810
     },
     {
       "epoch": 0.19203587269093922,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.5672,
       "step": 11820
     },
     {
       "epoch": 0.19219833958830887,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4368,
+      "loss": 1.5559,
       "step": 11830
     },
     {
       "epoch": 0.19236080648567855,
-      "grad_norm": 3.4375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4352,
+      "loss": 1.6185,
       "step": 11840
     },
     {
       "epoch": 0.1925232733830482,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.451,
+      "loss": 1.6002,
       "step": 11850
     },
     {
       "epoch": 0.19268574028041788,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4379,
+      "loss": 1.5527,
       "step": 11860
     },
     {
       "epoch": 0.19284820717778753,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4518,
+      "loss": 1.5548,
       "step": 11870
     },
     {
       "epoch": 0.19301067407515718,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4371,
+      "loss": 1.5461,
       "step": 11880
     },
     {
       "epoch": 0.19317314097252686,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.446,
+      "loss": 1.5905,
       "step": 11890
     },
     {
       "epoch": 0.1933356078698965,
-      "grad_norm": 3.765625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5334,
       "step": 11900
     },
     {
       "epoch": 0.19349807476726616,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4327,
+      "loss": 1.5428,
       "step": 11910
     },
     {
       "epoch": 0.19366054166463584,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4191,
+      "loss": 1.5317,
       "step": 11920
     },
     {
       "epoch": 0.1938230085620055,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4311,
+      "loss": 1.5915,
       "step": 11930
     },
     {
       "epoch": 0.19398547545937514,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4506,
+      "loss": 1.5475,
       "step": 11940
     },
     {
       "epoch": 0.19414794235674482,
-      "grad_norm": 2.0,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4369,
+      "loss": 1.5738,
       "step": 11950
     },
     {
       "epoch": 0.19431040925411447,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4348,
+      "loss": 1.5448,
       "step": 11960
     },
     {
       "epoch": 0.19447287615148415,
-      "grad_norm": 2.875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4261,
+      "loss": 1.5517,
       "step": 11970
     },
     {
       "epoch": 0.1946353430488538,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4307,
+      "loss": 1.5829,
       "step": 11980
     },
     {
       "epoch": 0.19479780994622345,
-      "grad_norm": 2.078125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4545,
+      "loss": 1.5892,
       "step": 11990
     },
     {
       "epoch": 0.19496027684359313,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.5661,
       "step": 12000
     },
     {
       "epoch": 0.19512274374096278,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4484,
+      "loss": 1.5611,
       "step": 12010
     },
     {
       "epoch": 0.19528521063833243,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4434,
+      "loss": 1.5613,
       "step": 12020
     },
     {
       "epoch": 0.1954476775357021,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.418,
+      "loss": 1.5734,
       "step": 12030
     },
     {
       "epoch": 0.19561014443307176,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4508,
+      "loss": 1.5719,
       "step": 12040
     },
     {
       "epoch": 0.1957726113304414,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4759,
+      "loss": 1.5725,
       "step": 12050
     },
     {
       "epoch": 0.1959350782278111,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4504,
+      "loss": 1.5821,
       "step": 12060
     },
     {
       "epoch": 0.19609754512518074,
-      "grad_norm": 2.625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.451,
+      "loss": 1.5295,
       "step": 12070
     },
     {
       "epoch": 0.19626001202255042,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4539,
+      "loss": 1.5623,
       "step": 12080
     },
     {
       "epoch": 0.19642247891992007,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4553,
+      "loss": 1.5906,
       "step": 12090
     },
     {
       "epoch": 0.19658494581728972,
-      "grad_norm": 2.296875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4455,
+      "loss": 1.576,
       "step": 12100
     },
     {
       "epoch": 0.1967474127146594,
-      "grad_norm": 5.8125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4486,
+      "loss": 1.5799,
       "step": 12110
     },
     {
       "epoch": 0.19690987961202905,
-      "grad_norm": 1.640625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4316,
+      "loss": 1.5813,
       "step": 12120
     },
     {
       "epoch": 0.1970723465093987,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.4919,
       "step": 12130
     },
     {
       "epoch": 0.19723481340676838,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4377,
+      "loss": 1.5419,
       "step": 12140
     },
     {
       "epoch": 0.19739728030413803,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4158,
+      "loss": 1.5881,
       "step": 12150
     },
     {
       "epoch": 0.19755974720150768,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.5644,
       "step": 12160
     },
     {
       "epoch": 0.19772221409887736,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4353,
+      "loss": 1.527,
       "step": 12170
     },
     {
       "epoch": 0.197884680996247,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4591,
+      "loss": 1.5429,
       "step": 12180
     },
     {
       "epoch": 0.1980471478936167,
-      "grad_norm": 2.375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4426,
+      "loss": 1.5201,
       "step": 12190
     },
     {
       "epoch": 0.19820961479098634,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.5388,
       "step": 12200
     },
     {
       "epoch": 0.198372081688356,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4441,
+      "loss": 1.5629,
       "step": 12210
     },
     {
       "epoch": 0.19853454858572567,
-      "grad_norm": 2.875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4386,
+      "loss": 1.6031,
       "step": 12220
     },
     {
       "epoch": 0.19869701548309532,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.446,
+      "loss": 1.5848,
       "step": 12230
     },
     {
       "epoch": 0.19885948238046497,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.558,
       "step": 12240
     },
     {
       "epoch": 0.19902194927783465,
-      "grad_norm": 3.125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4488,
+      "loss": 1.5594,
       "step": 12250
     },
     {
       "epoch": 0.1991844161752043,
-      "grad_norm": 4.25,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4481,
+      "loss": 1.574,
       "step": 12260
     },
     {
       "epoch": 0.19934688307257395,
-      "grad_norm": 2.625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.5561,
       "step": 12270
     },
     {
       "epoch": 0.19950934996994363,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.6061,
       "step": 12280
     },
     {
       "epoch": 0.19967181686731328,
-      "grad_norm": 2.5,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4465,
+      "loss": 1.5778,
       "step": 12290
     },
     {
       "epoch": 0.19983428376468296,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4376,
+      "loss": 1.5639,
       "step": 12300
     },
     {
       "epoch": 0.1999967506620526,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4288,
+      "loss": 1.5342,
       "step": 12310
     },
     {
       "epoch": 0.20015921755942226,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.5258,
       "step": 12320
     },
     {
       "epoch": 0.20032168445679194,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.5149,
       "step": 12330
     },
     {
       "epoch": 0.2004841513541616,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.5339,
       "step": 12340
     },
     {
       "epoch": 0.20064661825153124,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.5974,
       "step": 12350
     },
     {
       "epoch": 0.20080908514890092,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.447,
+      "loss": 1.5826,
       "step": 12360
     },
     {
       "epoch": 0.20097155204627057,
-      "grad_norm": 1.75,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.5239,
       "step": 12370
     },
     {
       "epoch": 0.20113401894364025,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4378,
+      "loss": 1.5602,
       "step": 12380
     },
     {
       "epoch": 0.2012964858410099,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4299,
+      "loss": 1.5931,
       "step": 12390
     },
     {
       "epoch": 0.20145895273837955,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4388,
+      "loss": 1.5505,
       "step": 12400
     },
     {
       "epoch": 0.20162141963574923,
-      "grad_norm": 2.75,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.5812,
       "step": 12410
     },
     {
       "epoch": 0.20178388653311888,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4254,
+      "loss": 1.5695,
       "step": 12420
     },
     {
       "epoch": 0.20194635343048853,
-      "grad_norm": 4.9375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4269,
+      "loss": 1.5237,
       "step": 12430
     },
     {
       "epoch": 0.2021088203278582,
-      "grad_norm": 3.8125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.5411,
       "step": 12440
     },
     {
       "epoch": 0.20227128722522786,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.5767,
       "step": 12450
     },
     {
       "epoch": 0.2024337541225975,
-      "grad_norm": 2.4375,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4495,
+      "loss": 1.5538,
       "step": 12460
     },
     {
       "epoch": 0.2025962210199672,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4321,
+      "loss": 1.5074,
       "step": 12470
     },
     {
       "epoch": 0.20275868791733684,
-      "grad_norm": 2.625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.5713,
       "step": 12480
     },
     {
       "epoch": 0.20292115481470652,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4318,
+      "loss": 1.5508,
       "step": 12490
     },
     {
       "epoch": 0.20308362171207617,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4343,
+      "loss": 1.5424,
       "step": 12500
     },
     {
       "epoch": 0.20324608860944582,
-      "grad_norm": 2.171875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4607,
+      "loss": 1.5601,
       "step": 12510
     },
     {
       "epoch": 0.2034085555068155,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4439,
+      "loss": 1.5486,
       "step": 12520
     },
     {
       "epoch": 0.20357102240418515,
-      "grad_norm": 2.875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4327,
+      "loss": 1.5955,
       "step": 12530
     },
     {
       "epoch": 0.2037334893015548,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4481,
+      "loss": 1.5974,
       "step": 12540
     },
     {
       "epoch": 0.20389595619892448,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4264,
+      "loss": 1.565,
       "step": 12550
     },
     {
       "epoch": 0.20405842309629413,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4361,
+      "loss": 1.5603,
       "step": 12560
     },
     {
       "epoch": 0.20422088999366378,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.433,
+      "loss": 1.5816,
       "step": 12570
     },
     {
       "epoch": 0.20438335689103346,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4326,
+      "loss": 1.5299,
       "step": 12580
     },
     {
       "epoch": 0.2045458237884031,
-      "grad_norm": 2.4375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.4413,
+      "loss": 1.6185,
       "step": 12590
     },
     {
       "epoch": 0.2047082906857728,
-      "grad_norm": 3.25,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4377,
+      "loss": 1.5718,
       "step": 12600
     },
     {
       "epoch": 0.20487075758314244,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4408,
+      "loss": 1.5976,
       "step": 12610
     },
     {
       "epoch": 0.2050332244805121,
-      "grad_norm": 2.25,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5367,
       "step": 12620
     },
     {
       "epoch": 0.20519569137788177,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.5422,
       "step": 12630
     },
     {
       "epoch": 0.20535815827525142,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.5074,
       "step": 12640
     },
     {
       "epoch": 0.20552062517262107,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.5782,
       "step": 12650
     },
     {
       "epoch": 0.20568309206999075,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4316,
+      "loss": 1.6024,
       "step": 12660
     },
     {
       "epoch": 0.2058455589673604,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.5418,
       "step": 12670
     },
     {
       "epoch": 0.20600802586473005,
-      "grad_norm": 2.25,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4526,
+      "loss": 1.5363,
       "step": 12680
     },
     {
       "epoch": 0.20617049276209973,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4358,
+      "loss": 1.5603,
       "step": 12690
     },
     {
       "epoch": 0.20633295965946938,
-      "grad_norm": 4.34375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4388,
+      "loss": 1.5067,
       "step": 12700
     },
     {
       "epoch": 0.20649542655683906,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.5693,
       "step": 12710
     },
     {
       "epoch": 0.2066578934542087,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4378,
+      "loss": 1.5218,
       "step": 12720
     },
     {
       "epoch": 0.20682036035157836,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.5173,
       "step": 12730
     },
     {
       "epoch": 0.20698282724894804,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4584,
+      "loss": 1.5761,
       "step": 12740
     },
     {
       "epoch": 0.2071452941463177,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4215,
+      "loss": 1.5427,
       "step": 12750
     },
     {
       "epoch": 0.20730776104368734,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4247,
+      "loss": 1.5193,
       "step": 12760
     },
     {
       "epoch": 0.20747022794105702,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.5966,
       "step": 12770
     },
     {
       "epoch": 0.20763269483842667,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4367,
+      "loss": 1.6111,
       "step": 12780
     },
     {
       "epoch": 0.20779516173579632,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.5583,
       "step": 12790
     },
     {
       "epoch": 0.207957628633166,
-      "grad_norm": 2.75,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4422,
+      "loss": 1.5449,
       "step": 12800
     },
     {
       "epoch": 0.20812009553053565,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.5946,
       "step": 12810
     },
     {
       "epoch": 0.20828256242790533,
-      "grad_norm": 2.515625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4302,
+      "loss": 1.596,
       "step": 12820
     },
     {
       "epoch": 0.20844502932527498,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4217,
+      "loss": 1.5693,
       "step": 12830
     },
     {
       "epoch": 0.20860749622264463,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.5763,
       "step": 12840
     },
     {
       "epoch": 0.2087699631200143,
-      "grad_norm": 2.875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.5432,
       "step": 12850
     },
     {
       "epoch": 0.20893243001738396,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4323,
+      "loss": 1.5426,
       "step": 12860
     },
     {
       "epoch": 0.2090948969147536,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4378,
+      "loss": 1.5184,
       "step": 12870
     },
     {
       "epoch": 0.2092573638121233,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4468,
+      "loss": 1.5405,
       "step": 12880
     },
     {
       "epoch": 0.20941983070949294,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.5935,
       "step": 12890
     },
     {
       "epoch": 0.2095822976068626,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4233,
+      "loss": 1.5428,
       "step": 12900
     },
     {
       "epoch": 0.20974476450423227,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.568,
       "step": 12910
     },
     {
       "epoch": 0.20990723140160192,
-      "grad_norm": 3.84375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.5881,
       "step": 12920
     },
     {
       "epoch": 0.2100696982989716,
-      "grad_norm": 2.125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.5128,
       "step": 12930
     },
     {
       "epoch": 0.21023216519634125,
-      "grad_norm": 2.3125,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.5317,
       "step": 12940
     },
     {
       "epoch": 0.2103946320937109,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.5407,
       "step": 12950
     },
     {
       "epoch": 0.21055709899108058,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4434,
+      "loss": 1.5104,
       "step": 12960
     },
     {
       "epoch": 0.21071956588845023,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4326,
+      "loss": 1.5349,
       "step": 12970
     },
     {
       "epoch": 0.21088203278581988,
-      "grad_norm": 2.65625,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4584,
+      "loss": 1.6088,
       "step": 12980
     },
     {
       "epoch": 0.21104449968318956,
-      "grad_norm": 3.0,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4284,
+      "loss": 1.5076,
       "step": 12990
     },
     {
       "epoch": 0.2112069665805592,
-      "grad_norm": 2.03125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.5835,
       "step": 13000
     },
     {
       "epoch": 0.21136943347792886,
-      "grad_norm": 2.0,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.5521,
       "step": 13010
     },
     {
       "epoch": 0.21153190037529854,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4446,
+      "loss": 1.5981,
       "step": 13020
     },
     {
       "epoch": 0.2116943672726682,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4325,
+      "loss": 1.5009,
       "step": 13030
     },
     {
       "epoch": 0.21185683417003787,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4406,
+      "loss": 1.4868,
       "step": 13040
     },
     {
       "epoch": 0.21201930106740752,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4484,
+      "loss": 1.5738,
       "step": 13050
     },
     {
       "epoch": 0.21218176796477717,
-      "grad_norm": 2.0,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.5363,
       "step": 13060
     },
     {
       "epoch": 0.21234423486214685,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4221,
+      "loss": 1.5823,
       "step": 13070
     },
     {
       "epoch": 0.2125067017595165,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4465,
+      "loss": 1.5669,
       "step": 13080
     },
     {
       "epoch": 0.21266916865688615,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.5662,
       "step": 13090
     },
     {
       "epoch": 0.21283163555425583,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.5684,
       "step": 13100
     },
     {
       "epoch": 0.21299410245162548,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4263,
+      "loss": 1.5333,
       "step": 13110
     },
     {
       "epoch": 0.21315656934899513,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4494,
+      "loss": 1.5792,
       "step": 13120
     },
     {
       "epoch": 0.2133190362463648,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4413,
+      "loss": 1.5424,
       "step": 13130
     },
     {
       "epoch": 0.21348150314373446,
-      "grad_norm": 1.9609375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.5279,
       "step": 13140
     },
     {
       "epoch": 0.21364397004110414,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.454,
+      "loss": 1.5204,
       "step": 13150
     },
     {
       "epoch": 0.2138064369384738,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4443,
+      "loss": 1.5349,
       "step": 13160
     },
     {
       "epoch": 0.21396890383584344,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4483,
+      "loss": 1.5225,
       "step": 13170
     },
     {
       "epoch": 0.21413137073321312,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4343,
+      "loss": 1.5671,
       "step": 13180
     },
     {
       "epoch": 0.21429383763058277,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4474,
+      "loss": 1.5126,
       "step": 13190
     },
     {
       "epoch": 0.21445630452795242,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4424,
+      "loss": 1.5537,
       "step": 13200
     },
     {
       "epoch": 0.2146187714253221,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4294,
+      "loss": 1.5054,
       "step": 13210
     },
     {
       "epoch": 0.21478123832269175,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.5661,
       "step": 13220
     },
     {
       "epoch": 0.21494370522006143,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4429,
+      "loss": 1.5309,
       "step": 13230
     },
     {
       "epoch": 0.21510617211743108,
-      "grad_norm": 2.75,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.57,
       "step": 13240
     },
     {
       "epoch": 0.21526863901480073,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4493,
+      "loss": 1.5834,
       "step": 13250
     },
     {
       "epoch": 0.2154311059121704,
-      "grad_norm": 3.375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.5554,
       "step": 13260
     },
     {
       "epoch": 0.21559357280954006,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.5343,
       "step": 13270
     },
     {
       "epoch": 0.2157560397069097,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4187,
+      "loss": 1.5778,
       "step": 13280
     },
     {
       "epoch": 0.2159185066042794,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4343,
+      "loss": 1.5551,
       "step": 13290
     },
     {
       "epoch": 0.21608097350164904,
-      "grad_norm": 3.484375,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4362,
+      "loss": 1.5394,
       "step": 13300
     },
     {
       "epoch": 0.2162434403990187,
-      "grad_norm": 1.53125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.433,
+      "loss": 1.5308,
       "step": 13310
     },
     {
       "epoch": 0.21640590729638837,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.437,
+      "loss": 1.5041,
       "step": 13320
     },
     {
       "epoch": 0.21656837419375802,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4424,
+      "loss": 1.5929,
       "step": 13330
     },
     {
       "epoch": 0.2167308410911277,
-      "grad_norm": 3.125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4454,
+      "loss": 1.5613,
       "step": 13340
     },
     {
       "epoch": 0.21689330798849735,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4478,
+      "loss": 1.5054,
       "step": 13350
     },
     {
       "epoch": 0.217055774885867,
-      "grad_norm": 2.5,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.5049,
       "step": 13360
     },
     {
       "epoch": 0.21721824178323668,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.443,
+      "loss": 1.5999,
       "step": 13370
     },
     {
       "epoch": 0.21738070868060633,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4415,
+      "loss": 1.5648,
       "step": 13380
     },
     {
       "epoch": 0.21754317557797598,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4191,
+      "loss": 1.5525,
       "step": 13390
     },
     {
       "epoch": 0.21770564247534566,
-      "grad_norm": 3.109375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4507,
+      "loss": 1.5237,
       "step": 13400
     },
     {
       "epoch": 0.2178681093727153,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.5539,
       "step": 13410
     },
     {
       "epoch": 0.21803057627008496,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4425,
+      "loss": 1.577,
       "step": 13420
     },
     {
       "epoch": 0.21819304316745464,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4505,
+      "loss": 1.5558,
       "step": 13430
     },
     {
       "epoch": 0.2183555100648243,
-      "grad_norm": 1.9140625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4324,
+      "loss": 1.5576,
       "step": 13440
     },
     {
       "epoch": 0.21851797696219397,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4492,
+      "loss": 1.5274,
       "step": 13450
     },
     {
       "epoch": 0.21868044385956362,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4365,
+      "loss": 1.5834,
       "step": 13460
     },
     {
       "epoch": 0.21884291075693327,
-      "grad_norm": 1.953125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4507,
+      "loss": 1.5326,
       "step": 13470
     },
     {
       "epoch": 0.21900537765430295,
-      "grad_norm": 1.7421875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.5601,
       "step": 13480
     },
     {
       "epoch": 0.2191678445516726,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4291,
+      "loss": 1.5652,
       "step": 13490
     },
     {
       "epoch": 0.21933031144904225,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.5759,
       "step": 13500
     },
     {
       "epoch": 0.21949277834641193,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.441,
+      "loss": 1.5053,
       "step": 13510
     },
     {
       "epoch": 0.21965524524378158,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4356,
+      "loss": 1.5699,
       "step": 13520
     },
     {
       "epoch": 0.21981771214115123,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4418,
+      "loss": 1.5411,
       "step": 13530
     },
     {
       "epoch": 0.2199801790385209,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.526,
       "step": 13540
     },
     {
       "epoch": 0.22014264593589056,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4355,
+      "loss": 1.5286,
       "step": 13550
     },
     {
       "epoch": 0.22030511283326024,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4445,
+      "loss": 1.5361,
       "step": 13560
     },
     {
       "epoch": 0.2204675797306299,
-      "grad_norm": 4.125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4314,
+      "loss": 1.5742,
       "step": 13570
     },
     {
       "epoch": 0.22063004662799954,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.5498,
       "step": 13580
     },
     {
       "epoch": 0.22079251352536922,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.5186,
       "step": 13590
     },
     {
       "epoch": 0.22095498042273887,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4529,
+      "loss": 1.4788,
       "step": 13600
     },
     {
       "epoch": 0.22111744732010852,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.43,
+      "loss": 1.5298,
       "step": 13610
     },
     {
       "epoch": 0.2212799142174782,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4459,
+      "loss": 1.6064,
       "step": 13620
     },
     {
       "epoch": 0.22144238111484785,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.5041,
       "step": 13630
     },
     {
       "epoch": 0.2216048480122175,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4464,
+      "loss": 1.5445,
       "step": 13640
     },
     {
       "epoch": 0.22176731490958718,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4327,
+      "loss": 1.5594,
       "step": 13650
     },
     {
       "epoch": 0.22192978180695683,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4595,
+      "loss": 1.5832,
       "step": 13660
     },
     {
       "epoch": 0.2220922487043265,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4431,
+      "loss": 1.546,
       "step": 13670
     },
     {
       "epoch": 0.22225471560169616,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4404,
+      "loss": 1.5999,
       "step": 13680
     },
     {
       "epoch": 0.2224171824990658,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4584,
+      "loss": 1.5238,
       "step": 13690
     },
     {
       "epoch": 0.22257964939643549,
-      "grad_norm": 2.5625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4411,
+      "loss": 1.5298,
       "step": 13700
     },
     {
       "epoch": 0.22274211629380514,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.5395,
       "step": 13710
     },
     {
       "epoch": 0.2229045831911748,
-      "grad_norm": 2.625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.446,
+      "loss": 1.5235,
       "step": 13720
     },
     {
       "epoch": 0.22306705008854447,
-      "grad_norm": 3.703125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4297,
+      "loss": 1.5215,
       "step": 13730
     },
     {
       "epoch": 0.22322951698591412,
-      "grad_norm": 3.890625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4327,
+      "loss": 1.5754,
       "step": 13740
     },
     {
       "epoch": 0.22339198388328377,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4334,
+      "loss": 1.5917,
       "step": 13750
     },
     {
       "epoch": 0.22355445078065345,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4409,
+      "loss": 1.5877,
       "step": 13760
     },
     {
       "epoch": 0.2237169176780231,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4287,
+      "loss": 1.5449,
       "step": 13770
     },
     {
       "epoch": 0.22387938457539278,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4311,
+      "loss": 1.5343,
       "step": 13780
     },
     {
       "epoch": 0.22404185147276243,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.428,
+      "loss": 1.53,
       "step": 13790
     },
     {
       "epoch": 0.22420431837013208,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.5284,
       "step": 13800
     },
     {
       "epoch": 0.22436678526750176,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.442,
+      "loss": 1.5441,
       "step": 13810
     },
     {
       "epoch": 0.2245292521648714,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.445,
+      "loss": 1.5479,
       "step": 13820
     },
     {
       "epoch": 0.22469171906224106,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4277,
+      "loss": 1.5845,
       "step": 13830
     },
     {
       "epoch": 0.22485418595961074,
-      "grad_norm": 2.25,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5273,
       "step": 13840
     },
     {
       "epoch": 0.2250166528569804,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.5499,
       "step": 13850
     },
     {
       "epoch": 0.22517911975435004,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.5252,
       "step": 13860
     },
     {
       "epoch": 0.22534158665171972,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4393,
+      "loss": 1.53,
       "step": 13870
     },
     {
       "epoch": 0.22550405354908937,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4282,
+      "loss": 1.5101,
       "step": 13880
     },
     {
       "epoch": 0.22566652044645905,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4273,
+      "loss": 1.5513,
       "step": 13890
     },
     {
       "epoch": 0.2258289873438287,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5541,
       "step": 13900
     },
     {
       "epoch": 0.22599145424119835,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4316,
+      "loss": 1.5856,
       "step": 13910
     },
     {
       "epoch": 0.22615392113856803,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.5368,
       "step": 13920
     },
     {
       "epoch": 0.22631638803593768,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4305,
+      "loss": 1.5605,
       "step": 13930
     },
     {
       "epoch": 0.22647885493330733,
-      "grad_norm": 1.8828125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4511,
+      "loss": 1.5306,
       "step": 13940
     },
     {
       "epoch": 0.226641321830677,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4496,
+      "loss": 1.5558,
       "step": 13950
     },
     {
       "epoch": 0.22680378872804666,
-      "grad_norm": 2.125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.5543,
       "step": 13960
     },
     {
       "epoch": 0.2269662556254163,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.445,
+      "loss": 1.5204,
       "step": 13970
     },
     {
       "epoch": 0.227128722522786,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4512,
+      "loss": 1.5146,
       "step": 13980
     },
     {
       "epoch": 0.22729118942015564,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4322,
+      "loss": 1.5173,
       "step": 13990
     },
     {
       "epoch": 0.22745365631752532,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4243,
+      "loss": 1.4996,
       "step": 14000
     },
     {
       "epoch": 0.22761612321489497,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4546,
+      "loss": 1.5538,
       "step": 14010
     },
     {
       "epoch": 0.22777859011226462,
-      "grad_norm": 2.5,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4285,
+      "loss": 1.5048,
       "step": 14020
     },
     {
       "epoch": 0.2279410570096343,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.4804,
       "step": 14030
     },
     {
       "epoch": 0.22810352390700395,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4495,
+      "loss": 1.5798,
       "step": 14040
     },
     {
       "epoch": 0.2282659908043736,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.5732,
       "step": 14050
     },
     {
       "epoch": 0.22842845770174328,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4397,
+      "loss": 1.5512,
       "step": 14060
     },
     {
       "epoch": 0.22859092459911293,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4303,
+      "loss": 1.5779,
       "step": 14070
     },
     {
       "epoch": 0.2287533914964826,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4502,
+      "loss": 1.4844,
       "step": 14080
     },
     {
       "epoch": 0.22891585839385226,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4543,
+      "loss": 1.5364,
       "step": 14090
     },
     {
       "epoch": 0.2290783252912219,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.5507,
       "step": 14100
     },
     {
       "epoch": 0.22924079218859159,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4585,
+      "loss": 1.5351,
       "step": 14110
     },
     {
       "epoch": 0.22940325908596124,
-      "grad_norm": 1.6171875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4181,
+      "loss": 1.5292,
       "step": 14120
     },
     {
       "epoch": 0.2295657259833309,
-      "grad_norm": 1.71875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.6123,
       "step": 14130
     },
     {
       "epoch": 0.22972819288070057,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4491,
+      "loss": 1.5574,
       "step": 14140
     },
     {
       "epoch": 0.22989065977807022,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4268,
+      "loss": 1.5748,
       "step": 14150
     },
     {
       "epoch": 0.23005312667543987,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.5661,
       "step": 14160
     },
     {
       "epoch": 0.23021559357280955,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4511,
+      "loss": 1.5273,
       "step": 14170
     },
     {
       "epoch": 0.2303780604701792,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4309,
+      "loss": 1.5346,
       "step": 14180
     },
     {
       "epoch": 0.23054052736754888,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4482,
+      "loss": 1.5057,
       "step": 14190
     },
     {
       "epoch": 0.23070299426491853,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.5432,
       "step": 14200
     },
     {
       "epoch": 0.23086546116228818,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.527,
       "step": 14210
     },
     {
       "epoch": 0.23102792805965786,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.45,
+      "loss": 1.5543,
       "step": 14220
     },
     {
       "epoch": 0.2311903949570275,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4414,
+      "loss": 1.5364,
       "step": 14230
     },
     {
       "epoch": 0.23135286185439716,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.5684,
       "step": 14240
     },
     {
       "epoch": 0.23151532875176684,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4394,
+      "loss": 1.5679,
       "step": 14250
     },
     {
       "epoch": 0.2316777956491365,
-      "grad_norm": 2.5,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.5055,
       "step": 14260
     },
     {
       "epoch": 0.23184026254650614,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4393,
+      "loss": 1.5406,
       "step": 14270
     },
     {
       "epoch": 0.23200272944387582,
-      "grad_norm": 2.625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4195,
+      "loss": 1.5527,
       "step": 14280
     },
     {
       "epoch": 0.23216519634124547,
-      "grad_norm": 1.8359375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4377,
+      "loss": 1.5406,
       "step": 14290
     },
     {
       "epoch": 0.23232766323861515,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4217,
+      "loss": 1.536,
       "step": 14300
     },
     {
       "epoch": 0.2324901301359848,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4263,
+      "loss": 1.5468,
       "step": 14310
     },
     {
       "epoch": 0.23265259703335445,
-      "grad_norm": 2.125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4219,
+      "loss": 1.5173,
       "step": 14320
     },
     {
       "epoch": 0.23281506393072413,
-      "grad_norm": 2.125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4322,
+      "loss": 1.5849,
       "step": 14330
     },
     {
       "epoch": 0.23297753082809378,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4303,
+      "loss": 1.4761,
       "step": 14340
     },
     {
       "epoch": 0.23313999772546343,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.5545,
       "step": 14350
     },
     {
       "epoch": 0.2333024646228331,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.5946,
       "step": 14360
     },
     {
       "epoch": 0.23346493152020276,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4451,
+      "loss": 1.5888,
       "step": 14370
     },
     {
       "epoch": 0.2336273984175724,
-      "grad_norm": 2.125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.5123,
       "step": 14380
     },
     {
       "epoch": 0.23378986531494209,
-      "grad_norm": 2.5,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4203,
+      "loss": 1.5754,
       "step": 14390
     },
     {
       "epoch": 0.23395233221231174,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4458,
+      "loss": 1.5266,
       "step": 14400
     },
     {
       "epoch": 0.23411479910968142,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4435,
+      "loss": 1.4815,
       "step": 14410
     },
     {
       "epoch": 0.23427726600705107,
-      "grad_norm": 2.625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4398,
+      "loss": 1.5642,
       "step": 14420
     },
     {
       "epoch": 0.23443973290442072,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.5094,
       "step": 14430
     },
     {
       "epoch": 0.2346021998017904,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.437,
+      "loss": 1.6077,
       "step": 14440
     },
     {
       "epoch": 0.23476466669916005,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.433,
+      "loss": 1.5465,
       "step": 14450
     },
     {
       "epoch": 0.2349271335965297,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4337,
+      "loss": 1.5087,
       "step": 14460
     },
     {
       "epoch": 0.23508960049389938,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.5097,
       "step": 14470
     },
     {
       "epoch": 0.23525206739126903,
-      "grad_norm": 4.375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4215,
+      "loss": 1.5286,
       "step": 14480
     },
     {
       "epoch": 0.23541453428863868,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4479,
+      "loss": 1.5406,
       "step": 14490
     },
     {
       "epoch": 0.23557700118600836,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4424,
+      "loss": 1.5139,
       "step": 14500
     },
     {
       "epoch": 0.235739468083378,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.4664,
       "step": 14510
     },
     {
       "epoch": 0.23590193498074769,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4404,
+      "loss": 1.5942,
       "step": 14520
     },
     {
       "epoch": 0.23606440187811734,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4287,
+      "loss": 1.5085,
       "step": 14530
     },
     {
       "epoch": 0.236226868775487,
-      "grad_norm": 1.828125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4373,
+      "loss": 1.5347,
       "step": 14540
     },
     {
       "epoch": 0.23638933567285667,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.5161,
       "step": 14550
     },
     {
       "epoch": 0.23655180257022632,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.515,
       "step": 14560
     },
     {
       "epoch": 0.23671426946759597,
-      "grad_norm": 3.609375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4375,
+      "loss": 1.5024,
       "step": 14570
     },
     {
       "epoch": 0.23687673636496565,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.5444,
       "step": 14580
     },
     {
       "epoch": 0.2370392032623353,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5395,
       "step": 14590
     },
     {
       "epoch": 0.23720167015970495,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.5129,
       "step": 14600
     },
     {
       "epoch": 0.23736413705707463,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4215,
+      "loss": 1.5287,
       "step": 14610
     },
     {
       "epoch": 0.23752660395444428,
-      "grad_norm": 4.9375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4449,
+      "loss": 1.4947,
       "step": 14620
     },
     {
       "epoch": 0.23768907085181395,
-      "grad_norm": 2.625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.583,
       "step": 14630
     },
     {
       "epoch": 0.2378515377491836,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4407,
+      "loss": 1.5387,
       "step": 14640
     },
     {
       "epoch": 0.23801400464655326,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.5224,
       "step": 14650
     },
     {
       "epoch": 0.23817647154392294,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.536,
       "step": 14660
     },
     {
       "epoch": 0.23833893844129259,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.5497,
       "step": 14670
     },
     {
       "epoch": 0.23850140533866224,
-      "grad_norm": 2.375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4287,
+      "loss": 1.4776,
       "step": 14680
     },
     {
       "epoch": 0.23866387223603192,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4263,
+      "loss": 1.5503,
       "step": 14690
     },
     {
       "epoch": 0.23882633913340157,
-      "grad_norm": 3.25,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4092,
+      "loss": 1.4898,
       "step": 14700
     },
     {
       "epoch": 0.23898880603077122,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4452,
+      "loss": 1.5612,
       "step": 14710
     },
     {
       "epoch": 0.2391512729281409,
-      "grad_norm": 2.234375,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4801,
       "step": 14720
     },
     {
       "epoch": 0.23931373982551055,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4373,
+      "loss": 1.5529,
       "step": 14730
     },
     {
       "epoch": 0.23947620672288022,
-      "grad_norm": 4.28125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.4429,
+      "loss": 1.5099,
       "step": 14740
     },
     {
       "epoch": 0.23963867362024988,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4365,
+      "loss": 1.5136,
       "step": 14750
     },
     {
       "epoch": 0.23980114051761953,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.5682,
       "step": 14760
     },
     {
       "epoch": 0.2399636074149892,
-      "grad_norm": 2.75,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.5768,
       "step": 14770
     },
     {
       "epoch": 0.24012607431235886,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4411,
+      "loss": 1.5395,
       "step": 14780
     },
     {
       "epoch": 0.2402885412097285,
-      "grad_norm": 2.375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4513,
+      "loss": 1.4854,
       "step": 14790
     },
     {
       "epoch": 0.24045100810709819,
-      "grad_norm": 2.375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4376,
+      "loss": 1.5313,
       "step": 14800
     },
     {
       "epoch": 0.24061347500446784,
-      "grad_norm": 2.75,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4535,
+      "loss": 1.5096,
       "step": 14810
     },
     {
       "epoch": 0.2407759419018375,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4217,
+      "loss": 1.4855,
       "step": 14820
     },
     {
       "epoch": 0.24093840879920717,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4315,
+      "loss": 1.5508,
       "step": 14830
     },
     {
       "epoch": 0.24110087569657682,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4319,
+      "loss": 1.5586,
       "step": 14840
     },
     {
       "epoch": 0.2412633425939465,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.5221,
       "step": 14850
     },
     {
       "epoch": 0.24142580949131615,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4185,
+      "loss": 1.5465,
       "step": 14860
     },
     {
       "epoch": 0.2415882763886858,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.5466,
       "step": 14870
     },
     {
       "epoch": 0.24175074328605548,
-      "grad_norm": 2.78125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4544,
+      "loss": 1.5373,
       "step": 14880
     },
     {
       "epoch": 0.24191321018342513,
-      "grad_norm": 2.40625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.5424,
       "step": 14890
     },
     {
       "epoch": 0.24207567708079478,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4198,
+      "loss": 1.542,
       "step": 14900
     },
     {
       "epoch": 0.24223814397816446,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4349,
+      "loss": 1.531,
       "step": 14910
     },
     {
       "epoch": 0.2424006108755341,
-      "grad_norm": 2.25,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4356,
+      "loss": 1.5426,
       "step": 14920
     },
     {
       "epoch": 0.24256307777290378,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4299,
+      "loss": 1.5599,
       "step": 14930
     },
     {
       "epoch": 0.24272554467027344,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.5389,
       "step": 14940
     },
     {
       "epoch": 0.2428880115676431,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4174,
+      "loss": 1.512,
       "step": 14950
     },
     {
       "epoch": 0.24305047846501276,
-      "grad_norm": 2.0,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.5083,
       "step": 14960
     },
     {
       "epoch": 0.24321294536238242,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.5195,
       "step": 14970
     },
     {
       "epoch": 0.24337541225975207,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.5015,
       "step": 14980
     },
     {
       "epoch": 0.24353787915712174,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4356,
+      "loss": 1.5272,
       "step": 14990
     },
     {
       "epoch": 0.2437003460544914,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.4424,
+      "loss": 1.5122,
       "step": 15000
     },
     {
       "epoch": 0.24386281295186105,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.5815,
       "step": 15010
     },
     {
       "epoch": 0.24402527984923073,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4064,
+      "loss": 1.5269,
       "step": 15020
     },
     {
       "epoch": 0.24418774674660038,
-      "grad_norm": 2.625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.506,
       "step": 15030
     },
     {
       "epoch": 0.24435021364397005,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.5105,
       "step": 15040
     },
     {
       "epoch": 0.2445126805413397,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.5692,
       "step": 15050
     },
     {
       "epoch": 0.24467514743870936,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.5315,
       "step": 15060
     },
     {
       "epoch": 0.24483761433607903,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.5208,
       "step": 15070
     },
     {
       "epoch": 0.24500008123344869,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.5299,
       "step": 15080
     },
     {
       "epoch": 0.24516254813081834,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.5493,
       "step": 15090
     },
     {
       "epoch": 0.24532501502818801,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4206,
+      "loss": 1.5531,
       "step": 15100
     },
     {
       "epoch": 0.24548748192555767,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4289,
+      "loss": 1.5264,
       "step": 15110
     },
     {
       "epoch": 0.24564994882292732,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.5534,
       "step": 15120
     },
     {
       "epoch": 0.245812415720297,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4346,
+      "loss": 1.5276,
       "step": 15130
     },
     {
       "epoch": 0.24597488261766665,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4205,
+      "loss": 1.5195,
       "step": 15140
     },
     {
       "epoch": 0.24613734951503632,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.5512,
       "step": 15150
     },
     {
       "epoch": 0.24629981641240598,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4365,
+      "loss": 1.5993,
       "step": 15160
     },
     {
       "epoch": 0.24646228330977563,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.4537,
       "step": 15170
     },
     {
       "epoch": 0.2466247502071453,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.5202,
       "step": 15180
     },
     {
       "epoch": 0.24678721710451496,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.5698,
       "step": 15190
     },
     {
       "epoch": 0.2469496840018846,
-      "grad_norm": 1.5390625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.5058,
       "step": 15200
     },
     {
       "epoch": 0.24711215089925428,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.427,
+      "loss": 1.5856,
       "step": 15210
     },
     {
       "epoch": 0.24727461779662394,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5391,
       "step": 15220
     },
     {
       "epoch": 0.2474370846939936,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.5268,
       "step": 15230
     },
     {
       "epoch": 0.24759955159136326,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4358,
+      "loss": 1.5236,
       "step": 15240
     },
     {
       "epoch": 0.24776201848873292,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4277,
+      "loss": 1.5121,
       "step": 15250
     },
     {
       "epoch": 0.2479244853861026,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4445,
+      "loss": 1.5505,
       "step": 15260
     },
     {
       "epoch": 0.24808695228347225,
-      "grad_norm": 1.7734375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4235,
+      "loss": 1.5112,
       "step": 15270
     },
     {
       "epoch": 0.2482494191808419,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4337,
+      "loss": 1.4983,
       "step": 15280
     },
     {
       "epoch": 0.24841188607821157,
-      "grad_norm": 2.46875,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4322,
+      "loss": 1.5201,
       "step": 15290
     },
     {
       "epoch": 0.24857435297558123,
-      "grad_norm": 3.640625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.5459,
       "step": 15300
     },
     {
       "epoch": 0.24873681987295088,
-      "grad_norm": 4.21875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.4818,
       "step": 15310
     },
     {
       "epoch": 0.24889928677032055,
-      "grad_norm": 3.5,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4221,
+      "loss": 1.5609,
       "step": 15320
     },
     {
       "epoch": 0.2490617536676902,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4337,
+      "loss": 1.5481,
       "step": 15330
     },
     {
       "epoch": 0.24922422056505986,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4355,
+      "loss": 1.5825,
       "step": 15340
     },
     {
       "epoch": 0.24938668746242953,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4388,
+      "loss": 1.5594,
       "step": 15350
     },
     {
       "epoch": 0.24954915435979919,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.5314,
       "step": 15360
     },
     {
       "epoch": 0.24971162125716886,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.5529,
       "step": 15370
     },
     {
       "epoch": 0.24987408815453852,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.5063,
       "step": 15380
     },
     {
       "epoch": 0.2500365550519082,
-      "grad_norm": 1.9921875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4296,
+      "loss": 1.4949,
       "step": 15390
     },
     {
       "epoch": 0.25019902194927784,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.5181,
       "step": 15400
     },
     {
       "epoch": 0.2503614888466475,
-      "grad_norm": 2.125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.5118,
       "step": 15410
     },
     {
       "epoch": 0.25052395574401715,
-      "grad_norm": 3.609375,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.5407,
       "step": 15420
     },
     {
       "epoch": 0.2506864226413868,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.5578,
       "step": 15430
     },
     {
       "epoch": 0.2508488895387565,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.5433,
       "step": 15440
     },
     {
       "epoch": 0.25101135643612615,
-      "grad_norm": 1.875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4432,
+      "loss": 1.5555,
       "step": 15450
     },
     {
       "epoch": 0.2511738233334958,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.5517,
       "step": 15460
     },
     {
       "epoch": 0.25133629023086546,
-      "grad_norm": 2.875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5424,
       "step": 15470
     },
     {
       "epoch": 0.2514987571282351,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.524,
       "step": 15480
     },
     {
       "epoch": 0.25166122402560476,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.5154,
       "step": 15490
     },
     {
       "epoch": 0.25182369092297446,
-      "grad_norm": 1.5625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.5004,
       "step": 15500
     },
     {
       "epoch": 0.2519861578203441,
-      "grad_norm": 3.875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.441,
+      "loss": 1.4947,
       "step": 15510
     },
     {
       "epoch": 0.25214862471771377,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4228,
+      "loss": 1.5078,
       "step": 15520
     },
     {
       "epoch": 0.2523110916150834,
-      "grad_norm": 2.75,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4413,
+      "loss": 1.5134,
       "step": 15530
     },
     {
       "epoch": 0.25247355851245307,
-      "grad_norm": 2.421875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.4417,
+      "loss": 1.5227,
       "step": 15540
     },
     {
       "epoch": 0.2526360254098228,
-      "grad_norm": 3.25,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4444,
+      "loss": 1.5166,
       "step": 15550
     },
     {
       "epoch": 0.2527984923071924,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4414,
+      "loss": 1.538,
       "step": 15560
     },
     {
       "epoch": 0.2529609592045621,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4472,
+      "loss": 1.5527,
       "step": 15570
     },
     {
       "epoch": 0.2531234261019317,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.5785,
       "step": 15580
     },
     {
       "epoch": 0.2532858929993014,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4224,
+      "loss": 1.5423,
       "step": 15590
     },
     {
       "epoch": 0.253448359896671,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4388,
+      "loss": 1.5099,
       "step": 15600
     },
     {
       "epoch": 0.25361082679404073,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4339,
+      "loss": 1.4593,
       "step": 15610
     },
     {
       "epoch": 0.2537732936914104,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4286,
+      "loss": 1.524,
       "step": 15620
     },
     {
       "epoch": 0.25393576058878004,
-      "grad_norm": 1.71875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.5421,
       "step": 15630
     },
     {
       "epoch": 0.2540982274861497,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4309,
+      "loss": 1.5335,
       "step": 15640
     },
     {
       "epoch": 0.25426069438351934,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4433,
+      "loss": 1.5284,
       "step": 15650
     },
     {
       "epoch": 0.25442316128088904,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4191,
+      "loss": 1.5366,
       "step": 15660
     },
     {
       "epoch": 0.2545856281782587,
-      "grad_norm": 2.875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.5282,
       "step": 15670
     },
     {
       "epoch": 0.25474809507562834,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.569,
       "step": 15680
     },
     {
       "epoch": 0.254910561972998,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4278,
+      "loss": 1.4904,
       "step": 15690
     },
     {
       "epoch": 0.25507302887036765,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.5381,
       "step": 15700
     },
     {
       "epoch": 0.2552354957677373,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4329,
+      "loss": 1.5603,
       "step": 15710
     },
     {
       "epoch": 0.255397962665107,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.5314,
       "step": 15720
     },
     {
       "epoch": 0.25556042956247665,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.5329,
       "step": 15730
     },
     {
       "epoch": 0.2557228964598463,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.5048,
       "step": 15740
     },
     {
       "epoch": 0.25588536335721596,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.5088,
       "step": 15750
     },
     {
       "epoch": 0.2560478302545856,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4273,
+      "loss": 1.5232,
       "step": 15760
     },
     {
       "epoch": 0.2562102971519553,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4314,
+      "loss": 1.5368,
       "step": 15770
     },
     {
       "epoch": 0.25637276404932496,
-      "grad_norm": 2.5,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.5247,
       "step": 15780
     },
     {
       "epoch": 0.2565352309466946,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4506,
+      "loss": 1.5182,
       "step": 15790
     },
     {
       "epoch": 0.25669769784406427,
-      "grad_norm": 2.25,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4379,
+      "loss": 1.5343,
       "step": 15800
     },
     {
       "epoch": 0.2568601647414339,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.5167,
       "step": 15810
     },
     {
       "epoch": 0.25702263163880357,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.532,
       "step": 15820
     },
     {
       "epoch": 0.2571850985361733,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.6029,
       "step": 15830
     },
     {
       "epoch": 0.2573475654335429,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.5029,
       "step": 15840
     },
     {
       "epoch": 0.2575100323309126,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5507,
       "step": 15850
     },
     {
       "epoch": 0.2576724992282822,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.5572,
       "step": 15860
     },
     {
       "epoch": 0.2578349661256519,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.5366,
       "step": 15870
     },
     {
       "epoch": 0.2579974330230216,
-      "grad_norm": 2.875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.5013,
       "step": 15880
     },
     {
       "epoch": 0.25815989992039123,
-      "grad_norm": 1.8046875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4311,
+      "loss": 1.5212,
       "step": 15890
     },
     {
       "epoch": 0.2583223668177609,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4318,
+      "loss": 1.4616,
       "step": 15900
     },
     {
       "epoch": 0.25848483371513054,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.5396,
       "step": 15910
     },
     {
       "epoch": 0.2586473006125002,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4358,
+      "loss": 1.5574,
       "step": 15920
     },
     {
       "epoch": 0.25880976750986984,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4433,
+      "loss": 1.4948,
       "step": 15930
     },
     {
       "epoch": 0.25897223440723954,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4508,
+      "loss": 1.5191,
       "step": 15940
     },
     {
       "epoch": 0.2591347013046092,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4356,
+      "loss": 1.5195,
       "step": 15950
     },
     {
       "epoch": 0.25929716820197884,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4219,
+      "loss": 1.5017,
       "step": 15960
     },
     {
       "epoch": 0.2594596350993485,
-      "grad_norm": 2.84375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4441,
+      "loss": 1.5465,
       "step": 15970
     },
     {
       "epoch": 0.25962210199671815,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4329,
+      "loss": 1.5179,
       "step": 15980
     },
     {
       "epoch": 0.25978456889408785,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.5626,
       "step": 15990
     },
     {
       "epoch": 0.2599470357914575,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.4892,
       "step": 16000
     },
     {
       "epoch": 0.26010950268882715,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.5016,
       "step": 16010
     },
     {
       "epoch": 0.2602719695861968,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.4941,
       "step": 16020
     },
     {
       "epoch": 0.26043443648356646,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4228,
+      "loss": 1.5692,
       "step": 16030
     },
     {
       "epoch": 0.2605969033809361,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.5246,
       "step": 16040
     },
     {
       "epoch": 0.2607593702783058,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4386,
+      "loss": 1.559,
       "step": 16050
     },
     {
       "epoch": 0.26092183717567546,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4235,
+      "loss": 1.4998,
       "step": 16060
     },
     {
       "epoch": 0.2610843040730451,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4389,
+      "loss": 1.5541,
       "step": 16070
     },
     {
       "epoch": 0.26124677097041477,
-      "grad_norm": 3.5,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.504,
       "step": 16080
     },
     {
       "epoch": 0.2614092378677844,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.5525,
       "step": 16090
     },
     {
       "epoch": 0.2615717047651541,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.5225,
       "step": 16100
     },
     {
       "epoch": 0.2617341716625238,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.536,
       "step": 16110
     },
     {
       "epoch": 0.2618966385598934,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.5109,
       "step": 16120
     },
     {
       "epoch": 0.2620591054572631,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4331,
+      "loss": 1.4979,
       "step": 16130
     },
     {
       "epoch": 0.2622215723546327,
-      "grad_norm": 2.25,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.4791,
       "step": 16140
     },
     {
       "epoch": 0.2623840392520024,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4323,
+      "loss": 1.4893,
       "step": 16150
     },
     {
       "epoch": 0.2625465061493721,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.5609,
       "step": 16160
     },
     {
       "epoch": 0.26270897304674173,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4355,
+      "loss": 1.5802,
       "step": 16170
     },
     {
       "epoch": 0.2628714399441114,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4331,
+      "loss": 1.4978,
       "step": 16180
     },
     {
       "epoch": 0.26303390684148104,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.5417,
       "step": 16190
     },
     {
       "epoch": 0.2631963737388507,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.5813,
       "step": 16200
     },
     {
       "epoch": 0.2633588406362204,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4398,
+      "loss": 1.507,
       "step": 16210
     },
     {
       "epoch": 0.26352130753359004,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4394,
+      "loss": 1.5124,
       "step": 16220
     },
     {
       "epoch": 0.2636837744309597,
-      "grad_norm": 2.375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4279,
+      "loss": 1.5117,
       "step": 16230
     },
     {
       "epoch": 0.26384624132832935,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4835,
       "step": 16240
     },
     {
       "epoch": 0.264008708225699,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4185,
+      "loss": 1.4915,
       "step": 16250
     },
     {
       "epoch": 0.2641711751230687,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.5635,
       "step": 16260
     },
     {
       "epoch": 0.26433364202043835,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.5371,
       "step": 16270
     },
     {
       "epoch": 0.264496108917808,
-      "grad_norm": 2.75,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.4821,
       "step": 16280
     },
     {
       "epoch": 0.26465857581517765,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4481,
+      "loss": 1.4776,
       "step": 16290
     },
     {
       "epoch": 0.2648210427125473,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.445,
+      "loss": 1.5053,
       "step": 16300
     },
     {
       "epoch": 0.26498350960991696,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4329,
+      "loss": 1.5333,
       "step": 16310
     },
     {
       "epoch": 0.26514597650728666,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4333,
+      "loss": 1.5416,
       "step": 16320
     },
     {
       "epoch": 0.2653084434046563,
-      "grad_norm": 4.5625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.474,
       "step": 16330
     },
     {
       "epoch": 0.26547091030202596,
-      "grad_norm": 2.046875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.5195,
       "step": 16340
     },
     {
       "epoch": 0.2656333771993956,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.5527,
       "step": 16350
     },
     {
       "epoch": 0.26579584409676527,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.5063,
       "step": 16360
     },
     {
       "epoch": 0.26595831099413497,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4296,
+      "loss": 1.5469,
       "step": 16370
     },
     {
       "epoch": 0.2661207778915046,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.491,
       "step": 16380
     },
     {
       "epoch": 0.2662832447888743,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.5239,
       "step": 16390
     },
     {
       "epoch": 0.2664457116862439,
-      "grad_norm": 2.625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4326,
+      "loss": 1.5296,
       "step": 16400
     },
     {
       "epoch": 0.2666081785836136,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4394,
+      "loss": 1.4955,
       "step": 16410
     },
     {
       "epoch": 0.2667706454809832,
-      "grad_norm": 3.125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.52,
       "step": 16420
     },
     {
       "epoch": 0.26693311237835293,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4459,
+      "loss": 1.5436,
       "step": 16430
     },
     {
       "epoch": 0.2670955792757226,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4304,
+      "loss": 1.4917,
       "step": 16440
     },
     {
       "epoch": 0.26725804617309223,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.5525,
       "step": 16450
     },
     {
       "epoch": 0.2674205130704619,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4303,
+      "loss": 1.5163,
       "step": 16460
     },
     {
       "epoch": 0.26758297996783154,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.5397,
       "step": 16470
     },
     {
       "epoch": 0.26774544686520124,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4315,
+      "loss": 1.5384,
       "step": 16480
     },
     {
       "epoch": 0.2679079137625709,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.5241,
       "step": 16490
     },
     {
       "epoch": 0.26807038065994054,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.5141,
       "step": 16500
     },
     {
       "epoch": 0.2682328475573102,
-      "grad_norm": 2.625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.6013,
       "step": 16510
     },
     {
       "epoch": 0.26839531445467985,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.4863,
       "step": 16520
     },
     {
       "epoch": 0.2685577813520495,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4461,
+      "loss": 1.5587,
       "step": 16530
     },
     {
       "epoch": 0.2687202482494192,
-      "grad_norm": 2.75,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.5537,
       "step": 16540
     },
     {
       "epoch": 0.26888271514678885,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4562,
+      "loss": 1.549,
       "step": 16550
     },
     {
       "epoch": 0.2690451820441585,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4865,
       "step": 16560
     },
     {
       "epoch": 0.26920764894152815,
-      "grad_norm": 1.8203125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4206,
+      "loss": 1.479,
       "step": 16570
     },
     {
       "epoch": 0.2693701158388978,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.5727,
       "step": 16580
     },
     {
       "epoch": 0.2695325827362675,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4415,
+      "loss": 1.5249,
       "step": 16590
     },
     {
       "epoch": 0.26969504963363716,
-      "grad_norm": 2.375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.5195,
       "step": 16600
     },
     {
       "epoch": 0.2698575165310068,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.5079,
       "step": 16610
     },
     {
       "epoch": 0.27001998342837646,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.424,
+      "loss": 1.5314,
       "step": 16620
     },
     {
       "epoch": 0.2701824503257461,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.5291,
       "step": 16630
     },
     {
       "epoch": 0.27034491722311577,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4431,
+      "loss": 1.4861,
       "step": 16640
     },
     {
       "epoch": 0.2705073841204855,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.5207,
       "step": 16650
     },
     {
       "epoch": 0.2706698510178551,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.5379,
       "step": 16660
     },
     {
       "epoch": 0.2708323179152248,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.5304,
       "step": 16670
     },
     {
       "epoch": 0.2709947848125944,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4503,
+      "loss": 1.545,
       "step": 16680
     },
     {
       "epoch": 0.2711572517099641,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4324,
+      "loss": 1.5168,
       "step": 16690
     },
     {
       "epoch": 0.2713197186073338,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.5543,
       "step": 16700
     },
     {
       "epoch": 0.27148218550470343,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.5341,
       "step": 16710
     },
     {
       "epoch": 0.2716446524020731,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.415,
+      "loss": 1.5556,
       "step": 16720
     },
     {
       "epoch": 0.27180711929944273,
-      "grad_norm": 2.375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.5085,
       "step": 16730
     },
     {
       "epoch": 0.2719695861968124,
-      "grad_norm": 2.8125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.5183,
       "step": 16740
     },
     {
       "epoch": 0.27213205309418204,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.5252,
       "step": 16750
     },
     {
       "epoch": 0.27229451999155174,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.484,
       "step": 16760
     },
     {
       "epoch": 0.2724569868889214,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4291,
+      "loss": 1.5553,
       "step": 16770
     },
     {
       "epoch": 0.27261945378629104,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4344,
+      "loss": 1.517,
       "step": 16780
     },
     {
       "epoch": 0.2727819206836607,
-      "grad_norm": 3.5,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4439,
+      "loss": 1.4988,
       "step": 16790
     },
     {
       "epoch": 0.27294438758103035,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4221,
+      "loss": 1.5332,
       "step": 16800
     },
     {
       "epoch": 0.27310685447840005,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.5193,
       "step": 16810
     },
     {
       "epoch": 0.2732693213757697,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.5944,
       "step": 16820
     },
     {
       "epoch": 0.27343178827313935,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.5146,
       "step": 16830
     },
     {
       "epoch": 0.273594255170509,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.5133,
       "step": 16840
     },
     {
       "epoch": 0.27375672206787866,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4306,
+      "loss": 1.5198,
       "step": 16850
     },
     {
       "epoch": 0.2739191889652483,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.5323,
       "step": 16860
     },
     {
       "epoch": 0.274081655862618,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4474,
+      "loss": 1.5054,
       "step": 16870
     },
     {
       "epoch": 0.27424412275998766,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.526,
       "step": 16880
     },
     {
       "epoch": 0.2744065896573573,
-      "grad_norm": 3.6875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.521,
       "step": 16890
     },
     {
       "epoch": 0.27456905655472696,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.454,
+      "loss": 1.4958,
       "step": 16900
     },
     {
       "epoch": 0.2747315234520966,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.4875,
       "step": 16910
     },
     {
       "epoch": 0.2748939903494663,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.479,
       "step": 16920
     },
     {
       "epoch": 0.275056457246836,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4903,
       "step": 16930
     },
     {
       "epoch": 0.2752189241442056,
-      "grad_norm": 3.25,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4263,
+      "loss": 1.524,
       "step": 16940
     },
     {
       "epoch": 0.2753813910415753,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.4344,
+      "loss": 1.4863,
       "step": 16950
     },
     {
       "epoch": 0.2755438579389449,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4267,
+      "loss": 1.5371,
       "step": 16960
     },
     {
       "epoch": 0.2757063248363146,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4321,
+      "loss": 1.5431,
       "step": 16970
     },
     {
       "epoch": 0.2758687917336843,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4168,
+      "loss": 1.494,
       "step": 16980
     },
     {
       "epoch": 0.27603125863105393,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.5381,
       "step": 16990
     },
     {
       "epoch": 0.2761937255284236,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.5166,
       "step": 17000
     },
     {
       "epoch": 0.27635619242579323,
-      "grad_norm": 1.78125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.4786,
       "step": 17010
     },
     {
       "epoch": 0.2765186593231629,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4379,
+      "loss": 1.532,
       "step": 17020
     },
     {
       "epoch": 0.2766811262205326,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.4961,
       "step": 17030
     },
     {
       "epoch": 0.27684359311790224,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4447,
+      "loss": 1.4555,
       "step": 17040
     },
     {
       "epoch": 0.2770060600152719,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.5127,
       "step": 17050
     },
     {
       "epoch": 0.27716852691264154,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.5262,
       "step": 17060
     },
     {
       "epoch": 0.2773309938100112,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.5033,
       "step": 17070
     },
     {
       "epoch": 0.27749346070738085,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4234,
+      "loss": 1.5263,
       "step": 17080
     },
     {
       "epoch": 0.27765592760475055,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.496,
       "step": 17090
     },
     {
       "epoch": 0.2778183945021202,
-      "grad_norm": 4.25,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.5303,
       "step": 17100
     },
     {
       "epoch": 0.27798086139948985,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4295,
+      "loss": 1.5036,
       "step": 17110
     },
     {
       "epoch": 0.2781433282968595,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4443,
+      "loss": 1.5728,
       "step": 17120
     },
     {
       "epoch": 0.27830579519422916,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.5562,
       "step": 17130
     },
     {
       "epoch": 0.27846826209159886,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.5411,
       "step": 17140
     },
     {
       "epoch": 0.2786307289889685,
-      "grad_norm": 3.421875,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4435,
+      "loss": 1.5297,
       "step": 17150
     },
     {
       "epoch": 0.27879319588633816,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.5071,
       "step": 17160
     },
     {
       "epoch": 0.2789556627837078,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4333,
+      "loss": 1.5347,
       "step": 17170
     },
     {
       "epoch": 0.27911812968107746,
-      "grad_norm": 2.125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.522,
       "step": 17180
     },
     {
       "epoch": 0.2792805965784471,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.522,
       "step": 17190
     },
     {
       "epoch": 0.2794430634758168,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.436,
+      "loss": 1.5131,
       "step": 17200
     },
     {
       "epoch": 0.2796055303731865,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4434,
+      "loss": 1.5377,
       "step": 17210
     },
     {
       "epoch": 0.2797679972705561,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4214,
+      "loss": 1.5149,
       "step": 17220
     },
     {
       "epoch": 0.2799304641679258,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.5312,
       "step": 17230
     },
     {
       "epoch": 0.2800929310652954,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.5499,
       "step": 17240
     },
     {
       "epoch": 0.28025539796266513,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.5497,
       "step": 17250
     },
     {
       "epoch": 0.2804178648600348,
-      "grad_norm": 2.375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.5417,
       "step": 17260
     },
     {
       "epoch": 0.28058033175740443,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4438,
+      "loss": 1.5672,
       "step": 17270
     },
     {
       "epoch": 0.2807427986547741,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.5358,
       "step": 17280
     },
     {
       "epoch": 0.28090526555214373,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.5178,
       "step": 17290
     },
     {
       "epoch": 0.2810677324495134,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4105,
+      "loss": 1.5339,
       "step": 17300
     },
     {
       "epoch": 0.2812301993468831,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4504,
+      "loss": 1.5137,
       "step": 17310
     },
     {
       "epoch": 0.28139266624425274,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.531,
       "step": 17320
     },
     {
       "epoch": 0.2815551331416224,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4406,
+      "loss": 1.4996,
       "step": 17330
     },
     {
       "epoch": 0.28171760003899204,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.5235,
       "step": 17340
     },
     {
       "epoch": 0.2818800669363617,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4256,
+      "loss": 1.551,
       "step": 17350
     },
     {
       "epoch": 0.2820425338337314,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.4922,
       "step": 17360
     },
     {
       "epoch": 0.28220500073110105,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.5614,
       "step": 17370
     },
     {
       "epoch": 0.2823674676284707,
-      "grad_norm": 2.25,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4137,
+      "loss": 1.5211,
       "step": 17380
     },
     {
       "epoch": 0.28252993452584035,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.5136,
       "step": 17390
     },
     {
       "epoch": 0.28269240142321,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4105,
+      "loss": 1.4982,
       "step": 17400
     },
     {
       "epoch": 0.28285486832057966,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4307,
+      "loss": 1.4852,
       "step": 17410
     },
     {
       "epoch": 0.28301733521794936,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.4782,
       "step": 17420
     },
     {
       "epoch": 0.283179802115319,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4377,
+      "loss": 1.5509,
       "step": 17430
     },
     {
       "epoch": 0.28334226901268866,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.4815,
       "step": 17440
     },
     {
       "epoch": 0.2835047359100583,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.5136,
       "step": 17450
     },
     {
       "epoch": 0.28366720280742797,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.5455,
       "step": 17460
     },
     {
       "epoch": 0.28382966970479767,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.5326,
       "step": 17470
     },
     {
       "epoch": 0.2839921366021673,
-      "grad_norm": 1.75,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.4668,
       "step": 17480
     },
     {
       "epoch": 0.284154603499537,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4208,
+      "loss": 1.5053,
       "step": 17490
     },
     {
       "epoch": 0.2843170703969066,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.5073,
       "step": 17500
     },
     {
       "epoch": 0.2844795372942763,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.4857,
       "step": 17510
     },
     {
       "epoch": 0.2846420041916459,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.5647,
       "step": 17520
     },
     {
       "epoch": 0.28480447108901563,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.5048,
       "step": 17530
     },
     {
       "epoch": 0.2849669379863853,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4282,
+      "loss": 1.5334,
       "step": 17540
     },
     {
       "epoch": 0.28512940488375493,
-      "grad_norm": 3.90625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.5397,
       "step": 17550
     },
     {
       "epoch": 0.2852918717811246,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4366,
+      "loss": 1.4937,
       "step": 17560
     },
     {
       "epoch": 0.28545433867849423,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.514,
       "step": 17570
     },
     {
       "epoch": 0.28561680557586394,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.5369,
       "step": 17580
     },
     {
       "epoch": 0.2857792724732336,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.5787,
       "step": 17590
     },
     {
       "epoch": 0.28594173937060324,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.425,
+      "loss": 1.5533,
       "step": 17600
     },
     {
       "epoch": 0.2861042062679729,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4406,
+      "loss": 1.5164,
       "step": 17610
     },
     {
       "epoch": 0.28626667316534254,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4402,
+      "loss": 1.4944,
       "step": 17620
     },
     {
       "epoch": 0.2864291400627122,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4397,
+      "loss": 1.5822,
       "step": 17630
     },
     {
       "epoch": 0.2865916069600819,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4442,
+      "loss": 1.5211,
       "step": 17640
     },
     {
       "epoch": 0.28675407385745155,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.5103,
       "step": 17650
     },
     {
       "epoch": 0.2869165407548212,
-      "grad_norm": 4.3125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4264,
+      "loss": 1.5129,
       "step": 17660
     },
     {
       "epoch": 0.28707900765219085,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.519,
       "step": 17670
     },
     {
       "epoch": 0.2872414745495605,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4439,
+      "loss": 1.4684,
       "step": 17680
     },
     {
       "epoch": 0.2874039414469302,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.5349,
       "step": 17690
     },
     {
       "epoch": 0.28756640834429986,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.425,
+      "loss": 1.5496,
       "step": 17700
     },
     {
       "epoch": 0.2877288752416695,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.5525,
       "step": 17710
     },
     {
       "epoch": 0.28789134213903916,
-      "grad_norm": 2.25,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4357,
+      "loss": 1.493,
       "step": 17720
     },
     {
       "epoch": 0.2880538090364088,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4457,
+      "loss": 1.5262,
       "step": 17730
     },
     {
       "epoch": 0.28821627593377847,
-      "grad_norm": 2.625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.5222,
       "step": 17740
     },
     {
       "epoch": 0.28837874283114817,
-      "grad_norm": 2.875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.5336,
       "step": 17750
     },
     {
       "epoch": 0.2885412097285178,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.424,
+      "loss": 1.5283,
       "step": 17760
     },
     {
       "epoch": 0.2887036766258875,
-      "grad_norm": 2.125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4523,
+      "loss": 1.5008,
       "step": 17770
     },
     {
       "epoch": 0.2888661435232571,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.5211,
       "step": 17780
     },
     {
       "epoch": 0.2890286104206268,
-      "grad_norm": 3.25,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.4838,
       "step": 17790
     },
     {
       "epoch": 0.2891910773179965,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.4836,
       "step": 17800
     },
     {
       "epoch": 0.28935354421536613,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.5155,
       "step": 17810
     },
     {
       "epoch": 0.2895160111127358,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4633,
       "step": 17820
     },
     {
       "epoch": 0.28967847801010543,
-      "grad_norm": 2.0,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.5324,
       "step": 17830
     },
     {
       "epoch": 0.2898409449074751,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4311,
+      "loss": 1.5295,
       "step": 17840
     },
     {
       "epoch": 0.29000341180484474,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4324,
+      "loss": 1.5047,
       "step": 17850
     },
     {
       "epoch": 0.29016587870221444,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.5369,
       "step": 17860
     },
     {
       "epoch": 0.2903283455995841,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.5021,
       "step": 17870
     },
     {
       "epoch": 0.29049081249695374,
-      "grad_norm": 1.8046875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4119,
+      "loss": 1.528,
       "step": 17880
     },
     {
       "epoch": 0.2906532793943234,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.4944,
       "step": 17890
     },
     {
       "epoch": 0.29081574629169304,
-      "grad_norm": 2.59375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.5106,
       "step": 17900
     },
     {
       "epoch": 0.29097821318906275,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4502,
+      "loss": 1.4968,
       "step": 17910
     },
     {
       "epoch": 0.2911406800864324,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.484,
       "step": 17920
     },
     {
       "epoch": 0.29130314698380205,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4326,
+      "loss": 1.4955,
       "step": 17930
     },
     {
       "epoch": 0.2914656138811717,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.5371,
       "step": 17940
     },
     {
       "epoch": 0.29162808077854135,
-      "grad_norm": 3.21875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4243,
+      "loss": 1.4806,
       "step": 17950
     },
     {
       "epoch": 0.29179054767591106,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4353,
+      "loss": 1.5333,
       "step": 17960
     },
     {
       "epoch": 0.2919530145732807,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.5127,
       "step": 17970
     },
     {
       "epoch": 0.29211548147065036,
-      "grad_norm": 3.71875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.5452,
       "step": 17980
     },
     {
       "epoch": 0.29227794836802,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.5333,
       "step": 17990
     },
     {
       "epoch": 0.29244041526538966,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4214,
+      "loss": 1.5302,
       "step": 18000
     },
     {
       "epoch": 0.2926028821627593,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4233,
+      "loss": 1.4856,
       "step": 18010
     },
     {
       "epoch": 0.292765349060129,
-      "grad_norm": 3.375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.5236,
       "step": 18020
     },
     {
       "epoch": 0.29292781595749867,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.478,
       "step": 18030
     },
     {
       "epoch": 0.2930902828548683,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4353,
+      "loss": 1.5213,
       "step": 18040
     },
     {
       "epoch": 0.293252749752238,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.4803,
       "step": 18050
     },
     {
       "epoch": 0.2934152166496076,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.393,
+      "loss": 1.4811,
       "step": 18060
     },
     {
       "epoch": 0.29357768354697733,
-      "grad_norm": 1.59375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4159,
+      "loss": 1.5118,
       "step": 18070
     },
     {
       "epoch": 0.293740150444347,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.5093,
       "step": 18080
     },
     {
       "epoch": 0.29390261734171663,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4203,
+      "loss": 1.525,
       "step": 18090
     },
     {
       "epoch": 0.2940650842390863,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4206,
+      "loss": 1.4621,
       "step": 18100
     },
     {
       "epoch": 0.29422755113645593,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4545,
+      "loss": 1.4777,
       "step": 18110
     },
     {
       "epoch": 0.2943900180338256,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.4814,
       "step": 18120
     },
     {
       "epoch": 0.2945524849311953,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.5259,
       "step": 18130
     },
     {
       "epoch": 0.29471495182856494,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4297,
+      "loss": 1.5393,
       "step": 18140
     },
     {
       "epoch": 0.2948774187259346,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4228,
+      "loss": 1.4838,
       "step": 18150
     },
     {
       "epoch": 0.29503988562330424,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4432,
+      "loss": 1.5044,
       "step": 18160
     },
     {
       "epoch": 0.2952023525206739,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4219,
+      "loss": 1.526,
       "step": 18170
     },
     {
       "epoch": 0.2953648194180436,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.5654,
       "step": 18180
     },
     {
       "epoch": 0.29552728631541325,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.5232,
       "step": 18190
     },
     {
       "epoch": 0.2956897532127829,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5091,
       "step": 18200
     },
     {
       "epoch": 0.29585222011015255,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4228,
+      "loss": 1.4817,
       "step": 18210
     },
     {
       "epoch": 0.2960146870075222,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4227,
+      "loss": 1.5033,
       "step": 18220
     },
     {
       "epoch": 0.29617715390489185,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.5554,
       "step": 18230
     },
     {
       "epoch": 0.29633962080226156,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.5393,
       "step": 18240
     },
     {
       "epoch": 0.2965020876996312,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.5616,
       "step": 18250
     },
     {
       "epoch": 0.29666455459700086,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4471,
+      "loss": 1.5079,
       "step": 18260
     },
     {
       "epoch": 0.2968270214943705,
-      "grad_norm": 1.6796875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4137,
+      "loss": 1.5349,
       "step": 18270
     },
     {
       "epoch": 0.29698948839174016,
-      "grad_norm": 2.5,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.4762,
       "step": 18280
     },
     {
       "epoch": 0.29715195528910987,
-      "grad_norm": 2.25,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.5289,
       "step": 18290
     },
     {
       "epoch": 0.2973144221864795,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.5587,
       "step": 18300
     },
     {
       "epoch": 0.29747688908384917,
-      "grad_norm": 3.5,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4371,
+      "loss": 1.5015,
       "step": 18310
     },
     {
       "epoch": 0.2976393559812188,
-      "grad_norm": 2.5,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.5248,
       "step": 18320
     },
     {
       "epoch": 0.2978018228785885,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.5225,
       "step": 18330
     },
     {
       "epoch": 0.2979642897759581,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4501,
+      "loss": 1.5273,
       "step": 18340
     },
     {
       "epoch": 0.29812675667332783,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.5023,
       "step": 18350
     },
     {
       "epoch": 0.2982892235706975,
-      "grad_norm": 2.75,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.5189,
       "step": 18360
     },
     {
       "epoch": 0.29845169046806713,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4985,
       "step": 18370
     },
     {
       "epoch": 0.2986141573654368,
-      "grad_norm": 2.875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4426,
+      "loss": 1.496,
       "step": 18380
     },
     {
       "epoch": 0.29877662426280643,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4362,
+      "loss": 1.5183,
       "step": 18390
     },
     {
       "epoch": 0.29893909116017614,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4283,
+      "loss": 1.4804,
       "step": 18400
     },
     {
       "epoch": 0.2991015580575458,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4278,
+      "loss": 1.521,
       "step": 18410
     },
     {
       "epoch": 0.29926402495491544,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4354,
+      "loss": 1.515,
       "step": 18420
     },
     {
       "epoch": 0.2994264918522851,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.4554,
       "step": 18430
     },
     {
       "epoch": 0.29958895874965474,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4333,
+      "loss": 1.5101,
       "step": 18440
     },
     {
       "epoch": 0.2997514256470244,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4873,
       "step": 18450
     },
     {
       "epoch": 0.2999138925443941,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4461,
+      "loss": 1.5057,
       "step": 18460
     },
     {
       "epoch": 0.30007635944176375,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4158,
+      "loss": 1.533,
       "step": 18470
     },
     {
       "epoch": 0.3002388263391334,
-      "grad_norm": 1.71875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4193,
+      "loss": 1.5433,
       "step": 18480
     },
     {
       "epoch": 0.30040129323650305,
-      "grad_norm": 2.125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4378,
+      "loss": 1.4884,
       "step": 18490
     },
     {
       "epoch": 0.3005637601338727,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.5094,
       "step": 18500
     },
     {
       "epoch": 0.3007262270312424,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.5286,
       "step": 18510
     },
     {
       "epoch": 0.30088869392861206,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.5146,
       "step": 18520
     },
     {
       "epoch": 0.3010511608259817,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4158,
+      "loss": 1.4747,
       "step": 18530
     },
     {
       "epoch": 0.30121362772335136,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4431,
+      "loss": 1.5002,
       "step": 18540
     },
     {
       "epoch": 0.301376094620721,
-      "grad_norm": 3.0625,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.5143,
       "step": 18550
     },
     {
       "epoch": 0.30153856151809066,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.4569,
       "step": 18560
     },
     {
       "epoch": 0.30170102841546037,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.5479,
       "step": 18570
     },
     {
       "epoch": 0.30186349531283,
-      "grad_norm": 2.109375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.494,
       "step": 18580
     },
     {
       "epoch": 0.30202596221019967,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4703,
       "step": 18590
     },
     {
       "epoch": 0.3021884291075693,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4433,
+      "loss": 1.523,
       "step": 18600
     },
     {
       "epoch": 0.302350896004939,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4318,
+      "loss": 1.5101,
       "step": 18610
     },
     {
       "epoch": 0.3025133629023087,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.4717,
       "step": 18620
     },
     {
       "epoch": 0.30267582979967833,
-      "grad_norm": 2.875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.4804,
       "step": 18630
     },
     {
       "epoch": 0.302838296697048,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.5629,
       "step": 18640
     },
     {
       "epoch": 0.30300076359441763,
-      "grad_norm": 3.703125,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4574,
       "step": 18650
     },
     {
       "epoch": 0.3031632304917873,
-      "grad_norm": 3.25,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4846,
       "step": 18660
     },
     {
       "epoch": 0.30332569738915693,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.5025,
       "step": 18670
     },
     {
       "epoch": 0.30348816428652664,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.5003,
       "step": 18680
     },
     {
       "epoch": 0.3036506311838963,
-      "grad_norm": 3.84375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.516,
       "step": 18690
     },
     {
       "epoch": 0.30381309808126594,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4347,
+      "loss": 1.4744,
       "step": 18700
     },
     {
       "epoch": 0.3039755649786356,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.414,
       "step": 18710
     },
     {
       "epoch": 0.30413803187600524,
-      "grad_norm": 3.53125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.4926,
       "step": 18720
     },
     {
       "epoch": 0.30430049877337495,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.538,
       "step": 18730
     },
     {
       "epoch": 0.3044629656707446,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4401,
+      "loss": 1.4644,
       "step": 18740
     },
     {
       "epoch": 0.30462543256811425,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4307,
+      "loss": 1.6009,
       "step": 18750
     },
     {
       "epoch": 0.3047878994654839,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4144,
+      "loss": 1.5288,
       "step": 18760
     },
     {
       "epoch": 0.30495036636285355,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4494,
+      "loss": 1.5133,
       "step": 18770
     },
     {
       "epoch": 0.3051128332602232,
-      "grad_norm": 1.90625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4368,
+      "loss": 1.4743,
       "step": 18780
     },
     {
       "epoch": 0.3052753001575929,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4493,
+      "loss": 1.4771,
       "step": 18790
     },
     {
       "epoch": 0.30543776705496256,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.4721,
       "step": 18800
     },
     {
       "epoch": 0.3056002339523322,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.5153,
       "step": 18810
     },
     {
       "epoch": 0.30576270084970186,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4376,
+      "loss": 1.4284,
       "step": 18820
     },
     {
       "epoch": 0.3059251677470715,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4108,
+      "loss": 1.4546,
       "step": 18830
     },
     {
       "epoch": 0.3060876346444412,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.5306,
       "step": 18840
     },
     {
       "epoch": 0.30625010154181087,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4144,
+      "loss": 1.5249,
       "step": 18850
     },
     {
       "epoch": 0.3064125684391805,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.5336,
       "step": 18860
     },
     {
       "epoch": 0.3065750353365502,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.5328,
       "step": 18870
     },
     {
       "epoch": 0.3067375022339198,
-      "grad_norm": 1.6875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.5015,
       "step": 18880
     },
     {
       "epoch": 0.3068999691312895,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4971,
       "step": 18890
     },
     {
       "epoch": 0.3070624360286592,
-      "grad_norm": 2.0,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4351,
+      "loss": 1.5071,
       "step": 18900
     },
     {
       "epoch": 0.30722490292602883,
-      "grad_norm": 2.625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.5246,
       "step": 18910
     },
     {
       "epoch": 0.3073873698233985,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4352,
+      "loss": 1.5315,
       "step": 18920
     },
     {
       "epoch": 0.30754983672076813,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.5349,
       "step": 18930
     },
     {
       "epoch": 0.3077123036181378,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.52,
       "step": 18940
     },
     {
       "epoch": 0.3078747705155075,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.5379,
       "step": 18950
     },
     {
       "epoch": 0.30803723741287714,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.5156,
       "step": 18960
     },
     {
       "epoch": 0.3081997043102468,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4234,
+      "loss": 1.5053,
       "step": 18970
     },
     {
       "epoch": 0.30836217120761644,
-      "grad_norm": 1.7734375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4233,
+      "loss": 1.4682,
       "step": 18980
     },
     {
       "epoch": 0.3085246381049861,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.5008,
       "step": 18990
     },
     {
       "epoch": 0.30868710500235574,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.5446,
       "step": 19000
     },
     {
       "epoch": 0.30884957189972545,
-      "grad_norm": 3.609375,
+      "grad_norm": 4.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.5021,
       "step": 19010
     },
     {
       "epoch": 0.3090120387970951,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.5159,
       "step": 19020
     },
     {
       "epoch": 0.30917450569446475,
-      "grad_norm": 2.75,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4875,
       "step": 19030
     },
     {
       "epoch": 0.3093369725918344,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.476,
       "step": 19040
     },
     {
       "epoch": 0.30949943948920405,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4212,
+      "loss": 1.4759,
       "step": 19050
     },
     {
       "epoch": 0.30966190638657376,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.5231,
       "step": 19060
     },
     {
       "epoch": 0.3098243732839434,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.4622,
       "step": 19070
     },
     {
       "epoch": 0.30998684018131306,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4452,
+      "loss": 1.5316,
       "step": 19080
     },
     {
       "epoch": 0.3101493070786827,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.4695,
       "step": 19090
     },
     {
       "epoch": 0.31031177397605236,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4378,
+      "loss": 1.5141,
       "step": 19100
     },
     {
       "epoch": 0.310474240873422,
-      "grad_norm": 3.671875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.476,
       "step": 19110
     },
     {
       "epoch": 0.3106367077707917,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.5165,
       "step": 19120
     },
     {
       "epoch": 0.31079917466816137,
-      "grad_norm": 2.0,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4339,
+      "loss": 1.4787,
       "step": 19130
     },
     {
       "epoch": 0.310961641565531,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.4933,
       "step": 19140
     },
     {
       "epoch": 0.3111241084629007,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.4744,
       "step": 19150
     },
     {
       "epoch": 0.3112865753602703,
-      "grad_norm": 3.59375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.5261,
       "step": 19160
     },
     {
       "epoch": 0.31144904225764003,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4371,
+      "loss": 1.5077,
       "step": 19170
     },
     {
       "epoch": 0.3116115091550097,
-      "grad_norm": 2.25,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.5224,
       "step": 19180
     },
     {
       "epoch": 0.31177397605237933,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4291,
+      "loss": 1.5158,
       "step": 19190
     },
     {
       "epoch": 0.311936442949749,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.5043,
       "step": 19200
     },
     {
       "epoch": 0.31209890984711863,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4656,
+      "loss": 1.5496,
       "step": 19210
     },
     {
       "epoch": 0.3122613767444883,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.5036,
       "step": 19220
     },
     {
       "epoch": 0.312423843641858,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4963,
       "step": 19230
     },
     {
       "epoch": 0.31258631053922764,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4273,
+      "loss": 1.486,
       "step": 19240
     },
     {
       "epoch": 0.3127487774365973,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.5598,
       "step": 19250
     },
     {
       "epoch": 0.31291124433396694,
-      "grad_norm": 4.6875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4299,
+      "loss": 1.5494,
       "step": 19260
     },
     {
       "epoch": 0.3130737112313366,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.5033,
       "step": 19270
     },
     {
       "epoch": 0.3132361781287063,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4839,
       "step": 19280
     },
     {
       "epoch": 0.31339864502607595,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.4819,
       "step": 19290
     },
     {
       "epoch": 0.3135611119234456,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4274,
+      "loss": 1.5094,
       "step": 19300
     },
     {
       "epoch": 0.31372357882081525,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4343,
+      "loss": 1.4923,
       "step": 19310
     },
     {
       "epoch": 0.3138860457181849,
-      "grad_norm": 2.125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4282,
+      "loss": 1.5556,
       "step": 19320
     },
     {
       "epoch": 0.31404851261555455,
-      "grad_norm": 2.125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4231,
+      "loss": 1.473,
       "step": 19330
     },
     {
       "epoch": 0.31421097951292426,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.4871,
       "step": 19340
     },
     {
       "epoch": 0.3143734464102939,
-      "grad_norm": 2.515625,
+      "grad_norm": 4.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.4635,
       "step": 19350
     },
     {
       "epoch": 0.31453591330766356,
-      "grad_norm": 3.375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4841,
       "step": 19360
     },
     {
       "epoch": 0.3146983802050332,
-      "grad_norm": 2.625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4123,
+      "loss": 1.4956,
       "step": 19370
     },
     {
       "epoch": 0.31486084710240286,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.5303,
       "step": 19380
     },
     {
       "epoch": 0.31502331399977257,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.4893,
       "step": 19390
     },
     {
       "epoch": 0.3151857808971422,
-      "grad_norm": 3.375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4414,
+      "loss": 1.5307,
       "step": 19400
     },
     {
       "epoch": 0.31534824779451187,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4369,
+      "loss": 1.4664,
       "step": 19410
     },
     {
       "epoch": 0.3155107146918815,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.4806,
       "step": 19420
     },
     {
       "epoch": 0.3156731815892512,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4193,
+      "loss": 1.4922,
       "step": 19430
     },
     {
       "epoch": 0.3158356484866208,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4407,
+      "loss": 1.464,
       "step": 19440
     },
     {
       "epoch": 0.31599811538399053,
-      "grad_norm": 2.25,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.4556,
       "step": 19450
     },
     {
       "epoch": 0.3161605822813602,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4886,
       "step": 19460
     },
     {
       "epoch": 0.31632304917872983,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.418,
+      "loss": 1.5011,
       "step": 19470
     },
     {
       "epoch": 0.3164855160760995,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4333,
+      "loss": 1.4622,
       "step": 19480
     },
     {
       "epoch": 0.31664798297346913,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.431,
+      "loss": 1.48,
       "step": 19490
     },
     {
       "epoch": 0.31681044987083884,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.4599,
       "step": 19500
     },
     {
       "epoch": 0.3169729167682085,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4305,
+      "loss": 1.5179,
       "step": 19510
     },
     {
       "epoch": 0.31713538366557814,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4307,
+      "loss": 1.5129,
       "step": 19520
     },
     {
       "epoch": 0.3172978505629478,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4389,
+      "loss": 1.4584,
       "step": 19530
     },
     {
       "epoch": 0.31746031746031744,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4195,
+      "loss": 1.5333,
       "step": 19540
     },
     {
       "epoch": 0.3176227843576871,
-      "grad_norm": 1.2890625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4809,
       "step": 19550
     },
     {
       "epoch": 0.3177852512550568,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4158,
+      "loss": 1.5012,
       "step": 19560
     },
     {
       "epoch": 0.31794771815242645,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.5023,
       "step": 19570
     },
     {
       "epoch": 0.3181101850497961,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5265,
       "step": 19580
     },
     {
       "epoch": 0.31827265194716575,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4287,
+      "loss": 1.467,
       "step": 19590
     },
     {
       "epoch": 0.3184351188445354,
-      "grad_norm": 1.8515625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4484,
       "step": 19600
     },
     {
       "epoch": 0.3185975857419051,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.4602,
       "step": 19610
     },
     {
       "epoch": 0.31876005263927476,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.4734,
       "step": 19620
     },
     {
       "epoch": 0.3189225195366444,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4258,
+      "loss": 1.5008,
       "step": 19630
     },
     {
       "epoch": 0.31908498643401406,
-      "grad_norm": 4.1875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4854,
       "step": 19640
     },
     {
       "epoch": 0.3192474533313837,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.4892,
       "step": 19650
     },
     {
       "epoch": 0.3194099202287534,
-      "grad_norm": 4.9375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4258,
+      "loss": 1.5012,
       "step": 19660
     },
     {
       "epoch": 0.31957238712612307,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4181,
+      "loss": 1.5101,
       "step": 19670
     },
     {
       "epoch": 0.3197348540234927,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4748,
       "step": 19680
     },
     {
       "epoch": 0.31989732092086237,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4861,
       "step": 19690
     },
     {
       "epoch": 0.320059787818232,
-      "grad_norm": 3.3125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4924,
       "step": 19700
     },
     {
       "epoch": 0.3202222547156017,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.523,
       "step": 19710
     },
     {
       "epoch": 0.3203847216129714,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4505,
       "step": 19720
     },
     {
       "epoch": 0.32054718851034103,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4846,
       "step": 19730
     },
     {
       "epoch": 0.3207096554077107,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.5093,
       "step": 19740
     },
     {
       "epoch": 0.32087212230508033,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.4938,
       "step": 19750
     },
     {
       "epoch": 0.32103458920245,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.4886,
       "step": 19760
     },
     {
       "epoch": 0.3211970560998197,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4243,
+      "loss": 1.4904,
       "step": 19770
     },
     {
       "epoch": 0.32135952299718934,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4185,
+      "loss": 1.4918,
       "step": 19780
     },
     {
       "epoch": 0.321521989894559,
-      "grad_norm": 2.25,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4159,
+      "loss": 1.4752,
       "step": 19790
     },
     {
       "epoch": 0.32168445679192864,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4349,
       "step": 19800
     },
     {
       "epoch": 0.3218469236892983,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4664,
       "step": 19810
     },
     {
       "epoch": 0.32200939058666794,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.4887,
       "step": 19820
     },
     {
       "epoch": 0.32217185748403765,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4347,
+      "loss": 1.5024,
       "step": 19830
     },
     {
       "epoch": 0.3223343243814073,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.437,
+      "loss": 1.483,
       "step": 19840
     },
     {
       "epoch": 0.32249679127877695,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4266,
+      "loss": 1.4987,
       "step": 19850
     },
     {
       "epoch": 0.3226592581761466,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.5019,
       "step": 19860
     },
     {
       "epoch": 0.32282172507351625,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.5101,
       "step": 19870
     },
     {
       "epoch": 0.32298419197088596,
-      "grad_norm": 4.21875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.5381,
       "step": 19880
     },
     {
       "epoch": 0.3231466588682556,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.4607,
       "step": 19890
     },
     {
       "epoch": 0.32330912576562526,
-      "grad_norm": 1.7109375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4119,
+      "loss": 1.5158,
       "step": 19900
     },
     {
       "epoch": 0.3234715926629949,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.4001,
       "step": 19910
     },
     {
       "epoch": 0.32363405956036456,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.4926,
       "step": 19920
     },
     {
       "epoch": 0.3237965264577342,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.5338,
       "step": 19930
     },
     {
       "epoch": 0.3239589933551039,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4641,
       "step": 19940
     },
     {
       "epoch": 0.32412146025247357,
-      "grad_norm": 3.921875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4681,
       "step": 19950
     },
     {
       "epoch": 0.3242839271498432,
-      "grad_norm": 2.75,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4737,
       "step": 19960
     },
     {
       "epoch": 0.32444639404721287,
-      "grad_norm": 1.96875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4413,
+      "loss": 1.4763,
       "step": 19970
     },
     {
       "epoch": 0.3246088609445825,
-      "grad_norm": 2.0,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4216,
+      "loss": 1.4823,
       "step": 19980
     },
     {
       "epoch": 0.32477132784195223,
-      "grad_norm": 4.65625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4323,
+      "loss": 1.4863,
       "step": 19990
     },
     {
       "epoch": 0.3249337947393219,
-      "grad_norm": 2.75,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.5141,
       "step": 20000
     },
     {
       "epoch": 0.32509626163669153,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.5129,
       "step": 20010
     },
     {
       "epoch": 0.3252587285340612,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.4814,
       "step": 20020
     },
     {
       "epoch": 0.32542119543143083,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4264,
+      "loss": 1.4944,
       "step": 20030
     },
     {
       "epoch": 0.3255836623288005,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.5181,
       "step": 20040
     },
     {
       "epoch": 0.3257461292261702,
-      "grad_norm": 7.125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.5122,
       "step": 20050
     },
     {
       "epoch": 0.32590859612353984,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4695,
+      "loss": 1.501,
       "step": 20060
     },
     {
       "epoch": 0.3260710630209095,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.545,
       "step": 20070
     },
     {
       "epoch": 0.32623352991827914,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4373,
+      "loss": 1.4933,
       "step": 20080
     },
     {
       "epoch": 0.3263959968156488,
-      "grad_norm": 4.34375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4223,
+      "loss": 1.4756,
       "step": 20090
     },
     {
       "epoch": 0.3265584637130185,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4234,
+      "loss": 1.5126,
       "step": 20100
     },
     {
       "epoch": 0.32672093061038815,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.4431,
       "step": 20110
     },
     {
       "epoch": 0.3268833975077578,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4294,
+      "loss": 1.4341,
       "step": 20120
     },
     {
       "epoch": 0.32704586440512745,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4828,
       "step": 20130
     },
     {
       "epoch": 0.3272083313024971,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.518,
       "step": 20140
     },
     {
       "epoch": 0.32737079819986675,
-      "grad_norm": 1.71875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.4479,
       "step": 20150
     },
     {
       "epoch": 0.32753326509723646,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4266,
+      "loss": 1.5113,
       "step": 20160
     },
     {
       "epoch": 0.3276957319946061,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.5381,
       "step": 20170
     },
     {
       "epoch": 0.32785819889197576,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.5197,
       "step": 20180
     },
     {
       "epoch": 0.3280206657893454,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4304,
+      "loss": 1.4751,
       "step": 20190
     },
     {
       "epoch": 0.32818313268671506,
-      "grad_norm": 1.9296875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.5462,
       "step": 20200
     },
     {
       "epoch": 0.32834559958408477,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4256,
+      "loss": 1.4884,
       "step": 20210
     },
     {
       "epoch": 0.3285080664814544,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.4932,
       "step": 20220
     },
     {
       "epoch": 0.32867053337882407,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4261,
+      "loss": 1.5091,
       "step": 20230
     },
     {
       "epoch": 0.3288330002761937,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.5317,
       "step": 20240
     },
     {
       "epoch": 0.32899546717356337,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4163,
+      "loss": 1.5383,
       "step": 20250
     },
     {
       "epoch": 0.329157934070933,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4217,
+      "loss": 1.4721,
       "step": 20260
     },
     {
       "epoch": 0.32932040096830273,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4219,
+      "loss": 1.4688,
       "step": 20270
     },
     {
       "epoch": 0.3294828678656724,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.469,
       "step": 20280
     },
     {
       "epoch": 0.32964533476304203,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.4807,
       "step": 20290
     },
     {
       "epoch": 0.3298078016604117,
-      "grad_norm": 1.8125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4304,
+      "loss": 1.4681,
       "step": 20300
     },
     {
       "epoch": 0.32997026855778133,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4261,
+      "loss": 1.4476,
       "step": 20310
     },
     {
       "epoch": 0.33013273545515104,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.4211,
+      "loss": 1.4345,
       "step": 20320
     },
     {
       "epoch": 0.3302952023525207,
-      "grad_norm": 2.25,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4945,
       "step": 20330
     },
     {
       "epoch": 0.33045766924989034,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4303,
+      "loss": 1.5037,
       "step": 20340
     },
     {
       "epoch": 0.33062013614726,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4282,
+      "loss": 1.5322,
       "step": 20350
     },
     {
       "epoch": 0.33078260304462964,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4258,
+      "loss": 1.5023,
       "step": 20360
     },
     {
       "epoch": 0.3309450699419993,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.5049,
       "step": 20370
     },
     {
       "epoch": 0.331107536839369,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4148,
+      "loss": 1.5381,
       "step": 20380
     },
     {
       "epoch": 0.33127000373673865,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.5194,
       "step": 20390
     },
     {
       "epoch": 0.3314324706341083,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.479,
       "step": 20400
     },
     {
       "epoch": 0.33159493753147795,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4282,
+      "loss": 1.501,
       "step": 20410
     },
     {
       "epoch": 0.3317574044288476,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4797,
       "step": 20420
     },
     {
       "epoch": 0.3319198713262173,
-      "grad_norm": 1.5390625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4121,
+      "loss": 1.4415,
       "step": 20430
     },
     {
       "epoch": 0.33208233822358696,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4214,
+      "loss": 1.5148,
       "step": 20440
     },
     {
       "epoch": 0.3322448051209566,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.5377,
       "step": 20450
     },
     {
       "epoch": 0.33240727201832626,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4359,
+      "loss": 1.5087,
       "step": 20460
     },
     {
       "epoch": 0.3325697389156959,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4553,
+      "loss": 1.4686,
       "step": 20470
     },
     {
       "epoch": 0.33273220581306556,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.5613,
       "step": 20480
     },
     {
       "epoch": 0.33289467271043527,
-      "grad_norm": 2.5,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4412,
+      "loss": 1.4718,
       "step": 20490
     },
     {
       "epoch": 0.3330571396078049,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4195,
+      "loss": 1.5074,
       "step": 20500
     },
     {
       "epoch": 0.33321960650517457,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4317,
+      "loss": 1.4934,
       "step": 20510
     },
     {
       "epoch": 0.3333820734025442,
-      "grad_norm": 3.421875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4453,
+      "loss": 1.5028,
       "step": 20520
     },
     {
       "epoch": 0.33354454029991387,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.396,
+      "loss": 1.5056,
       "step": 20530
     },
     {
       "epoch": 0.3337070071972836,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4754,
       "step": 20540
     },
     {
       "epoch": 0.33386947409465323,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.5144,
       "step": 20550
     },
     {
       "epoch": 0.3340319409920229,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4306,
+      "loss": 1.5303,
       "step": 20560
     },
     {
       "epoch": 0.33419440788939253,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4365,
+      "loss": 1.5082,
       "step": 20570
     },
     {
       "epoch": 0.3343568747867622,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4293,
+      "loss": 1.4515,
       "step": 20580
     },
     {
       "epoch": 0.33451934168413183,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.5226,
       "step": 20590
     },
     {
       "epoch": 0.33468180858150154,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4267,
+      "loss": 1.5051,
       "step": 20600
     },
     {
       "epoch": 0.3348442754788712,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.4786,
       "step": 20610
     },
     {
       "epoch": 0.33500674237624084,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4996,
       "step": 20620
     },
     {
       "epoch": 0.3351692092736105,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.5143,
       "step": 20630
     },
     {
       "epoch": 0.33533167617098014,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.485,
       "step": 20640
     },
     {
       "epoch": 0.33549414306834985,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4291,
+      "loss": 1.5079,
       "step": 20650
     },
     {
       "epoch": 0.3356566099657195,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4267,
+      "loss": 1.5038,
       "step": 20660
     },
     {
       "epoch": 0.33581907686308915,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.431,
+      "loss": 1.458,
       "step": 20670
     },
     {
       "epoch": 0.3359815437604588,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.514,
       "step": 20680
     },
     {
       "epoch": 0.33614401065782845,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4386,
+      "loss": 1.5098,
       "step": 20690
     },
     {
       "epoch": 0.3363064775551981,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4459,
+      "loss": 1.4936,
       "step": 20700
     },
     {
       "epoch": 0.3364689444525678,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.4999,
       "step": 20710
     },
     {
       "epoch": 0.33663141134993746,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.4436,
       "step": 20720
     },
     {
       "epoch": 0.3367938782473071,
-      "grad_norm": 3.890625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4238,
+      "loss": 1.4751,
       "step": 20730
     },
     {
       "epoch": 0.33695634514467676,
-      "grad_norm": 2.3125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.5132,
       "step": 20740
     },
     {
       "epoch": 0.3371188120420464,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.5011,
       "step": 20750
     },
     {
       "epoch": 0.3372812789394161,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.5013,
       "step": 20760
     },
     {
       "epoch": 0.33744374583678577,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.512,
       "step": 20770
     },
     {
       "epoch": 0.3376062127341554,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4236,
+      "loss": 1.4859,
       "step": 20780
     },
     {
       "epoch": 0.33776867963152507,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4123,
+      "loss": 1.5325,
       "step": 20790
     },
     {
       "epoch": 0.3379311465288947,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.4853,
       "step": 20800
     },
     {
       "epoch": 0.33809361342626437,
-      "grad_norm": 4.34375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4691,
       "step": 20810
     },
     {
       "epoch": 0.3382560803236341,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.4604,
       "step": 20820
     },
     {
       "epoch": 0.33841854722100373,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.5002,
       "step": 20830
     },
     {
       "epoch": 0.3385810141183734,
-      "grad_norm": 3.25,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.4567,
       "step": 20840
     },
     {
       "epoch": 0.33874348101574303,
-      "grad_norm": 2.21875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5133,
       "step": 20850
     },
     {
       "epoch": 0.3389059479131127,
-      "grad_norm": 2.625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.5085,
       "step": 20860
     },
     {
       "epoch": 0.3390684148104824,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4594,
       "step": 20870
     },
     {
       "epoch": 0.33923088170785204,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4562,
       "step": 20880
     },
     {
       "epoch": 0.3393933486052217,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.462,
       "step": 20890
     },
     {
       "epoch": 0.33955581550259134,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.4973,
       "step": 20900
     },
     {
       "epoch": 0.339718282399961,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4341,
+      "loss": 1.4875,
       "step": 20910
     },
     {
       "epoch": 0.33988074929733064,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.505,
       "step": 20920
     },
     {
       "epoch": 0.34004321619470035,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.5019,
       "step": 20930
     },
     {
       "epoch": 0.34020568309207,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4138,
+      "loss": 1.5019,
       "step": 20940
     },
     {
       "epoch": 0.34036814998943965,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.5097,
       "step": 20950
     },
     {
       "epoch": 0.3405306168868093,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.5129,
       "step": 20960
     },
     {
       "epoch": 0.34069308378417895,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.5022,
       "step": 20970
     },
     {
       "epoch": 0.34085555068154866,
-      "grad_norm": 3.59375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4181,
+      "loss": 1.4911,
       "step": 20980
     },
     {
       "epoch": 0.3410180175789183,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.5298,
       "step": 20990
     },
     {
       "epoch": 0.34118048447628796,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.4876,
       "step": 21000
     },
     {
       "epoch": 0.3413429513736576,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.4683,
       "step": 21010
     },
     {
       "epoch": 0.34150541827102726,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.529,
       "step": 21020
     },
     {
       "epoch": 0.3416678851683969,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.4699,
       "step": 21030
     },
     {
       "epoch": 0.3418303520657666,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4193,
+      "loss": 1.4868,
       "step": 21040
     },
     {
       "epoch": 0.34199281896313627,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.4809,
       "step": 21050
     },
     {
       "epoch": 0.3421552858605059,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4961,
       "step": 21060
     },
     {
       "epoch": 0.34231775275787557,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.4536,
       "step": 21070
     },
     {
       "epoch": 0.3424802196552452,
-      "grad_norm": 3.0,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.5046,
       "step": 21080
     },
     {
       "epoch": 0.34264268655261493,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.4523,
       "step": 21090
     },
     {
       "epoch": 0.3428051534499846,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.5018,
       "step": 21100
     },
     {
       "epoch": 0.34296762034735423,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.427,
+      "loss": 1.4565,
       "step": 21110
     },
     {
       "epoch": 0.3431300872447239,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4064,
+      "loss": 1.4818,
       "step": 21120
     },
     {
       "epoch": 0.34329255414209353,
-      "grad_norm": 1.9609375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4472,
       "step": 21130
     },
     {
       "epoch": 0.3434550210394632,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.441,
+      "loss": 1.5413,
       "step": 21140
     },
     {
       "epoch": 0.3436174879368329,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4275,
+      "loss": 1.5244,
       "step": 21150
     },
     {
       "epoch": 0.34377995483420254,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.4684,
       "step": 21160
     },
     {
       "epoch": 0.3439424217315722,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4144,
+      "loss": 1.461,
       "step": 21170
     },
     {
       "epoch": 0.34410488862894184,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4627,
       "step": 21180
     },
     {
       "epoch": 0.3442673555263115,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.542,
       "step": 21190
     },
     {
       "epoch": 0.3444298224236812,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.444,
+      "loss": 1.5327,
       "step": 21200
     },
     {
       "epoch": 0.34459228932105085,
-      "grad_norm": 4.375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.4343,
       "step": 21210
     },
     {
       "epoch": 0.3447547562184205,
-      "grad_norm": 1.46875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.5094,
       "step": 21220
     },
     {
       "epoch": 0.34491722311579015,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.5008,
       "step": 21230
     },
     {
       "epoch": 0.3450796900131598,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.4592,
       "step": 21240
     },
     {
       "epoch": 0.34524215691052945,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4527,
+      "loss": 1.4436,
       "step": 21250
     },
     {
       "epoch": 0.34540462380789916,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.4801,
       "step": 21260
     },
     {
       "epoch": 0.3455670907052688,
-      "grad_norm": 2.625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4612,
       "step": 21270
     },
     {
       "epoch": 0.34572955760263846,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.4928,
       "step": 21280
     },
     {
       "epoch": 0.3458920245000081,
-      "grad_norm": 1.6015625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.5026,
       "step": 21290
     },
     {
       "epoch": 0.34605449139737776,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.4879,
       "step": 21300
     },
     {
       "epoch": 0.34621695829474747,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.5021,
       "step": 21310
     },
     {
       "epoch": 0.3463794251921171,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4966,
       "step": 21320
     },
     {
       "epoch": 0.34654189208948677,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4159,
+      "loss": 1.4864,
       "step": 21330
     },
     {
       "epoch": 0.3467043589868564,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4482,
+      "loss": 1.4396,
       "step": 21340
     },
     {
       "epoch": 0.34686682588422607,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.4902,
       "step": 21350
     },
     {
       "epoch": 0.3470292927815958,
-      "grad_norm": 2.625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4829,
       "step": 21360
     },
     {
       "epoch": 0.34719175967896543,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4404,
+      "loss": 1.4754,
       "step": 21370
     },
     {
       "epoch": 0.3473542265763351,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.5046,
       "step": 21380
     },
     {
       "epoch": 0.34751669347370473,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4215,
+      "loss": 1.4944,
       "step": 21390
     },
     {
       "epoch": 0.3476791603710744,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.43,
+      "loss": 1.5031,
       "step": 21400
     },
     {
       "epoch": 0.34784162726844403,
-      "grad_norm": 3.671875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4469,
       "step": 21410
     },
     {
       "epoch": 0.34800409416581374,
-      "grad_norm": 2.375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.4934,
       "step": 21420
     },
     {
       "epoch": 0.3481665610631834,
-      "grad_norm": 3.84375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4881,
       "step": 21430
     },
     {
       "epoch": 0.34832902796055304,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.5258,
       "step": 21440
     },
     {
       "epoch": 0.3484914948579227,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.4782,
       "step": 21450
     },
     {
       "epoch": 0.34865396175529234,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.4469,
       "step": 21460
     },
     {
       "epoch": 0.34881642865266205,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.5003,
       "step": 21470
     },
     {
       "epoch": 0.3489788955500317,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.5035,
       "step": 21480
     },
     {
       "epoch": 0.34914136244740135,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.4952,
       "step": 21490
     },
     {
       "epoch": 0.349303829344771,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.4975,
       "step": 21500
     },
     {
       "epoch": 0.34946629624214065,
-      "grad_norm": 2.34375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.4777,
       "step": 21510
     },
     {
       "epoch": 0.3496287631395103,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4655,
       "step": 21520
     },
     {
       "epoch": 0.34979123003688,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4587,
       "step": 21530
     },
     {
       "epoch": 0.34995369693424966,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4761,
       "step": 21540
     },
     {
       "epoch": 0.3501161638316193,
-      "grad_norm": 1.7734375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.517,
       "step": 21550
     },
     {
       "epoch": 0.35027863072898896,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.5153,
       "step": 21560
     },
     {
       "epoch": 0.3504410976263586,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4634,
       "step": 21570
     },
     {
       "epoch": 0.3506035645237283,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.426,
+      "loss": 1.5017,
       "step": 21580
     },
     {
       "epoch": 0.35076603142109797,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4295,
+      "loss": 1.4872,
       "step": 21590
     },
     {
       "epoch": 0.3509284983184676,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.4606,
       "step": 21600
     },
     {
       "epoch": 0.35109096521583727,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4277,
+      "loss": 1.4806,
       "step": 21610
     },
     {
       "epoch": 0.3512534321132069,
-      "grad_norm": 2.75,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.4835,
       "step": 21620
     },
     {
       "epoch": 0.35141589901057657,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4268,
+      "loss": 1.4852,
       "step": 21630
     },
     {
       "epoch": 0.3515783659079463,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4249,
+      "loss": 1.5433,
       "step": 21640
     },
     {
       "epoch": 0.35174083280531593,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.4878,
       "step": 21650
     },
     {
       "epoch": 0.3519032997026856,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.5451,
       "step": 21660
     },
     {
       "epoch": 0.35206576660005523,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4708,
       "step": 21670
     },
     {
       "epoch": 0.3522282334974249,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.4604,
       "step": 21680
     },
     {
       "epoch": 0.3523907003947946,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.5407,
       "step": 21690
     },
     {
       "epoch": 0.35255316729216424,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.5011,
       "step": 21700
     },
     {
       "epoch": 0.3527156341895339,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.4999,
       "step": 21710
     },
     {
       "epoch": 0.35287810108690354,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4219,
+      "loss": 1.5036,
       "step": 21720
     },
     {
       "epoch": 0.3530405679842732,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.5197,
       "step": 21730
     },
     {
       "epoch": 0.35320303488164284,
-      "grad_norm": 1.7109375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.4807,
       "step": 21740
     },
     {
       "epoch": 0.35336550177901255,
-      "grad_norm": 2.75,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.504,
       "step": 21750
     },
     {
       "epoch": 0.3535279686763822,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.4844,
       "step": 21760
     },
     {
       "epoch": 0.35369043557375185,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.4923,
       "step": 21770
     },
     {
       "epoch": 0.3538529024711215,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4174,
+      "loss": 1.4997,
       "step": 21780
     },
     {
       "epoch": 0.35401536936849115,
-      "grad_norm": 4.46875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.441,
       "step": 21790
     },
     {
       "epoch": 0.35417783626586086,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.488,
       "step": 21800
     },
     {
       "epoch": 0.3543403031632305,
-      "grad_norm": 3.203125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.5336,
       "step": 21810
     },
     {
       "epoch": 0.35450277006060016,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.5116,
       "step": 21820
     },
     {
       "epoch": 0.3546652369579698,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4803,
       "step": 21830
     },
     {
       "epoch": 0.35482770385533946,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.4925,
       "step": 21840
     },
     {
       "epoch": 0.3549901707527091,
-      "grad_norm": 2.21875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4731,
       "step": 21850
     },
     {
       "epoch": 0.3551526376500788,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4268,
+      "loss": 1.4963,
       "step": 21860
     },
     {
       "epoch": 0.35531510454744847,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4216,
+      "loss": 1.4373,
       "step": 21870
     },
     {
       "epoch": 0.3554775714448181,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.4761,
       "step": 21880
     },
     {
       "epoch": 0.35564003834218777,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.4521,
       "step": 21890
     },
     {
       "epoch": 0.3558025052395574,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.402,
+      "loss": 1.5091,
       "step": 21900
     },
     {
       "epoch": 0.3559649721369271,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.5009,
       "step": 21910
     },
     {
       "epoch": 0.3561274390342968,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.522,
       "step": 21920
     },
     {
       "epoch": 0.35628990593166643,
-      "grad_norm": 4.0,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.5002,
       "step": 21930
     },
     {
       "epoch": 0.3564523728290361,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.4394,
       "step": 21940
     },
     {
       "epoch": 0.35661483972640573,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.4791,
       "step": 21950
     },
     {
       "epoch": 0.3567773066237754,
-      "grad_norm": 2.28125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.5103,
       "step": 21960
     },
     {
       "epoch": 0.3569397735211451,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4308,
+      "loss": 1.4515,
       "step": 21970
     },
     {
       "epoch": 0.35710224041851474,
-      "grad_norm": 2.578125,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4235,
+      "loss": 1.4953,
       "step": 21980
     },
     {
       "epoch": 0.3572647073158844,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.4816,
       "step": 21990
     },
     {
       "epoch": 0.35742717421325404,
-      "grad_norm": 2.0,
+      "grad_norm": 13.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4992,
       "step": 22000
     },
     {
       "epoch": 0.3575896411106237,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.5149,
       "step": 22010
     },
     {
       "epoch": 0.3577521080079934,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.515,
       "step": 22020
     },
     {
       "epoch": 0.35791457490536305,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4316,
+      "loss": 1.4852,
       "step": 22030
     },
     {
       "epoch": 0.3580770418027327,
-      "grad_norm": 3.75,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.4627,
       "step": 22040
     },
     {
       "epoch": 0.35823950870010235,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.5163,
       "step": 22050
     },
     {
       "epoch": 0.358401975597472,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.5002,
       "step": 22060
     },
     {
       "epoch": 0.35856444249484165,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4627,
       "step": 22070
     },
     {
       "epoch": 0.35872690939221136,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.5151,
       "step": 22080
     },
     {
       "epoch": 0.358889376289581,
-      "grad_norm": 1.96875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.4821,
       "step": 22090
     },
     {
       "epoch": 0.35905184318695066,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4432,
+      "loss": 1.4634,
       "step": 22100
     },
     {
       "epoch": 0.3592143100843203,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4849,
       "step": 22110
     },
     {
       "epoch": 0.35937677698168996,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4656,
       "step": 22120
     },
     {
       "epoch": 0.35953924387905967,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.5094,
       "step": 22130
     },
     {
       "epoch": 0.3597017107764293,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4382,
+      "loss": 1.4484,
       "step": 22140
     },
     {
       "epoch": 0.35986417767379897,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.44,
       "step": 22150
     },
     {
       "epoch": 0.3600266445711686,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.4921,
       "step": 22160
     },
     {
       "epoch": 0.36018911146853827,
-      "grad_norm": 2.5,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4339,
+      "loss": 1.4929,
       "step": 22170
     },
     {
       "epoch": 0.3603515783659079,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.5144,
       "step": 22180
     },
     {
       "epoch": 0.3605140452632776,
-      "grad_norm": 1.765625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.5094,
       "step": 22190
     },
     {
       "epoch": 0.3606765121606473,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.5511,
       "step": 22200
     },
     {
       "epoch": 0.36083897905801693,
-      "grad_norm": 1.921875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.4726,
       "step": 22210
     },
     {
       "epoch": 0.3610014459553866,
-      "grad_norm": 3.84375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4255,
+      "loss": 1.4795,
       "step": 22220
     },
     {
       "epoch": 0.36116391285275623,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4216,
+      "loss": 1.4784,
       "step": 22230
     },
     {
       "epoch": 0.36132637975012594,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4632,
       "step": 22240
     },
     {
       "epoch": 0.3614888466474956,
-      "grad_norm": 2.125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4252,
+      "loss": 1.513,
       "step": 22250
     },
     {
       "epoch": 0.36165131354486524,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.4765,
       "step": 22260
     },
     {
       "epoch": 0.3618137804422349,
-      "grad_norm": 2.25,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.5198,
       "step": 22270
     },
     {
       "epoch": 0.36197624733960454,
-      "grad_norm": 4.09375,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.434,
+      "loss": 1.4551,
       "step": 22280
     },
     {
       "epoch": 0.3621387142369742,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4887,
       "step": 22290
     },
     {
       "epoch": 0.3623011811343439,
-      "grad_norm": 3.0,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4238,
+      "loss": 1.537,
       "step": 22300
     },
     {
       "epoch": 0.36246364803171355,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.433,
+      "loss": 1.5092,
       "step": 22310
     },
     {
       "epoch": 0.3626261149290832,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4828,
       "step": 22320
     },
     {
       "epoch": 0.36278858182645285,
-      "grad_norm": 3.296875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4436,
+      "loss": 1.4811,
       "step": 22330
     },
     {
       "epoch": 0.3629510487238225,
-      "grad_norm": 3.90625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4707,
       "step": 22340
     },
     {
       "epoch": 0.3631135156211922,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.5046,
       "step": 22350
     },
     {
       "epoch": 0.36327598251856186,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4158,
+      "loss": 1.4598,
       "step": 22360
     },
     {
       "epoch": 0.3634384494159315,
-      "grad_norm": 3.0,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4353,
+      "loss": 1.4649,
       "step": 22370
     },
     {
       "epoch": 0.36360091631330116,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.4506,
       "step": 22380
     },
     {
       "epoch": 0.3637633832106708,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4219,
+      "loss": 1.5106,
       "step": 22390
     },
     {
       "epoch": 0.36392585010804046,
-      "grad_norm": 1.84375,
+      "grad_norm": 4.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4311,
       "step": 22400
     },
     {
       "epoch": 0.36408831700541017,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.4825,
       "step": 22410
     },
     {
       "epoch": 0.3642507839027798,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4689,
       "step": 22420
     },
     {
       "epoch": 0.36441325080014947,
-      "grad_norm": 2.125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.4924,
       "step": 22430
     },
     {
       "epoch": 0.3645757176975191,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4286,
       "step": 22440
     },
     {
       "epoch": 0.36473818459488877,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4092,
+      "loss": 1.5314,
       "step": 22450
     },
     {
       "epoch": 0.3649006514922585,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4314,
+      "loss": 1.4861,
       "step": 22460
     },
     {
       "epoch": 0.3650631183896281,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.478,
       "step": 22470
     },
     {
       "epoch": 0.3652255852869978,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.4841,
       "step": 22480
     },
     {
       "epoch": 0.36538805218436743,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.4231,
       "step": 22490
     },
     {
       "epoch": 0.3655505190817371,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.4937,
       "step": 22500
     },
     {
       "epoch": 0.36571298597910673,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.5094,
       "step": 22510
     },
     {
       "epoch": 0.36587545287647644,
-      "grad_norm": 3.3125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4318,
+      "loss": 1.4383,
       "step": 22520
     },
     {
       "epoch": 0.3660379197738461,
-      "grad_norm": 2.875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4615,
       "step": 22530
     },
     {
       "epoch": 0.36620038667121574,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4472,
       "step": 22540
     },
     {
       "epoch": 0.3663628535685854,
-      "grad_norm": 4.375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.5527,
       "step": 22550
     },
     {
       "epoch": 0.36652532046595504,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4709,
       "step": 22560
     },
     {
       "epoch": 0.36668778736332475,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4936,
       "step": 22570
     },
     {
       "epoch": 0.3668502542606944,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.4592,
       "step": 22580
     },
     {
       "epoch": 0.36701272115806405,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.4719,
       "step": 22590
     },
     {
       "epoch": 0.3671751880554337,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4383,
+      "loss": 1.4692,
       "step": 22600
     },
     {
       "epoch": 0.36733765495280335,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4512,
       "step": 22610
     },
     {
       "epoch": 0.367500121850173,
-      "grad_norm": 5.40625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4369,
+      "loss": 1.4793,
       "step": 22620
     },
     {
       "epoch": 0.3676625887475427,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4692,
       "step": 22630
     },
     {
       "epoch": 0.36782505564491236,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.5043,
       "step": 22640
     },
     {
       "epoch": 0.367987522542282,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4182,
+      "loss": 1.4758,
       "step": 22650
     },
     {
       "epoch": 0.36814998943965166,
-      "grad_norm": 2.375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.4122,
       "step": 22660
     },
     {
       "epoch": 0.3683124563370213,
-      "grad_norm": 1.6015625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3872,
+      "loss": 1.4827,
       "step": 22670
     },
     {
       "epoch": 0.368474923234391,
-      "grad_norm": 2.75,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.5229,
       "step": 22680
     },
     {
       "epoch": 0.36863739013176067,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.5239,
       "step": 22690
     },
     {
       "epoch": 0.3687998570291303,
-      "grad_norm": 2.625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4685,
       "step": 22700
     },
     {
       "epoch": 0.36896232392649997,
-      "grad_norm": 1.78125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.5,
       "step": 22710
     },
     {
       "epoch": 0.3691247908238696,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.5158,
       "step": 22720
     },
     {
       "epoch": 0.36928725772123927,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.515,
       "step": 22730
     },
     {
       "epoch": 0.369449724618609,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.4995,
       "step": 22740
     },
     {
       "epoch": 0.3696121915159786,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.4809,
       "step": 22750
     },
     {
       "epoch": 0.3697746584133483,
-      "grad_norm": 1.7734375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4231,
+      "loss": 1.4475,
       "step": 22760
     },
     {
       "epoch": 0.36993712531071793,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.4465,
       "step": 22770
     },
     {
       "epoch": 0.3700995922080876,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.4956,
       "step": 22780
     },
     {
       "epoch": 0.3702620591054573,
-      "grad_norm": 3.5625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.463,
       "step": 22790
     },
     {
       "epoch": 0.37042452600282694,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4258,
+      "loss": 1.419,
       "step": 22800
     },
     {
       "epoch": 0.3705869929001966,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4681,
       "step": 22810
     },
     {
       "epoch": 0.37074945979756624,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.4715,
       "step": 22820
     },
     {
       "epoch": 0.3709119266949359,
-      "grad_norm": 1.5078125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.4949,
       "step": 22830
     },
     {
       "epoch": 0.37107439359230554,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4346,
+      "loss": 1.5096,
       "step": 22840
     },
     {
       "epoch": 0.37123686048967525,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4261,
+      "loss": 1.4395,
       "step": 22850
     },
     {
       "epoch": 0.3713993273870449,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4454,
       "step": 22860
     },
     {
       "epoch": 0.37156179428441455,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.5131,
       "step": 22870
     },
     {
       "epoch": 0.3717242611817842,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4856,
       "step": 22880
     },
     {
       "epoch": 0.37188672807915385,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.5145,
       "step": 22890
     },
     {
       "epoch": 0.37204919497652356,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.441,
+      "loss": 1.4899,
       "step": 22900
     },
     {
       "epoch": 0.3722116618738932,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4606,
       "step": 22910
     },
     {
       "epoch": 0.37237412877126286,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.5163,
       "step": 22920
     },
     {
       "epoch": 0.3725365956686325,
-      "grad_norm": 3.203125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.467,
       "step": 22930
     },
     {
       "epoch": 0.37269906256600216,
-      "grad_norm": 4.59375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4229,
       "step": 22940
     },
     {
       "epoch": 0.3728615294633718,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.5118,
       "step": 22950
     },
     {
       "epoch": 0.3730239963607415,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4947,
       "step": 22960
     },
     {
       "epoch": 0.37318646325811117,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4703,
       "step": 22970
     },
     {
       "epoch": 0.3733489301554808,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4327,
+      "loss": 1.4595,
       "step": 22980
     },
     {
       "epoch": 0.37351139705285047,
-      "grad_norm": 4.03125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4926,
       "step": 22990
     },
     {
       "epoch": 0.3736738639502201,
-      "grad_norm": 2.25,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.4467,
       "step": 23000
     },
     {
       "epoch": 0.3738363308475898,
-      "grad_norm": 2.75,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.4943,
       "step": 23010
     },
     {
       "epoch": 0.3739987977449595,
-      "grad_norm": 2.25,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4508,
       "step": 23020
     },
     {
       "epoch": 0.3741612646423291,
-      "grad_norm": 2.28125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.4433,
       "step": 23030
     },
     {
       "epoch": 0.3743237315396988,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4973,
       "step": 23040
     },
     {
       "epoch": 0.37448619843706843,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.4842,
       "step": 23050
     },
     {
       "epoch": 0.37464866533443814,
-      "grad_norm": 3.125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.4563,
       "step": 23060
     },
     {
       "epoch": 0.3748111322318078,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4298,
+      "loss": 1.4824,
       "step": 23070
     },
     {
       "epoch": 0.37497359912917744,
-      "grad_norm": 3.71875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4697,
       "step": 23080
     },
     {
       "epoch": 0.3751360660265471,
-      "grad_norm": 2.25,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4872,
       "step": 23090
     },
     {
       "epoch": 0.37529853292391674,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4768,
       "step": 23100
     },
     {
       "epoch": 0.3754609998212864,
-      "grad_norm": 2.75,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4163,
+      "loss": 1.5121,
       "step": 23110
     },
     {
       "epoch": 0.3756234667186561,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4765,
       "step": 23120
     },
     {
       "epoch": 0.37578593361602575,
-      "grad_norm": 3.765625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.4551,
       "step": 23130
     },
     {
       "epoch": 0.3759484005133954,
-      "grad_norm": 2.625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.5403,
       "step": 23140
     },
     {
       "epoch": 0.37611086741076505,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4784,
       "step": 23150
     },
     {
       "epoch": 0.3762733343081347,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.4619,
       "step": 23160
     },
     {
       "epoch": 0.3764358012055044,
-      "grad_norm": 1.7734375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4723,
       "step": 23170
     },
     {
       "epoch": 0.37659826810287406,
-      "grad_norm": 3.0,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.4744,
       "step": 23180
     },
     {
       "epoch": 0.3767607350002437,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4108,
+      "loss": 1.4413,
       "step": 23190
     },
     {
       "epoch": 0.37692320189761336,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.4892,
       "step": 23200
     },
     {
       "epoch": 0.377085668794983,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4135,
+      "loss": 1.522,
       "step": 23210
     },
     {
       "epoch": 0.37724813569235266,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4941,
       "step": 23220
     },
     {
       "epoch": 0.37741060258972237,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.4754,
       "step": 23230
     },
     {
       "epoch": 0.377573069487092,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3791,
+      "loss": 1.5097,
       "step": 23240
     },
     {
       "epoch": 0.37773553638446167,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4238,
+      "loss": 1.465,
       "step": 23250
     },
     {
       "epoch": 0.3778980032818313,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4685,
       "step": 23260
     },
     {
       "epoch": 0.37806047017920097,
-      "grad_norm": 1.71875,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4914,
       "step": 23270
     },
     {
       "epoch": 0.3782229370765707,
-      "grad_norm": 3.15625,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4137,
+      "loss": 1.4937,
       "step": 23280
     },
     {
       "epoch": 0.3783854039739403,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.5094,
       "step": 23290
     },
     {
       "epoch": 0.37854787087131,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.4677,
       "step": 23300
     },
     {
       "epoch": 0.37871033776867963,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4751,
       "step": 23310
     },
     {
       "epoch": 0.3788728046660493,
-      "grad_norm": 1.609375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4933,
       "step": 23320
     },
     {
       "epoch": 0.37903527156341893,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.4804,
       "step": 23330
     },
     {
       "epoch": 0.37919773846078864,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.4878,
       "step": 23340
     },
     {
       "epoch": 0.3793602053581583,
-      "grad_norm": 2.625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.4533,
       "step": 23350
     },
     {
       "epoch": 0.37952267225552794,
-      "grad_norm": 3.140625,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.4443,
       "step": 23360
     },
     {
       "epoch": 0.3796851391528976,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.5184,
       "step": 23370
     },
     {
       "epoch": 0.37984760605026724,
-      "grad_norm": 3.828125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4838,
       "step": 23380
     },
     {
       "epoch": 0.38001007294763695,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4227,
+      "loss": 1.4481,
       "step": 23390
     },
     {
       "epoch": 0.3801725398450066,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.485,
       "step": 23400
     },
     {
       "epoch": 0.38033500674237625,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.4245,
       "step": 23410
     },
     {
       "epoch": 0.3804974736397459,
-      "grad_norm": 2.125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.4789,
       "step": 23420
     },
     {
       "epoch": 0.38065994053711555,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3808,
+      "loss": 1.5191,
       "step": 23430
     },
     {
       "epoch": 0.3808224074344852,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4105,
+      "loss": 1.4832,
       "step": 23440
     },
     {
       "epoch": 0.3809848743318549,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.4865,
       "step": 23450
     },
     {
       "epoch": 0.38114734122922456,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.5058,
       "step": 23460
     },
     {
       "epoch": 0.3813098081265942,
-      "grad_norm": 2.375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4264,
+      "loss": 1.4469,
       "step": 23470
     },
     {
       "epoch": 0.38147227502396386,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4714,
       "step": 23480
     },
     {
       "epoch": 0.3816347419213335,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.4646,
       "step": 23490
     },
     {
       "epoch": 0.3817972088187032,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.5087,
       "step": 23500
     },
     {
       "epoch": 0.38195967571607287,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4831,
       "step": 23510
     },
     {
       "epoch": 0.3821221426134425,
-      "grad_norm": 2.0,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4247,
+      "loss": 1.5174,
       "step": 23520
     },
     {
       "epoch": 0.38228460951081217,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4779,
       "step": 23530
     },
     {
       "epoch": 0.3824470764081818,
-      "grad_norm": 2.625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4517,
       "step": 23540
     },
     {
       "epoch": 0.38260954330555147,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.4905,
       "step": 23550
     },
     {
       "epoch": 0.3827720102029212,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.5061,
       "step": 23560
     },
     {
       "epoch": 0.3829344771002908,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.4955,
       "step": 23570
     },
     {
       "epoch": 0.3830969439976605,
-      "grad_norm": 2.375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.4614,
       "step": 23580
     },
     {
       "epoch": 0.38325941089503013,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.4763,
       "step": 23590
     },
     {
       "epoch": 0.3834218777923998,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4769,
       "step": 23600
     },
     {
       "epoch": 0.3835843446897695,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.4416,
       "step": 23610
     },
     {
       "epoch": 0.38374681158713914,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4347,
+      "loss": 1.4362,
       "step": 23620
     },
     {
       "epoch": 0.3839092784845088,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4323,
+      "loss": 1.4504,
       "step": 23630
     },
     {
       "epoch": 0.38407174538187844,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4265,
+      "loss": 1.4884,
       "step": 23640
     },
     {
       "epoch": 0.3842342122792481,
-      "grad_norm": 4.5,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4702,
       "step": 23650
     },
     {
       "epoch": 0.38439667917661774,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4748,
       "step": 23660
     },
     {
       "epoch": 0.38455914607398745,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.452,
       "step": 23670
     },
     {
       "epoch": 0.3847216129713571,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.465,
       "step": 23680
     },
     {
       "epoch": 0.38488407986872675,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.4436,
       "step": 23690
     },
     {
       "epoch": 0.3850465467660964,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.4783,
       "step": 23700
     },
     {
       "epoch": 0.38520901366346605,
-      "grad_norm": 2.75,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4402,
+      "loss": 1.4218,
       "step": 23710
     },
     {
       "epoch": 0.38537148056083576,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.5118,
       "step": 23720
     },
     {
       "epoch": 0.3855339474582054,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.4697,
       "step": 23730
     },
     {
       "epoch": 0.38569641435557506,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.4973,
       "step": 23740
     },
     {
       "epoch": 0.3858588812529447,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.4934,
       "step": 23750
     },
     {
       "epoch": 0.38602134815031436,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4438,
+      "loss": 1.4759,
       "step": 23760
     },
     {
       "epoch": 0.386183815047684,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4553,
       "step": 23770
     },
     {
       "epoch": 0.3863462819450537,
-      "grad_norm": 4.71875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.4258,
       "step": 23780
     },
     {
       "epoch": 0.38650874884242337,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4421,
+      "loss": 1.4413,
       "step": 23790
     },
     {
       "epoch": 0.386671215739793,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.5094,
       "step": 23800
     },
     {
       "epoch": 0.38683368263716267,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.4652,
       "step": 23810
     },
     {
       "epoch": 0.3869961495345323,
-      "grad_norm": 4.46875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.4604,
       "step": 23820
     },
     {
       "epoch": 0.387158616431902,
-      "grad_norm": 2.671875,
+      "grad_norm": 4.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4486,
       "step": 23830
     },
     {
       "epoch": 0.3873210833292717,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.4988,
       "step": 23840
     },
     {
       "epoch": 0.3874835502266413,
-      "grad_norm": 1.90625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4852,
       "step": 23850
     },
     {
       "epoch": 0.387646017124011,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4223,
+      "loss": 1.447,
       "step": 23860
     },
     {
       "epoch": 0.38780848402138063,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4344,
+      "loss": 1.4202,
       "step": 23870
     },
     {
       "epoch": 0.3879709509187503,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.5014,
       "step": 23880
     },
     {
       "epoch": 0.38813341781612,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.5383,
       "step": 23890
     },
     {
       "epoch": 0.38829588471348964,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.4812,
       "step": 23900
     },
     {
       "epoch": 0.3884583516108593,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3972,
+      "loss": 1.4563,
       "step": 23910
     },
     {
       "epoch": 0.38862081850822894,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4243,
+      "loss": 1.4879,
       "step": 23920
     },
     {
       "epoch": 0.3887832854055986,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.497,
       "step": 23930
     },
     {
       "epoch": 0.3889457523029683,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4703,
       "step": 23940
     },
     {
       "epoch": 0.38910821920033795,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.423,
       "step": 23950
     },
     {
       "epoch": 0.3892706860977076,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4269,
+      "loss": 1.451,
       "step": 23960
     },
     {
       "epoch": 0.38943315299507725,
-      "grad_norm": 3.75,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4187,
+      "loss": 1.5052,
       "step": 23970
     },
     {
       "epoch": 0.3895956198924469,
-      "grad_norm": 2.75,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.4763,
       "step": 23980
     },
     {
       "epoch": 0.38975808678981655,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4277,
+      "loss": 1.4702,
       "step": 23990
     },
     {
       "epoch": 0.38992055368718626,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4276,
+      "loss": 1.4751,
       "step": 24000
     },
     {
       "epoch": 0.3900830205845559,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4323,
       "step": 24010
     },
     {
       "epoch": 0.39024548748192556,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.4591,
       "step": 24020
     },
     {
       "epoch": 0.3904079543792952,
-      "grad_norm": 4.65625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4301,
+      "loss": 1.5014,
       "step": 24030
     },
     {
       "epoch": 0.39057042127666486,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4269,
+      "loss": 1.4686,
       "step": 24040
     },
     {
       "epoch": 0.39073288817403456,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.4974,
       "step": 24050
     },
     {
       "epoch": 0.3908953550714042,
-      "grad_norm": 2.125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.4624,
       "step": 24060
     },
     {
       "epoch": 0.39105782196877387,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4367,
+      "loss": 1.5464,
       "step": 24070
     },
     {
       "epoch": 0.3912202888661435,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4535,
       "step": 24080
     },
     {
       "epoch": 0.39138275576351317,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.5168,
       "step": 24090
     },
     {
       "epoch": 0.3915452226608828,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.4682,
       "step": 24100
     },
     {
       "epoch": 0.3917076895582525,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.4882,
       "step": 24110
     },
     {
       "epoch": 0.3918701564556222,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4587,
       "step": 24120
     },
     {
       "epoch": 0.3920326233529918,
-      "grad_norm": 3.375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.5447,
       "step": 24130
     },
     {
       "epoch": 0.3921950902503615,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4718,
       "step": 24140
     },
     {
       "epoch": 0.39235755714773113,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.4773,
       "step": 24150
     },
     {
       "epoch": 0.39252002404510083,
-      "grad_norm": 4.03125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.4844,
       "step": 24160
     },
     {
       "epoch": 0.3926824909424705,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.4701,
       "step": 24170
     },
     {
       "epoch": 0.39284495783984014,
-      "grad_norm": 1.8046875,
+      "grad_norm": 4.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4108,
+      "loss": 1.4824,
       "step": 24180
     },
     {
       "epoch": 0.3930074247372098,
-      "grad_norm": 5.34375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4374,
+      "loss": 1.4574,
       "step": 24190
     },
     {
       "epoch": 0.39316989163457944,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4379,
+      "loss": 1.449,
       "step": 24200
     },
     {
       "epoch": 0.3933323585319491,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.5063,
       "step": 24210
     },
     {
       "epoch": 0.3934948254293188,
-      "grad_norm": 3.953125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4578,
       "step": 24220
     },
     {
       "epoch": 0.39365729232668845,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.5364,
       "step": 24230
     },
     {
       "epoch": 0.3938197592240581,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4743,
       "step": 24240
     },
     {
       "epoch": 0.39398222612142775,
-      "grad_norm": 2.078125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.4677,
       "step": 24250
     },
     {
       "epoch": 0.3941446930187974,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4343,
+      "loss": 1.4983,
       "step": 24260
     },
     {
       "epoch": 0.3943071599161671,
-      "grad_norm": 2.390625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.4395,
       "step": 24270
     },
     {
       "epoch": 0.39446962681353676,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.4175,
       "step": 24280
     },
     {
       "epoch": 0.3946320937109064,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.4963,
       "step": 24290
     },
     {
       "epoch": 0.39479456060827606,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4373,
+      "loss": 1.4937,
       "step": 24300
     },
     {
       "epoch": 0.3949570275056457,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4138,
+      "loss": 1.4889,
       "step": 24310
     },
     {
       "epoch": 0.39511949440301536,
-      "grad_norm": 2.203125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.5164,
       "step": 24320
     },
     {
       "epoch": 0.39528196130038507,
-      "grad_norm": 2.875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.437,
+      "loss": 1.4972,
       "step": 24330
     },
     {
       "epoch": 0.3954444281977547,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.474,
       "step": 24340
     },
     {
       "epoch": 0.39560689509512437,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4254,
+      "loss": 1.4634,
       "step": 24350
     },
     {
       "epoch": 0.395769361992494,
-      "grad_norm": 1.875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3816,
+      "loss": 1.519,
       "step": 24360
     },
     {
       "epoch": 0.39593182888986367,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.4857,
       "step": 24370
     },
     {
       "epoch": 0.3960942957872334,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.4937,
       "step": 24380
     },
     {
       "epoch": 0.396256762684603,
-      "grad_norm": 4.875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.448,
       "step": 24390
     },
     {
       "epoch": 0.3964192295819727,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.4426,
       "step": 24400
     },
     {
       "epoch": 0.3965816964793423,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.5127,
       "step": 24410
     },
     {
       "epoch": 0.396744163376712,
-      "grad_norm": 2.375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4651,
       "step": 24420
     },
     {
       "epoch": 0.39690663027408163,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4283,
+      "loss": 1.46,
       "step": 24430
     },
     {
       "epoch": 0.39706909717145134,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.438,
       "step": 24440
     },
     {
       "epoch": 0.397231564068821,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4177,
+      "loss": 1.4768,
       "step": 24450
     },
     {
       "epoch": 0.39739403096619064,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4681,
       "step": 24460
     },
     {
       "epoch": 0.3975564978635603,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.5307,
       "step": 24470
     },
     {
       "epoch": 0.39771896476092994,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4412,
+      "loss": 1.5242,
       "step": 24480
     },
     {
       "epoch": 0.39788143165829964,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.4403,
+      "loss": 1.5155,
       "step": 24490
     },
     {
       "epoch": 0.3980438985556693,
-      "grad_norm": 2.5,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3889,
+      "loss": 1.4901,
       "step": 24500
     },
     {
       "epoch": 0.39820636545303895,
-      "grad_norm": 2.25,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.4922,
       "step": 24510
     },
     {
       "epoch": 0.3983688323504086,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4703,
       "step": 24520
     },
     {
       "epoch": 0.39853129924777825,
-      "grad_norm": 3.78125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4587,
       "step": 24530
     },
     {
       "epoch": 0.3986937661451479,
-      "grad_norm": 3.484375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4523,
       "step": 24540
     },
     {
       "epoch": 0.3988562330425176,
-      "grad_norm": 2.015625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.474,
       "step": 24550
     },
     {
       "epoch": 0.39901869993988726,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.5017,
       "step": 24560
     },
     {
       "epoch": 0.3991811668372569,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.491,
       "step": 24570
     },
     {
       "epoch": 0.39934363373462656,
-      "grad_norm": 2.125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.4784,
       "step": 24580
     },
     {
       "epoch": 0.3995061006319962,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.43,
+      "loss": 1.4829,
       "step": 24590
     },
     {
       "epoch": 0.3996685675293659,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4207,
+      "loss": 1.4564,
       "step": 24600
     },
     {
       "epoch": 0.39983103442673557,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.4435,
       "step": 24610
     },
     {
       "epoch": 0.3999935013241052,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.4507,
       "step": 24620
     },
     {
       "epoch": 0.40015596822147487,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.5028,
       "step": 24630
     },
     {
       "epoch": 0.4003184351188445,
-      "grad_norm": 2.75,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4372,
+      "loss": 1.4782,
       "step": 24640
     },
     {
       "epoch": 0.40048090201621417,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4149,
       "step": 24650
     },
     {
       "epoch": 0.4006433689135839,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4903,
       "step": 24660
     },
     {
       "epoch": 0.4008058358109535,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.4924,
       "step": 24670
     },
     {
       "epoch": 0.4009683027083232,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4265,
+      "loss": 1.4697,
       "step": 24680
     },
     {
       "epoch": 0.4011307696056928,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.4606,
       "step": 24690
     },
     {
       "epoch": 0.4012932365030625,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.4194,
       "step": 24700
     },
     {
       "epoch": 0.4014557034004322,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.436,
       "step": 24710
     },
     {
       "epoch": 0.40161817029780184,
-      "grad_norm": 3.84375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4945,
       "step": 24720
     },
     {
       "epoch": 0.4017806371951715,
-      "grad_norm": 2.578125,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.466,
       "step": 24730
     },
     {
       "epoch": 0.40194310409254114,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.5159,
       "step": 24740
     },
     {
       "epoch": 0.4021055709899108,
-      "grad_norm": 2.265625,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.4302,
       "step": 24750
     },
     {
       "epoch": 0.4022680378872805,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.4517,
       "step": 24760
     },
     {
       "epoch": 0.40243050478465014,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4216,
+      "loss": 1.4319,
       "step": 24770
     },
     {
       "epoch": 0.4025929716820198,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.4588,
       "step": 24780
     },
     {
       "epoch": 0.40275543857938945,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4215,
+      "loss": 1.4629,
       "step": 24790
     },
     {
       "epoch": 0.4029179054767591,
-      "grad_norm": 2.125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.5375,
       "step": 24800
     },
     {
       "epoch": 0.40308037237412875,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4972,
       "step": 24810
     },
     {
       "epoch": 0.40324283927149845,
-      "grad_norm": 3.953125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4719,
       "step": 24820
     },
     {
       "epoch": 0.4034053061688681,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4855,
       "step": 24830
     },
     {
       "epoch": 0.40356777306623776,
-      "grad_norm": 1.8046875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.4489,
       "step": 24840
     },
     {
       "epoch": 0.4037302399636074,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4941,
       "step": 24850
     },
     {
       "epoch": 0.40389270686097706,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.4603,
       "step": 24860
     },
     {
       "epoch": 0.40405517375834676,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4313,
+      "loss": 1.4629,
       "step": 24870
     },
     {
       "epoch": 0.4042176406557164,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.496,
       "step": 24880
     },
     {
       "epoch": 0.40438010755308607,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4653,
       "step": 24890
     },
     {
       "epoch": 0.4045425744504557,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.4894,
       "step": 24900
     },
     {
       "epoch": 0.40470504134782537,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4253,
+      "loss": 1.4535,
       "step": 24910
     },
     {
       "epoch": 0.404867508245195,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4334,
+      "loss": 1.5177,
       "step": 24920
     },
     {
       "epoch": 0.4050299751425647,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4288,
+      "loss": 1.4477,
       "step": 24930
     },
     {
       "epoch": 0.4051924420399344,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.5117,
       "step": 24940
     },
     {
       "epoch": 0.405354908937304,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4618,
       "step": 24950
     },
     {
       "epoch": 0.4055173758346737,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4774,
       "step": 24960
     },
     {
       "epoch": 0.4056798427320433,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4345,
+      "loss": 1.5064,
       "step": 24970
     },
     {
       "epoch": 0.40584230962941303,
-      "grad_norm": 2.75,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4231,
+      "loss": 1.4463,
       "step": 24980
     },
     {
       "epoch": 0.4060047765267827,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4174,
+      "loss": 1.4936,
       "step": 24990
     },
     {
       "epoch": 0.40616724342415234,
-      "grad_norm": 2.125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.4413,
       "step": 25000
     },
     {
       "epoch": 0.406329710321522,
-      "grad_norm": 1.8515625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.4001,
       "step": 25010
     },
     {
       "epoch": 0.40649217721889164,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.4774,
       "step": 25020
     },
     {
       "epoch": 0.4066546441162613,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.4695,
       "step": 25030
     },
     {
       "epoch": 0.406817111013631,
-      "grad_norm": 2.25,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.459,
       "step": 25040
     },
     {
       "epoch": 0.40697957791100065,
-      "grad_norm": 2.625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4255,
+      "loss": 1.4139,
       "step": 25050
     },
     {
       "epoch": 0.4071420448083703,
-      "grad_norm": 1.703125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4755,
       "step": 25060
     },
     {
       "epoch": 0.40730451170573995,
-      "grad_norm": 2.0,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.4889,
       "step": 25070
     },
     {
       "epoch": 0.4074669786031096,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4459,
+      "loss": 1.4686,
       "step": 25080
     },
     {
       "epoch": 0.4076294455004793,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.459,
       "step": 25090
     },
     {
       "epoch": 0.40779191239784895,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4177,
+      "loss": 1.4237,
       "step": 25100
     },
     {
       "epoch": 0.4079543792952186,
-      "grad_norm": 3.0,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.43,
+      "loss": 1.4974,
       "step": 25110
     },
     {
       "epoch": 0.40811684619258826,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4254,
+      "loss": 1.4892,
       "step": 25120
     },
     {
       "epoch": 0.4082793130899579,
-      "grad_norm": 1.9296875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4306,
+      "loss": 1.4749,
       "step": 25130
     },
     {
       "epoch": 0.40844177998732756,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4193,
+      "loss": 1.4401,
       "step": 25140
     },
     {
       "epoch": 0.40860424688469726,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.4614,
       "step": 25150
     },
     {
       "epoch": 0.4087667137820669,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.4817,
       "step": 25160
     },
     {
       "epoch": 0.40892918067943657,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.4259,
       "step": 25170
     },
     {
       "epoch": 0.4090916475768062,
-      "grad_norm": 2.25,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4404,
       "step": 25180
     },
     {
       "epoch": 0.40925411447417587,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4322,
+      "loss": 1.466,
       "step": 25190
     },
     {
       "epoch": 0.4094165813715456,
-      "grad_norm": 1.875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4767,
       "step": 25200
     },
     {
       "epoch": 0.4095790482689152,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.5004,
       "step": 25210
     },
     {
       "epoch": 0.4097415151662849,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4542,
       "step": 25220
     },
     {
       "epoch": 0.4099039820636545,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4331,
+      "loss": 1.4743,
       "step": 25230
     },
     {
       "epoch": 0.4100664489610242,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4786,
       "step": 25240
     },
     {
       "epoch": 0.41022891585839383,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.4549,
       "step": 25250
     },
     {
       "epoch": 0.41039138275576353,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.4825,
       "step": 25260
     },
     {
       "epoch": 0.4105538496531332,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.4925,
       "step": 25270
     },
     {
       "epoch": 0.41071631655050284,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4472,
       "step": 25280
     },
     {
       "epoch": 0.4108787834478725,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4168,
+      "loss": 1.4972,
       "step": 25290
     },
     {
       "epoch": 0.41104125034524214,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4231,
+      "loss": 1.4657,
       "step": 25300
     },
     {
       "epoch": 0.41120371724261184,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4865,
       "step": 25310
     },
     {
       "epoch": 0.4113661841399815,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4711,
       "step": 25320
     },
     {
       "epoch": 0.41152865103735115,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.488,
       "step": 25330
     },
     {
       "epoch": 0.4116911179347208,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.4658,
       "step": 25340
     },
     {
       "epoch": 0.41185358483209045,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4861,
       "step": 25350
     },
     {
       "epoch": 0.4120160517294601,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.4935,
       "step": 25360
     },
     {
       "epoch": 0.4121785186268298,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4843,
       "step": 25370
     },
     {
       "epoch": 0.41234098552419945,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.4711,
       "step": 25380
     },
     {
       "epoch": 0.4125034524215691,
-      "grad_norm": 1.6796875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4926,
       "step": 25390
     },
     {
       "epoch": 0.41266591931893876,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4815,
       "step": 25400
     },
     {
       "epoch": 0.4128283862163084,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.4993,
       "step": 25410
     },
     {
       "epoch": 0.4129908531136781,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.439,
+      "loss": 1.449,
       "step": 25420
     },
     {
       "epoch": 0.41315332001104776,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.4279,
       "step": 25430
     },
     {
       "epoch": 0.4133157869084174,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4234,
+      "loss": 1.4626,
       "step": 25440
     },
     {
       "epoch": 0.41347825380578707,
-      "grad_norm": 2.125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.4813,
       "step": 25450
     },
     {
       "epoch": 0.4136407207031567,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.3994,
       "step": 25460
     },
     {
       "epoch": 0.41380318760052637,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.461,
       "step": 25470
     },
     {
       "epoch": 0.4139656544978961,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.4357,
       "step": 25480
     },
     {
       "epoch": 0.4141281213952657,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4237,
+      "loss": 1.5137,
       "step": 25490
     },
     {
       "epoch": 0.4142905882926354,
-      "grad_norm": 3.90625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4743,
       "step": 25500
     },
     {
       "epoch": 0.414453055190005,
-      "grad_norm": 2.625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.5217,
       "step": 25510
     },
     {
       "epoch": 0.4146155220873747,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.5034,
       "step": 25520
     },
     {
       "epoch": 0.4147779889847444,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4208,
+      "loss": 1.4152,
       "step": 25530
     },
     {
       "epoch": 0.41494045588211403,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.4265,
       "step": 25540
     },
     {
       "epoch": 0.4151029227794837,
-      "grad_norm": 1.484375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.4176,
       "step": 25550
     },
     {
       "epoch": 0.41526538967685334,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.4025,
       "step": 25560
     },
     {
       "epoch": 0.415427856574223,
-      "grad_norm": 3.0,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.4717,
       "step": 25570
     },
     {
       "epoch": 0.41559032347159264,
-      "grad_norm": 5.71875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4391,
+      "loss": 1.4684,
       "step": 25580
     },
     {
       "epoch": 0.41575279036896234,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.402,
+      "loss": 1.4335,
       "step": 25590
     },
     {
       "epoch": 0.415915257266332,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.501,
       "step": 25600
     },
     {
       "epoch": 0.41607772416370165,
-      "grad_norm": 4.34375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.4735,
       "step": 25610
     },
     {
       "epoch": 0.4162401910610713,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4243,
       "step": 25620
     },
     {
       "epoch": 0.41640265795844095,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4567,
       "step": 25630
     },
     {
       "epoch": 0.41656512485581065,
-      "grad_norm": 4.125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4108,
+      "loss": 1.4696,
       "step": 25640
     },
     {
       "epoch": 0.4167275917531803,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4489,
       "step": 25650
     },
     {
       "epoch": 0.41689005865054996,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.4625,
       "step": 25660
     },
     {
       "epoch": 0.4170525255479196,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.4652,
       "step": 25670
     },
     {
       "epoch": 0.41721499244528926,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.4631,
       "step": 25680
     },
     {
       "epoch": 0.4173774593426589,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.5057,
       "step": 25690
     },
     {
       "epoch": 0.4175399262400286,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.5138,
       "step": 25700
     },
     {
       "epoch": 0.41770239313739826,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.4899,
       "step": 25710
     },
     {
       "epoch": 0.4178648600347679,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4706,
       "step": 25720
     },
     {
       "epoch": 0.41802732693213757,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4764,
       "step": 25730
     },
     {
       "epoch": 0.4181897938295072,
-      "grad_norm": 2.25,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4959,
       "step": 25740
     },
     {
       "epoch": 0.4183522607268769,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.4311,
       "step": 25750
     },
     {
       "epoch": 0.4185147276242466,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4327,
       "step": 25760
     },
     {
       "epoch": 0.4186771945216162,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4336,
+      "loss": 1.5248,
       "step": 25770
     },
     {
       "epoch": 0.4188396614189859,
-      "grad_norm": 3.609375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4705,
       "step": 25780
     },
     {
       "epoch": 0.4190021283163555,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4451,
       "step": 25790
     },
     {
       "epoch": 0.4191645952137252,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4233,
+      "loss": 1.4631,
       "step": 25800
     },
     {
       "epoch": 0.4193270621110949,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4533,
       "step": 25810
     },
     {
       "epoch": 0.41948952900846453,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.4425,
       "step": 25820
     },
     {
       "epoch": 0.4196519959058342,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.4735,
       "step": 25830
     },
     {
       "epoch": 0.41981446280320384,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.46,
       "step": 25840
     },
     {
       "epoch": 0.4199769297005735,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4227,
+      "loss": 1.477,
       "step": 25850
     },
     {
       "epoch": 0.4201393965979432,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.4529,
       "step": 25860
     },
     {
       "epoch": 0.42030186349531284,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4174,
+      "loss": 1.4429,
       "step": 25870
     },
     {
       "epoch": 0.4204643303926825,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4734,
       "step": 25880
     },
     {
       "epoch": 0.42062679729005215,
-      "grad_norm": 2.75,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4159,
+      "loss": 1.4958,
       "step": 25890
     },
     {
       "epoch": 0.4207892641874218,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4258,
       "step": 25900
     },
     {
       "epoch": 0.42095173108479145,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4214,
       "step": 25910
     },
     {
       "epoch": 0.42111419798216115,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.4996,
       "step": 25920
     },
     {
       "epoch": 0.4212766648795308,
-      "grad_norm": 1.9140625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4439,
       "step": 25930
     },
     {
       "epoch": 0.42143913177690046,
-      "grad_norm": 2.625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.4182,
       "step": 25940
     },
     {
       "epoch": 0.4216015986742701,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3798,
+      "loss": 1.4926,
       "step": 25950
     },
     {
       "epoch": 0.42176406557163976,
-      "grad_norm": 1.96875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4581,
       "step": 25960
     },
     {
       "epoch": 0.42192653246900946,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.4384,
       "step": 25970
     },
     {
       "epoch": 0.4220889993663791,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4087,
       "step": 25980
     },
     {
       "epoch": 0.42225146626374876,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.455,
       "step": 25990
     },
     {
       "epoch": 0.4224139331611184,
-      "grad_norm": 2.75,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4159,
+      "loss": 1.4712,
       "step": 26000
     },
     {
       "epoch": 0.42257640005848807,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4715,
       "step": 26010
     },
     {
       "epoch": 0.4227388669558577,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.4587,
       "step": 26020
     },
     {
       "epoch": 0.4229013338532274,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4346,
+      "loss": 1.4735,
       "step": 26030
     },
     {
       "epoch": 0.4230638007505971,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.508,
       "step": 26040
     },
     {
       "epoch": 0.4232262676479667,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.5006,
       "step": 26050
     },
     {
       "epoch": 0.4233887345453364,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4406,
       "step": 26060
     },
     {
       "epoch": 0.423551201442706,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.4873,
       "step": 26070
     },
     {
       "epoch": 0.42371366834007573,
-      "grad_norm": 1.7578125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4965,
       "step": 26080
     },
     {
       "epoch": 0.4238761352374454,
-      "grad_norm": 3.859375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.5404,
       "step": 26090
     },
     {
       "epoch": 0.42403860213481503,
-      "grad_norm": 2.609375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.4145,
       "step": 26100
     },
     {
       "epoch": 0.4242010690321847,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4255,
+      "loss": 1.4372,
       "step": 26110
     },
     {
       "epoch": 0.42436353592955434,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.4873,
       "step": 26120
     },
     {
       "epoch": 0.424526002826924,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.4854,
       "step": 26130
     },
     {
       "epoch": 0.4246884697242937,
-      "grad_norm": 1.7578125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.4469,
       "step": 26140
     },
     {
       "epoch": 0.42485093662166334,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.4744,
       "step": 26150
     },
     {
       "epoch": 0.425013403519033,
-      "grad_norm": 3.84375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4481,
       "step": 26160
     },
     {
       "epoch": 0.42517587041640265,
-      "grad_norm": 4.34375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4751,
       "step": 26170
     },
     {
       "epoch": 0.4253383373137723,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4401,
+      "loss": 1.4581,
       "step": 26180
     },
     {
       "epoch": 0.425500804211142,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4417,
       "step": 26190
     },
     {
       "epoch": 0.42566327110851165,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4795,
       "step": 26200
     },
     {
       "epoch": 0.4258257380058813,
-      "grad_norm": 2.421875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.4506,
       "step": 26210
     },
     {
       "epoch": 0.42598820490325096,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.4531,
       "step": 26220
     },
     {
       "epoch": 0.4261506718006206,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.5267,
       "step": 26230
     },
     {
       "epoch": 0.42631313869799026,
-      "grad_norm": 2.53125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.4601,
       "step": 26240
     },
     {
       "epoch": 0.42647560559535996,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4282,
+      "loss": 1.4302,
       "step": 26250
     },
     {
       "epoch": 0.4266380724927296,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.5108,
       "step": 26260
     },
     {
       "epoch": 0.42680053939009927,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4662,
       "step": 26270
     },
     {
       "epoch": 0.4269630062874689,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.47,
       "step": 26280
     },
     {
       "epoch": 0.42712547318483857,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4647,
       "step": 26290
     },
     {
       "epoch": 0.4272879400822083,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4092,
+      "loss": 1.3805,
       "step": 26300
     },
     {
       "epoch": 0.4274504069795779,
-      "grad_norm": 3.0,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.4691,
       "step": 26310
     },
     {
       "epoch": 0.4276128738769476,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.4694,
       "step": 26320
     },
     {
       "epoch": 0.4277753407743172,
-      "grad_norm": 3.921875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.4475,
       "step": 26330
     },
     {
       "epoch": 0.4279378076716869,
-      "grad_norm": 3.125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.5073,
       "step": 26340
     },
     {
       "epoch": 0.4281002745690565,
-      "grad_norm": 2.5,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4363,
+      "loss": 1.4429,
       "step": 26350
     },
     {
       "epoch": 0.42826274146642623,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4288,
+      "loss": 1.4869,
       "step": 26360
     },
     {
       "epoch": 0.4284252083637959,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.4498,
       "step": 26370
     },
     {
       "epoch": 0.42858767526116553,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4285,
+      "loss": 1.4739,
       "step": 26380
     },
     {
       "epoch": 0.4287501421585352,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.4144,
       "step": 26390
     },
     {
       "epoch": 0.42891260905590484,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3815,
+      "loss": 1.449,
       "step": 26400
     },
     {
       "epoch": 0.42907507595327454,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.484,
       "step": 26410
     },
     {
       "epoch": 0.4292375428506442,
-      "grad_norm": 1.8515625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3808,
+      "loss": 1.4595,
       "step": 26420
     },
     {
       "epoch": 0.42940000974801384,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4781,
       "step": 26430
     },
     {
       "epoch": 0.4295624766453835,
-      "grad_norm": 3.515625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.4667,
       "step": 26440
     },
     {
       "epoch": 0.42972494354275315,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4676,
       "step": 26450
     },
     {
       "epoch": 0.42988741044012285,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4495,
       "step": 26460
     },
     {
       "epoch": 0.4300498773374925,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.4489,
       "step": 26470
     },
     {
       "epoch": 0.43021234423486215,
-      "grad_norm": 1.609375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4224,
+      "loss": 1.4889,
       "step": 26480
     },
     {
       "epoch": 0.4303748111322318,
-      "grad_norm": 2.0,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4274,
+      "loss": 1.4871,
       "step": 26490
     },
     {
       "epoch": 0.43053727802960146,
-      "grad_norm": 2.421875,
+      "grad_norm": 4.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.5001,
       "step": 26500
     },
     {
       "epoch": 0.4306997449269711,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.4117,
       "step": 26510
     },
     {
       "epoch": 0.4308622118243408,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.4993,
       "step": 26520
     },
     {
       "epoch": 0.43102467872171046,
-      "grad_norm": 2.078125,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.4257,
       "step": 26530
     },
     {
       "epoch": 0.4311871456190801,
-      "grad_norm": 3.9375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4653,
       "step": 26540
     },
     {
       "epoch": 0.43134961251644977,
-      "grad_norm": 2.25,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4564,
       "step": 26550
     },
     {
       "epoch": 0.4315120794138194,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.425,
       "step": 26560
     },
     {
       "epoch": 0.4316745463111891,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.4705,
       "step": 26570
     },
     {
       "epoch": 0.4318370132085588,
-      "grad_norm": 1.890625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4185,
+      "loss": 1.4317,
       "step": 26580
     },
     {
       "epoch": 0.4319994801059284,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4829,
       "step": 26590
     },
     {
       "epoch": 0.4321619470032981,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.4532,
       "step": 26600
     },
     {
       "epoch": 0.4323244139006677,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.4678,
       "step": 26610
     },
     {
       "epoch": 0.4324868807980374,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.4882,
       "step": 26620
     },
     {
       "epoch": 0.4326493476954071,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4321,
+      "loss": 1.4734,
       "step": 26630
     },
     {
       "epoch": 0.43281181459277673,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4246,
+      "loss": 1.4595,
       "step": 26640
     },
     {
       "epoch": 0.4329742814901464,
-      "grad_norm": 3.0,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4221,
+      "loss": 1.4807,
       "step": 26650
     },
     {
       "epoch": 0.43313674838751604,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4933,
       "step": 26660
     },
     {
       "epoch": 0.4332992152848857,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4649,
       "step": 26670
     },
     {
       "epoch": 0.4334616821822554,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4917,
       "step": 26680
     },
     {
       "epoch": 0.43362414907962504,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4329,
       "step": 26690
     },
     {
       "epoch": 0.4337866159769947,
-      "grad_norm": 2.875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.5236,
       "step": 26700
     },
     {
       "epoch": 0.43394908287436434,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4789,
       "step": 26710
     },
     {
       "epoch": 0.434111549771734,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4284,
+      "loss": 1.4657,
       "step": 26720
     },
     {
       "epoch": 0.43427401666910365,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4144,
+      "loss": 1.4663,
       "step": 26730
     },
     {
       "epoch": 0.43443648356647335,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.4504,
       "step": 26740
     },
     {
       "epoch": 0.434598950463843,
-      "grad_norm": 4.125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4796,
       "step": 26750
     },
     {
       "epoch": 0.43476141736121265,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4839,
       "step": 26760
     },
     {
       "epoch": 0.4349238842585823,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.4836,
       "step": 26770
     },
     {
       "epoch": 0.43508635115595196,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.5016,
       "step": 26780
     },
     {
       "epoch": 0.43524881805332166,
-      "grad_norm": 2.125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4255,
+      "loss": 1.5141,
       "step": 26790
     },
     {
       "epoch": 0.4354112849506913,
-      "grad_norm": 3.625,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4912,
       "step": 26800
     },
     {
       "epoch": 0.43557375184806096,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.447,
       "step": 26810
     },
     {
       "epoch": 0.4357362187454306,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4341,
+      "loss": 1.4971,
       "step": 26820
     },
     {
       "epoch": 0.43589868564280027,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.458,
       "step": 26830
     },
     {
       "epoch": 0.4360611525401699,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.479,
       "step": 26840
     },
     {
       "epoch": 0.4362236194375396,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.496,
       "step": 26850
     },
     {
       "epoch": 0.4363860863349093,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.486,
       "step": 26860
     },
     {
       "epoch": 0.4365485532322789,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4206,
+      "loss": 1.4979,
       "step": 26870
     },
     {
       "epoch": 0.4367110201296486,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4821,
       "step": 26880
     },
     {
       "epoch": 0.4368734870270182,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.4502,
       "step": 26890
     },
     {
       "epoch": 0.43703595392438793,
-      "grad_norm": 2.46875,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.5036,
       "step": 26900
     },
     {
       "epoch": 0.4371984208217576,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.4989,
       "step": 26910
     },
     {
       "epoch": 0.43736088771912723,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.396,
+      "loss": 1.5183,
       "step": 26920
     },
     {
       "epoch": 0.4375233546164969,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4294,
+      "loss": 1.4087,
       "step": 26930
     },
     {
       "epoch": 0.43768582151386654,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.427,
       "step": 26940
     },
     {
       "epoch": 0.4378482884112362,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4696,
       "step": 26950
     },
     {
       "epoch": 0.4380107553086059,
-      "grad_norm": 2.75,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.415,
+      "loss": 1.5088,
       "step": 26960
     },
     {
       "epoch": 0.43817322220597554,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4214,
+      "loss": 1.4461,
       "step": 26970
     },
     {
       "epoch": 0.4383356891033452,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.484,
       "step": 26980
     },
     {
       "epoch": 0.43849815600071484,
-      "grad_norm": 2.75,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.4843,
       "step": 26990
     },
     {
       "epoch": 0.4386606228980845,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4254,
+      "loss": 1.4379,
       "step": 27000
     },
     {
       "epoch": 0.4388230897954542,
-      "grad_norm": 1.9765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.4493,
       "step": 27010
     },
     {
       "epoch": 0.43898555669282385,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4505,
+      "loss": 1.465,
       "step": 27020
     },
     {
       "epoch": 0.4391480235901935,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.473,
       "step": 27030
     },
     {
       "epoch": 0.43931049048756315,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.4631,
       "step": 27040
     },
     {
       "epoch": 0.4394729573849328,
-      "grad_norm": 1.75,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4476,
       "step": 27050
     },
     {
       "epoch": 0.43963542428230246,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.4437,
       "step": 27060
     },
     {
       "epoch": 0.43979789117967216,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4254,
+      "loss": 1.4758,
       "step": 27070
     },
     {
       "epoch": 0.4399603580770418,
-      "grad_norm": 2.625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.4843,
       "step": 27080
     },
     {
       "epoch": 0.44012282497441146,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4221,
+      "loss": 1.4228,
       "step": 27090
     },
     {
       "epoch": 0.4402852918717811,
-      "grad_norm": 3.265625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.477,
       "step": 27100
     },
     {
       "epoch": 0.44044775876915077,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.5041,
       "step": 27110
     },
     {
       "epoch": 0.44061022566652047,
-      "grad_norm": 2.75,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.4757,
       "step": 27120
     },
     {
       "epoch": 0.4407726925638901,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4121,
+      "loss": 1.4352,
       "step": 27130
     },
     {
       "epoch": 0.4409351594612598,
-      "grad_norm": 3.0,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4358,
+      "loss": 1.466,
       "step": 27140
     },
     {
       "epoch": 0.4410976263586294,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.429,
+      "loss": 1.4465,
       "step": 27150
     },
     {
       "epoch": 0.4412600932559991,
-      "grad_norm": 2.25,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4542,
       "step": 27160
     },
     {
       "epoch": 0.4414225601533687,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.4357,
       "step": 27170
     },
     {
       "epoch": 0.44158502705073843,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4342,
       "step": 27180
     },
     {
       "epoch": 0.4417474939481081,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4262,
+      "loss": 1.4348,
       "step": 27190
     },
     {
       "epoch": 0.44190996084547773,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4799,
       "step": 27200
     },
     {
       "epoch": 0.4420724277428474,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4205,
+      "loss": 1.4794,
       "step": 27210
     },
     {
       "epoch": 0.44223489464021704,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.443,
       "step": 27220
     },
     {
       "epoch": 0.44239736153758674,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4808,
       "step": 27230
     },
     {
       "epoch": 0.4425598284349564,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4772,
       "step": 27240
     },
     {
       "epoch": 0.44272229533232604,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.4235,
       "step": 27250
     },
     {
       "epoch": 0.4428847622296957,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4572,
       "step": 27260
     },
     {
       "epoch": 0.44304722912706535,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4351,
       "step": 27270
     },
     {
       "epoch": 0.443209696024435,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4703,
       "step": 27280
     },
     {
       "epoch": 0.4433721629218047,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4867,
       "step": 27290
     },
     {
       "epoch": 0.44353462981917435,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.4665,
       "step": 27300
     },
     {
       "epoch": 0.443697096716544,
-      "grad_norm": 1.78125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3803,
+      "loss": 1.4984,
       "step": 27310
     },
     {
       "epoch": 0.44385956361391365,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.4516,
       "step": 27320
     },
     {
       "epoch": 0.4440220305112833,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4559,
       "step": 27330
     },
     {
       "epoch": 0.444184497408653,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.418,
       "step": 27340
     },
     {
       "epoch": 0.44434696430602266,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4749,
       "step": 27350
     },
     {
       "epoch": 0.4445094312033923,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4537,
       "step": 27360
     },
     {
       "epoch": 0.44467189810076196,
-      "grad_norm": 4.03125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4353,
+      "loss": 1.4435,
       "step": 27370
     },
     {
       "epoch": 0.4448343649981316,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4349,
+      "loss": 1.4119,
       "step": 27380
     },
     {
       "epoch": 0.44499683189550127,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.5098,
       "step": 27390
     },
     {
       "epoch": 0.44515929879287097,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4295,
+      "loss": 1.4361,
       "step": 27400
     },
     {
       "epoch": 0.4453217656902406,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.5032,
       "step": 27410
     },
     {
       "epoch": 0.4454842325876103,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4573,
       "step": 27420
     },
     {
       "epoch": 0.4456466994849799,
-      "grad_norm": 3.875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.5029,
       "step": 27430
     },
     {
       "epoch": 0.4458091663823496,
-      "grad_norm": 1.8359375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.4576,
       "step": 27440
     },
     {
       "epoch": 0.4459716332797193,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.5079,
       "step": 27450
     },
     {
       "epoch": 0.44613410017708893,
-      "grad_norm": 1.84375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.4276,
       "step": 27460
     },
     {
       "epoch": 0.4462965670744586,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.4559,
       "step": 27470
     },
     {
       "epoch": 0.44645903397182823,
-      "grad_norm": 2.625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4595,
       "step": 27480
     },
     {
       "epoch": 0.4466215008691979,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.5022,
       "step": 27490
     },
     {
       "epoch": 0.44678396776656754,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4691,
       "step": 27500
     },
     {
       "epoch": 0.44694643466393724,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.4404,
       "step": 27510
     },
     {
       "epoch": 0.4471089015613069,
-      "grad_norm": 3.125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4496,
       "step": 27520
     },
     {
       "epoch": 0.44727136845867654,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4402,
+      "loss": 1.4506,
       "step": 27530
     },
     {
       "epoch": 0.4474338353560462,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4138,
+      "loss": 1.514,
       "step": 27540
     },
     {
       "epoch": 0.44759630225341585,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4515,
       "step": 27550
     },
     {
       "epoch": 0.44775876915078555,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.431,
+      "loss": 1.4629,
       "step": 27560
     },
     {
       "epoch": 0.4479212360481552,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4673,
       "step": 27570
     },
     {
       "epoch": 0.44808370294552485,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.494,
       "step": 27580
     },
     {
       "epoch": 0.4482461698428945,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4307,
+      "loss": 1.464,
       "step": 27590
     },
     {
       "epoch": 0.44840863674026415,
-      "grad_norm": 3.0,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.442,
       "step": 27600
     },
     {
       "epoch": 0.4485711036376338,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.3885,
       "step": 27610
     },
     {
       "epoch": 0.4487335705350035,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.5091,
       "step": 27620
     },
     {
       "epoch": 0.44889603743237316,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.4486,
       "step": 27630
     },
     {
       "epoch": 0.4490585043297428,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4392,
       "step": 27640
     },
     {
       "epoch": 0.44922097122711246,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.4781,
       "step": 27650
     },
     {
       "epoch": 0.4493834381244821,
-      "grad_norm": 2.875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.447,
       "step": 27660
     },
     {
       "epoch": 0.4495459050218518,
-      "grad_norm": 1.7578125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.466,
       "step": 27670
     },
     {
       "epoch": 0.4497083719192215,
-      "grad_norm": 1.4296875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4841,
       "step": 27680
     },
     {
       "epoch": 0.4498708388165911,
-      "grad_norm": 2.5,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4261,
+      "loss": 1.4651,
       "step": 27690
     },
     {
       "epoch": 0.4500333057139608,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.5113,
       "step": 27700
     },
     {
       "epoch": 0.4501957726113304,
-      "grad_norm": 1.9296875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4474,
       "step": 27710
     },
     {
       "epoch": 0.4503582395087001,
-      "grad_norm": 1.84375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4316,
       "step": 27720
     },
     {
       "epoch": 0.4505207064060698,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4375,
       "step": 27730
     },
     {
       "epoch": 0.45068317330343943,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4063,
       "step": 27740
     },
     {
       "epoch": 0.4508456402008091,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4399,
       "step": 27750
     },
     {
       "epoch": 0.45100810709817873,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4808,
       "step": 27760
     },
     {
       "epoch": 0.4511705739955484,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.4419,
       "step": 27770
     },
     {
       "epoch": 0.4513330408929181,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4034,
+      "loss": 1.4357,
       "step": 27780
     },
     {
       "epoch": 0.45149550779028774,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4311,
+      "loss": 1.4508,
       "step": 27790
     },
     {
       "epoch": 0.4516579746876574,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4333,
+      "loss": 1.4866,
       "step": 27800
     },
     {
       "epoch": 0.45182044158502704,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.4598,
       "step": 27810
     },
     {
       "epoch": 0.4519829084823967,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4821,
       "step": 27820
     },
     {
       "epoch": 0.45214537537976635,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.4003,
       "step": 27830
     },
     {
       "epoch": 0.45230784227713605,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.402,
+      "loss": 1.4707,
       "step": 27840
     },
     {
       "epoch": 0.4524703091745057,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.463,
       "step": 27850
     },
     {
       "epoch": 0.45263277607187535,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.4598,
       "step": 27860
     },
     {
       "epoch": 0.452795242969245,
-      "grad_norm": 2.125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.4316,
       "step": 27870
     },
     {
       "epoch": 0.45295770986661466,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.5065,
       "step": 27880
     },
     {
       "epoch": 0.45312017676398436,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4582,
       "step": 27890
     },
     {
       "epoch": 0.453282643661354,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.4108,
       "step": 27900
     },
     {
       "epoch": 0.45344511055872366,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.4478,
       "step": 27910
     },
     {
       "epoch": 0.4536075774560933,
-      "grad_norm": 1.875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.4672,
       "step": 27920
     },
     {
       "epoch": 0.45377004435346296,
-      "grad_norm": 1.4765625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.519,
       "step": 27930
     },
     {
       "epoch": 0.4539325112508326,
-      "grad_norm": 1.796875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.4767,
       "step": 27940
     },
     {
       "epoch": 0.4540949781482023,
-      "grad_norm": 3.125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4558,
       "step": 27950
     },
     {
       "epoch": 0.454257445045572,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.5049,
       "step": 27960
     },
     {
       "epoch": 0.4544199119429416,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4064,
+      "loss": 1.4982,
       "step": 27970
     },
     {
       "epoch": 0.4545823788403113,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4357,
       "step": 27980
     },
     {
       "epoch": 0.4547448457376809,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.4596,
       "step": 27990
     },
     {
       "epoch": 0.45490731263505063,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4324,
+      "loss": 1.4714,
       "step": 28000
     },
     {
       "epoch": 0.4550697795324203,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.4502,
       "step": 28010
     },
     {
       "epoch": 0.45523224642978993,
-      "grad_norm": 1.4140625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4566,
       "step": 28020
     },
     {
       "epoch": 0.4553947133271596,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4115,
       "step": 28030
     },
     {
       "epoch": 0.45555718022452923,
-      "grad_norm": 1.875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4531,
       "step": 28040
     },
     {
       "epoch": 0.4557196471218989,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4671,
       "step": 28050
     },
     {
       "epoch": 0.4558821140192686,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4229,
       "step": 28060
     },
     {
       "epoch": 0.45604458091663824,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4633,
       "step": 28070
     },
     {
       "epoch": 0.4562070478140079,
-      "grad_norm": 1.96875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4785,
       "step": 28080
     },
     {
       "epoch": 0.45636951471137754,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.423,
       "step": 28090
     },
     {
       "epoch": 0.4565319816087472,
-      "grad_norm": 2.09375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4144,
+      "loss": 1.4436,
       "step": 28100
     },
     {
       "epoch": 0.4566944485061169,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4435,
       "step": 28110
     },
     {
       "epoch": 0.45685691540348655,
-      "grad_norm": 3.25,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3994,
+      "loss": 1.4796,
       "step": 28120
     },
     {
       "epoch": 0.4570193823008562,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4206,
+      "loss": 1.4682,
       "step": 28130
     },
     {
       "epoch": 0.45718184919822585,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.5019,
       "step": 28140
     },
     {
       "epoch": 0.4573443160955955,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4664,
       "step": 28150
     },
     {
       "epoch": 0.4575067829929652,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4494,
       "step": 28160
     },
     {
       "epoch": 0.45766924989033486,
-      "grad_norm": 2.375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3967,
+      "loss": 1.508,
       "step": 28170
     },
     {
       "epoch": 0.4578317167877045,
-      "grad_norm": 1.5625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.4531,
       "step": 28180
     },
     {
       "epoch": 0.45799418368507416,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4206,
+      "loss": 1.4386,
       "step": 28190
     },
     {
       "epoch": 0.4581566505824438,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.466,
       "step": 28200
     },
     {
       "epoch": 0.45831911747981346,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4191,
+      "loss": 1.5022,
       "step": 28210
     },
     {
       "epoch": 0.45848158437718317,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3809,
+      "loss": 1.3998,
       "step": 28220
     },
     {
       "epoch": 0.4586440512745528,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.4511,
       "step": 28230
     },
     {
       "epoch": 0.4588065181719225,
-      "grad_norm": 2.875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.5038,
       "step": 28240
     },
     {
       "epoch": 0.4589689850692921,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4952,
       "step": 28250
     },
     {
       "epoch": 0.4591314519666618,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4714,
       "step": 28260
     },
     {
       "epoch": 0.4592939188640315,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4702,
       "step": 28270
     },
     {
       "epoch": 0.45945638576140113,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4103,
+      "loss": 1.4633,
       "step": 28280
     },
     {
       "epoch": 0.4596188526587708,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.4909,
       "step": 28290
     },
     {
       "epoch": 0.45978131955614043,
-      "grad_norm": 3.25,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.5247,
       "step": 28300
     },
     {
       "epoch": 0.4599437864535101,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4145,
       "step": 28310
     },
     {
       "epoch": 0.46010625335087973,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.479,
       "step": 28320
     },
     {
       "epoch": 0.46026872024824944,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.4616,
       "step": 28330
     },
     {
       "epoch": 0.4604311871456191,
-      "grad_norm": 3.5,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4324,
+      "loss": 1.5002,
       "step": 28340
     },
     {
       "epoch": 0.46059365404298874,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.3845,
       "step": 28350
     },
     {
       "epoch": 0.4607561209403584,
-      "grad_norm": 5.9375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.4375,
       "step": 28360
     },
     {
       "epoch": 0.46091858783772804,
-      "grad_norm": 1.7734375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4664,
       "step": 28370
     },
     {
       "epoch": 0.46108105473509775,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.4556,
       "step": 28380
     },
     {
       "epoch": 0.4612435216324674,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.4163,
       "step": 28390
     },
     {
       "epoch": 0.46140598852983705,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4078,
       "step": 28400
     },
     {
       "epoch": 0.4615684554272067,
-      "grad_norm": 2.609375,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4182,
+      "loss": 1.4632,
       "step": 28410
     },
     {
       "epoch": 0.46173092232457635,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.5143,
       "step": 28420
     },
     {
       "epoch": 0.461893389221946,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.4316,
       "step": 28430
     },
     {
       "epoch": 0.4620558561193157,
-      "grad_norm": 5.25,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4449,
       "step": 28440
     },
     {
       "epoch": 0.46221832301668536,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.4747,
       "step": 28450
     },
     {
       "epoch": 0.462380789914055,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.4614,
       "step": 28460
     },
     {
       "epoch": 0.46254325681142466,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.3881,
       "step": 28470
     },
     {
       "epoch": 0.4627057237087943,
-      "grad_norm": 2.03125,
+      "grad_norm": 4.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.4733,
       "step": 28480
     },
     {
       "epoch": 0.462868190606164,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.5109,
       "step": 28490
     },
     {
       "epoch": 0.46303065750353367,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4914,
       "step": 28500
     },
     {
       "epoch": 0.4631931244009033,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4123,
+      "loss": 1.447,
       "step": 28510
     },
     {
       "epoch": 0.463355591298273,
-      "grad_norm": 1.6015625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.4202,
       "step": 28520
     },
     {
       "epoch": 0.4635180581956426,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4547,
       "step": 28530
     },
     {
       "epoch": 0.4636805250930123,
-      "grad_norm": 2.125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.491,
       "step": 28540
     },
     {
       "epoch": 0.463842991990382,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.4764,
       "step": 28550
     },
     {
       "epoch": 0.46400545888775163,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.5319,
       "step": 28560
     },
     {
       "epoch": 0.4641679257851213,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.4689,
       "step": 28570
     },
     {
       "epoch": 0.46433039268249093,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.4233,
       "step": 28580
     },
     {
       "epoch": 0.4644928595798606,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3714,
+      "loss": 1.45,
       "step": 28590
     },
     {
       "epoch": 0.4646553264772303,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.4259,
       "step": 28600
     },
     {
       "epoch": 0.46481779337459994,
-      "grad_norm": 1.6875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4595,
       "step": 28610
     },
     {
       "epoch": 0.4649802602719696,
-      "grad_norm": 4.40625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.4613,
       "step": 28620
     },
     {
       "epoch": 0.46514272716933924,
-      "grad_norm": 5.59375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4237,
+      "loss": 1.4262,
       "step": 28630
     },
     {
       "epoch": 0.4653051940667089,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.4945,
       "step": 28640
     },
     {
       "epoch": 0.46546766096407854,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.4743,
       "step": 28650
     },
     {
       "epoch": 0.46563012786144825,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4951,
       "step": 28660
     },
     {
       "epoch": 0.4657925947588179,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.4325,
       "step": 28670
     },
     {
       "epoch": 0.46595506165618755,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.3879,
       "step": 28680
     },
     {
       "epoch": 0.4661175285535572,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.4914,
       "step": 28690
     },
     {
       "epoch": 0.46627999545092685,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4209,
       "step": 28700
     },
     {
       "epoch": 0.46644246234829656,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.3859,
       "step": 28710
     },
     {
       "epoch": 0.4666049292456662,
-      "grad_norm": 3.0,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.4626,
       "step": 28720
     },
     {
       "epoch": 0.46676739614303586,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.4973,
       "step": 28730
     },
     {
       "epoch": 0.4669298630404055,
-      "grad_norm": 2.625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.4392,
       "step": 28740
     },
     {
       "epoch": 0.46709232993777516,
-      "grad_norm": 1.8828125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4224,
+      "loss": 1.4628,
       "step": 28750
     },
     {
       "epoch": 0.4672547968351448,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4211,
       "step": 28760
     },
     {
       "epoch": 0.4674172637325145,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4285,
+      "loss": 1.4816,
       "step": 28770
     },
     {
       "epoch": 0.46757973062988417,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.485,
       "step": 28780
     },
     {
       "epoch": 0.4677421975272538,
-      "grad_norm": 1.875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4253,
+      "loss": 1.4364,
       "step": 28790
     },
     {
       "epoch": 0.4679046644246235,
-      "grad_norm": 1.6953125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4577,
       "step": 28800
     },
     {
       "epoch": 0.4680671313219931,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4416,
       "step": 28810
     },
     {
       "epoch": 0.46822959821936283,
-      "grad_norm": 1.578125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.4321,
       "step": 28820
     },
     {
       "epoch": 0.4683920651167325,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4234,
+      "loss": 1.4555,
       "step": 28830
     },
     {
       "epoch": 0.46855453201410213,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.466,
       "step": 28840
     },
     {
       "epoch": 0.4687169989114718,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.408,
       "step": 28850
     },
     {
       "epoch": 0.46887946580884143,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.438,
       "step": 28860
     },
     {
       "epoch": 0.4690419327062111,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.429,
       "step": 28870
     },
     {
       "epoch": 0.4692043996035808,
-      "grad_norm": 3.515625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4545,
       "step": 28880
     },
     {
       "epoch": 0.46936686650095044,
-      "grad_norm": 2.75,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4665,
       "step": 28890
     },
     {
       "epoch": 0.4695293333983201,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.4505,
       "step": 28900
     },
     {
       "epoch": 0.46969180029568974,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.4253,
       "step": 28910
     },
     {
       "epoch": 0.4698542671930594,
-      "grad_norm": 2.25,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.5,
       "step": 28920
     },
     {
       "epoch": 0.4700167340904291,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.461,
       "step": 28930
     },
     {
       "epoch": 0.47017920098779875,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4307,
+      "loss": 1.4496,
       "step": 28940
     },
     {
       "epoch": 0.4703416678851684,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4306,
+      "loss": 1.3986,
       "step": 28950
     },
     {
       "epoch": 0.47050413478253805,
-      "grad_norm": 2.25,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4204,
       "step": 28960
     },
     {
       "epoch": 0.4706666016799077,
-      "grad_norm": 2.328125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.4999,
       "step": 28970
     },
     {
       "epoch": 0.47082906857727735,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.4193,
       "step": 28980
     },
     {
       "epoch": 0.47099153547464706,
-      "grad_norm": 1.515625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4585,
       "step": 28990
     },
     {
       "epoch": 0.4711540023720167,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.4392,
       "step": 29000
     },
     {
       "epoch": 0.47131646926938636,
-      "grad_norm": 2.125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.4825,
       "step": 29010
     },
     {
       "epoch": 0.471478936166756,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3822,
+      "loss": 1.474,
       "step": 29020
     },
     {
       "epoch": 0.47164140306412566,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.4799,
       "step": 29030
     },
     {
       "epoch": 0.47180386996149537,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.4822,
       "step": 29040
     },
     {
       "epoch": 0.471966336858865,
-      "grad_norm": 3.78125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.457,
       "step": 29050
     },
     {
       "epoch": 0.47212880375623467,
-      "grad_norm": 2.125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.4547,
       "step": 29060
     },
     {
       "epoch": 0.4722912706536043,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.4448,
       "step": 29070
     },
     {
       "epoch": 0.472453737550974,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4396,
+      "loss": 1.4344,
       "step": 29080
     },
     {
       "epoch": 0.4726162044483436,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4137,
+      "loss": 1.5121,
       "step": 29090
     },
     {
       "epoch": 0.47277867134571333,
-      "grad_norm": 2.28125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.4302,
       "step": 29100
     },
     {
       "epoch": 0.472941138243083,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.4803,
       "step": 29110
     },
     {
       "epoch": 0.47310360514045263,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3778,
+      "loss": 1.4989,
       "step": 29120
     },
     {
       "epoch": 0.4732660720378223,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.4542,
       "step": 29130
     },
     {
       "epoch": 0.47342853893519193,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.4866,
       "step": 29140
     },
     {
       "epoch": 0.47359100583256164,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.425,
       "step": 29150
     },
     {
       "epoch": 0.4737534727299313,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.4361,
       "step": 29160
     },
     {
       "epoch": 0.47391593962730094,
-      "grad_norm": 2.375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4162,
       "step": 29170
     },
     {
       "epoch": 0.4740784065246706,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4638,
       "step": 29180
     },
     {
       "epoch": 0.47424087342204024,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4205,
+      "loss": 1.479,
       "step": 29190
     },
     {
       "epoch": 0.4744033403194099,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4947,
       "step": 29200
     },
     {
       "epoch": 0.4745658072167796,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4191,
+      "loss": 1.4174,
       "step": 29210
     },
     {
       "epoch": 0.47472827411414925,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.374,
+      "loss": 1.4254,
       "step": 29220
     },
     {
       "epoch": 0.4748907410115189,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.4129,
       "step": 29230
     },
     {
       "epoch": 0.47505320790888855,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4454,
       "step": 29240
     },
     {
       "epoch": 0.4752156748062582,
-      "grad_norm": 2.625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.4788,
       "step": 29250
     },
     {
       "epoch": 0.4753781417036279,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4752,
       "step": 29260
     },
     {
       "epoch": 0.47554060860099756,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3784,
+      "loss": 1.489,
       "step": 29270
     },
     {
       "epoch": 0.4757030754983672,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.4036,
       "step": 29280
     },
     {
       "epoch": 0.47586554239573686,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3873,
+      "loss": 1.4594,
       "step": 29290
     },
     {
       "epoch": 0.4760280092931065,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4882,
       "step": 29300
     },
     {
       "epoch": 0.47619047619047616,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4312,
       "step": 29310
     },
     {
       "epoch": 0.47635294308784587,
-      "grad_norm": 2.25,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.5333,
       "step": 29320
     },
     {
       "epoch": 0.4765154099852155,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4247,
+      "loss": 1.4027,
       "step": 29330
     },
     {
       "epoch": 0.47667787688258517,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4462,
       "step": 29340
     },
     {
       "epoch": 0.4768403437799548,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4138,
       "step": 29350
     },
     {
       "epoch": 0.4770028106773245,
-      "grad_norm": 1.984375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4418,
       "step": 29360
     },
     {
       "epoch": 0.4771652775746942,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.4404,
       "step": 29370
     },
     {
       "epoch": 0.47732774447206383,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4121,
+      "loss": 1.4656,
       "step": 29380
     },
     {
       "epoch": 0.4774902113694335,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.3745,
       "step": 29390
     },
     {
       "epoch": 0.47765267826680313,
-      "grad_norm": 2.875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.5056,
       "step": 29400
     },
     {
       "epoch": 0.4778151451641728,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4305,
       "step": 29410
     },
     {
       "epoch": 0.47797761206154243,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.4509,
       "step": 29420
     },
     {
       "epoch": 0.47814007895891214,
-      "grad_norm": 1.890625,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.4466,
       "step": 29430
     },
     {
       "epoch": 0.4783025458562818,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4447,
       "step": 29440
     },
     {
       "epoch": 0.47846501275365144,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.4723,
       "step": 29450
     },
     {
       "epoch": 0.4786274796510211,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4448,
       "step": 29460
     },
     {
       "epoch": 0.47878994654839074,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4343,
+      "loss": 1.4617,
       "step": 29470
     },
     {
       "epoch": 0.47895241344576045,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.4595,
       "step": 29480
     },
     {
       "epoch": 0.4791148803431301,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.4976,
       "step": 29490
     },
     {
       "epoch": 0.47927734724049975,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.4655,
       "step": 29500
     },
     {
       "epoch": 0.4794398141378694,
-      "grad_norm": 2.734375,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.4834,
       "step": 29510
     },
     {
       "epoch": 0.47960228103523905,
-      "grad_norm": 1.6875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4525,
       "step": 29520
     },
     {
       "epoch": 0.4797647479326087,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4524,
       "step": 29530
     },
     {
       "epoch": 0.4799272148299784,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4369,
+      "loss": 1.4823,
       "step": 29540
     },
     {
       "epoch": 0.48008968172734806,
-      "grad_norm": 2.484375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4871,
       "step": 29550
     },
     {
       "epoch": 0.4802521486247177,
-      "grad_norm": 2.125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.4315,
       "step": 29560
     },
     {
       "epoch": 0.48041461552208736,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.5133,
       "step": 29570
     },
     {
       "epoch": 0.480577082419457,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4742,
       "step": 29580
     },
     {
       "epoch": 0.4807395493168267,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.4592,
       "step": 29590
     },
     {
       "epoch": 0.48090201621419637,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4207,
+      "loss": 1.4453,
       "step": 29600
     },
     {
       "epoch": 0.481064483111566,
-      "grad_norm": 1.7109375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4251,
       "step": 29610
     },
     {
       "epoch": 0.48122695000893567,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4554,
       "step": 29620
     },
     {
       "epoch": 0.4813894169063053,
-      "grad_norm": 2.28125,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.4725,
       "step": 29630
     },
     {
       "epoch": 0.481551883803675,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4367,
       "step": 29640
     },
     {
       "epoch": 0.4817143507010447,
-      "grad_norm": 1.71875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4451,
       "step": 29650
     },
     {
       "epoch": 0.48187681759841433,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4261,
+      "loss": 1.4565,
       "step": 29660
     },
     {
       "epoch": 0.482039284495784,
-      "grad_norm": 1.8359375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4457,
       "step": 29670
     },
     {
       "epoch": 0.48220175139315363,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3908,
+      "loss": 1.4609,
       "step": 29680
     },
     {
       "epoch": 0.4823642182905233,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.4562,
       "step": 29690
     },
     {
       "epoch": 0.482526685187893,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4267,
       "step": 29700
     },
     {
       "epoch": 0.48268915208526264,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.4455,
       "step": 29710
     },
     {
       "epoch": 0.4828516189826323,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4572,
       "step": 29720
     },
     {
       "epoch": 0.48301408588000194,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4435,
       "step": 29730
     },
     {
       "epoch": 0.4831765527773716,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.425,
+      "loss": 1.3993,
       "step": 29740
     },
     {
       "epoch": 0.48333901967474124,
-      "grad_norm": 1.75,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4875,
       "step": 29750
     },
     {
       "epoch": 0.48350148657211095,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4457,
       "step": 29760
     },
     {
       "epoch": 0.4836639534694806,
-      "grad_norm": 2.296875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.3829,
       "step": 29770
     },
     {
       "epoch": 0.48382642036685025,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4919,
       "step": 29780
     },
     {
       "epoch": 0.4839888872642199,
-      "grad_norm": 2.625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.407,
       "step": 29790
     },
     {
       "epoch": 0.48415135416158955,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4182,
+      "loss": 1.4689,
       "step": 29800
     },
     {
       "epoch": 0.48431382105895926,
-      "grad_norm": 2.671875,
+      "grad_norm": 4.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4168,
+      "loss": 1.3889,
       "step": 29810
     },
     {
       "epoch": 0.4844762879563289,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.4619,
       "step": 29820
     },
     {
       "epoch": 0.48463875485369856,
-      "grad_norm": 2.8125,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4244,
       "step": 29830
     },
     {
       "epoch": 0.4848012217510682,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4224,
+      "loss": 1.4707,
       "step": 29840
     },
     {
       "epoch": 0.48496368864843786,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3967,
+      "loss": 1.448,
       "step": 29850
     },
     {
       "epoch": 0.48512615554580757,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.475,
       "step": 29860
     },
     {
       "epoch": 0.4852886224431772,
-      "grad_norm": 4.5625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4307,
       "step": 29870
     },
     {
       "epoch": 0.48545108934054687,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4414,
       "step": 29880
     },
     {
       "epoch": 0.4856135562379165,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.4535,
       "step": 29890
     },
     {
       "epoch": 0.4857760231352862,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.4534,
       "step": 29900
     },
     {
       "epoch": 0.4859384900326558,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.4487,
       "step": 29910
     },
     {
       "epoch": 0.48610095693002553,
-      "grad_norm": 1.9609375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4437,
       "step": 29920
     },
     {
       "epoch": 0.4862634238273952,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.4632,
       "step": 29930
     },
     {
       "epoch": 0.48642589072476483,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.4568,
       "step": 29940
     },
     {
       "epoch": 0.4865883576221345,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4485,
       "step": 29950
     },
     {
       "epoch": 0.48675082451950413,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4236,
+      "loss": 1.4645,
       "step": 29960
     },
     {
       "epoch": 0.48691329141687384,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.407,
       "step": 29970
     },
     {
       "epoch": 0.4870757583142435,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.425,
+      "loss": 1.4691,
       "step": 29980
     },
     {
       "epoch": 0.48723822521161314,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.4369,
       "step": 29990
     },
     {
       "epoch": 0.4874006921089828,
-      "grad_norm": 3.125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.455,
       "step": 30000
     },
     {
       "epoch": 0.48756315900635244,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4598,
       "step": 30010
     },
     {
       "epoch": 0.4877256259037221,
-      "grad_norm": 3.828125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4472,
       "step": 30020
     },
     {
       "epoch": 0.4878880928010918,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4318,
       "step": 30030
     },
     {
       "epoch": 0.48805055969846145,
-      "grad_norm": 4.0,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4198,
       "step": 30040
     },
     {
       "epoch": 0.4882130265958311,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.4701,
       "step": 30050
     },
     {
       "epoch": 0.48837549349320075,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3791,
+      "loss": 1.5099,
       "step": 30060
     },
     {
       "epoch": 0.4885379603905704,
-      "grad_norm": 4.03125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.4479,
       "step": 30070
     },
     {
       "epoch": 0.4887004272879401,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4855,
       "step": 30080
     },
     {
       "epoch": 0.48886289418530976,
-      "grad_norm": 2.125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4237,
+      "loss": 1.4671,
       "step": 30090
     },
     {
       "epoch": 0.4890253610826794,
-      "grad_norm": 1.7578125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.424,
+      "loss": 1.4203,
       "step": 30100
     },
     {
       "epoch": 0.48918782798004906,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.4866,
       "step": 30110
     },
     {
       "epoch": 0.4893502948774187,
-      "grad_norm": 1.703125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4294,
+      "loss": 1.4133,
       "step": 30120
     },
     {
       "epoch": 0.48951276177478836,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.435,
       "step": 30130
     },
     {
       "epoch": 0.48967522867215807,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.5078,
       "step": 30140
     },
     {
       "epoch": 0.4898376955695277,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4311,
       "step": 30150
     },
     {
       "epoch": 0.49000016246689737,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4463,
       "step": 30160
     },
     {
       "epoch": 0.490162629364267,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.4625,
       "step": 30170
     },
     {
       "epoch": 0.4903250962616367,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3732,
+      "loss": 1.435,
       "step": 30180
     },
     {
       "epoch": 0.4904875631590064,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.4373,
       "step": 30190
     },
     {
       "epoch": 0.49065003005637603,
-      "grad_norm": 3.765625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.4622,
       "step": 30200
     },
     {
       "epoch": 0.4908124969537457,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4267,
+      "loss": 1.4631,
       "step": 30210
     },
     {
       "epoch": 0.49097496385111533,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4073,
       "step": 30220
     },
     {
       "epoch": 0.491137430748485,
-      "grad_norm": 2.28125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4539,
       "step": 30230
     },
     {
       "epoch": 0.49129989764585463,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.436,
       "step": 30240
     },
     {
       "epoch": 0.49146236454322434,
-      "grad_norm": 3.640625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.5033,
       "step": 30250
     },
     {
       "epoch": 0.491624831440594,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.4392,
       "step": 30260
     },
     {
       "epoch": 0.49178729833796364,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.4844,
       "step": 30270
     },
     {
       "epoch": 0.4919497652353333,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4927,
       "step": 30280
     },
     {
       "epoch": 0.49211223213270294,
-      "grad_norm": 2.125,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3994,
+      "loss": 1.4396,
       "step": 30290
     },
     {
       "epoch": 0.49227469903007265,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.4602,
       "step": 30300
     },
     {
       "epoch": 0.4924371659274423,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4522,
       "step": 30310
     },
     {
       "epoch": 0.49259963282481195,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.4317,
       "step": 30320
     },
     {
       "epoch": 0.4927620997221816,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4837,
       "step": 30330
     },
     {
       "epoch": 0.49292456661955125,
-      "grad_norm": 1.875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4561,
       "step": 30340
     },
     {
       "epoch": 0.4930870335169209,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4503,
       "step": 30350
     },
     {
       "epoch": 0.4932495004142906,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4822,
       "step": 30360
     },
     {
       "epoch": 0.49341196731166026,
-      "grad_norm": 4.34375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.4998,
       "step": 30370
     },
     {
       "epoch": 0.4935744342090299,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4393,
       "step": 30380
     },
     {
       "epoch": 0.49373690110639956,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.4786,
       "step": 30390
     },
     {
       "epoch": 0.4938993680037692,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4279,
+      "loss": 1.4496,
       "step": 30400
     },
     {
       "epoch": 0.4940618349011389,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4504,
       "step": 30410
     },
     {
       "epoch": 0.49422430179850857,
-      "grad_norm": 1.625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.4551,
       "step": 30420
     },
     {
       "epoch": 0.4943867686958782,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4558,
       "step": 30430
     },
     {
       "epoch": 0.49454923559324787,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.417,
       "step": 30440
     },
     {
       "epoch": 0.4947117024906175,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.4418,
       "step": 30450
     },
     {
       "epoch": 0.4948741693879872,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.4308,
       "step": 30460
     },
     {
       "epoch": 0.4950366362853569,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.4689,
       "step": 30470
     },
     {
       "epoch": 0.49519910318272653,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.4621,
       "step": 30480
     },
     {
       "epoch": 0.4953615700800962,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.473,
       "step": 30490
     },
     {
       "epoch": 0.49552403697746583,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.3961,
       "step": 30500
     },
     {
       "epoch": 0.4956865038748355,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4328,
+      "loss": 1.4182,
       "step": 30510
     },
     {
       "epoch": 0.4958489707722052,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4386,
+      "loss": 1.4337,
       "step": 30520
     },
     {
       "epoch": 0.49601143766957484,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.4511,
       "step": 30530
     },
     {
       "epoch": 0.4961739045669445,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4289,
       "step": 30540
     },
     {
       "epoch": 0.49633637146431414,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.3771,
       "step": 30550
     },
     {
       "epoch": 0.4964988383616838,
-      "grad_norm": 1.796875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.4307,
       "step": 30560
     },
     {
       "epoch": 0.49666130525905344,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.494,
       "step": 30570
     },
     {
       "epoch": 0.49682377215642315,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4205,
+      "loss": 1.4085,
       "step": 30580
     },
     {
       "epoch": 0.4969862390537928,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.4279,
       "step": 30590
     },
     {
       "epoch": 0.49714870595116245,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.438,
       "step": 30600
     },
     {
       "epoch": 0.4973111728485321,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4794,
       "step": 30610
     },
     {
       "epoch": 0.49747363974590175,
-      "grad_norm": 2.625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.4442,
       "step": 30620
     },
     {
       "epoch": 0.49763610664327146,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4696,
       "step": 30630
     },
     {
       "epoch": 0.4977985735406411,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.4171,
       "step": 30640
     },
     {
       "epoch": 0.49796104043801076,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4623,
       "step": 30650
     },
     {
       "epoch": 0.4981235073353804,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4427,
       "step": 30660
     },
     {
       "epoch": 0.49828597423275006,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4754,
       "step": 30670
     },
     {
       "epoch": 0.4984484411301197,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.4135,
       "step": 30680
     },
     {
       "epoch": 0.4986109080274894,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4064,
+      "loss": 1.4773,
       "step": 30690
     },
     {
       "epoch": 0.49877337492485907,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.412,
       "step": 30700
     },
     {
       "epoch": 0.4989358418222287,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4205,
+      "loss": 1.4631,
       "step": 30710
     },
     {
       "epoch": 0.49909830871959837,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.4122,
       "step": 30720
     },
     {
       "epoch": 0.499260775616968,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.505,
       "step": 30730
     },
     {
       "epoch": 0.49942324251433773,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.4878,
       "step": 30740
     },
     {
       "epoch": 0.4995857094117074,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4723,
       "step": 30750
     },
     {
       "epoch": 0.49974817630907703,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.4464,
       "step": 30760
     },
     {
       "epoch": 0.4999106432064467,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.4856,
       "step": 30770
     },
     {
       "epoch": 0.5000731101038164,
-      "grad_norm": 2.375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.3997,
       "step": 30780
     },
     {
       "epoch": 0.500235577001186,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4103,
+      "loss": 1.4309,
       "step": 30790
     },
     {
       "epoch": 0.5003980438985557,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.4494,
       "step": 30800
     },
     {
       "epoch": 0.5005605107959253,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.4659,
       "step": 30810
     },
     {
       "epoch": 0.500722977693295,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4168,
+      "loss": 1.4592,
       "step": 30820
     },
     {
       "epoch": 0.5008854445906646,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4224,
+      "loss": 1.448,
       "step": 30830
     },
     {
       "epoch": 0.5010479114880343,
-      "grad_norm": 1.703125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4539,
       "step": 30840
     },
     {
       "epoch": 0.5012103783854039,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4142,
       "step": 30850
     },
     {
       "epoch": 0.5013728452827736,
-      "grad_norm": 2.0,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.449,
       "step": 30860
     },
     {
       "epoch": 0.5015353121801432,
-      "grad_norm": 2.25,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.4658,
       "step": 30870
     },
     {
       "epoch": 0.501697779077513,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4861,
       "step": 30880
     },
     {
       "epoch": 0.5018602459748827,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.4397,
       "step": 30890
     },
     {
       "epoch": 0.5020227128722523,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4058,
       "step": 30900
     },
     {
       "epoch": 0.502185179769622,
-      "grad_norm": 2.25,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4361,
       "step": 30910
     },
     {
       "epoch": 0.5023476466669916,
-      "grad_norm": 4.5,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.42,
+      "loss": 1.4675,
       "step": 30920
     },
     {
       "epoch": 0.5025101135643613,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.411,
       "step": 30930
     },
     {
       "epoch": 0.5026725804617309,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.4451,
       "step": 30940
     },
     {
       "epoch": 0.5028350473591006,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.396,
+      "loss": 1.4615,
       "step": 30950
     },
     {
       "epoch": 0.5029975142564702,
-      "grad_norm": 3.609375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.4457,
       "step": 30960
     },
     {
       "epoch": 0.5031599811538399,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.4576,
       "step": 30970
     },
     {
       "epoch": 0.5033224480512095,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.475,
       "step": 30980
     },
     {
       "epoch": 0.5034849149485793,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.5275,
       "step": 30990
     },
     {
       "epoch": 0.5036473818459489,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.4533,
       "step": 31000
     },
     {
       "epoch": 0.5038098487433186,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4285,
       "step": 31010
     },
     {
       "epoch": 0.5039723156406882,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.4883,
       "step": 31020
     },
     {
       "epoch": 0.5041347825380579,
-      "grad_norm": 2.359375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.4571,
       "step": 31030
     },
     {
       "epoch": 0.5042972494354275,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4444,
       "step": 31040
     },
     {
       "epoch": 0.5044597163327972,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.5073,
       "step": 31050
     },
     {
       "epoch": 0.5046221832301668,
-      "grad_norm": 2.375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.3907,
       "step": 31060
     },
     {
       "epoch": 0.5047846501275365,
-      "grad_norm": 3.421875,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3898,
+      "loss": 1.4606,
       "step": 31070
     },
     {
       "epoch": 0.5049471170249061,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4691,
       "step": 31080
     },
     {
       "epoch": 0.5051095839222758,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3841,
+      "loss": 1.4624,
       "step": 31090
     },
     {
       "epoch": 0.5052720508196455,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4909,
       "step": 31100
     },
     {
       "epoch": 0.5054345177170152,
-      "grad_norm": 2.5,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.4384,
       "step": 31110
     },
     {
       "epoch": 0.5055969846143848,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.451,
       "step": 31120
     },
     {
       "epoch": 0.5057594515117545,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4195,
+      "loss": 1.4384,
       "step": 31130
     },
     {
       "epoch": 0.5059219184091241,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.4676,
       "step": 31140
     },
     {
       "epoch": 0.5060843853064938,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4506,
       "step": 31150
     },
     {
       "epoch": 0.5062468522038635,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4086,
       "step": 31160
     },
     {
       "epoch": 0.5064093191012331,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4436,
       "step": 31170
     },
     {
       "epoch": 0.5065717859986028,
-      "grad_norm": 2.5,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4524,
       "step": 31180
     },
     {
       "epoch": 0.5067342528959724,
-      "grad_norm": 3.578125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4189,
       "step": 31190
     },
     {
       "epoch": 0.506896719793342,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.4744,
       "step": 31200
     },
     {
       "epoch": 0.5070591866907118,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.4408,
       "step": 31210
     },
     {
       "epoch": 0.5072216535880815,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3807,
+      "loss": 1.4551,
       "step": 31220
     },
     {
       "epoch": 0.5073841204854511,
-      "grad_norm": 3.25,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4752,
       "step": 31230
     },
     {
       "epoch": 0.5075465873828208,
-      "grad_norm": 2.125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.4725,
       "step": 31240
     },
     {
       "epoch": 0.5077090542801904,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.3916,
       "step": 31250
     },
     {
       "epoch": 0.5078715211775601,
-      "grad_norm": 3.25,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4685,
       "step": 31260
     },
     {
       "epoch": 0.5080339880749297,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4253,
+      "loss": 1.4445,
       "step": 31270
     },
     {
       "epoch": 0.5081964549722994,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4622,
       "step": 31280
     },
     {
       "epoch": 0.508358921869669,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4158,
+      "loss": 1.4924,
       "step": 31290
     },
     {
       "epoch": 0.5085213887670387,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4629,
       "step": 31300
     },
     {
       "epoch": 0.5086838556644083,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4242,
+      "loss": 1.4301,
       "step": 31310
     },
     {
       "epoch": 0.5088463225617781,
-      "grad_norm": 2.234375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4469,
       "step": 31320
     },
     {
       "epoch": 0.5090087894591477,
-      "grad_norm": 1.734375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.4735,
       "step": 31330
     },
     {
       "epoch": 0.5091712563565174,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4279,
       "step": 31340
     },
     {
       "epoch": 0.509333723253887,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3805,
+      "loss": 1.4926,
       "step": 31350
     },
     {
       "epoch": 0.5094961901512567,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.4537,
       "step": 31360
     },
     {
       "epoch": 0.5096586570486263,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4309,
       "step": 31370
     },
     {
       "epoch": 0.509821123945996,
-      "grad_norm": 2.328125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4221,
       "step": 31380
     },
     {
       "epoch": 0.5099835908433656,
-      "grad_norm": 2.5,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4169,
       "step": 31390
     },
     {
       "epoch": 0.5101460577407353,
-      "grad_norm": 1.890625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.4353,
       "step": 31400
     },
     {
       "epoch": 0.5103085246381049,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.499,
       "step": 31410
     },
     {
       "epoch": 0.5104709915354746,
-      "grad_norm": 1.8203125,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.4541,
       "step": 31420
     },
     {
       "epoch": 0.5106334584328444,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4649,
       "step": 31430
     },
     {
       "epoch": 0.510795925330214,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.3869,
       "step": 31440
     },
     {
       "epoch": 0.5109583922275837,
-      "grad_norm": 3.5,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4256,
+      "loss": 1.408,
       "step": 31450
     },
     {
       "epoch": 0.5111208591249533,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4934,
       "step": 31460
     },
     {
       "epoch": 0.511283326022323,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.4025,
       "step": 31470
     },
     {
       "epoch": 0.5114457929196926,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.4631,
       "step": 31480
     },
     {
       "epoch": 0.5116082598170623,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4051,
+      "loss": 1.4916,
       "step": 31490
     },
     {
       "epoch": 0.5117707267144319,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3949,
       "step": 31500
     },
     {
       "epoch": 0.5119331936118016,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4608,
       "step": 31510
     },
     {
       "epoch": 0.5120956605091712,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.4282,
       "step": 31520
     },
     {
       "epoch": 0.5122581274065409,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3776,
+      "loss": 1.4816,
       "step": 31530
     },
     {
       "epoch": 0.5124205943039106,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.513,
       "step": 31540
     },
     {
       "epoch": 0.5125830612012803,
-      "grad_norm": 2.328125,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.4377,
       "step": 31550
     },
     {
       "epoch": 0.5127455280986499,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4439,
       "step": 31560
     },
     {
       "epoch": 0.5129079949960196,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4215,
+      "loss": 1.4692,
       "step": 31570
     },
     {
       "epoch": 0.5130704618933892,
-      "grad_norm": 2.625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4355,
       "step": 31580
     },
     {
       "epoch": 0.5132329287907589,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.4175,
       "step": 31590
     },
     {
       "epoch": 0.5133953956881285,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4266,
       "step": 31600
     },
     {
       "epoch": 0.5135578625854982,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4119,
+      "loss": 1.4658,
       "step": 31610
     },
     {
       "epoch": 0.5137203294828678,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.4403,
       "step": 31620
     },
     {
       "epoch": 0.5138827963802375,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4177,
       "step": 31630
     },
     {
       "epoch": 0.5140452632776071,
-      "grad_norm": 2.671875,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4878,
       "step": 31640
     },
     {
       "epoch": 0.5142077301749769,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4246,
+      "loss": 1.4452,
       "step": 31650
     },
     {
       "epoch": 0.5143701970723465,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4043,
       "step": 31660
     },
     {
       "epoch": 0.5145326639697162,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.4305,
       "step": 31670
     },
     {
       "epoch": 0.5146951308670858,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4148,
       "step": 31680
     },
     {
       "epoch": 0.5148575977644555,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.3988,
       "step": 31690
     },
     {
       "epoch": 0.5150200646618251,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.4689,
       "step": 31700
     },
     {
       "epoch": 0.5151825315591948,
-      "grad_norm": 1.890625,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.467,
       "step": 31710
     },
     {
       "epoch": 0.5153449984565645,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.463,
       "step": 31720
     },
     {
       "epoch": 0.5155074653539341,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4262,
       "step": 31730
     },
     {
       "epoch": 0.5156699322513038,
-      "grad_norm": 3.125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.4235,
       "step": 31740
     },
     {
       "epoch": 0.5158323991486734,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4287,
+      "loss": 1.4352,
       "step": 31750
     },
     {
       "epoch": 0.5159948660460432,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4618,
       "step": 31760
     },
     {
       "epoch": 0.5161573329434128,
-      "grad_norm": 1.78125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.4512,
       "step": 31770
     },
     {
       "epoch": 0.5163197998407825,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.4596,
       "step": 31780
     },
     {
       "epoch": 0.5164822667381521,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4152,
+      "loss": 1.4794,
       "step": 31790
     },
     {
       "epoch": 0.5166447336355218,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4183,
       "step": 31800
     },
     {
       "epoch": 0.5168072005328914,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3823,
+      "loss": 1.4746,
       "step": 31810
     },
     {
       "epoch": 0.5169696674302611,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.4975,
       "step": 31820
     },
     {
       "epoch": 0.5171321343276307,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.4364,
       "step": 31830
     },
     {
       "epoch": 0.5172946012250004,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4277,
+      "loss": 1.4961,
       "step": 31840
     },
     {
       "epoch": 0.51745706812237,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4372,
+      "loss": 1.4772,
       "step": 31850
     },
     {
       "epoch": 0.5176195350197397,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.3592,
       "step": 31860
     },
     {
       "epoch": 0.5177820019171094,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.458,
       "step": 31870
     },
     {
       "epoch": 0.5179444688144791,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4574,
       "step": 31880
     },
     {
       "epoch": 0.5181069357118487,
-      "grad_norm": 3.265625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4631,
       "step": 31890
     },
     {
       "epoch": 0.5182694026092184,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.485,
       "step": 31900
     },
     {
       "epoch": 0.518431869506588,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4034,
+      "loss": 1.4425,
       "step": 31910
     },
     {
       "epoch": 0.5185943364039577,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.4731,
       "step": 31920
     },
     {
       "epoch": 0.5187568033013273,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.4586,
       "step": 31930
     },
     {
       "epoch": 0.518919270198697,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4321,
       "step": 31940
     },
     {
       "epoch": 0.5190817370960666,
-      "grad_norm": 3.578125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.426,
       "step": 31950
     },
     {
       "epoch": 0.5192442039934363,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.4085,
       "step": 31960
     },
     {
       "epoch": 0.5194066708908059,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4852,
       "step": 31970
     },
     {
       "epoch": 0.5195691377881757,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.4243,
       "step": 31980
     },
     {
       "epoch": 0.5197316046855454,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.4288,
       "step": 31990
     },
     {
       "epoch": 0.519894071582915,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4329,
       "step": 32000
     },
     {
       "epoch": 0.5200565384802847,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.3996,
       "step": 32010
     },
     {
       "epoch": 0.5202190053776543,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.4383,
       "step": 32020
     },
     {
       "epoch": 0.520381472275024,
-      "grad_norm": 4.59375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.4132,
       "step": 32030
     },
     {
       "epoch": 0.5205439391723936,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4773,
       "step": 32040
     },
     {
       "epoch": 0.5207064060697633,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3889,
+      "loss": 1.4502,
       "step": 32050
     },
     {
       "epoch": 0.5208688729671329,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.4241,
       "step": 32060
     },
     {
       "epoch": 0.5210313398645026,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4322,
       "step": 32070
     },
     {
       "epoch": 0.5211938067618722,
-      "grad_norm": 2.25,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4641,
       "step": 32080
     },
     {
       "epoch": 0.521356273659242,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4401,
       "step": 32090
     },
     {
       "epoch": 0.5215187405566116,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4312,
+      "loss": 1.441,
       "step": 32100
     },
     {
       "epoch": 0.5216812074539813,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.421,
       "step": 32110
     },
     {
       "epoch": 0.5218436743513509,
-      "grad_norm": 2.0,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.4362,
       "step": 32120
     },
     {
       "epoch": 0.5220061412487206,
-      "grad_norm": 2.5625,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.3995,
       "step": 32130
     },
     {
       "epoch": 0.5221686081460902,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4243,
+      "loss": 1.4807,
       "step": 32140
     },
     {
       "epoch": 0.5223310750434599,
-      "grad_norm": 2.75,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4137,
+      "loss": 1.4297,
       "step": 32150
     },
     {
       "epoch": 0.5224935419408295,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4272,
+      "loss": 1.4034,
       "step": 32160
     },
     {
       "epoch": 0.5226560088381992,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.452,
       "step": 32170
     },
     {
       "epoch": 0.5228184757355688,
-      "grad_norm": 1.9296875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.5063,
       "step": 32180
     },
     {
       "epoch": 0.5229809426329385,
-      "grad_norm": 1.5859375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.3826,
       "step": 32190
     },
     {
       "epoch": 0.5231434095303082,
-      "grad_norm": 2.234375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.4163,
+      "loss": 1.4272,
       "step": 32200
     },
     {
       "epoch": 0.5233058764276779,
-      "grad_norm": 1.5,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.4263,
       "step": 32210
     },
     {
       "epoch": 0.5234683433250475,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.4861,
       "step": 32220
     },
     {
       "epoch": 0.5236308102224172,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.4057,
       "step": 32230
     },
     {
       "epoch": 0.5237932771197868,
-      "grad_norm": 1.96875,
+      "grad_norm": 4.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4591,
       "step": 32240
     },
     {
       "epoch": 0.5239557440171565,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.415,
+      "loss": 1.4054,
       "step": 32250
     },
     {
       "epoch": 0.5241182109145262,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4428,
       "step": 32260
     },
     {
       "epoch": 0.5242806778118958,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4318,
       "step": 32270
     },
     {
       "epoch": 0.5244431447092655,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.4668,
       "step": 32280
     },
     {
       "epoch": 0.5246056116066351,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.3943,
       "step": 32290
     },
     {
       "epoch": 0.5247680785040048,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.4205,
       "step": 32300
     },
     {
       "epoch": 0.5249305454013745,
-      "grad_norm": 1.6796875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.3821,
       "step": 32310
     },
     {
       "epoch": 0.5250930122987442,
-      "grad_norm": 2.234375,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.4278,
       "step": 32320
     },
     {
       "epoch": 0.5252554791961138,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.4836,
       "step": 32330
     },
     {
       "epoch": 0.5254179460934835,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4823,
       "step": 32340
     },
     {
       "epoch": 0.5255804129908531,
-      "grad_norm": 1.453125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3765,
+      "loss": 1.4603,
       "step": 32350
     },
     {
       "epoch": 0.5257428798882228,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.384,
+      "loss": 1.4013,
       "step": 32360
     },
     {
       "epoch": 0.5259053467855924,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4457,
       "step": 32370
     },
     {
       "epoch": 0.5260678136829621,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.4563,
       "step": 32380
     },
     {
       "epoch": 0.5262302805803317,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4308,
       "step": 32390
     },
     {
       "epoch": 0.5263927474777014,
-      "grad_norm": 2.125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4384,
+      "loss": 1.4445,
       "step": 32400
     },
     {
       "epoch": 0.5265552143750711,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4681,
       "step": 32410
     },
     {
       "epoch": 0.5267176812724408,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.436,
       "step": 32420
     },
     {
       "epoch": 0.5268801481698104,
-      "grad_norm": 2.125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.427,
+      "loss": 1.4572,
       "step": 32430
     },
     {
       "epoch": 0.5270426150671801,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4199,
+      "loss": 1.4235,
       "step": 32440
     },
     {
       "epoch": 0.5272050819645497,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4473,
       "step": 32450
     },
     {
       "epoch": 0.5273675488619194,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.4859,
       "step": 32460
     },
     {
       "epoch": 0.527530015759289,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.4095,
       "step": 32470
     },
     {
       "epoch": 0.5276924826566587,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3794,
+      "loss": 1.4655,
       "step": 32480
     },
     {
       "epoch": 0.5278549495540283,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4783,
       "step": 32490
     },
     {
       "epoch": 0.528017416451398,
-      "grad_norm": 3.25,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4211,
       "step": 32500
     },
     {
       "epoch": 0.5281798833487676,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4411,
       "step": 32510
     },
     {
       "epoch": 0.5283423502461374,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4051,
+      "loss": 1.4261,
       "step": 32520
     },
     {
       "epoch": 0.528504817143507,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4463,
       "step": 32530
     },
     {
       "epoch": 0.5286672840408767,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4923,
       "step": 32540
     },
     {
       "epoch": 0.5288297509382464,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.4531,
       "step": 32550
     },
     {
       "epoch": 0.528992217835616,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.5021,
       "step": 32560
     },
     {
       "epoch": 0.5291546847329857,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.4524,
       "step": 32570
     },
     {
       "epoch": 0.5293171516303553,
-      "grad_norm": 3.609375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.4064,
       "step": 32580
     },
     {
       "epoch": 0.529479618527725,
-      "grad_norm": 2.78125,
+      "grad_norm": 4.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4494,
       "step": 32590
     },
     {
       "epoch": 0.5296420854250946,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.4207,
       "step": 32600
     },
     {
       "epoch": 0.5298045523224643,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.3872,
       "step": 32610
     },
     {
       "epoch": 0.5299670192198339,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4356,
       "step": 32620
     },
     {
       "epoch": 0.5301294861172037,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.414,
       "step": 32630
     },
     {
       "epoch": 0.5302919530145733,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4135,
+      "loss": 1.4129,
       "step": 32640
     },
     {
       "epoch": 0.530454419911943,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.4473,
       "step": 32650
     },
     {
       "epoch": 0.5306168868093126,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4489,
       "step": 32660
     },
     {
       "epoch": 0.5307793537066823,
-      "grad_norm": 2.84375,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4322,
       "step": 32670
     },
     {
       "epoch": 0.5309418206040519,
-      "grad_norm": 1.6875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4378,
       "step": 32680
     },
     {
       "epoch": 0.5311042875014216,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4059,
       "step": 32690
     },
     {
       "epoch": 0.5312667543987912,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4137,
       "step": 32700
     },
     {
       "epoch": 0.5314292212961609,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.4309,
       "step": 32710
     },
     {
       "epoch": 0.5315916881935305,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4529,
       "step": 32720
     },
     {
       "epoch": 0.5317541550909002,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.466,
       "step": 32730
     },
     {
       "epoch": 0.5319166219882699,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.4469,
       "step": 32740
     },
     {
       "epoch": 0.5320790888856396,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.4461,
       "step": 32750
     },
     {
       "epoch": 0.5322415557830092,
-      "grad_norm": 1.875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4319,
       "step": 32760
     },
     {
       "epoch": 0.5324040226803789,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4265,
+      "loss": 1.3821,
       "step": 32770
     },
     {
       "epoch": 0.5325664895777485,
-      "grad_norm": 1.7265625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.4897,
       "step": 32780
     },
     {
       "epoch": 0.5327289564751182,
-      "grad_norm": 1.71875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.4406,
       "step": 32790
     },
     {
       "epoch": 0.5328914233724878,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4341,
       "step": 32800
     },
     {
       "epoch": 0.5330538902698575,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3747,
+      "loss": 1.4284,
       "step": 32810
     },
     {
       "epoch": 0.5332163571672272,
-      "grad_norm": 2.5,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4309,
       "step": 32820
     },
     {
       "epoch": 0.5333788240645968,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4504,
       "step": 32830
     },
     {
       "epoch": 0.5335412909619665,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4127,
       "step": 32840
     },
     {
       "epoch": 0.5337037578593362,
-      "grad_norm": 1.7890625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.4865,
       "step": 32850
     },
     {
       "epoch": 0.5338662247567059,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.453,
       "step": 32860
     },
     {
       "epoch": 0.5340286916540755,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4218,
       "step": 32870
     },
     {
       "epoch": 0.5341911585514452,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.4342,
       "step": 32880
     },
     {
       "epoch": 0.5343536254488148,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4187,
+      "loss": 1.4477,
       "step": 32890
     },
     {
       "epoch": 0.5345160923461845,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.4334,
       "step": 32900
     },
     {
       "epoch": 0.5346785592435541,
-      "grad_norm": 2.25,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4267,
       "step": 32910
     },
     {
       "epoch": 0.5348410261409238,
-      "grad_norm": 1.578125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.4708,
       "step": 32920
     },
     {
       "epoch": 0.5350034930382934,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.4136,
       "step": 32930
     },
     {
       "epoch": 0.5351659599356631,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4054,
       "step": 32940
     },
     {
       "epoch": 0.5353284268330327,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4654,
       "step": 32950
     },
     {
       "epoch": 0.5354908937304025,
-      "grad_norm": 2.625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4197,
       "step": 32960
     },
     {
       "epoch": 0.5356533606277721,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.4375,
       "step": 32970
     },
     {
       "epoch": 0.5358158275251418,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.4674,
       "step": 32980
     },
     {
       "epoch": 0.5359782944225114,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.4455,
       "step": 32990
     },
     {
       "epoch": 0.5361407613198811,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.4818,
       "step": 33000
     },
     {
       "epoch": 0.5363032282172507,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.4531,
       "step": 33010
     },
     {
       "epoch": 0.5364656951146204,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.5005,
       "step": 33020
     },
     {
       "epoch": 0.53662816201199,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4154,
       "step": 33030
     },
     {
       "epoch": 0.5367906289093597,
-      "grad_norm": 2.0,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.4529,
       "step": 33040
     },
     {
       "epoch": 0.5369530958067293,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.479,
       "step": 33050
     },
     {
       "epoch": 0.537115562704099,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3833,
+      "loss": 1.4623,
       "step": 33060
     },
     {
       "epoch": 0.5372780296014688,
-      "grad_norm": 1.890625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4324,
       "step": 33070
     },
     {
       "epoch": 0.5374404964988384,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.4762,
       "step": 33080
     },
     {
       "epoch": 0.5376029633962081,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4934,
       "step": 33090
     },
     {
       "epoch": 0.5377654302935777,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.418,
+      "loss": 1.4436,
       "step": 33100
     },
     {
       "epoch": 0.5379278971909474,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3756,
+      "loss": 1.4548,
       "step": 33110
     },
     {
       "epoch": 0.538090364088317,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.381,
+      "loss": 1.4459,
       "step": 33120
     },
     {
       "epoch": 0.5382528309856867,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.4523,
       "step": 33130
     },
     {
       "epoch": 0.5384152978830563,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4247,
       "step": 33140
     },
     {
       "epoch": 0.538577764780426,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.4423,
       "step": 33150
     },
     {
       "epoch": 0.5387402316777956,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.4277,
       "step": 33160
     },
     {
       "epoch": 0.5389026985751653,
-      "grad_norm": 1.8828125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4446,
       "step": 33170
     },
     {
       "epoch": 0.539065165472535,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4437,
       "step": 33180
     },
     {
       "epoch": 0.5392276323699047,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.408,
       "step": 33190
     },
     {
       "epoch": 0.5393900992672743,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.4197,
       "step": 33200
     },
     {
       "epoch": 0.539552566164644,
-      "grad_norm": 1.703125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4525,
       "step": 33210
     },
     {
       "epoch": 0.5397150330620136,
-      "grad_norm": 2.484375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4696,
       "step": 33220
     },
     {
       "epoch": 0.5398774999593833,
-      "grad_norm": 2.4375,
+      "grad_norm": 4.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3972,
+      "loss": 1.4143,
       "step": 33230
     },
     {
       "epoch": 0.5400399668567529,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.4521,
       "step": 33240
     },
     {
       "epoch": 0.5402024337541226,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4126,
       "step": 33250
     },
     {
       "epoch": 0.5403649006514922,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4042,
       "step": 33260
     },
     {
       "epoch": 0.5405273675488619,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4562,
       "step": 33270
     },
     {
       "epoch": 0.5406898344462315,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4415,
       "step": 33280
     },
     {
       "epoch": 0.5408523013436013,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4332,
       "step": 33290
     },
     {
       "epoch": 0.541014768240971,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3807,
+      "loss": 1.4315,
       "step": 33300
     },
     {
       "epoch": 0.5411772351383406,
-      "grad_norm": 4.3125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4258,
+      "loss": 1.4157,
       "step": 33310
     },
     {
       "epoch": 0.5413397020357102,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4179,
       "step": 33320
     },
     {
       "epoch": 0.5415021689330799,
-      "grad_norm": 4.75,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.3995,
       "step": 33330
     },
     {
       "epoch": 0.5416646358304495,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.4515,
       "step": 33340
     },
     {
       "epoch": 0.5418271027278192,
-      "grad_norm": 2.796875,
+      "grad_norm": 4.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.4171,
       "step": 33350
     },
     {
       "epoch": 0.5419895696251888,
-      "grad_norm": 2.078125,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.4332,
       "step": 33360
     },
     {
       "epoch": 0.5421520365225585,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.4382,
       "step": 33370
     },
     {
       "epoch": 0.5423145034199282,
-      "grad_norm": 3.0,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4257,
       "step": 33380
     },
     {
       "epoch": 0.5424769703172978,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.4374,
       "step": 33390
     },
     {
       "epoch": 0.5426394372146676,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.4575,
       "step": 33400
     },
     {
       "epoch": 0.5428019041120372,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.4775,
       "step": 33410
     },
     {
       "epoch": 0.5429643710094069,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4146,
+      "loss": 1.4523,
       "step": 33420
     },
     {
       "epoch": 0.5431268379067765,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.4373,
       "step": 33430
     },
     {
       "epoch": 0.5432893048041462,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.4458,
       "step": 33440
     },
     {
       "epoch": 0.5434517717015158,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4536,
       "step": 33450
     },
     {
       "epoch": 0.5436142385988855,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.4674,
       "step": 33460
     },
     {
       "epoch": 0.5437767054962551,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.3473,
       "step": 33470
     },
     {
       "epoch": 0.5439391723936248,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4364,
       "step": 33480
     },
     {
       "epoch": 0.5441016392909944,
-      "grad_norm": 1.6953125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4345,
       "step": 33490
     },
     {
       "epoch": 0.5442641061883641,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.4021,
       "step": 33500
     },
     {
       "epoch": 0.5444265730857338,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.4724,
       "step": 33510
     },
     {
       "epoch": 0.5445890399831035,
-      "grad_norm": 1.8671875,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.4469,
       "step": 33520
     },
     {
       "epoch": 0.5447515068804731,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3883,
+      "loss": 1.4433,
       "step": 33530
     },
     {
       "epoch": 0.5449139737778428,
-      "grad_norm": 2.234375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.4888,
       "step": 33540
     },
     {
       "epoch": 0.5450764406752124,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4263,
       "step": 33550
     },
     {
       "epoch": 0.5452389075725821,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3888,
       "step": 33560
     },
     {
       "epoch": 0.5454013744699517,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4325,
+      "loss": 1.4358,
       "step": 33570
     },
     {
       "epoch": 0.5455638413673214,
-      "grad_norm": 3.0,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.4685,
       "step": 33580
     },
     {
       "epoch": 0.545726308264691,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.4554,
       "step": 33590
     },
     {
       "epoch": 0.5458887751620607,
-      "grad_norm": 2.625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4188,
+      "loss": 1.4406,
       "step": 33600
     },
     {
       "epoch": 0.5460512420594303,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4065,
       "step": 33610
     },
     {
       "epoch": 0.5462137089568001,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.406,
       "step": 33620
     },
     {
       "epoch": 0.5463761758541698,
-      "grad_norm": 2.265625,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.465,
       "step": 33630
     },
     {
       "epoch": 0.5465386427515394,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4554,
       "step": 33640
     },
     {
       "epoch": 0.5467011096489091,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4176,
+      "loss": 1.4338,
       "step": 33650
     },
     {
       "epoch": 0.5468635765462787,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.454,
       "step": 33660
     },
     {
       "epoch": 0.5470260434436484,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4423,
+      "loss": 1.46,
       "step": 33670
     },
     {
       "epoch": 0.547188510341018,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4238,
       "step": 33680
     },
     {
       "epoch": 0.5473509772383877,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3777,
+      "loss": 1.3773,
       "step": 33690
     },
     {
       "epoch": 0.5475134441357573,
-      "grad_norm": 2.25,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4356,
       "step": 33700
     },
     {
       "epoch": 0.547675911033127,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.4548,
       "step": 33710
     },
     {
       "epoch": 0.5478383779304966,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4048,
+      "loss": 1.4808,
       "step": 33720
     },
     {
       "epoch": 0.5480008448278664,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.4714,
       "step": 33730
     },
     {
       "epoch": 0.548163311725236,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.4213,
       "step": 33740
     },
     {
       "epoch": 0.5483257786226057,
-      "grad_norm": 1.765625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.4601,
       "step": 33750
     },
     {
       "epoch": 0.5484882455199753,
-      "grad_norm": 1.703125,
+      "grad_norm": 4.5,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4456,
       "step": 33760
     },
     {
       "epoch": 0.548650712417345,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4098,
       "step": 33770
     },
     {
       "epoch": 0.5488131793147146,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4077,
       "step": 33780
     },
     {
       "epoch": 0.5489756462120843,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4304,
+      "loss": 1.4576,
       "step": 33790
     },
     {
       "epoch": 0.5491381131094539,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4434,
       "step": 33800
     },
     {
       "epoch": 0.5493005800068236,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.437,
       "step": 33810
     },
     {
       "epoch": 0.5494630469041932,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4253,
+      "loss": 1.402,
       "step": 33820
     },
     {
       "epoch": 0.5496255138015629,
-      "grad_norm": 4.3125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.4785,
       "step": 33830
     },
     {
       "epoch": 0.5497879806989326,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.451,
       "step": 33840
     },
     {
       "epoch": 0.5499504475963023,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4616,
       "step": 33850
     },
     {
       "epoch": 0.550112914493672,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4122,
       "step": 33860
     },
     {
       "epoch": 0.5502753813910416,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.4701,
       "step": 33870
     },
     {
       "epoch": 0.5504378482884112,
-      "grad_norm": 2.625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3798,
+      "loss": 1.4439,
       "step": 33880
     },
     {
       "epoch": 0.5506003151857809,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.3841,
       "step": 33890
     },
     {
       "epoch": 0.5507627820831505,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.4164,
       "step": 33900
     },
     {
       "epoch": 0.5509252489805202,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4313,
       "step": 33910
     },
     {
       "epoch": 0.5510877158778898,
-      "grad_norm": 2.3125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4719,
       "step": 33920
     },
     {
       "epoch": 0.5512501827752595,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.4725,
       "step": 33930
     },
     {
       "epoch": 0.5514126496726292,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4702,
       "step": 33940
     },
     {
       "epoch": 0.5515751165699989,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3788,
+      "loss": 1.4106,
       "step": 33950
     },
     {
       "epoch": 0.5517375834673686,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.4115,
       "step": 33960
     },
     {
       "epoch": 0.5519000503647382,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4638,
       "step": 33970
     },
     {
       "epoch": 0.5520625172621079,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.4136,
       "step": 33980
     },
     {
       "epoch": 0.5522249841594775,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4976,
       "step": 33990
     },
     {
       "epoch": 0.5523874510568472,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.4257,
       "step": 34000
     },
     {
       "epoch": 0.5525499179542168,
-      "grad_norm": 4.5625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4821,
       "step": 34010
     },
     {
       "epoch": 0.5527123848515865,
-      "grad_norm": 4.34375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.384,
+      "loss": 1.4892,
       "step": 34020
     },
     {
       "epoch": 0.5528748517489561,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4142,
       "step": 34030
     },
     {
       "epoch": 0.5530373186463258,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.4495,
       "step": 34040
     },
     {
       "epoch": 0.5531997855436954,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4535,
       "step": 34050
     },
     {
       "epoch": 0.5533622524410652,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4692,
       "step": 34060
     },
     {
       "epoch": 0.5535247193384348,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4687,
       "step": 34070
     },
     {
       "epoch": 0.5536871862358045,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3798,
+      "loss": 1.4576,
       "step": 34080
     },
     {
       "epoch": 0.5538496531331741,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.4427,
       "step": 34090
     },
     {
       "epoch": 0.5540121200305438,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4791,
       "step": 34100
     },
     {
       "epoch": 0.5541745869279134,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.4499,
       "step": 34110
     },
     {
       "epoch": 0.5543370538252831,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4333,
       "step": 34120
     },
     {
       "epoch": 0.5544995207226527,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.4283,
       "step": 34130
     },
     {
       "epoch": 0.5546619876200224,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.4446,
       "step": 34140
     },
     {
       "epoch": 0.554824454517392,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.4283,
       "step": 34150
     },
     {
       "epoch": 0.5549869214147617,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.4343,
       "step": 34160
     },
     {
       "epoch": 0.5551493883121315,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.3754,
       "step": 34170
     },
     {
       "epoch": 0.5553118552095011,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4108,
+      "loss": 1.45,
       "step": 34180
     },
     {
       "epoch": 0.5554743221068708,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4243,
       "step": 34190
     },
     {
       "epoch": 0.5556367890042404,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.4081,
       "step": 34200
     },
     {
       "epoch": 0.5557992559016101,
-      "grad_norm": 1.9296875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4358,
       "step": 34210
     },
     {
       "epoch": 0.5559617227989797,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.4631,
       "step": 34220
     },
     {
       "epoch": 0.5561241896963494,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.4413,
       "step": 34230
     },
     {
       "epoch": 0.556286656593719,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.4423,
       "step": 34240
     },
     {
       "epoch": 0.5564491234910887,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.4198,
       "step": 34250
     },
     {
       "epoch": 0.5566115903884583,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.4608,
       "step": 34260
     },
     {
       "epoch": 0.556774057285828,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4212,
+      "loss": 1.4409,
       "step": 34270
     },
     {
       "epoch": 0.5569365241831977,
-      "grad_norm": 2.8125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4244,
+      "loss": 1.3937,
       "step": 34280
     },
     {
       "epoch": 0.5570989910805674,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4821,
       "step": 34290
     },
     {
       "epoch": 0.557261457977937,
-      "grad_norm": 2.375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3885,
       "step": 34300
     },
     {
       "epoch": 0.5574239248753067,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.4367,
       "step": 34310
     },
     {
       "epoch": 0.5575863917726763,
-      "grad_norm": 2.0,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4408,
       "step": 34320
     },
     {
       "epoch": 0.557748858670046,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.401,
       "step": 34330
     },
     {
       "epoch": 0.5579113255674156,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.374,
+      "loss": 1.4785,
       "step": 34340
     },
     {
       "epoch": 0.5580737924647853,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.5038,
       "step": 34350
     },
     {
       "epoch": 0.5582362593621549,
-      "grad_norm": 2.875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4351,
       "step": 34360
     },
     {
       "epoch": 0.5583987262595246,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4372,
       "step": 34370
     },
     {
       "epoch": 0.5585611931568942,
-      "grad_norm": 1.765625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.444,
       "step": 34380
     },
     {
       "epoch": 0.558723660054264,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4299,
+      "loss": 1.399,
       "step": 34390
     },
     {
       "epoch": 0.5588861269516336,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3867,
+      "loss": 1.4413,
       "step": 34400
     },
     {
       "epoch": 0.5590485938490033,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.4196,
       "step": 34410
     },
     {
       "epoch": 0.559211060746373,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.4289,
       "step": 34420
     },
     {
       "epoch": 0.5593735276437426,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4224,
+      "loss": 1.4285,
       "step": 34430
     },
     {
       "epoch": 0.5595359945411122,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.446,
       "step": 34440
     },
     {
       "epoch": 0.5596984614384819,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.4204,
       "step": 34450
     },
     {
       "epoch": 0.5598609283358515,
-      "grad_norm": 1.59375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4515,
       "step": 34460
     },
     {
       "epoch": 0.5600233952332212,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.4183,
       "step": 34470
     },
     {
       "epoch": 0.5601858621305909,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.4251,
       "step": 34480
     },
     {
       "epoch": 0.5603483290279605,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3773,
+      "loss": 1.3951,
       "step": 34490
     },
     {
       "epoch": 0.5605107959253303,
-      "grad_norm": 2.0,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4648,
       "step": 34500
     },
     {
       "epoch": 0.5606732628226999,
-      "grad_norm": 2.75,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.4586,
       "step": 34510
     },
     {
       "epoch": 0.5608357297200696,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.4187,
       "step": 34520
     },
     {
       "epoch": 0.5609981966174392,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4004,
       "step": 34530
     },
     {
       "epoch": 0.5611606635148089,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.415,
+      "loss": 1.4906,
       "step": 34540
     },
     {
       "epoch": 0.5613231304121785,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.4287,
       "step": 34550
     },
     {
       "epoch": 0.5614855973095482,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.4804,
       "step": 34560
     },
     {
       "epoch": 0.5616480642069178,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4375,
       "step": 34570
     },
     {
       "epoch": 0.5618105311042875,
-      "grad_norm": 2.625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.451,
       "step": 34580
     },
     {
       "epoch": 0.5619729980016571,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.4354,
       "step": 34590
     },
     {
       "epoch": 0.5621354648990268,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.3764,
       "step": 34600
     },
     {
       "epoch": 0.5622979317963965,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4734,
       "step": 34610
     },
     {
       "epoch": 0.5624603986937662,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.4037,
       "step": 34620
     },
     {
       "epoch": 0.5626228655911358,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.4807,
       "step": 34630
     },
     {
       "epoch": 0.5627853324885055,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4208,
+      "loss": 1.4933,
       "step": 34640
     },
     {
       "epoch": 0.5629477993858751,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4331,
       "step": 34650
     },
     {
       "epoch": 0.5631102662832448,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4321,
       "step": 34660
     },
     {
       "epoch": 0.5632727331806144,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.4578,
       "step": 34670
     },
     {
       "epoch": 0.5634352000779841,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4592,
       "step": 34680
     },
     {
       "epoch": 0.5635976669753537,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4392,
       "step": 34690
     },
     {
       "epoch": 0.5637601338727234,
-      "grad_norm": 1.6171875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4315,
       "step": 34700
     },
     {
       "epoch": 0.563922600770093,
-      "grad_norm": 2.34375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.4573,
       "step": 34710
     },
     {
       "epoch": 0.5640850676674628,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.4403,
       "step": 34720
     },
     {
       "epoch": 0.5642475345648325,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.4609,
       "step": 34730
     },
     {
       "epoch": 0.5644100014622021,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.3901,
       "step": 34740
     },
     {
       "epoch": 0.5645724683595718,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4187,
+      "loss": 1.4132,
       "step": 34750
     },
     {
       "epoch": 0.5647349352569414,
-      "grad_norm": 1.8515625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4157,
       "step": 34760
     },
     {
       "epoch": 0.5648974021543111,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3808,
+      "loss": 1.468,
       "step": 34770
     },
     {
       "epoch": 0.5650598690516807,
-      "grad_norm": 1.6328125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3738,
+      "loss": 1.4278,
       "step": 34780
     },
     {
       "epoch": 0.5652223359490504,
-      "grad_norm": 2.0,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.451,
       "step": 34790
     },
     {
       "epoch": 0.56538480284642,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4743,
       "step": 34800
     },
     {
       "epoch": 0.5655472697437897,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4182,
       "step": 34810
     },
     {
       "epoch": 0.5657097366411593,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.3824,
       "step": 34820
     },
     {
       "epoch": 0.5658722035385291,
-      "grad_norm": 1.609375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4154,
+      "loss": 1.4347,
       "step": 34830
     },
     {
       "epoch": 0.5660346704358987,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.431,
+      "loss": 1.4034,
       "step": 34840
     },
     {
       "epoch": 0.5661971373332684,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4647,
       "step": 34850
     },
     {
       "epoch": 0.566359604230638,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4541,
       "step": 34860
     },
     {
       "epoch": 0.5665220711280077,
-      "grad_norm": 2.125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4759,
       "step": 34870
     },
     {
       "epoch": 0.5666845380253773,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3829,
+      "loss": 1.4266,
       "step": 34880
     },
     {
       "epoch": 0.566847004922747,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4544,
       "step": 34890
     },
     {
       "epoch": 0.5670094718201166,
-      "grad_norm": 2.46875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.434,
       "step": 34900
     },
     {
       "epoch": 0.5671719387174863,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.416,
       "step": 34910
     },
     {
       "epoch": 0.5673344056148559,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.467,
       "step": 34920
     },
     {
       "epoch": 0.5674968725122256,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.4357,
       "step": 34930
     },
     {
       "epoch": 0.5676593394095953,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.4945,
       "step": 34940
     },
     {
       "epoch": 0.567821806306965,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4271,
+      "loss": 1.4259,
       "step": 34950
     },
     {
       "epoch": 0.5679842732043346,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.4588,
       "step": 34960
     },
     {
       "epoch": 0.5681467401017043,
-      "grad_norm": 1.6953125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.4345,
       "step": 34970
     },
     {
       "epoch": 0.568309206999074,
-      "grad_norm": 1.7421875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.3893,
       "step": 34980
     },
     {
       "epoch": 0.5684716738964436,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.3922,
       "step": 34990
     },
     {
       "epoch": 0.5686341407938132,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4492,
       "step": 35000
     },
     {
       "epoch": 0.5687966076911829,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.471,
       "step": 35010
     },
     {
       "epoch": 0.5689590745885525,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4569,
       "step": 35020
     },
     {
       "epoch": 0.5691215414859222,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4923,
       "step": 35030
     },
     {
       "epoch": 0.5692840083832919,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.4389,
       "step": 35040
     },
     {
       "epoch": 0.5694464752806616,
-      "grad_norm": 2.375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4547,
       "step": 35050
     },
     {
       "epoch": 0.5696089421780313,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.474,
       "step": 35060
     },
     {
       "epoch": 0.5697714090754009,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4524,
       "step": 35070
     },
     {
       "epoch": 0.5699338759727706,
-      "grad_norm": 1.6640625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4674,
       "step": 35080
     },
     {
       "epoch": 0.5700963428701402,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4207,
+      "loss": 1.4259,
       "step": 35090
     },
     {
       "epoch": 0.5702588097675099,
-      "grad_norm": 2.75,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4098,
       "step": 35100
     },
     {
       "epoch": 0.5704212766648795,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.4241,
       "step": 35110
     },
     {
       "epoch": 0.5705837435622492,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4583,
       "step": 35120
     },
     {
       "epoch": 0.5707462104596188,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3828,
+      "loss": 1.4244,
       "step": 35130
     },
     {
       "epoch": 0.5709086773569885,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.4188,
       "step": 35140
     },
     {
       "epoch": 0.5710711442543581,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.457,
       "step": 35150
     },
     {
       "epoch": 0.5712336111517279,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.4063,
       "step": 35160
     },
     {
       "epoch": 0.5713960780490975,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.4252,
       "step": 35170
     },
     {
       "epoch": 0.5715585449464672,
-      "grad_norm": 2.75,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3805,
+      "loss": 1.3947,
       "step": 35180
     },
     {
       "epoch": 0.5717210118438368,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4457,
       "step": 35190
     },
     {
       "epoch": 0.5718834787412065,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4371,
       "step": 35200
     },
     {
       "epoch": 0.5720459456385761,
-      "grad_norm": 1.875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.443,
       "step": 35210
     },
     {
       "epoch": 0.5722084125359458,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.4034,
       "step": 35220
     },
     {
       "epoch": 0.5723708794333154,
-      "grad_norm": 4.875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4004,
       "step": 35230
     },
     {
       "epoch": 0.5725333463306851,
-      "grad_norm": 2.25,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.3782,
       "step": 35240
     },
     {
       "epoch": 0.5726958132280547,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4192,
       "step": 35250
     },
     {
       "epoch": 0.5728582801254244,
-      "grad_norm": 2.5,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4283,
+      "loss": 1.4075,
       "step": 35260
     },
     {
       "epoch": 0.5730207470227942,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.3916,
       "step": 35270
     },
     {
       "epoch": 0.5731832139201638,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.4052,
       "step": 35280
     },
     {
       "epoch": 0.5733456808175335,
-      "grad_norm": 1.9296875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4373,
       "step": 35290
     },
     {
       "epoch": 0.5735081477149031,
-      "grad_norm": 1.578125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.382,
+      "loss": 1.4424,
       "step": 35300
     },
     {
       "epoch": 0.5736706146122728,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4007,
       "step": 35310
     },
     {
       "epoch": 0.5738330815096424,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.448,
       "step": 35320
     },
     {
       "epoch": 0.5739955484070121,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.4314,
       "step": 35330
     },
     {
       "epoch": 0.5741580153043817,
-      "grad_norm": 3.203125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.472,
       "step": 35340
     },
     {
       "epoch": 0.5743204822017514,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4357,
       "step": 35350
     },
     {
       "epoch": 0.574482949099121,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4451,
       "step": 35360
     },
     {
       "epoch": 0.5746454159964907,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.4063,
       "step": 35370
     },
     {
       "epoch": 0.5748078828938604,
-      "grad_norm": 3.5,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4497,
       "step": 35380
     },
     {
       "epoch": 0.5749703497912301,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3834,
+      "loss": 1.416,
       "step": 35390
     },
     {
       "epoch": 0.5751328166885997,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.4355,
       "step": 35400
     },
     {
       "epoch": 0.5752952835859694,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4248,
+      "loss": 1.4534,
       "step": 35410
     },
     {
       "epoch": 0.575457750483339,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4302,
       "step": 35420
     },
     {
       "epoch": 0.5756202173807087,
-      "grad_norm": 1.875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4921,
       "step": 35430
     },
     {
       "epoch": 0.5757826842780783,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4614,
       "step": 35440
     },
     {
       "epoch": 0.575945151175448,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4034,
+      "loss": 1.435,
       "step": 35450
     },
     {
       "epoch": 0.5761076180728176,
-      "grad_norm": 3.890625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.4406,
       "step": 35460
     },
     {
       "epoch": 0.5762700849701873,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.4534,
       "step": 35470
     },
     {
       "epoch": 0.5764325518675569,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.5064,
       "step": 35480
     },
     {
       "epoch": 0.5765950187649267,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.4095,
       "step": 35490
     },
     {
       "epoch": 0.5767574856622963,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.4876,
       "step": 35500
     },
     {
       "epoch": 0.576919952559666,
-      "grad_norm": 4.4375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4267,
+      "loss": 1.3881,
       "step": 35510
     },
     {
       "epoch": 0.5770824194570356,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4493,
       "step": 35520
     },
     {
       "epoch": 0.5772448863544053,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.4143,
       "step": 35530
     },
     {
       "epoch": 0.577407353251775,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3895,
+      "loss": 1.4526,
       "step": 35540
     },
     {
       "epoch": 0.5775698201491446,
-      "grad_norm": 2.0,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4168,
       "step": 35550
     },
     {
       "epoch": 0.5777322870465142,
-      "grad_norm": 1.7578125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4075,
       "step": 35560
     },
     {
       "epoch": 0.5778947539438839,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3737,
+      "loss": 1.4299,
       "step": 35570
     },
     {
       "epoch": 0.5780572208412535,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.4489,
       "step": 35580
     },
     {
       "epoch": 0.5782196877386232,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.434,
       "step": 35590
     },
     {
       "epoch": 0.578382154635993,
-      "grad_norm": 4.8125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4077,
       "step": 35600
     },
     {
       "epoch": 0.5785446215333626,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4435,
       "step": 35610
     },
     {
       "epoch": 0.5787070884307323,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.3966,
       "step": 35620
     },
     {
       "epoch": 0.5788695553281019,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4394,
       "step": 35630
     },
     {
       "epoch": 0.5790320222254716,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.4414,
       "step": 35640
     },
     {
       "epoch": 0.5791944891228412,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4216,
       "step": 35650
     },
     {
       "epoch": 0.5793569560202109,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.4376,
       "step": 35660
     },
     {
       "epoch": 0.5795194229175805,
-      "grad_norm": 1.4140625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4261,
       "step": 35670
     },
     {
       "epoch": 0.5796818898149502,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3734,
+      "loss": 1.4516,
       "step": 35680
     },
     {
       "epoch": 0.5798443567123198,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.3857,
       "step": 35690
     },
     {
       "epoch": 0.5800068236096895,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.4119,
       "step": 35700
     },
     {
       "epoch": 0.5801692905070592,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.37,
+      "loss": 1.4909,
       "step": 35710
     },
     {
       "epoch": 0.5803317574044289,
-      "grad_norm": 2.75,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3863,
+      "loss": 1.3701,
       "step": 35720
     },
     {
       "epoch": 0.5804942243017985,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.4288,
       "step": 35730
     },
     {
       "epoch": 0.5806566911991682,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.4556,
       "step": 35740
     },
     {
       "epoch": 0.5808191580965378,
-      "grad_norm": 2.375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.4366,
       "step": 35750
     },
     {
       "epoch": 0.5809816249939075,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4582,
       "step": 35760
     },
     {
       "epoch": 0.5811440918912771,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4042,
       "step": 35770
     },
     {
       "epoch": 0.5813065587886468,
-      "grad_norm": 1.4140625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.4669,
       "step": 35780
     },
     {
       "epoch": 0.5814690256860164,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4443,
       "step": 35790
     },
     {
       "epoch": 0.5816314925833861,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.4357,
       "step": 35800
     },
     {
       "epoch": 0.5817939594807559,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4491,
       "step": 35810
     },
     {
       "epoch": 0.5819564263781255,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4259,
+      "loss": 1.4234,
       "step": 35820
     },
     {
       "epoch": 0.5821188932754952,
-      "grad_norm": 1.78125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.4949,
       "step": 35830
     },
     {
       "epoch": 0.5822813601728648,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4605,
       "step": 35840
     },
     {
       "epoch": 0.5824438270702345,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4373,
       "step": 35850
     },
     {
       "epoch": 0.5826062939676041,
-      "grad_norm": 1.796875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.3746,
       "step": 35860
     },
     {
       "epoch": 0.5827687608649738,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.4107,
       "step": 35870
     },
     {
       "epoch": 0.5829312277623434,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4301,
       "step": 35880
     },
     {
       "epoch": 0.5830936946597131,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4197,
       "step": 35890
     },
     {
       "epoch": 0.5832561615570827,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4243,
       "step": 35900
     },
     {
       "epoch": 0.5834186284544524,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3994,
+      "loss": 1.4425,
       "step": 35910
     },
     {
       "epoch": 0.5835810953518221,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.466,
       "step": 35920
     },
     {
       "epoch": 0.5837435622491918,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3877,
+      "loss": 1.4377,
       "step": 35930
     },
     {
       "epoch": 0.5839060291465614,
-      "grad_norm": 3.21875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.4338,
       "step": 35940
     },
     {
       "epoch": 0.5840684960439311,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4193,
+      "loss": 1.4234,
       "step": 35950
     },
     {
       "epoch": 0.5842309629413007,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.398,
       "step": 35960
     },
     {
       "epoch": 0.5843934298386704,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.4152,
       "step": 35970
     },
     {
       "epoch": 0.58455589673604,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4192,
       "step": 35980
     },
     {
       "epoch": 0.5847183636334097,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4265,
+      "loss": 1.422,
       "step": 35990
     },
     {
       "epoch": 0.5848808305307793,
-      "grad_norm": 2.734375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4057,
       "step": 36000
     },
     {
       "epoch": 0.585043297428149,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.3998,
       "step": 36010
     },
     {
       "epoch": 0.5852057643255186,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.438,
       "step": 36020
     },
     {
       "epoch": 0.5853682312228884,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.4203,
       "step": 36030
     },
     {
       "epoch": 0.585530698120258,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.5001,
       "step": 36040
     },
     {
       "epoch": 0.5856931650176277,
-      "grad_norm": 3.5625,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4492,
       "step": 36050
     },
     {
       "epoch": 0.5858556319149973,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4566,
       "step": 36060
     },
     {
       "epoch": 0.586018098812367,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4031,
       "step": 36070
     },
     {
       "epoch": 0.5861805657097366,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4535,
       "step": 36080
     },
     {
       "epoch": 0.5863430326071063,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.4995,
       "step": 36090
     },
     {
       "epoch": 0.586505499504476,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4086,
       "step": 36100
     },
     {
       "epoch": 0.5866679664018456,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4534,
       "step": 36110
     },
     {
       "epoch": 0.5868304332992152,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.4454,
       "step": 36120
     },
     {
       "epoch": 0.5869929001965849,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.3996,
       "step": 36130
     },
     {
       "epoch": 0.5871553670939547,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.3995,
       "step": 36140
     },
     {
       "epoch": 0.5873178339913243,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4516,
       "step": 36150
     },
     {
       "epoch": 0.587480300888694,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.3955,
       "step": 36160
     },
     {
       "epoch": 0.5876427677860636,
-      "grad_norm": 2.015625,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.4375,
       "step": 36170
     },
     {
       "epoch": 0.5878052346834333,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4515,
       "step": 36180
     },
     {
       "epoch": 0.5879677015808029,
-      "grad_norm": 2.84375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.4484,
       "step": 36190
     },
     {
       "epoch": 0.5881301684781726,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4131,
       "step": 36200
     },
     {
       "epoch": 0.5882926353755422,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4195,
+      "loss": 1.4752,
       "step": 36210
     },
     {
       "epoch": 0.5884551022729119,
-      "grad_norm": 1.2734375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3779,
+      "loss": 1.4369,
       "step": 36220
     },
     {
       "epoch": 0.5886175691702815,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4236,
       "step": 36230
     },
     {
       "epoch": 0.5887800360676512,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3816,
+      "loss": 1.4181,
       "step": 36240
     },
     {
       "epoch": 0.5889425029650209,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4461,
       "step": 36250
     },
     {
       "epoch": 0.5891049698623906,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4436,
       "step": 36260
     },
     {
       "epoch": 0.5892674367597602,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4365,
       "step": 36270
     },
     {
       "epoch": 0.5894299036571299,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.3844,
       "step": 36280
     },
     {
       "epoch": 0.5895923705544995,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.4354,
       "step": 36290
     },
     {
       "epoch": 0.5897548374518692,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4252,
+      "loss": 1.4205,
       "step": 36300
     },
     {
       "epoch": 0.5899173043492388,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4154,
       "step": 36310
     },
     {
       "epoch": 0.5900797712466085,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4077,
       "step": 36320
     },
     {
       "epoch": 0.5902422381439781,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.4149,
       "step": 36330
     },
     {
       "epoch": 0.5904047050413478,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4021,
       "step": 36340
     },
     {
       "epoch": 0.5905671719387174,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.3952,
       "step": 36350
     },
     {
       "epoch": 0.5907296388360872,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.4234,
       "step": 36360
     },
     {
       "epoch": 0.5908921057334569,
-      "grad_norm": 3.25,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.396,
+      "loss": 1.4018,
       "step": 36370
     },
     {
       "epoch": 0.5910545726308265,
-      "grad_norm": 1.7890625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4228,
+      "loss": 1.4319,
       "step": 36380
     },
     {
       "epoch": 0.5912170395281962,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4225,
+      "loss": 1.4588,
       "step": 36390
     },
     {
       "epoch": 0.5913795064255658,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.3854,
       "step": 36400
     },
     {
       "epoch": 0.5915419733229355,
-      "grad_norm": 3.78125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.4226,
       "step": 36410
     },
     {
       "epoch": 0.5917044402203051,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.4403,
       "step": 36420
     },
     {
       "epoch": 0.5918669071176748,
-      "grad_norm": 3.78125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.4369,
       "step": 36430
     },
     {
       "epoch": 0.5920293740150444,
-      "grad_norm": 1.921875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.418,
+      "loss": 1.4534,
       "step": 36440
     },
     {
       "epoch": 0.5921918409124141,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3729,
+      "loss": 1.464,
       "step": 36450
     },
     {
       "epoch": 0.5923543078097837,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4393,
       "step": 36460
     },
     {
       "epoch": 0.5925167747071535,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.4332,
       "step": 36470
     },
     {
       "epoch": 0.5926792416045231,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4801,
       "step": 36480
     },
     {
       "epoch": 0.5928417085018928,
-      "grad_norm": 1.546875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.4198,
       "step": 36490
     },
     {
       "epoch": 0.5930041753992624,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.4216,
       "step": 36500
     },
     {
       "epoch": 0.5931666422966321,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.4343,
       "step": 36510
     },
     {
       "epoch": 0.5933291091940017,
-      "grad_norm": 2.625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4302,
       "step": 36520
     },
     {
       "epoch": 0.5934915760913714,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.4468,
       "step": 36530
     },
     {
       "epoch": 0.593654042988741,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4371,
       "step": 36540
     },
     {
       "epoch": 0.5938165098861107,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.4358,
       "step": 36550
     },
     {
       "epoch": 0.5939789767834803,
-      "grad_norm": 2.71875,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.4449,
       "step": 36560
     },
     {
       "epoch": 0.59414144368085,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4435,
       "step": 36570
     },
     {
       "epoch": 0.5943039105782197,
-      "grad_norm": 3.25,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.4236,
       "step": 36580
     },
     {
       "epoch": 0.5944663774755894,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4587,
       "step": 36590
     },
     {
       "epoch": 0.594628844372959,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4002,
+      "loss": 1.4179,
       "step": 36600
     },
     {
       "epoch": 0.5947913112703287,
-      "grad_norm": 2.875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.413,
       "step": 36610
     },
     {
       "epoch": 0.5949537781676983,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4243,
       "step": 36620
     },
     {
       "epoch": 0.595116245065068,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4138,
+      "loss": 1.3959,
       "step": 36630
     },
     {
       "epoch": 0.5952787119624376,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4197,
       "step": 36640
     },
     {
       "epoch": 0.5954411788598073,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.423,
       "step": 36650
     },
     {
       "epoch": 0.595603645757177,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4249,
       "step": 36660
     },
     {
       "epoch": 0.5957661126545466,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.3955,
       "step": 36670
     },
     {
       "epoch": 0.5959285795519162,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4362,
       "step": 36680
     },
     {
       "epoch": 0.596091046449286,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4064,
+      "loss": 1.4236,
       "step": 36690
     },
     {
       "epoch": 0.5962535133466557,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.3956,
       "step": 36700
     },
     {
       "epoch": 0.5964159802440253,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.4496,
       "step": 36710
     },
     {
       "epoch": 0.596578447141395,
-      "grad_norm": 1.6953125,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4245,
       "step": 36720
     },
     {
       "epoch": 0.5967409140387646,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4034,
       "step": 36730
     },
     {
       "epoch": 0.5969033809361343,
-      "grad_norm": 1.8984375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.3683,
       "step": 36740
     },
     {
       "epoch": 0.5970658478335039,
-      "grad_norm": 1.7734375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3825,
+      "loss": 1.4286,
       "step": 36750
     },
     {
       "epoch": 0.5972283147308736,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.4386,
       "step": 36760
     },
     {
       "epoch": 0.5973907816282432,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4177,
+      "loss": 1.443,
       "step": 36770
     },
     {
       "epoch": 0.5975532485256129,
-      "grad_norm": 3.78125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4488,
       "step": 36780
     },
     {
       "epoch": 0.5977157154229825,
-      "grad_norm": 1.8203125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3807,
+      "loss": 1.4152,
       "step": 36790
     },
     {
       "epoch": 0.5978781823203523,
-      "grad_norm": 1.71875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.4479,
       "step": 36800
     },
     {
       "epoch": 0.5980406492177219,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4627,
       "step": 36810
     },
     {
       "epoch": 0.5982031161150916,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4269,
+      "loss": 1.4551,
       "step": 36820
     },
     {
       "epoch": 0.5983655830124612,
-      "grad_norm": 1.875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4623,
       "step": 36830
     },
     {
       "epoch": 0.5985280499098309,
-      "grad_norm": 1.609375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.4108,
       "step": 36840
     },
     {
       "epoch": 0.5986905168072005,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4002,
+      "loss": 1.4446,
       "step": 36850
     },
     {
       "epoch": 0.5988529837045702,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4418,
       "step": 36860
     },
     {
       "epoch": 0.5990154506019398,
-      "grad_norm": 2.625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4093,
+      "loss": 1.4351,
       "step": 36870
     },
     {
       "epoch": 0.5991779174993095,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.4279,
       "step": 36880
     },
     {
       "epoch": 0.5993403843966791,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.4689,
       "step": 36890
     },
     {
       "epoch": 0.5995028512940488,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4284,
       "step": 36900
     },
     {
       "epoch": 0.5996653181914186,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.42,
       "step": 36910
     },
     {
       "epoch": 0.5998277850887882,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.4244,
       "step": 36920
     },
     {
       "epoch": 0.5999902519861579,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.43,
       "step": 36930
     },
     {
       "epoch": 0.6001527188835275,
-      "grad_norm": 3.0,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4523,
       "step": 36940
     },
     {
       "epoch": 0.6003151857808972,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.411,
       "step": 36950
     },
     {
       "epoch": 0.6004776526782668,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.3922,
       "step": 36960
     },
     {
       "epoch": 0.6006401195756365,
-      "grad_norm": 1.75,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4321,
       "step": 36970
     },
     {
       "epoch": 0.6008025864730061,
-      "grad_norm": 1.6171875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4221,
       "step": 36980
     },
     {
       "epoch": 0.6009650533703758,
-      "grad_norm": 2.25,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4141,
       "step": 36990
     },
     {
       "epoch": 0.6011275202677454,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.4806,
       "step": 37000
     },
     {
       "epoch": 0.6012899871651151,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.4249,
       "step": 37010
     },
     {
       "epoch": 0.6014524540624848,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.3885,
       "step": 37020
     },
     {
       "epoch": 0.6016149209598545,
-      "grad_norm": 1.9140625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.3972,
       "step": 37030
     },
     {
       "epoch": 0.6017773878572241,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.394,
       "step": 37040
     },
     {
       "epoch": 0.6019398547545938,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.3784,
       "step": 37050
     },
     {
       "epoch": 0.6021023216519634,
-      "grad_norm": 2.375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4029,
+      "loss": 1.4082,
       "step": 37060
     },
     {
       "epoch": 0.6022647885493331,
-      "grad_norm": 1.7421875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4224,
       "step": 37070
     },
     {
       "epoch": 0.6024272554467027,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.3491,
       "step": 37080
     },
     {
       "epoch": 0.6025897223440724,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4089,
       "step": 37090
     },
     {
       "epoch": 0.602752189241442,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4068,
       "step": 37100
     },
     {
       "epoch": 0.6029146561388117,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4557,
       "step": 37110
     },
     {
       "epoch": 0.6030771230361813,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.471,
       "step": 37120
     },
     {
       "epoch": 0.6032395899335511,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.3979,
       "step": 37130
     },
     {
       "epoch": 0.6034020568309207,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.4487,
       "step": 37140
     },
     {
       "epoch": 0.6035645237282904,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4036,
       "step": 37150
     },
     {
       "epoch": 0.60372699062566,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.38,
       "step": 37160
     },
     {
       "epoch": 0.6038894575230297,
-      "grad_norm": 2.375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4558,
       "step": 37170
     },
     {
       "epoch": 0.6040519244203993,
-      "grad_norm": 1.453125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3774,
+      "loss": 1.4557,
       "step": 37180
     },
     {
       "epoch": 0.604214391317769,
-      "grad_norm": 2.53125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.4482,
       "step": 37190
     },
     {
       "epoch": 0.6043768582151386,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4104,
       "step": 37200
     },
     {
       "epoch": 0.6045393251125083,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4256,
+      "loss": 1.439,
       "step": 37210
     },
     {
       "epoch": 0.604701792009878,
-      "grad_norm": 2.4375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4029,
+      "loss": 1.4011,
       "step": 37220
     },
     {
       "epoch": 0.6048642589072476,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.4207,
       "step": 37230
     },
     {
       "epoch": 0.6050267258046174,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.387,
+      "loss": 1.4027,
       "step": 37240
     },
     {
       "epoch": 0.605189192701987,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.4009,
       "step": 37250
     },
     {
       "epoch": 0.6053516595993567,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4194,
       "step": 37260
     },
     {
       "epoch": 0.6055141264967263,
-      "grad_norm": 4.09375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.3958,
       "step": 37270
     },
     {
       "epoch": 0.605676593394096,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.4446,
       "step": 37280
     },
     {
       "epoch": 0.6058390602914656,
-      "grad_norm": 2.5,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3637,
       "step": 37290
     },
     {
       "epoch": 0.6060015271888353,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.453,
       "step": 37300
     },
     {
       "epoch": 0.6061639940862049,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.3846,
       "step": 37310
     },
     {
       "epoch": 0.6063264609835746,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.4939,
       "step": 37320
     },
     {
       "epoch": 0.6064889278809442,
-      "grad_norm": 2.125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4595,
       "step": 37330
     },
     {
       "epoch": 0.6066513947783139,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.4221,
       "step": 37340
     },
     {
       "epoch": 0.6068138616756836,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3863,
+      "loss": 1.4233,
       "step": 37350
     },
     {
       "epoch": 0.6069763285730533,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4215,
       "step": 37360
     },
     {
       "epoch": 0.6071387954704229,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.4607,
       "step": 37370
     },
     {
       "epoch": 0.6073012623677926,
-      "grad_norm": 2.375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.476,
       "step": 37380
     },
     {
       "epoch": 0.6074637292651622,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4624,
       "step": 37390
     },
     {
       "epoch": 0.6076261961625319,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4012,
       "step": 37400
     },
     {
       "epoch": 0.6077886630599015,
-      "grad_norm": 1.890625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4227,
+      "loss": 1.413,
       "step": 37410
     },
     {
       "epoch": 0.6079511299572712,
-      "grad_norm": 1.9921875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3822,
+      "loss": 1.4384,
       "step": 37420
     },
     {
       "epoch": 0.6081135968546408,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4035,
       "step": 37430
     },
     {
       "epoch": 0.6082760637520105,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4138,
       "step": 37440
     },
     {
       "epoch": 0.6084385306493801,
-      "grad_norm": 2.375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.3855,
       "step": 37450
     },
     {
       "epoch": 0.6086009975467499,
-      "grad_norm": 1.84375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4183,
       "step": 37460
     },
     {
       "epoch": 0.6087634644441196,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4486,
       "step": 37470
     },
     {
       "epoch": 0.6089259313414892,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3743,
+      "loss": 1.4277,
       "step": 37480
     },
     {
       "epoch": 0.6090883982388589,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.4119,
       "step": 37490
     },
     {
       "epoch": 0.6092508651362285,
-      "grad_norm": 4.5625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4584,
       "step": 37500
     },
     {
       "epoch": 0.6094133320335982,
-      "grad_norm": 1.875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4579,
       "step": 37510
     },
     {
       "epoch": 0.6095757989309678,
-      "grad_norm": 2.1875,
+      "grad_norm": 4.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.4431,
       "step": 37520
     },
     {
       "epoch": 0.6097382658283375,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4429,
       "step": 37530
     },
     {
       "epoch": 0.6099007327257071,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.4279,
       "step": 37540
     },
     {
       "epoch": 0.6100631996230768,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.469,
       "step": 37550
     },
     {
       "epoch": 0.6102256665204464,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.4248,
       "step": 37560
     },
     {
       "epoch": 0.6103881334178162,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3749,
+      "loss": 1.4504,
       "step": 37570
     },
     {
       "epoch": 0.6105506003151858,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.4244,
       "step": 37580
     },
     {
       "epoch": 0.6107130672125555,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4177,
       "step": 37590
     },
     {
       "epoch": 0.6108755341099251,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.4245,
       "step": 37600
     },
     {
       "epoch": 0.6110380010072948,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4148,
       "step": 37610
     },
     {
       "epoch": 0.6112004679046644,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4025,
       "step": 37620
     },
     {
       "epoch": 0.6113629348020341,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.4453,
       "step": 37630
     },
     {
       "epoch": 0.6115254016994037,
-      "grad_norm": 1.65625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.4231,
       "step": 37640
     },
     {
       "epoch": 0.6116878685967734,
-      "grad_norm": 2.0,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4198,
+      "loss": 1.3774,
       "step": 37650
     },
     {
       "epoch": 0.611850335494143,
-      "grad_norm": 2.125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4687,
       "step": 37660
     },
     {
       "epoch": 0.6120128023915127,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4492,
       "step": 37670
     },
     {
       "epoch": 0.6121752692888824,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4872,
       "step": 37680
     },
     {
       "epoch": 0.6123377361862521,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.4188,
       "step": 37690
     },
     {
       "epoch": 0.6125002030836217,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.4452,
       "step": 37700
     },
     {
       "epoch": 0.6126626699809914,
-      "grad_norm": 3.0,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4134,
+      "loss": 1.4165,
       "step": 37710
     },
     {
       "epoch": 0.612825136878361,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4412,
       "step": 37720
     },
     {
       "epoch": 0.6129876037757307,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3715,
+      "loss": 1.3569,
       "step": 37730
     },
     {
       "epoch": 0.6131500706731003,
-      "grad_norm": 2.75,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4373,
       "step": 37740
     },
     {
       "epoch": 0.61331253757047,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4351,
       "step": 37750
     },
     {
       "epoch": 0.6134750044678396,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4573,
       "step": 37760
     },
     {
       "epoch": 0.6136374713652093,
-      "grad_norm": 1.8203125,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4369,
       "step": 37770
     },
     {
       "epoch": 0.613799938262579,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.4613,
       "step": 37780
     },
     {
       "epoch": 0.6139624051599487,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.421,
       "step": 37790
     },
     {
       "epoch": 0.6141248720573184,
-      "grad_norm": 1.8203125,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3856,
+      "loss": 1.4808,
       "step": 37800
     },
     {
       "epoch": 0.614287338954688,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4511,
       "step": 37810
     },
     {
       "epoch": 0.6144498058520577,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.429,
       "step": 37820
     },
     {
       "epoch": 0.6146122727494273,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.4488,
       "step": 37830
     },
     {
       "epoch": 0.614774739646797,
-      "grad_norm": 2.125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.4208,
       "step": 37840
     },
     {
       "epoch": 0.6149372065441666,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.4159,
       "step": 37850
     },
     {
       "epoch": 0.6150996734415363,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.3986,
       "step": 37860
     },
     {
       "epoch": 0.6152621403389059,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3811,
+      "loss": 1.3819,
       "step": 37870
     },
     {
       "epoch": 0.6154246072362756,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.3389,
       "step": 37880
     },
     {
       "epoch": 0.6155870741336452,
-      "grad_norm": 2.25,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.4273,
+      "loss": 1.4269,
       "step": 37890
     },
     {
       "epoch": 0.615749541031015,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.4189,
       "step": 37900
     },
     {
       "epoch": 0.6159120079283846,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.4322,
       "step": 37910
     },
     {
       "epoch": 0.6160744748257543,
-      "grad_norm": 1.7890625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4927,
       "step": 37920
     },
     {
       "epoch": 0.6162369417231239,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.4041,
       "step": 37930
     },
     {
       "epoch": 0.6163994086204936,
-      "grad_norm": 2.125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.4721,
       "step": 37940
     },
     {
       "epoch": 0.6165618755178632,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4363,
       "step": 37950
     },
     {
       "epoch": 0.6167243424152329,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.4423,
       "step": 37960
     },
     {
       "epoch": 0.6168868093126025,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4419,
       "step": 37970
     },
     {
       "epoch": 0.6170492762099722,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4678,
       "step": 37980
     },
     {
       "epoch": 0.6172117431073418,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3972,
+      "loss": 1.3982,
       "step": 37990
     },
     {
       "epoch": 0.6173742100047115,
-      "grad_norm": 1.625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.4436,
       "step": 38000
     },
     {
       "epoch": 0.6175366769020812,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.4157,
       "step": 38010
     },
     {
       "epoch": 0.6176991437994509,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4426,
       "step": 38020
     },
     {
       "epoch": 0.6178616106968206,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4061,
       "step": 38030
     },
     {
       "epoch": 0.6180240775941902,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.418,
       "step": 38040
     },
     {
       "epoch": 0.6181865444915599,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4645,
       "step": 38050
     },
     {
       "epoch": 0.6183490113889295,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4242,
       "step": 38060
     },
     {
       "epoch": 0.6185114782862992,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4139,
       "step": 38070
     },
     {
       "epoch": 0.6186739451836688,
-      "grad_norm": 1.75,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.3828,
       "step": 38080
     },
     {
       "epoch": 0.6188364120810385,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4282,
       "step": 38090
     },
     {
       "epoch": 0.6189988789784081,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4579,
       "step": 38100
     },
     {
       "epoch": 0.6191613458757778,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.4381,
       "step": 38110
     },
     {
       "epoch": 0.6193238127731475,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.3809,
       "step": 38120
     },
     {
       "epoch": 0.6194862796705172,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4119,
       "step": 38130
     },
     {
       "epoch": 0.6196487465678868,
-      "grad_norm": 1.8515625,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.4567,
       "step": 38140
     },
     {
       "epoch": 0.6198112134652565,
-      "grad_norm": 1.75,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4251,
       "step": 38150
     },
     {
       "epoch": 0.6199736803626261,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4421,
       "step": 38160
     },
     {
       "epoch": 0.6201361472599958,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.4112,
       "step": 38170
     },
     {
       "epoch": 0.6202986141573654,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4069,
       "step": 38180
     },
     {
       "epoch": 0.6204610810547351,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.4231,
       "step": 38190
     },
     {
       "epoch": 0.6206235479521047,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4609,
       "step": 38200
     },
     {
       "epoch": 0.6207860148494744,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.4149,
       "step": 38210
     },
     {
       "epoch": 0.620948481746844,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3895,
+      "loss": 1.453,
       "step": 38220
     },
     {
       "epoch": 0.6211109486442138,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4611,
       "step": 38230
     },
     {
       "epoch": 0.6212734155415834,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.409,
       "step": 38240
     },
     {
       "epoch": 0.6214358824389531,
-      "grad_norm": 1.828125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.415,
+      "loss": 1.3964,
       "step": 38250
     },
     {
       "epoch": 0.6215983493363227,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4226,
       "step": 38260
     },
     {
       "epoch": 0.6217608162336924,
-      "grad_norm": 2.0625,
+      "grad_norm": 4.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3816,
+      "loss": 1.4662,
       "step": 38270
     },
     {
       "epoch": 0.621923283131062,
-      "grad_norm": 1.7890625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.4627,
       "step": 38280
     },
     {
       "epoch": 0.6220857500284317,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.4065,
       "step": 38290
     },
     {
       "epoch": 0.6222482169258013,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.439,
       "step": 38300
     },
     {
       "epoch": 0.622410683823171,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4286,
       "step": 38310
     },
     {
       "epoch": 0.6225731507205406,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4341,
       "step": 38320
     },
     {
       "epoch": 0.6227356176179103,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.4535,
       "step": 38330
     },
     {
       "epoch": 0.6228980845152801,
-      "grad_norm": 2.4375,
+      "grad_norm": 4.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.4126,
       "step": 38340
     },
     {
       "epoch": 0.6230605514126497,
-      "grad_norm": 2.75,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.4491,
       "step": 38350
     },
     {
       "epoch": 0.6232230183100194,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.44,
       "step": 38360
     },
     {
       "epoch": 0.623385485207389,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.452,
       "step": 38370
     },
     {
       "epoch": 0.6235479521047587,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4283,
       "step": 38380
     },
     {
       "epoch": 0.6237104190021283,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.3736,
       "step": 38390
     },
     {
       "epoch": 0.623872885899498,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4522,
       "step": 38400
     },
     {
       "epoch": 0.6240353527968676,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.3983,
       "step": 38410
     },
     {
       "epoch": 0.6241978196942373,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4159,
+      "loss": 1.4442,
       "step": 38420
     },
     {
       "epoch": 0.6243602865916069,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.387,
+      "loss": 1.4295,
       "step": 38430
     },
     {
       "epoch": 0.6245227534889766,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.389,
       "step": 38440
     },
     {
       "epoch": 0.6246852203863463,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4107,
+      "loss": 1.3847,
       "step": 38450
     },
     {
       "epoch": 0.624847687283716,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.396,
       "step": 38460
     },
     {
       "epoch": 0.6250101541810856,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.4688,
       "step": 38470
     },
     {
       "epoch": 0.6251726210784553,
-      "grad_norm": 1.953125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.443,
       "step": 38480
     },
     {
       "epoch": 0.6253350879758249,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4217,
       "step": 38490
     },
     {
       "epoch": 0.6254975548731946,
-      "grad_norm": 1.765625,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.4239,
       "step": 38500
     },
     {
       "epoch": 0.6256600217705642,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4418,
       "step": 38510
     },
     {
       "epoch": 0.6258224886679339,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4293,
       "step": 38520
     },
     {
       "epoch": 0.6259849555653035,
-      "grad_norm": 2.375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4187,
       "step": 38530
     },
     {
       "epoch": 0.6261474224626732,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.4452,
       "step": 38540
     },
     {
       "epoch": 0.6263098893600428,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.4503,
       "step": 38550
     },
     {
       "epoch": 0.6264723562574126,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4214,
+      "loss": 1.4314,
       "step": 38560
     },
     {
       "epoch": 0.6266348231547823,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.4203,
       "step": 38570
     },
     {
       "epoch": 0.6267972900521519,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4117,
       "step": 38580
     },
     {
       "epoch": 0.6269597569495216,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4555,
       "step": 38590
     },
     {
       "epoch": 0.6271222238468912,
-      "grad_norm": 2.75,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.4276,
       "step": 38600
     },
     {
       "epoch": 0.6272846907442609,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.3899,
       "step": 38610
     },
     {
       "epoch": 0.6274471576416305,
-      "grad_norm": 2.5,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.4606,
       "step": 38620
     },
     {
       "epoch": 0.6276096245390002,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.4251,
       "step": 38630
     },
     {
       "epoch": 0.6277720914363698,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.4211,
       "step": 38640
     },
     {
       "epoch": 0.6279345583337395,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4297,
       "step": 38650
     },
     {
       "epoch": 0.6280970252311091,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.4491,
       "step": 38660
     },
     {
       "epoch": 0.6282594921284789,
-      "grad_norm": 2.5,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.3832,
       "step": 38670
     },
     {
       "epoch": 0.6284219590258485,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.4586,
       "step": 38680
     },
     {
       "epoch": 0.6285844259232182,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.4134,
       "step": 38690
     },
     {
       "epoch": 0.6287468928205878,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4437,
       "step": 38700
     },
     {
       "epoch": 0.6289093597179575,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4255,
+      "loss": 1.42,
       "step": 38710
     },
     {
       "epoch": 0.6290718266153271,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4333,
       "step": 38720
     },
     {
       "epoch": 0.6292342935126968,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3825,
+      "loss": 1.42,
       "step": 38730
     },
     {
       "epoch": 0.6293967604100664,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.4319,
       "step": 38740
     },
     {
       "epoch": 0.6295592273074361,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.399,
       "step": 38750
     },
     {
       "epoch": 0.6297216942048057,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4283,
       "step": 38760
     },
     {
       "epoch": 0.6298841611021754,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.4416,
       "step": 38770
     },
     {
       "epoch": 0.6300466279995451,
-      "grad_norm": 1.8046875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4387,
       "step": 38780
     },
     {
       "epoch": 0.6302090948969148,
-      "grad_norm": 1.7578125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.416,
       "step": 38790
     },
     {
       "epoch": 0.6303715617942844,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.3872,
       "step": 38800
     },
     {
       "epoch": 0.6305340286916541,
-      "grad_norm": 3.84375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.4279,
       "step": 38810
     },
     {
       "epoch": 0.6306964955890237,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4071,
+      "loss": 1.4413,
       "step": 38820
     },
     {
       "epoch": 0.6308589624863934,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4299,
       "step": 38830
     },
     {
       "epoch": 0.631021429383763,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4712,
       "step": 38840
     },
     {
       "epoch": 0.6311838962811327,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.3902,
       "step": 38850
     },
     {
       "epoch": 0.6313463631785023,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.466,
       "step": 38860
     },
     {
       "epoch": 0.631508830075872,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.4484,
       "step": 38870
     },
     {
       "epoch": 0.6316712969732416,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4236,
+      "loss": 1.4204,
       "step": 38880
     },
     {
       "epoch": 0.6318337638706114,
-      "grad_norm": 2.71875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4034,
+      "loss": 1.4281,
       "step": 38890
     },
     {
       "epoch": 0.6319962307679811,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3701,
+      "loss": 1.4292,
       "step": 38900
     },
     {
       "epoch": 0.6321586976653507,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.4342,
       "step": 38910
     },
     {
       "epoch": 0.6323211645627204,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4256,
       "step": 38920
     },
     {
       "epoch": 0.63248363146009,
-      "grad_norm": 1.625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.43,
       "step": 38930
     },
     {
       "epoch": 0.6326460983574597,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.4295,
       "step": 38940
     },
     {
       "epoch": 0.6328085652548293,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.4104,
       "step": 38950
     },
     {
       "epoch": 0.632971032152199,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4002,
+      "loss": 1.4628,
       "step": 38960
     },
     {
       "epoch": 0.6331334990495686,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4256,
       "step": 38970
     },
     {
       "epoch": 0.6332959659469383,
-      "grad_norm": 2.125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.4454,
       "step": 38980
     },
     {
       "epoch": 0.6334584328443079,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.3928,
       "step": 38990
     },
     {
       "epoch": 0.6336208997416777,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4396,
       "step": 39000
     },
     {
       "epoch": 0.6337833666390473,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4137,
+      "loss": 1.3978,
       "step": 39010
     },
     {
       "epoch": 0.633945833536417,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4224,
       "step": 39020
     },
     {
       "epoch": 0.6341083004337866,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.4327,
       "step": 39030
     },
     {
       "epoch": 0.6342707673311563,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4107,
       "step": 39040
     },
     {
       "epoch": 0.6344332342285259,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.3959,
       "step": 39050
     },
     {
       "epoch": 0.6345957011258956,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.404,
       "step": 39060
     },
     {
       "epoch": 0.6347581680232652,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3779,
+      "loss": 1.4512,
       "step": 39070
     },
     {
       "epoch": 0.6349206349206349,
-      "grad_norm": 2.875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4038,
       "step": 39080
     },
     {
       "epoch": 0.6350831018180045,
-      "grad_norm": 2.375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.4615,
       "step": 39090
     },
     {
       "epoch": 0.6352455687153742,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.4223,
       "step": 39100
     },
     {
       "epoch": 0.635408035612744,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.4465,
       "step": 39110
     },
     {
       "epoch": 0.6355705025101136,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4432,
       "step": 39120
     },
     {
       "epoch": 0.6357329694074833,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4194,
       "step": 39130
     },
     {
       "epoch": 0.6358954363048529,
-      "grad_norm": 1.7578125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.3789,
       "step": 39140
     },
     {
       "epoch": 0.6360579032022226,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4047,
       "step": 39150
     },
     {
       "epoch": 0.6362203700995922,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4191,
+      "loss": 1.4435,
       "step": 39160
     },
     {
       "epoch": 0.6363828369969619,
-      "grad_norm": 1.71875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3994,
+      "loss": 1.3764,
       "step": 39170
     },
     {
       "epoch": 0.6365453038943315,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.4223,
       "step": 39180
     },
     {
       "epoch": 0.6367077707917012,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.4339,
       "step": 39190
     },
     {
       "epoch": 0.6368702376890708,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4154,
       "step": 39200
     },
     {
       "epoch": 0.6370327045864406,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4499,
       "step": 39210
     },
     {
       "epoch": 0.6371951714838102,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4295,
+      "loss": 1.4498,
       "step": 39220
     },
     {
       "epoch": 0.6373576383811799,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4148,
+      "loss": 1.4203,
       "step": 39230
     },
     {
       "epoch": 0.6375201052785495,
-      "grad_norm": 1.640625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.419,
+      "loss": 1.4547,
       "step": 39240
     },
     {
       "epoch": 0.6376825721759192,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4359,
       "step": 39250
     },
     {
       "epoch": 0.6378450390732888,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.408,
+      "loss": 1.4897,
       "step": 39260
     },
     {
       "epoch": 0.6380075059706585,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4169,
+      "loss": 1.3967,
       "step": 39270
     },
     {
       "epoch": 0.6381699728680281,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4292,
       "step": 39280
     },
     {
       "epoch": 0.6383324397653978,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.441,
       "step": 39290
     },
     {
       "epoch": 0.6384949066627674,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.409,
       "step": 39300
     },
     {
       "epoch": 0.6386573735601371,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.3922,
       "step": 39310
     },
     {
       "epoch": 0.6388198404575068,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.443,
       "step": 39320
     },
     {
       "epoch": 0.6389823073548765,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.4057,
       "step": 39330
     },
     {
       "epoch": 0.6391447742522461,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.4403,
       "step": 39340
     },
     {
       "epoch": 0.6393072411496158,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3836,
+      "loss": 1.3961,
       "step": 39350
     },
     {
       "epoch": 0.6394697080469854,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.3991,
       "step": 39360
     },
     {
       "epoch": 0.6396321749443551,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.4163,
       "step": 39370
     },
     {
       "epoch": 0.6397946418417247,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4314,
       "step": 39380
     },
     {
       "epoch": 0.6399571087390944,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.4243,
       "step": 39390
     },
     {
       "epoch": 0.640119575636464,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.3761,
       "step": 39400
     },
     {
       "epoch": 0.6402820425338337,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4148,
       "step": 39410
     },
     {
       "epoch": 0.6404445094312033,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4112,
       "step": 39420
     },
     {
       "epoch": 0.6406069763285731,
-      "grad_norm": 3.25,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4045,
       "step": 39430
     },
     {
       "epoch": 0.6407694432259428,
-      "grad_norm": 3.453125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4466,
       "step": 39440
     },
     {
       "epoch": 0.6409319101233124,
-      "grad_norm": 2.34375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3881,
       "step": 39450
     },
     {
       "epoch": 0.6410943770206821,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.3739,
       "step": 39460
     },
     {
       "epoch": 0.6412568439180517,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4243,
+      "loss": 1.4226,
       "step": 39470
     },
     {
       "epoch": 0.6414193108154214,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.3996,
       "step": 39480
     },
     {
       "epoch": 0.641581777712791,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4502,
       "step": 39490
     },
     {
       "epoch": 0.6417442446101607,
-      "grad_norm": 2.21875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3873,
+      "loss": 1.4052,
       "step": 39500
     },
     {
       "epoch": 0.6419067115075303,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4454,
       "step": 39510
     },
     {
       "epoch": 0.6420691784049,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4485,
       "step": 39520
     },
     {
       "epoch": 0.6422316453022696,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.4095,
       "step": 39530
     },
     {
       "epoch": 0.6423941121996394,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4074,
       "step": 39540
     },
     {
       "epoch": 0.642556579097009,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.3824,
       "step": 39550
     },
     {
       "epoch": 0.6427190459943787,
-      "grad_norm": 2.125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4249,
       "step": 39560
     },
     {
       "epoch": 0.6428815128917483,
-      "grad_norm": 1.8203125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4196,
+      "loss": 1.3887,
       "step": 39570
     },
     {
       "epoch": 0.643043979789118,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.3812,
       "step": 39580
     },
     {
       "epoch": 0.6432064466864876,
-      "grad_norm": 3.03125,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3925,
       "step": 39590
     },
     {
       "epoch": 0.6433689135838573,
-      "grad_norm": 2.546875,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4575,
       "step": 39600
     },
     {
       "epoch": 0.6435313804812269,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.4503,
       "step": 39610
     },
     {
       "epoch": 0.6436938473785966,
-      "grad_norm": 2.375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.4265,
       "step": 39620
     },
     {
       "epoch": 0.6438563142759662,
-      "grad_norm": 1.4375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.439,
       "step": 39630
     },
     {
       "epoch": 0.6440187811733359,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.466,
       "step": 39640
     },
     {
       "epoch": 0.6441812480707056,
-      "grad_norm": 4.375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3867,
+      "loss": 1.4167,
       "step": 39650
     },
     {
       "epoch": 0.6443437149680753,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.3976,
       "step": 39660
     },
     {
       "epoch": 0.644506181865445,
-      "grad_norm": 2.421875,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.409,
       "step": 39670
     },
     {
       "epoch": 0.6446686487628146,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.416,
       "step": 39680
     },
     {
       "epoch": 0.6448311156601843,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4309,
+      "loss": 1.3811,
       "step": 39690
     },
     {
       "epoch": 0.6449935825575539,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4608,
       "step": 39700
     },
     {
       "epoch": 0.6451560494549236,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.3654,
       "step": 39710
     },
     {
       "epoch": 0.6453185163522932,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3747,
+      "loss": 1.3642,
       "step": 39720
     },
     {
       "epoch": 0.6454809832496629,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.3729,
       "step": 39730
     },
     {
       "epoch": 0.6456434501470325,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.435,
       "step": 39740
     },
     {
       "epoch": 0.6458059170444022,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4189,
       "step": 39750
     },
     {
       "epoch": 0.6459683839417719,
-      "grad_norm": 2.5,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.414,
       "step": 39760
     },
     {
       "epoch": 0.6461308508391416,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3872,
+      "loss": 1.4268,
       "step": 39770
     },
     {
       "epoch": 0.6462933177365112,
-      "grad_norm": 4.46875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.3688,
       "step": 39780
     },
     {
       "epoch": 0.6464557846338809,
-      "grad_norm": 1.6796875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.4258,
       "step": 39790
     },
     {
       "epoch": 0.6466182515312505,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.3873,
       "step": 39800
     },
     {
       "epoch": 0.6467807184286202,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4085,
       "step": 39810
     },
     {
       "epoch": 0.6469431853259898,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4384,
       "step": 39820
     },
     {
       "epoch": 0.6471056522233595,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.4354,
       "step": 39830
     },
     {
       "epoch": 0.6472681191207291,
-      "grad_norm": 2.0,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3924,
+      "loss": 1.4334,
       "step": 39840
     },
     {
       "epoch": 0.6474305860180988,
-      "grad_norm": 3.0,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.447,
       "step": 39850
     },
     {
       "epoch": 0.6475930529154684,
-      "grad_norm": 1.75,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4254,
       "step": 39860
     },
     {
       "epoch": 0.6477555198128382,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.4313,
       "step": 39870
     },
     {
       "epoch": 0.6479179867102078,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.3835,
       "step": 39880
     },
     {
       "epoch": 0.6480804536075775,
-      "grad_norm": 2.890625,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4029,
+      "loss": 1.3958,
       "step": 39890
     },
     {
       "epoch": 0.6482429205049471,
-      "grad_norm": 2.375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.4348,
       "step": 39900
     },
     {
       "epoch": 0.6484053874023168,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4284,
       "step": 39910
     },
     {
       "epoch": 0.6485678542996864,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4257,
       "step": 39920
     },
     {
       "epoch": 0.6487303211970561,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4298,
+      "loss": 1.4508,
       "step": 39930
     },
     {
       "epoch": 0.6488927880944257,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4779,
       "step": 39940
     },
     {
       "epoch": 0.6490552549917954,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.4355,
       "step": 39950
     },
     {
       "epoch": 0.649217721889165,
-      "grad_norm": 2.0,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.3953,
       "step": 39960
     },
     {
       "epoch": 0.6493801887865347,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4447,
       "step": 39970
     },
     {
       "epoch": 0.6495426556839045,
-      "grad_norm": 1.703125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4388,
       "step": 39980
     },
     {
       "epoch": 0.6497051225812741,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.4204,
       "step": 39990
     },
     {
       "epoch": 0.6498675894786438,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.4024,
       "step": 40000
     },
     {
       "epoch": 0.6500300563760134,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.4477,
       "step": 40010
     },
     {
       "epoch": 0.6501925232733831,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.4089,
       "step": 40020
     },
     {
       "epoch": 0.6503549901707527,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4138,
+      "loss": 1.4555,
       "step": 40030
     },
     {
       "epoch": 0.6505174570681224,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.3764,
       "step": 40040
     },
     {
       "epoch": 0.650679923965492,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.4364,
       "step": 40050
     },
     {
       "epoch": 0.6508423908628617,
-      "grad_norm": 2.25,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.4185,
       "step": 40060
     },
     {
       "epoch": 0.6510048577602313,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3973,
       "step": 40070
     },
     {
       "epoch": 0.651167324657601,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4505,
       "step": 40080
     },
     {
       "epoch": 0.6513297915549707,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3797,
+      "loss": 1.451,
       "step": 40090
     },
     {
       "epoch": 0.6514922584523404,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3824,
+      "loss": 1.4045,
       "step": 40100
     },
     {
       "epoch": 0.65165472534971,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4294,
       "step": 40110
     },
     {
       "epoch": 0.6518171922470797,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.3999,
       "step": 40120
     },
     {
       "epoch": 0.6519796591444493,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4414,
       "step": 40130
     },
     {
       "epoch": 0.652142126041819,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4171,
       "step": 40140
     },
     {
       "epoch": 0.6523045929391886,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4167,
       "step": 40150
     },
     {
       "epoch": 0.6524670598365583,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4195,
       "step": 40160
     },
     {
       "epoch": 0.6526295267339279,
-      "grad_norm": 2.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.4047,
       "step": 40170
     },
     {
       "epoch": 0.6527919936312976,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.4329,
       "step": 40180
     },
     {
       "epoch": 0.6529544605286672,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.372,
+      "loss": 1.4369,
       "step": 40190
     },
     {
       "epoch": 0.653116927426037,
-      "grad_norm": 2.0,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4264,
       "step": 40200
     },
     {
       "epoch": 0.6532793943234066,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4293,
       "step": 40210
     },
     {
       "epoch": 0.6534418612207763,
-      "grad_norm": 2.25,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.4685,
       "step": 40220
     },
     {
       "epoch": 0.653604328118146,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.3952,
       "step": 40230
     },
     {
       "epoch": 0.6537667950155156,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4122,
       "step": 40240
     },
     {
       "epoch": 0.6539292619128853,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.4068,
       "step": 40250
     },
     {
       "epoch": 0.6540917288102549,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3744,
+      "loss": 1.4789,
       "step": 40260
     },
     {
       "epoch": 0.6542541957076246,
-      "grad_norm": 2.75,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4006,
       "step": 40270
     },
     {
       "epoch": 0.6544166626049942,
-      "grad_norm": 2.75,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.3982,
       "step": 40280
     },
     {
       "epoch": 0.6545791295023639,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.4215,
       "step": 40290
     },
     {
       "epoch": 0.6547415963997335,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.3985,
       "step": 40300
     },
     {
       "epoch": 0.6549040632971033,
-      "grad_norm": 2.75,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4183,
+      "loss": 1.4256,
       "step": 40310
     },
     {
       "epoch": 0.6550665301944729,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3801,
+      "loss": 1.3536,
       "step": 40320
     },
     {
       "epoch": 0.6552289970918426,
-      "grad_norm": 1.7890625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4524,
       "step": 40330
     },
     {
       "epoch": 0.6553914639892122,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.4401,
       "step": 40340
     },
     {
       "epoch": 0.6555539308865819,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4879,
       "step": 40350
     },
     {
       "epoch": 0.6557163977839515,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4245,
       "step": 40360
     },
     {
       "epoch": 0.6558788646813212,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4387,
       "step": 40370
     },
     {
       "epoch": 0.6560413315786908,
-      "grad_norm": 1.921875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4253,
       "step": 40380
     },
     {
       "epoch": 0.6562037984760605,
-      "grad_norm": 1.8203125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3711,
+      "loss": 1.4334,
       "step": 40390
     },
     {
       "epoch": 0.6563662653734301,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3768,
+      "loss": 1.4523,
       "step": 40400
     },
     {
       "epoch": 0.6565287322707998,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4236,
       "step": 40410
     },
     {
       "epoch": 0.6566911991681695,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4025,
       "step": 40420
     },
     {
       "epoch": 0.6568536660655392,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.4367,
       "step": 40430
     },
     {
       "epoch": 0.6570161329629088,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.4092,
       "step": 40440
     },
     {
       "epoch": 0.6571785998602785,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.4128,
       "step": 40450
     },
     {
       "epoch": 0.6573410667576481,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4144,
+      "loss": 1.419,
       "step": 40460
     },
     {
       "epoch": 0.6575035336550178,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4186,
       "step": 40470
     },
     {
       "epoch": 0.6576660005523874,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4755,
       "step": 40480
     },
     {
       "epoch": 0.6578284674497571,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4207,
       "step": 40490
     },
     {
       "epoch": 0.6579909343471267,
-      "grad_norm": 5.21875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.3966,
       "step": 40500
     },
     {
       "epoch": 0.6581534012444964,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.437,
       "step": 40510
     },
     {
       "epoch": 0.658315868141866,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.4533,
       "step": 40520
     },
     {
       "epoch": 0.6584783350392358,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4866,
       "step": 40530
     },
     {
       "epoch": 0.6586408019366055,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.4311,
       "step": 40540
     },
     {
       "epoch": 0.6588032688339751,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.3984,
       "step": 40550
     },
     {
       "epoch": 0.6589657357313448,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3788,
+      "loss": 1.4609,
       "step": 40560
     },
     {
       "epoch": 0.6591282026287144,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.4102,
       "step": 40570
     },
     {
       "epoch": 0.6592906695260841,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3829,
+      "loss": 1.4424,
       "step": 40580
     },
     {
       "epoch": 0.6594531364234537,
-      "grad_norm": 3.0,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4269,
       "step": 40590
     },
     {
       "epoch": 0.6596156033208234,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.4355,
       "step": 40600
     },
     {
       "epoch": 0.659778070218193,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4335,
       "step": 40610
     },
     {
       "epoch": 0.6599405371155627,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4019,
       "step": 40620
     },
     {
       "epoch": 0.6601030040129323,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4239,
+      "loss": 1.4366,
       "step": 40630
     },
     {
       "epoch": 0.6602654709103021,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.4123,
       "step": 40640
     },
     {
       "epoch": 0.6604279378076717,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4207,
       "step": 40650
     },
     {
       "epoch": 0.6605904047050414,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.4012,
       "step": 40660
     },
     {
       "epoch": 0.660752871602411,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.3976,
       "step": 40670
     },
     {
       "epoch": 0.6609153384997807,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.4327,
       "step": 40680
     },
     {
       "epoch": 0.6610778053971503,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.415,
       "step": 40690
     },
     {
       "epoch": 0.66124027229452,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4135,
+      "loss": 1.4162,
       "step": 40700
     },
     {
       "epoch": 0.6614027391918896,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4203,
       "step": 40710
     },
     {
       "epoch": 0.6615652060892593,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.37,
       "step": 40720
     },
     {
       "epoch": 0.6617276729866289,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4393,
       "step": 40730
     },
     {
       "epoch": 0.6618901398839986,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.4159,
       "step": 40740
     },
     {
       "epoch": 0.6620526067813683,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4028,
       "step": 40750
     },
     {
       "epoch": 0.662215073678738,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.4074,
       "step": 40760
     },
     {
       "epoch": 0.6623775405761076,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.4167,
       "step": 40770
     },
     {
       "epoch": 0.6625400074734773,
-      "grad_norm": 2.5,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4478,
       "step": 40780
     },
     {
       "epoch": 0.662702474370847,
-      "grad_norm": 1.8515625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.4421,
       "step": 40790
     },
     {
       "epoch": 0.6628649412682166,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.3748,
       "step": 40800
     },
     {
       "epoch": 0.6630274081655863,
-      "grad_norm": 1.7734375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.3728,
       "step": 40810
     },
     {
       "epoch": 0.6631898750629559,
-      "grad_norm": 1.828125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.4288,
+      "loss": 1.4297,
       "step": 40820
     },
     {
       "epoch": 0.6633523419603256,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3797,
+      "loss": 1.4059,
       "step": 40830
     },
     {
       "epoch": 0.6635148088576952,
-      "grad_norm": 2.0,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4445,
       "step": 40840
     },
     {
       "epoch": 0.6636772757550649,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3803,
+      "loss": 1.4426,
       "step": 40850
     },
     {
       "epoch": 0.6638397426524346,
-      "grad_norm": 1.921875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.431,
       "step": 40860
     },
     {
       "epoch": 0.6640022095498043,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4391,
       "step": 40870
     },
     {
       "epoch": 0.6641646764471739,
-      "grad_norm": 1.7578125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4195,
       "step": 40880
     },
     {
       "epoch": 0.6643271433445436,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4203,
+      "loss": 1.3851,
       "step": 40890
     },
     {
       "epoch": 0.6644896102419132,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.4181,
       "step": 40900
     },
     {
       "epoch": 0.6646520771392829,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4085,
       "step": 40910
     },
     {
       "epoch": 0.6648145440366525,
-      "grad_norm": 1.8203125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4032,
       "step": 40920
     },
     {
       "epoch": 0.6649770109340222,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.4147,
       "step": 40930
     },
     {
       "epoch": 0.6651394778313918,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.5207,
       "step": 40940
     },
     {
       "epoch": 0.6653019447287615,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4519,
       "step": 40950
     },
     {
       "epoch": 0.6654644116261311,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4201,
       "step": 40960
     },
     {
       "epoch": 0.6656268785235009,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.385,
+      "loss": 1.4645,
       "step": 40970
     },
     {
       "epoch": 0.6657893454208705,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4196,
       "step": 40980
     },
     {
       "epoch": 0.6659518123182402,
-      "grad_norm": 2.3125,
+      "grad_norm": 4.71875,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.4292,
       "step": 40990
     },
     {
       "epoch": 0.6661142792156098,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.4138,
       "step": 41000
     },
     {
       "epoch": 0.6662767461129795,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4001,
       "step": 41010
     },
     {
       "epoch": 0.6664392130103491,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.396,
+      "loss": 1.4449,
       "step": 41020
     },
     {
       "epoch": 0.6666016799077188,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.43,
+      "loss": 1.478,
       "step": 41030
     },
     {
       "epoch": 0.6667641468050884,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3764,
+      "loss": 1.411,
       "step": 41040
     },
     {
       "epoch": 0.6669266137024581,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.4169,
       "step": 41050
     },
     {
       "epoch": 0.6670890805998277,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3688,
       "step": 41060
     },
     {
       "epoch": 0.6672515474971974,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3798,
+      "loss": 1.4077,
       "step": 41070
     },
     {
       "epoch": 0.6674140143945672,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3836,
+      "loss": 1.4101,
       "step": 41080
     },
     {
       "epoch": 0.6675764812919368,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4346,
       "step": 41090
     },
     {
       "epoch": 0.6677389481893065,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3714,
+      "loss": 1.4187,
       "step": 41100
     },
     {
       "epoch": 0.6679014150866761,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.4591,
       "step": 41110
     },
     {
       "epoch": 0.6680638819840458,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4333,
+      "loss": 1.4567,
       "step": 41120
     },
     {
       "epoch": 0.6682263488814154,
-      "grad_norm": 3.046875,
+      "grad_norm": 4.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3823,
+      "loss": 1.3676,
       "step": 41130
     },
     {
       "epoch": 0.6683888157787851,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.4219,
       "step": 41140
     },
     {
       "epoch": 0.6685512826761547,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4259,
       "step": 41150
     },
     {
       "epoch": 0.6687137495735244,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.3811,
       "step": 41160
     },
     {
       "epoch": 0.668876216470894,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.4089,
       "step": 41170
     },
     {
       "epoch": 0.6690386833682637,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4023,
       "step": 41180
     },
     {
       "epoch": 0.6692011502656334,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.3892,
       "step": 41190
     },
     {
       "epoch": 0.6693636171630031,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3815,
+      "loss": 1.3968,
       "step": 41200
     },
     {
       "epoch": 0.6695260840603727,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3758,
+      "loss": 1.4605,
       "step": 41210
     },
     {
       "epoch": 0.6696885509577424,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4394,
       "step": 41220
     },
     {
       "epoch": 0.669851017855112,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3734,
+      "loss": 1.4156,
       "step": 41230
     },
     {
       "epoch": 0.6700134847524817,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4653,
       "step": 41240
     },
     {
       "epoch": 0.6701759516498513,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.4578,
       "step": 41250
     },
     {
       "epoch": 0.670338418547221,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.4291,
       "step": 41260
     },
     {
       "epoch": 0.6705008854445906,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.432,
       "step": 41270
     },
     {
       "epoch": 0.6706633523419603,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.4158,
       "step": 41280
     },
     {
       "epoch": 0.6708258192393299,
-      "grad_norm": 1.8671875,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.4361,
       "step": 41290
     },
     {
       "epoch": 0.6709882861366997,
-      "grad_norm": 2.0,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3757,
+      "loss": 1.4029,
       "step": 41300
     },
     {
       "epoch": 0.6711507530340693,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.3844,
       "step": 41310
     },
     {
       "epoch": 0.671313219931439,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3705,
+      "loss": 1.3977,
       "step": 41320
     },
     {
       "epoch": 0.6714756868288086,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4167,
       "step": 41330
     },
     {
       "epoch": 0.6716381537261783,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4009,
       "step": 41340
     },
     {
       "epoch": 0.671800620623548,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3823,
+      "loss": 1.3734,
       "step": 41350
     },
     {
       "epoch": 0.6719630875209176,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3767,
+      "loss": 1.4548,
       "step": 41360
     },
     {
       "epoch": 0.6721255544182873,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3869,
+      "loss": 1.4117,
       "step": 41370
     },
     {
       "epoch": 0.6722880213156569,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4157,
       "step": 41380
     },
     {
       "epoch": 0.6724504882130266,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.3692,
       "step": 41390
     },
     {
       "epoch": 0.6726129551103962,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4402,
       "step": 41400
     },
     {
       "epoch": 0.672775422007766,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4252,
       "step": 41410
     },
     {
       "epoch": 0.6729378889051356,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4128,
+      "loss": 1.3831,
       "step": 41420
     },
     {
       "epoch": 0.6731003558025053,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4394,
       "step": 41430
     },
     {
       "epoch": 0.6732628226998749,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4082,
       "step": 41440
     },
     {
       "epoch": 0.6734252895972446,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.451,
       "step": 41450
     },
     {
       "epoch": 0.6735877564946142,
-      "grad_norm": 3.25,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4268,
+      "loss": 1.4346,
       "step": 41460
     },
     {
       "epoch": 0.6737502233919839,
-      "grad_norm": 1.71875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3716,
+      "loss": 1.429,
       "step": 41470
     },
     {
       "epoch": 0.6739126902893535,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.405,
       "step": 41480
     },
     {
       "epoch": 0.6740751571867232,
-      "grad_norm": 1.765625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.4748,
       "step": 41490
     },
     {
       "epoch": 0.6742376240840928,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.4257,
       "step": 41500
     },
     {
       "epoch": 0.6744000909814625,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.4311,
       "step": 41510
     },
     {
       "epoch": 0.6745625578788322,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4329,
+      "loss": 1.3944,
       "step": 41520
     },
     {
       "epoch": 0.6747250247762019,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4252,
       "step": 41530
     },
     {
       "epoch": 0.6748874916735715,
-      "grad_norm": 1.328125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.3844,
       "step": 41540
     },
     {
       "epoch": 0.6750499585709412,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4405,
       "step": 41550
     },
     {
       "epoch": 0.6752124254683108,
-      "grad_norm": 2.875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4201,
       "step": 41560
     },
     {
       "epoch": 0.6753748923656805,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4535,
       "step": 41570
     },
     {
       "epoch": 0.6755373592630501,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.4423,
       "step": 41580
     },
     {
       "epoch": 0.6756998261604198,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4536,
       "step": 41590
     },
     {
       "epoch": 0.6758622930577894,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4382,
       "step": 41600
     },
     {
       "epoch": 0.6760247599551591,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4152,
       "step": 41610
     },
     {
       "epoch": 0.6761872268525287,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.4688,
       "step": 41620
     },
     {
       "epoch": 0.6763496937498985,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4131,
       "step": 41630
     },
     {
       "epoch": 0.6765121606472682,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.4009,
       "step": 41640
     },
     {
       "epoch": 0.6766746275446378,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4171,
       "step": 41650
     },
     {
       "epoch": 0.6768370944420075,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3811,
+      "loss": 1.4682,
       "step": 41660
     },
     {
       "epoch": 0.6769995613393771,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3782,
+      "loss": 1.3866,
       "step": 41670
     },
     {
       "epoch": 0.6771620282367468,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.424,
       "step": 41680
     },
     {
       "epoch": 0.6773244951341164,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.409,
       "step": 41690
     },
     {
       "epoch": 0.6774869620314861,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.4298,
       "step": 41700
     },
     {
       "epoch": 0.6776494289288557,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3908,
+      "loss": 1.3885,
       "step": 41710
     },
     {
       "epoch": 0.6778118958262254,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4042,
       "step": 41720
     },
     {
       "epoch": 0.677974362723595,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3846,
       "step": 41730
     },
     {
       "epoch": 0.6781368296209648,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3812,
+      "loss": 1.4065,
       "step": 41740
     },
     {
       "epoch": 0.6782992965183344,
-      "grad_norm": 5.90625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4118,
       "step": 41750
     },
     {
       "epoch": 0.6784617634157041,
-      "grad_norm": 2.0,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.4179,
       "step": 41760
     },
     {
       "epoch": 0.6786242303130737,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4183,
       "step": 41770
     },
     {
       "epoch": 0.6787866972104434,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4077,
       "step": 41780
     },
     {
       "epoch": 0.678949164107813,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.4342,
       "step": 41790
     },
     {
       "epoch": 0.6791116310051827,
-      "grad_norm": 2.75,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4336,
       "step": 41800
     },
     {
       "epoch": 0.6792740979025523,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.3981,
       "step": 41810
     },
     {
       "epoch": 0.679436564799922,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4123,
+      "loss": 1.4836,
       "step": 41820
     },
     {
       "epoch": 0.6795990316972916,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.4194,
       "step": 41830
     },
     {
       "epoch": 0.6797614985946613,
-      "grad_norm": 2.625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3908,
+      "loss": 1.4111,
       "step": 41840
     },
     {
       "epoch": 0.679923965492031,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3812,
+      "loss": 1.4221,
       "step": 41850
     },
     {
       "epoch": 0.6800864323894007,
-      "grad_norm": 2.265625,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.3753,
+      "loss": 1.4021,
       "step": 41860
     },
     {
       "epoch": 0.6802488992867703,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.4165,
       "step": 41870
     },
     {
       "epoch": 0.68041136618414,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3742,
+      "loss": 1.4665,
       "step": 41880
     },
     {
       "epoch": 0.6805738330815096,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.3922,
       "step": 41890
     },
     {
       "epoch": 0.6807362999788793,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.425,
+      "loss": 1.4635,
       "step": 41900
     },
     {
       "epoch": 0.680898766876249,
-      "grad_norm": 1.9765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.418,
       "step": 41910
     },
     {
       "epoch": 0.6810612337736186,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.4426,
       "step": 41920
     },
     {
       "epoch": 0.6812237006709883,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.361,
       "step": 41930
     },
     {
       "epoch": 0.6813861675683579,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4177,
+      "loss": 1.4374,
       "step": 41940
     },
     {
       "epoch": 0.6815486344657276,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4127,
       "step": 41950
     },
     {
       "epoch": 0.6817111013630973,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.4444,
       "step": 41960
     },
     {
       "epoch": 0.681873568260467,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3794,
+      "loss": 1.4001,
       "step": 41970
     },
     {
       "epoch": 0.6820360351578366,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4195,
+      "loss": 1.408,
       "step": 41980
     },
     {
       "epoch": 0.6821985020552063,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4355,
       "step": 41990
     },
     {
       "epoch": 0.6823609689525759,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4509,
       "step": 42000
     },
     {
       "epoch": 0.6825234358499456,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4266,
       "step": 42010
     },
     {
       "epoch": 0.6826859027473152,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.4189,
       "step": 42020
     },
     {
       "epoch": 0.6828483696446849,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4213,
       "step": 42030
     },
     {
       "epoch": 0.6830108365420545,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.4042,
       "step": 42040
     },
     {
       "epoch": 0.6831733034394242,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.415,
       "step": 42050
     },
     {
       "epoch": 0.6833357703367938,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4079,
       "step": 42060
     },
     {
       "epoch": 0.6834982372341636,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.4104,
       "step": 42070
     },
     {
       "epoch": 0.6836607041315332,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4347,
       "step": 42080
     },
     {
       "epoch": 0.6838231710289029,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.411,
       "step": 42090
     },
     {
       "epoch": 0.6839856379262725,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.4063,
       "step": 42100
     },
     {
       "epoch": 0.6841481048236422,
-      "grad_norm": 2.875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.4143,
       "step": 42110
     },
     {
       "epoch": 0.6843105717210118,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4003,
       "step": 42120
     },
     {
       "epoch": 0.6844730386183815,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4541,
       "step": 42130
     },
     {
       "epoch": 0.6846355055157511,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.3663,
       "step": 42140
     },
     {
       "epoch": 0.6847979724131208,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.4092,
+      "loss": 1.3597,
       "step": 42150
     },
     {
       "epoch": 0.6849604393104904,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.4169,
       "step": 42160
     },
     {
       "epoch": 0.6851229062078601,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.4117,
       "step": 42170
     },
     {
       "epoch": 0.6852853731052299,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.4549,
       "step": 42180
     },
     {
       "epoch": 0.6854478400025995,
-      "grad_norm": 2.0,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4249,
       "step": 42190
     },
     {
       "epoch": 0.6856103068999692,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.376,
+      "loss": 1.4484,
       "step": 42200
     },
     {
       "epoch": 0.6857727737973388,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4036,
       "step": 42210
     },
     {
       "epoch": 0.6859352406947085,
-      "grad_norm": 1.6640625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.4238,
       "step": 42220
     },
     {
       "epoch": 0.6860977075920781,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.4508,
       "step": 42230
     },
     {
       "epoch": 0.6862601744894478,
-      "grad_norm": 2.5,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4249,
+      "loss": 1.4207,
       "step": 42240
     },
     {
       "epoch": 0.6864226413868174,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4399,
       "step": 42250
     },
     {
       "epoch": 0.6865851082841871,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4538,
       "step": 42260
     },
     {
       "epoch": 0.6867475751815567,
-      "grad_norm": 2.625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.396,
+      "loss": 1.3881,
       "step": 42270
     },
     {
       "epoch": 0.6869100420789264,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.426,
       "step": 42280
     },
     {
       "epoch": 0.6870725089762961,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3767,
+      "loss": 1.381,
       "step": 42290
     },
     {
       "epoch": 0.6872349758736658,
-      "grad_norm": 1.546875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4216,
       "step": 42300
     },
     {
       "epoch": 0.6873974427710354,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.3909,
       "step": 42310
     },
     {
       "epoch": 0.6875599096684051,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4219,
       "step": 42320
     },
     {
       "epoch": 0.6877223765657747,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3792,
+      "loss": 1.4232,
       "step": 42330
     },
     {
       "epoch": 0.6878848434631444,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.418,
       "step": 42340
     },
     {
       "epoch": 0.688047310360514,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.4143,
       "step": 42350
     },
     {
       "epoch": 0.6882097772578837,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.3825,
       "step": 42360
     },
     {
       "epoch": 0.6883722441552533,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4292,
       "step": 42370
     },
     {
       "epoch": 0.688534711052623,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.427,
+      "loss": 1.3887,
       "step": 42380
     },
     {
       "epoch": 0.6886971779499926,
-      "grad_norm": 2.125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4324,
       "step": 42390
     },
     {
       "epoch": 0.6888596448473624,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3667,
+      "loss": 1.4475,
       "step": 42400
     },
     {
       "epoch": 0.689022111744732,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4235,
       "step": 42410
     },
     {
       "epoch": 0.6891845786421017,
-      "grad_norm": 3.734375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4265,
+      "loss": 1.3938,
       "step": 42420
     },
     {
       "epoch": 0.6893470455394713,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3924,
+      "loss": 1.4319,
       "step": 42430
     },
     {
       "epoch": 0.689509512436841,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4144,
       "step": 42440
     },
     {
       "epoch": 0.6896719793342106,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4156,
       "step": 42450
     },
     {
       "epoch": 0.6898344462315803,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.4251,
       "step": 42460
     },
     {
       "epoch": 0.68999691312895,
-      "grad_norm": 1.859375,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4338,
       "step": 42470
     },
     {
       "epoch": 0.6901593800263196,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.4177,
       "step": 42480
     },
     {
       "epoch": 0.6903218469236893,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4222,
+      "loss": 1.3964,
       "step": 42490
     },
     {
       "epoch": 0.6904843138210589,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3842,
+      "loss": 1.4185,
       "step": 42500
     },
     {
       "epoch": 0.6906467807184287,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4172,
       "step": 42510
     },
     {
       "epoch": 0.6908092476157983,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.4497,
       "step": 42520
     },
     {
       "epoch": 0.690971714513168,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.385,
+      "loss": 1.4413,
       "step": 42530
     },
     {
       "epoch": 0.6911341814105376,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.4259,
       "step": 42540
     },
     {
       "epoch": 0.6912966483079073,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4069,
       "step": 42550
     },
     {
       "epoch": 0.6914591152052769,
-      "grad_norm": 1.7265625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.4282,
       "step": 42560
     },
     {
       "epoch": 0.6916215821026466,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3645,
+      "loss": 1.4023,
       "step": 42570
     },
     {
       "epoch": 0.6917840490000162,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4213,
       "step": 42580
     },
     {
       "epoch": 0.6919465158973859,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.3885,
       "step": 42590
     },
     {
       "epoch": 0.6921089827947555,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.3883,
       "step": 42600
     },
     {
       "epoch": 0.6922714496921253,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4411,
       "step": 42610
     },
     {
       "epoch": 0.6924339165894949,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.398,
       "step": 42620
     },
     {
       "epoch": 0.6925963834868646,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.3568,
       "step": 42630
     },
     {
       "epoch": 0.6927588503842342,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.3706,
       "step": 42640
     },
     {
       "epoch": 0.6929213172816039,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4038,
+      "loss": 1.3801,
       "step": 42650
     },
     {
       "epoch": 0.6930837841789735,
-      "grad_norm": 2.5,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.4365,
       "step": 42660
     },
     {
       "epoch": 0.6932462510763432,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4654,
       "step": 42670
     },
     {
       "epoch": 0.6934087179737128,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.3915,
       "step": 42680
     },
     {
       "epoch": 0.6935711848710825,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3801,
+      "loss": 1.4162,
       "step": 42690
     },
     {
       "epoch": 0.6937336517684521,
-      "grad_norm": 1.8984375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.3823,
       "step": 42700
     },
     {
       "epoch": 0.6938961186658218,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4413,
       "step": 42710
     },
     {
       "epoch": 0.6940585855631916,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4457,
       "step": 42720
     },
     {
       "epoch": 0.6942210524605612,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.3752,
       "step": 42730
     },
     {
       "epoch": 0.6943835193579309,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.3968,
       "step": 42740
     },
     {
       "epoch": 0.6945459862553005,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.3829,
       "step": 42750
     },
     {
       "epoch": 0.6947084531526702,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4135,
       "step": 42760
     },
     {
       "epoch": 0.6948709200500398,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3873,
+      "loss": 1.4111,
       "step": 42770
     },
     {
       "epoch": 0.6950333869474095,
-      "grad_norm": 1.6875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4147,
       "step": 42780
     },
     {
       "epoch": 0.6951958538447791,
-      "grad_norm": 2.0,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4256,
       "step": 42790
     },
     {
       "epoch": 0.6953583207421488,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.4274,
       "step": 42800
     },
     {
       "epoch": 0.6955207876395184,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.3745,
       "step": 42810
     },
     {
       "epoch": 0.6956832545368881,
-      "grad_norm": 2.078125,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.4263,
       "step": 42820
     },
     {
       "epoch": 0.6958457214342578,
-      "grad_norm": 2.5,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.3716,
       "step": 42830
     },
     {
       "epoch": 0.6960081883316275,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3782,
+      "loss": 1.417,
       "step": 42840
     },
     {
       "epoch": 0.6961706552289971,
-      "grad_norm": 2.75,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4292,
+      "loss": 1.3935,
       "step": 42850
     },
     {
       "epoch": 0.6963331221263668,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4043,
       "step": 42860
     },
     {
       "epoch": 0.6964955890237364,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3829,
+      "loss": 1.4409,
       "step": 42870
     },
     {
       "epoch": 0.6966580559211061,
-      "grad_norm": 2.9375,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3726,
+      "loss": 1.3706,
       "step": 42880
     },
     {
       "epoch": 0.6968205228184757,
-      "grad_norm": 1.5625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3867,
+      "loss": 1.4036,
       "step": 42890
     },
     {
       "epoch": 0.6969829897158454,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.4587,
       "step": 42900
     },
     {
       "epoch": 0.697145456613215,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4113,
+      "loss": 1.4109,
       "step": 42910
     },
     {
       "epoch": 0.6973079235105847,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.396,
       "step": 42920
     },
     {
       "epoch": 0.6974703904079543,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.397,
       "step": 42930
     },
     {
       "epoch": 0.6976328573053241,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4505,
       "step": 42940
     },
     {
       "epoch": 0.6977953242026937,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.4299,
       "step": 42950
     },
     {
       "epoch": 0.6979577911000634,
-      "grad_norm": 2.859375,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.4367,
       "step": 42960
     },
     {
       "epoch": 0.698120257997433,
-      "grad_norm": 2.625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4163,
       "step": 42970
     },
     {
       "epoch": 0.6982827248948027,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.4186,
       "step": 42980
     },
     {
       "epoch": 0.6984451917921723,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4156,
       "step": 42990
     },
     {
       "epoch": 0.698607658689542,
-      "grad_norm": 1.7265625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.4148,
       "step": 43000
     },
     {
       "epoch": 0.6987701255869117,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.4253,
       "step": 43010
     },
     {
       "epoch": 0.6989325924842813,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4284,
+      "loss": 1.3987,
       "step": 43020
     },
     {
       "epoch": 0.699095059381651,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4316,
+      "loss": 1.4319,
       "step": 43030
     },
     {
       "epoch": 0.6992575262790206,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.3948,
       "step": 43040
     },
     {
       "epoch": 0.6994199931763904,
-      "grad_norm": 1.8515625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4203,
       "step": 43050
     },
     {
       "epoch": 0.69958246007376,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4371,
       "step": 43060
     },
     {
       "epoch": 0.6997449269711297,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.402,
+      "loss": 1.4275,
       "step": 43070
     },
     {
       "epoch": 0.6999073938684993,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.4015,
       "step": 43080
     },
     {
       "epoch": 0.700069860765869,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4045,
       "step": 43090
     },
     {
       "epoch": 0.7002323276632386,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4179,
       "step": 43100
     },
     {
       "epoch": 0.7003947945606083,
-      "grad_norm": 3.859375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.3885,
       "step": 43110
     },
     {
       "epoch": 0.7005572614579779,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4505,
       "step": 43120
     },
     {
       "epoch": 0.7007197283553476,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.3517,
       "step": 43130
     },
     {
       "epoch": 0.7008821952527172,
-      "grad_norm": 2.78125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.422,
       "step": 43140
     },
     {
       "epoch": 0.7010446621500869,
-      "grad_norm": 2.640625,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.4017,
       "step": 43150
     },
     {
       "epoch": 0.7012071290474566,
-      "grad_norm": 3.875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4304,
       "step": 43160
     },
     {
       "epoch": 0.7013695959448263,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4216,
       "step": 43170
     },
     {
       "epoch": 0.7015320628421959,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.4508,
       "step": 43180
     },
     {
       "epoch": 0.7016945297395656,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.4624,
       "step": 43190
     },
     {
       "epoch": 0.7018569966369352,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.4436,
       "step": 43200
     },
     {
       "epoch": 0.7020194635343049,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3599,
+      "loss": 1.404,
       "step": 43210
     },
     {
       "epoch": 0.7021819304316745,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.4483,
       "step": 43220
     },
     {
       "epoch": 0.7023443973290442,
-      "grad_norm": 2.4375,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4105,
       "step": 43230
     },
     {
       "epoch": 0.7025068642264138,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.4051,
       "step": 43240
     },
     {
       "epoch": 0.7026693311237835,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3781,
       "step": 43250
     },
     {
       "epoch": 0.7028317980211531,
-      "grad_norm": 2.875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.427,
       "step": 43260
     },
     {
       "epoch": 0.7029942649185229,
-      "grad_norm": 1.8046875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3858,
+      "loss": 1.4736,
       "step": 43270
     },
     {
       "epoch": 0.7031567318158926,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4285,
+      "loss": 1.4264,
       "step": 43280
     },
     {
       "epoch": 0.7033191987132622,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4069,
       "step": 43290
     },
     {
       "epoch": 0.7034816656106319,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4209,
+      "loss": 1.3975,
       "step": 43300
     },
     {
       "epoch": 0.7036441325080015,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.3658,
       "step": 43310
     },
     {
       "epoch": 0.7038065994053712,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4175,
       "step": 43320
     },
     {
       "epoch": 0.7039690663027408,
-      "grad_norm": 2.625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4132,
+      "loss": 1.4295,
       "step": 43330
     },
     {
       "epoch": 0.7041315332001105,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.44,
       "step": 43340
     },
     {
       "epoch": 0.7042940000974801,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.4251,
       "step": 43350
     },
     {
       "epoch": 0.7044564669948498,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.4409,
       "step": 43360
     },
     {
       "epoch": 0.7046189338922194,
-      "grad_norm": 1.6953125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3732,
+      "loss": 1.4394,
       "step": 43370
     },
     {
       "epoch": 0.7047814007895892,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3804,
+      "loss": 1.397,
       "step": 43380
     },
     {
       "epoch": 0.7049438676869588,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4172,
+      "loss": 1.3453,
       "step": 43390
     },
     {
       "epoch": 0.7051063345843285,
-      "grad_norm": 3.25,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4092,
+      "loss": 1.3933,
       "step": 43400
     },
     {
       "epoch": 0.7052688014816981,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.3723,
       "step": 43410
     },
     {
       "epoch": 0.7054312683790678,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.4651,
       "step": 43420
     },
     {
       "epoch": 0.7055937352764374,
-      "grad_norm": 2.125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4185,
       "step": 43430
     },
     {
       "epoch": 0.7057562021738071,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4233,
       "step": 43440
     },
     {
       "epoch": 0.7059186690711767,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4301,
       "step": 43450
     },
     {
       "epoch": 0.7060811359685464,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.3945,
       "step": 43460
     },
     {
       "epoch": 0.706243602865916,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.4574,
       "step": 43470
     },
     {
       "epoch": 0.7064060697632857,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4386,
       "step": 43480
     },
     {
       "epoch": 0.7065685366606554,
-      "grad_norm": 3.75,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3784,
+      "loss": 1.4153,
       "step": 43490
     },
     {
       "epoch": 0.7067310035580251,
-      "grad_norm": 3.421875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.422,
+      "loss": 1.432,
       "step": 43500
     },
     {
       "epoch": 0.7068934704553947,
-      "grad_norm": 3.484375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3727,
+      "loss": 1.4283,
       "step": 43510
     },
     {
       "epoch": 0.7070559373527644,
-      "grad_norm": 1.8515625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.4289,
       "step": 43520
     },
     {
       "epoch": 0.707218404250134,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.3898,
       "step": 43530
     },
     {
       "epoch": 0.7073808711475037,
-      "grad_norm": 1.640625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.4558,
       "step": 43540
     },
     {
       "epoch": 0.7075433380448733,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3626,
       "step": 43550
     },
     {
       "epoch": 0.707705804942243,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.372,
+      "loss": 1.4466,
       "step": 43560
     },
     {
       "epoch": 0.7078682718396127,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3797,
+      "loss": 1.4595,
       "step": 43570
     },
     {
       "epoch": 0.7080307387369823,
-      "grad_norm": 2.5,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.3958,
       "step": 43580
     },
     {
       "epoch": 0.708193205634352,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4289,
       "step": 43590
     },
     {
       "epoch": 0.7083556725317217,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.4402,
       "step": 43600
     },
     {
       "epoch": 0.7085181394290914,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.3613,
       "step": 43610
     },
     {
       "epoch": 0.708680606326461,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3703,
+      "loss": 1.3881,
       "step": 43620
     },
     {
       "epoch": 0.7088430732238307,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.4048,
       "step": 43630
     },
     {
       "epoch": 0.7090055401212003,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.376,
       "step": 43640
     },
     {
       "epoch": 0.70916800701857,
-      "grad_norm": 1.765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.402,
       "step": 43650
     },
     {
       "epoch": 0.7093304739159396,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.382,
+      "loss": 1.4314,
       "step": 43660
     },
     {
       "epoch": 0.7094929408133093,
-      "grad_norm": 3.25,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.4256,
       "step": 43670
     },
     {
       "epoch": 0.7096554077106789,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3771,
+      "loss": 1.393,
       "step": 43680
     },
     {
       "epoch": 0.7098178746080486,
-      "grad_norm": 2.25,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4166,
+      "loss": 1.3435,
       "step": 43690
     },
     {
       "epoch": 0.7099803415054182,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.404,
       "step": 43700
     },
     {
       "epoch": 0.710142808402788,
-      "grad_norm": 1.890625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.4128,
       "step": 43710
     },
     {
       "epoch": 0.7103052753001576,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4071,
       "step": 43720
     },
     {
       "epoch": 0.7104677421975273,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.4192,
       "step": 43730
     },
     {
       "epoch": 0.7106302090948969,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4067,
       "step": 43740
     },
     {
       "epoch": 0.7107926759922666,
-      "grad_norm": 2.0,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3758,
+      "loss": 1.4745,
       "step": 43750
     },
     {
       "epoch": 0.7109551428896362,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4005,
       "step": 43760
     },
     {
       "epoch": 0.7111176097870059,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4176,
       "step": 43770
     },
     {
       "epoch": 0.7112800766843755,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3872,
+      "loss": 1.4051,
       "step": 43780
     },
     {
       "epoch": 0.7114425435817452,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.4241,
       "step": 43790
     },
     {
       "epoch": 0.7116050104791148,
-      "grad_norm": 2.015625,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4197,
+      "loss": 1.4318,
       "step": 43800
     },
     {
       "epoch": 0.7117674773764845,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4255,
+      "loss": 1.4209,
       "step": 43810
     },
     {
       "epoch": 0.7119299442738543,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4055,
+      "loss": 1.3838,
       "step": 43820
     },
     {
       "epoch": 0.7120924111712239,
-      "grad_norm": 2.125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.374,
       "step": 43830
     },
     {
       "epoch": 0.7122548780685936,
-      "grad_norm": 3.765625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4303,
       "step": 43840
     },
     {
       "epoch": 0.7124173449659632,
-      "grad_norm": 2.125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4346,
       "step": 43850
     },
     {
       "epoch": 0.7125798118633329,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.4273,
       "step": 43860
     },
     {
       "epoch": 0.7127422787607025,
-      "grad_norm": 2.375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.4039,
       "step": 43870
     },
     {
       "epoch": 0.7129047456580722,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4051,
+      "loss": 1.3863,
       "step": 43880
     },
     {
       "epoch": 0.7130672125554418,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4075,
       "step": 43890
     },
     {
       "epoch": 0.7132296794528115,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4539,
       "step": 43900
     },
     {
       "epoch": 0.7133921463501811,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.3864,
       "step": 43910
     },
     {
       "epoch": 0.7135546132475508,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3541,
       "step": 43920
     },
     {
       "epoch": 0.7137170801449205,
-      "grad_norm": 3.125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.4216,
       "step": 43930
     },
     {
       "epoch": 0.7138795470422902,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.422,
       "step": 43940
     },
     {
       "epoch": 0.7140420139396598,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4147,
       "step": 43950
     },
     {
       "epoch": 0.7142044808370295,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4252,
+      "loss": 1.4515,
       "step": 43960
     },
     {
       "epoch": 0.7143669477343991,
-      "grad_norm": 2.03125,
+      "grad_norm": 12.875,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.3487,
       "step": 43970
     },
     {
       "epoch": 0.7145294146317688,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3738,
+      "loss": 1.4359,
       "step": 43980
     },
     {
       "epoch": 0.7146918815291384,
-      "grad_norm": 2.125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4181,
+      "loss": 1.4155,
       "step": 43990
     },
     {
       "epoch": 0.7148543484265081,
-      "grad_norm": 2.5,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4319,
       "step": 44000
     },
     {
       "epoch": 0.7150168153238777,
-      "grad_norm": 2.3125,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4023,
       "step": 44010
     },
     {
       "epoch": 0.7151792822212474,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.3927,
       "step": 44020
     },
     {
       "epoch": 0.715341749118617,
-      "grad_norm": 2.25,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.3894,
       "step": 44030
     },
     {
       "epoch": 0.7155042160159868,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.3725,
       "step": 44040
     },
     {
       "epoch": 0.7156666829133564,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4397,
       "step": 44050
     },
     {
       "epoch": 0.7158291498107261,
-      "grad_norm": 3.59375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.3988,
       "step": 44060
     },
     {
       "epoch": 0.7159916167080957,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.42,
       "step": 44070
     },
     {
       "epoch": 0.7161540836054654,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.407,
+      "loss": 1.393,
       "step": 44080
     },
     {
       "epoch": 0.716316550502835,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.3998,
       "step": 44090
     },
     {
       "epoch": 0.7164790174002047,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.3852,
       "step": 44100
     },
     {
       "epoch": 0.7166414842975743,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4481,
       "step": 44110
     },
     {
       "epoch": 0.716803951194944,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4256,
       "step": 44120
     },
     {
       "epoch": 0.7169664180923137,
-      "grad_norm": 3.890625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4206,
       "step": 44130
     },
     {
       "epoch": 0.7171288849896833,
-      "grad_norm": 1.5703125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.3883,
       "step": 44140
     },
     {
       "epoch": 0.7172913518870531,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4263,
+      "loss": 1.3964,
       "step": 44150
     },
     {
       "epoch": 0.7174538187844227,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.4465,
       "step": 44160
     },
     {
       "epoch": 0.7176162856817924,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.393,
+      "loss": 1.4559,
       "step": 44170
     },
     {
       "epoch": 0.717778752579162,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.426,
       "step": 44180
     },
     {
       "epoch": 0.7179412194765317,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3699,
+      "loss": 1.4042,
       "step": 44190
     },
     {
       "epoch": 0.7181036863739013,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.387,
+      "loss": 1.4271,
       "step": 44200
     },
     {
       "epoch": 0.718266153271271,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4126,
       "step": 44210
     },
     {
       "epoch": 0.7184286201686406,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.4578,
       "step": 44220
     },
     {
       "epoch": 0.7185910870660103,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4087,
       "step": 44230
     },
     {
       "epoch": 0.7187535539633799,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.373,
       "step": 44240
     },
     {
       "epoch": 0.7189160208607496,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4255,
       "step": 44250
     },
     {
       "epoch": 0.7190784877581193,
-      "grad_norm": 2.25,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.4261,
       "step": 44260
     },
     {
       "epoch": 0.719240954655489,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.3843,
       "step": 44270
     },
     {
       "epoch": 0.7194034215528586,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.4278,
       "step": 44280
     },
     {
       "epoch": 0.7195658884502283,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4123,
+      "loss": 1.3845,
       "step": 44290
     },
     {
       "epoch": 0.7197283553475979,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4706,
       "step": 44300
     },
     {
       "epoch": 0.7198908222449676,
-      "grad_norm": 2.484375,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4109,
       "step": 44310
     },
     {
       "epoch": 0.7200532891423372,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4105,
+      "loss": 1.414,
       "step": 44320
     },
     {
       "epoch": 0.7202157560397069,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3775,
+      "loss": 1.3987,
       "step": 44330
     },
     {
       "epoch": 0.7203782229370765,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3863,
+      "loss": 1.4207,
       "step": 44340
     },
     {
       "epoch": 0.7205406898344462,
-      "grad_norm": 2.375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.4287,
       "step": 44350
     },
     {
       "epoch": 0.7207031567318158,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4101,
       "step": 44360
     },
     {
       "epoch": 0.7208656236291856,
-      "grad_norm": 1.765625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4103,
+      "loss": 1.4256,
       "step": 44370
     },
     {
       "epoch": 0.7210280905265553,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.3846,
       "step": 44380
     },
     {
       "epoch": 0.7211905574239249,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3802,
+      "loss": 1.4299,
       "step": 44390
     },
     {
       "epoch": 0.7213530243212946,
-      "grad_norm": 4.125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.442,
       "step": 44400
     },
     {
       "epoch": 0.7215154912186642,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.4463,
       "step": 44410
     },
     {
       "epoch": 0.7216779581160339,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.3919,
       "step": 44420
     },
     {
       "epoch": 0.7218404250134035,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.3973,
       "step": 44430
     },
     {
       "epoch": 0.7220028919107732,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.4002,
       "step": 44440
     },
     {
       "epoch": 0.7221653588081428,
-      "grad_norm": 3.828125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3733,
+      "loss": 1.3644,
       "step": 44450
     },
     {
       "epoch": 0.7223278257055125,
-      "grad_norm": 1.4140625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.382,
+      "loss": 1.3651,
       "step": 44460
     },
     {
       "epoch": 0.7224902926028821,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.4084,
       "step": 44470
     },
     {
       "epoch": 0.7226527595002519,
-      "grad_norm": 2.125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.3741,
       "step": 44480
     },
     {
       "epoch": 0.7228152263976215,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.3654,
       "step": 44490
     },
     {
       "epoch": 0.7229776932949912,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3877,
+      "loss": 1.3912,
       "step": 44500
     },
     {
       "epoch": 0.7231401601923608,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.3762,
       "step": 44510
     },
     {
       "epoch": 0.7233026270897305,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.3797,
       "step": 44520
     },
     {
       "epoch": 0.7234650939871001,
-      "grad_norm": 1.78125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4059,
       "step": 44530
     },
     {
       "epoch": 0.7236275608844698,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4509,
       "step": 44540
     },
     {
       "epoch": 0.7237900277818394,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.3762,
       "step": 44550
     },
     {
       "epoch": 0.7239524946792091,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4133,
+      "loss": 1.4605,
       "step": 44560
     },
     {
       "epoch": 0.7241149615765787,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.3859,
       "step": 44570
     },
     {
       "epoch": 0.7242774284739484,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4073,
       "step": 44580
     },
     {
       "epoch": 0.7244398953713181,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.4328,
       "step": 44590
     },
     {
       "epoch": 0.7246023622686878,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4176,
       "step": 44600
     },
     {
       "epoch": 0.7247648291660574,
-      "grad_norm": 1.5859375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.406,
       "step": 44610
     },
     {
       "epoch": 0.7249272960634271,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.4202,
       "step": 44620
     },
     {
       "epoch": 0.7250897629607967,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4279,
+      "loss": 1.4348,
       "step": 44630
     },
     {
       "epoch": 0.7252522298581664,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.4262,
       "step": 44640
     },
     {
       "epoch": 0.725414696755536,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4216,
       "step": 44650
     },
     {
       "epoch": 0.7255771636529057,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4218,
+      "loss": 1.3767,
       "step": 44660
     },
     {
       "epoch": 0.7257396305502753,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4179,
       "step": 44670
     },
     {
       "epoch": 0.725902097447645,
-      "grad_norm": 2.25,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.3972,
       "step": 44680
     },
     {
       "epoch": 0.7260645643450147,
-      "grad_norm": 1.7578125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 1.3956,
       "step": 44690
     },
     {
       "epoch": 0.7262270312423844,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.4278,
       "step": 44700
     },
     {
       "epoch": 0.7263894981397541,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4267,
       "step": 44710
     },
     {
       "epoch": 0.7265519650371237,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.3887,
       "step": 44720
     },
     {
       "epoch": 0.7267144319344934,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3924,
+      "loss": 1.3865,
       "step": 44730
     },
     {
       "epoch": 0.726876898831863,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.383,
+      "loss": 1.4087,
       "step": 44740
     },
     {
       "epoch": 0.7270393657292327,
-      "grad_norm": 3.0,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.3961,
       "step": 44750
     },
     {
       "epoch": 0.7272018326266023,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4017,
+      "loss": 1.4293,
       "step": 44760
     },
     {
       "epoch": 0.727364299523972,
-      "grad_norm": 2.375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3994,
+      "loss": 1.4558,
       "step": 44770
     },
     {
       "epoch": 0.7275267664213416,
-      "grad_norm": 1.7578125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3752,
+      "loss": 1.4281,
       "step": 44780
     },
     {
       "epoch": 0.7276892333187113,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.3985,
       "step": 44790
     },
     {
       "epoch": 0.7278517002160809,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.3854,
       "step": 44800
     },
     {
       "epoch": 0.7280141671134507,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4116,
+      "loss": 1.4587,
       "step": 44810
     },
     {
       "epoch": 0.7281766340108203,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4157,
       "step": 44820
     },
     {
       "epoch": 0.72833910090819,
-      "grad_norm": 3.8125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4232,
+      "loss": 1.4325,
       "step": 44830
     },
     {
       "epoch": 0.7285015678055596,
-      "grad_norm": 3.0,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.4403,
       "step": 44840
     },
     {
       "epoch": 0.7286640347029293,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.4356,
       "step": 44850
     },
     {
       "epoch": 0.7288265016002989,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.375,
+      "loss": 1.4394,
       "step": 44860
     },
     {
       "epoch": 0.7289889684976686,
-      "grad_norm": 3.625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4133,
       "step": 44870
     },
     {
       "epoch": 0.7291514353950382,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3979,
       "step": 44880
     },
     {
       "epoch": 0.7293139022924079,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.3972,
       "step": 44890
     },
     {
       "epoch": 0.7294763691897775,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.4114,
       "step": 44900
     },
     {
       "epoch": 0.7296388360871472,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.3768,
       "step": 44910
     },
     {
       "epoch": 0.729801302984517,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4155,
+      "loss": 1.4121,
       "step": 44920
     },
     {
       "epoch": 0.7299637698818866,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.3931,
       "step": 44930
     },
     {
       "epoch": 0.7301262367792563,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3867,
+      "loss": 1.4441,
       "step": 44940
     },
     {
       "epoch": 0.7302887036766259,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4037,
       "step": 44950
     },
     {
       "epoch": 0.7304511705739956,
-      "grad_norm": 3.375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.3885,
       "step": 44960
     },
     {
       "epoch": 0.7306136374713652,
-      "grad_norm": 2.625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.3982,
       "step": 44970
     },
     {
       "epoch": 0.7307761043687349,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.3842,
       "step": 44980
     },
     {
       "epoch": 0.7309385712661045,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4446,
       "step": 44990
     },
     {
       "epoch": 0.7311010381634742,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.467,
       "step": 45000
     },
     {
       "epoch": 0.7312635050608438,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3774,
+      "loss": 1.4008,
       "step": 45010
     },
     {
       "epoch": 0.7314259719582135,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4606,
       "step": 45020
     },
     {
       "epoch": 0.7315884388555832,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.3843,
       "step": 45030
     },
     {
       "epoch": 0.7317509057529529,
-      "grad_norm": 2.515625,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.3668,
       "step": 45040
     },
     {
       "epoch": 0.7319133726503225,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.4239,
       "step": 45050
     },
     {
       "epoch": 0.7320758395476922,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4744,
       "step": 45060
     },
     {
       "epoch": 0.7322383064450618,
-      "grad_norm": 4.59375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.4298,
       "step": 45070
     },
     {
       "epoch": 0.7324007733424315,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.3673,
       "step": 45080
     },
     {
       "epoch": 0.7325632402398011,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.3928,
       "step": 45090
     },
     {
       "epoch": 0.7327257071371708,
-      "grad_norm": 2.0,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.3744,
       "step": 45100
     },
     {
       "epoch": 0.7328881740345404,
-      "grad_norm": 1.75,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3626,
+      "loss": 1.3571,
       "step": 45110
     },
     {
       "epoch": 0.7330506409319101,
-      "grad_norm": 2.0,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.3909,
       "step": 45120
     },
     {
       "epoch": 0.7332131078292797,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4268,
       "step": 45130
     },
     {
       "epoch": 0.7333755747266495,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.4098,
       "step": 45140
     },
     {
       "epoch": 0.7335380416240191,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3803,
+      "loss": 1.3874,
       "step": 45150
     },
     {
       "epoch": 0.7337005085213888,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4168,
+      "loss": 1.3759,
       "step": 45160
     },
     {
       "epoch": 0.7338629754187584,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.3747,
       "step": 45170
     },
     {
       "epoch": 0.7340254423161281,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.4049,
       "step": 45180
     },
     {
       "epoch": 0.7341879092134977,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.3703,
       "step": 45190
     },
     {
       "epoch": 0.7343503761108674,
-      "grad_norm": 2.03125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4062,
+      "loss": 1.4311,
       "step": 45200
     },
     {
       "epoch": 0.734512843008237,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4486,
       "step": 45210
     },
     {
       "epoch": 0.7346753099056067,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3746,
+      "loss": 1.395,
       "step": 45220
     },
     {
       "epoch": 0.7348377768029764,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.417,
       "step": 45230
     },
     {
       "epoch": 0.735000243700346,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.4363,
       "step": 45240
     },
     {
       "epoch": 0.7351627105977158,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4321,
       "step": 45250
     },
     {
       "epoch": 0.7353251774950854,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.411,
       "step": 45260
     },
     {
       "epoch": 0.7354876443924551,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4764,
       "step": 45270
     },
     {
       "epoch": 0.7356501112898247,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.3887,
       "step": 45280
     },
     {
       "epoch": 0.7358125781871944,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.434,
       "step": 45290
     },
     {
       "epoch": 0.735975045084564,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3761,
+      "loss": 1.4448,
       "step": 45300
     },
     {
       "epoch": 0.7361375119819337,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.3902,
       "step": 45310
     },
     {
       "epoch": 0.7362999788793033,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3772,
+      "loss": 1.3742,
       "step": 45320
     },
     {
       "epoch": 0.736462445776673,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.4058,
       "step": 45330
     },
     {
       "epoch": 0.7366249126740426,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3959,
       "step": 45340
     },
     {
       "epoch": 0.7367873795714123,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4443,
       "step": 45350
     },
     {
       "epoch": 0.736949846468782,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.4105,
       "step": 45360
     },
     {
       "epoch": 0.7371123133661517,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4229,
+      "loss": 1.4023,
       "step": 45370
     },
     {
       "epoch": 0.7372747802635213,
-      "grad_norm": 1.8984375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.3959,
       "step": 45380
     },
     {
       "epoch": 0.737437247160891,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3825,
+      "loss": 1.3565,
       "step": 45390
     },
     {
       "epoch": 0.7375997140582606,
-      "grad_norm": 1.5234375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.4014,
       "step": 45400
     },
     {
       "epoch": 0.7377621809556303,
-      "grad_norm": 4.21875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.3766,
       "step": 45410
     },
     {
       "epoch": 0.7379246478529999,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4269,
+      "loss": 1.4321,
       "step": 45420
     },
     {
       "epoch": 0.7380871147503696,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3646,
+      "loss": 1.3872,
       "step": 45430
     },
     {
       "epoch": 0.7382495816477392,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.3976,
       "step": 45440
     },
     {
       "epoch": 0.7384120485451089,
-      "grad_norm": 2.125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3765,
       "step": 45450
     },
     {
       "epoch": 0.7385745154424785,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4031,
+      "loss": 1.4369,
       "step": 45460
     },
     {
       "epoch": 0.7387369823398483,
-      "grad_norm": 2.21875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.4026,
       "step": 45470
     },
     {
       "epoch": 0.738899449237218,
-      "grad_norm": 1.640625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3747,
+      "loss": 1.4177,
       "step": 45480
     },
     {
       "epoch": 0.7390619161345876,
-      "grad_norm": 3.296875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3853,
+      "loss": 1.3873,
       "step": 45490
     },
     {
       "epoch": 0.7392243830319573,
-      "grad_norm": 2.3125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4383,
       "step": 45500
     },
     {
       "epoch": 0.7393868499293269,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.4147,
       "step": 45510
     },
     {
       "epoch": 0.7395493168266966,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.4066,
       "step": 45520
     },
     {
       "epoch": 0.7397117837240662,
-      "grad_norm": 3.625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.3908,
       "step": 45530
     },
     {
       "epoch": 0.7398742506214359,
-      "grad_norm": 2.625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4029,
       "step": 45540
     },
     {
       "epoch": 0.7400367175188055,
-      "grad_norm": 3.515625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.4286,
       "step": 45550
     },
     {
       "epoch": 0.7401991844161752,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.4108,
       "step": 45560
     },
     {
       "epoch": 0.7403616513135448,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.4271,
       "step": 45570
     },
     {
       "epoch": 0.7405241182109146,
-      "grad_norm": 1.6953125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.42,
       "step": 45580
     },
     {
       "epoch": 0.7406865851082842,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4393,
       "step": 45590
     },
     {
       "epoch": 0.7408490520056539,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.3985,
       "step": 45600
     },
     {
       "epoch": 0.7410115189030235,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.4093,
       "step": 45610
     },
     {
       "epoch": 0.7411739858003932,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4281,
+      "loss": 1.378,
       "step": 45620
     },
     {
       "epoch": 0.7413364526977628,
-      "grad_norm": 1.765625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4519,
       "step": 45630
     },
     {
       "epoch": 0.7414989195951325,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.3989,
       "step": 45640
     },
     {
       "epoch": 0.7416613864925021,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.4141,
       "step": 45650
     },
     {
       "epoch": 0.7418238533898718,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3796,
+      "loss": 1.3858,
       "step": 45660
     },
     {
       "epoch": 0.7419863202872414,
-      "grad_norm": 1.9921875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3913,
       "step": 45670
     },
     {
       "epoch": 0.7421487871846111,
-      "grad_norm": 2.0,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.3799,
       "step": 45680
     },
     {
       "epoch": 0.7423112540819808,
-      "grad_norm": 3.640625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.3756,
       "step": 45690
     },
     {
       "epoch": 0.7424737209793505,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4004,
       "step": 45700
     },
     {
       "epoch": 0.7426361878767201,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4167,
       "step": 45710
     },
     {
       "epoch": 0.7427986547740898,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3773,
+      "loss": 1.4078,
       "step": 45720
     },
     {
       "epoch": 0.7429611216714594,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.4029,
       "step": 45730
     },
     {
       "epoch": 0.7431235885688291,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.3936,
       "step": 45740
     },
     {
       "epoch": 0.7432860554661987,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3647,
+      "loss": 1.4326,
       "step": 45750
     },
     {
       "epoch": 0.7434485223635684,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.4092,
       "step": 45760
     },
     {
       "epoch": 0.743610989260938,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.3955,
       "step": 45770
     },
     {
       "epoch": 0.7437734561583077,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4006,
       "step": 45780
     },
     {
       "epoch": 0.7439359230556774,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4002,
       "step": 45790
     },
     {
       "epoch": 0.7440983899530471,
-      "grad_norm": 1.7890625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.4179,
       "step": 45800
     },
     {
       "epoch": 0.7442608568504168,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.4242,
       "step": 45810
     },
     {
       "epoch": 0.7444233237477864,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3834,
+      "loss": 1.3818,
       "step": 45820
     },
     {
       "epoch": 0.7445857906451561,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3775,
+      "loss": 1.4235,
       "step": 45830
     },
     {
       "epoch": 0.7447482575425257,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4258,
       "step": 45840
     },
     {
       "epoch": 0.7449107244398954,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.3974,
       "step": 45850
     },
     {
       "epoch": 0.745073191337265,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.393,
+      "loss": 1.4138,
       "step": 45860
     },
     {
       "epoch": 0.7452356582346347,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.4236,
       "step": 45870
     },
     {
       "epoch": 0.7453981251320043,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3705,
       "step": 45880
     },
     {
       "epoch": 0.745560592029374,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3808,
+      "loss": 1.4225,
       "step": 45890
     },
     {
       "epoch": 0.7457230589267436,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4322,
       "step": 45900
     },
     {
       "epoch": 0.7458855258241134,
-      "grad_norm": 1.703125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.416,
       "step": 45910
     },
     {
       "epoch": 0.746047992721483,
-      "grad_norm": 1.78125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4022,
       "step": 45920
     },
     {
       "epoch": 0.7462104596188527,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4048,
+      "loss": 1.3999,
       "step": 45930
     },
     {
       "epoch": 0.7463729265162223,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.3883,
       "step": 45940
     },
     {
       "epoch": 0.746535393413592,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.4158,
       "step": 45950
     },
     {
       "epoch": 0.7466978603109616,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.3759,
       "step": 45960
     },
     {
       "epoch": 0.7468603272083313,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4189,
       "step": 45970
     },
     {
       "epoch": 0.7470227941057009,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.446,
       "step": 45980
     },
     {
       "epoch": 0.7471852610030706,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3931,
+      "loss": 1.4159,
       "step": 45990
     },
     {
       "epoch": 0.7473477279004402,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4155,
       "step": 46000
     },
     {
       "epoch": 0.74751019479781,
-      "grad_norm": 2.640625,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4379,
       "step": 46010
     },
     {
       "epoch": 0.7476726616951797,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.469,
       "step": 46020
     },
     {
       "epoch": 0.7478351285925493,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4137,
       "step": 46030
     },
     {
       "epoch": 0.747997595489919,
-      "grad_norm": 2.8125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.4095,
       "step": 46040
     },
     {
       "epoch": 0.7481600623872886,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.421,
+      "loss": 1.4285,
       "step": 46050
     },
     {
       "epoch": 0.7483225292846583,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3904,
+      "loss": 1.4118,
       "step": 46060
     },
     {
       "epoch": 0.7484849961820279,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.3887,
       "step": 46070
     },
     {
       "epoch": 0.7486474630793976,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4234,
+      "loss": 1.4472,
       "step": 46080
     },
     {
       "epoch": 0.7488099299767672,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3735,
+      "loss": 1.4352,
       "step": 46090
     },
     {
       "epoch": 0.7489723968741369,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.425,
+      "loss": 1.4734,
       "step": 46100
     },
     {
       "epoch": 0.7491348637715065,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.4344,
       "step": 46110
     },
     {
       "epoch": 0.7492973306688763,
-      "grad_norm": 1.734375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.4015,
       "step": 46120
     },
     {
       "epoch": 0.7494597975662459,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.3544,
       "step": 46130
     },
     {
       "epoch": 0.7496222644636156,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.382,
+      "loss": 1.4356,
       "step": 46140
     },
     {
       "epoch": 0.7497847313609852,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.4345,
       "step": 46150
     },
     {
       "epoch": 0.7499471982583549,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.3892,
       "step": 46160
     },
     {
       "epoch": 0.7501096651557245,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.4444,
       "step": 46170
     },
     {
       "epoch": 0.7502721320530942,
-      "grad_norm": 4.21875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.4592,
       "step": 46180
     },
     {
       "epoch": 0.7504345989504638,
-      "grad_norm": 2.859375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4157,
+      "loss": 1.435,
       "step": 46190
     },
     {
       "epoch": 0.7505970658478335,
-      "grad_norm": 2.375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.3621,
       "step": 46200
     },
     {
       "epoch": 0.7507595327452031,
-      "grad_norm": 2.0,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3797,
+      "loss": 1.4034,
       "step": 46210
     },
     {
       "epoch": 0.7509219996425728,
-      "grad_norm": 1.734375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.3883,
       "step": 46220
     },
     {
       "epoch": 0.7510844665399425,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4281,
       "step": 46230
     },
     {
       "epoch": 0.7512469334373122,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4101,
       "step": 46240
     },
     {
       "epoch": 0.7514094003346818,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.4359,
       "step": 46250
     },
     {
       "epoch": 0.7515718672320515,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.415,
       "step": 46260
     },
     {
       "epoch": 0.7517343341294211,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4141,
+      "loss": 1.4204,
       "step": 46270
     },
     {
       "epoch": 0.7518968010267908,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.3817,
       "step": 46280
     },
     {
       "epoch": 0.7520592679241604,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4284,
       "step": 46290
     },
     {
       "epoch": 0.7522217348215301,
-      "grad_norm": 2.75,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.3878,
       "step": 46300
     },
     {
       "epoch": 0.7523842017188997,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.3714,
       "step": 46310
     },
     {
       "epoch": 0.7525466686162694,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.4335,
       "step": 46320
     },
     {
       "epoch": 0.752709135513639,
-      "grad_norm": 2.125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3833,
+      "loss": 1.3954,
       "step": 46330
     },
     {
       "epoch": 0.7528716024110088,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.3734,
       "step": 46340
     },
     {
       "epoch": 0.7530340693083785,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.3711,
       "step": 46350
     },
     {
       "epoch": 0.7531965362057481,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.4125,
       "step": 46360
     },
     {
       "epoch": 0.7533590031031178,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.3948,
       "step": 46370
     },
     {
       "epoch": 0.7535214700004874,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.424,
       "step": 46380
     },
     {
       "epoch": 0.7536839368978571,
-      "grad_norm": 2.375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.4092,
       "step": 46390
     },
     {
       "epoch": 0.7538464037952267,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4051,
+      "loss": 1.4009,
       "step": 46400
     },
     {
       "epoch": 0.7540088706925964,
-      "grad_norm": 4.0,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.4247,
       "step": 46410
     },
     {
       "epoch": 0.754171337589966,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4051,
       "step": 46420
     },
     {
       "epoch": 0.7543338044873357,
-      "grad_norm": 2.90625,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.3943,
       "step": 46430
     },
     {
       "epoch": 0.7544962713847053,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3784,
+      "loss": 1.4332,
       "step": 46440
     },
     {
       "epoch": 0.7546587382820751,
-      "grad_norm": 2.0,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.3937,
       "step": 46450
     },
     {
       "epoch": 0.7548212051794447,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.4335,
       "step": 46460
     },
     {
       "epoch": 0.7549836720768144,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4109,
       "step": 46470
     },
     {
       "epoch": 0.755146138974184,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4124,
       "step": 46480
     },
     {
       "epoch": 0.7553086058715537,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4046,
       "step": 46490
     },
     {
       "epoch": 0.7554710727689233,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4141,
       "step": 46500
     },
     {
       "epoch": 0.755633539666293,
-      "grad_norm": 1.9375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.3778,
       "step": 46510
     },
     {
       "epoch": 0.7557960065636626,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.416,
       "step": 46520
     },
     {
       "epoch": 0.7559584734610323,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4028,
       "step": 46530
     },
     {
       "epoch": 0.7561209403584019,
-      "grad_norm": 1.5703125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.4355,
       "step": 46540
     },
     {
       "epoch": 0.7562834072557716,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.3931,
       "step": 46550
     },
     {
       "epoch": 0.7564458741531414,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.307,
       "step": 46560
     },
     {
       "epoch": 0.756608341050511,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.4078,
       "step": 46570
     },
     {
       "epoch": 0.7567708079478807,
-      "grad_norm": 4.1875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4066,
+      "loss": 1.4027,
       "step": 46580
     },
     {
       "epoch": 0.7569332748452503,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3734,
+      "loss": 1.4159,
       "step": 46590
     },
     {
       "epoch": 0.75709574174262,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.3677,
       "step": 46600
     },
     {
       "epoch": 0.7572582086399896,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3697,
+      "loss": 1.4055,
       "step": 46610
     },
     {
       "epoch": 0.7574206755373593,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.3761,
       "step": 46620
     },
     {
       "epoch": 0.7575831424347289,
-      "grad_norm": 4.53125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.3642,
       "step": 46630
     },
     {
       "epoch": 0.7577456093320986,
-      "grad_norm": 3.640625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.4056,
       "step": 46640
     },
     {
       "epoch": 0.7579080762294682,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3628,
+      "loss": 1.3857,
       "step": 46650
     },
     {
       "epoch": 0.7580705431268379,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3785,
+      "loss": 1.4106,
       "step": 46660
     },
     {
       "epoch": 0.7582330100242076,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.415,
+      "loss": 1.4072,
       "step": 46670
     },
     {
       "epoch": 0.7583954769215773,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4142,
       "step": 46680
     },
     {
       "epoch": 0.7585579438189469,
-      "grad_norm": 1.6640625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.3905,
       "step": 46690
     },
     {
       "epoch": 0.7587204107163166,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3869,
+      "loss": 1.4136,
       "step": 46700
     },
     {
       "epoch": 0.7588828776136862,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.3868,
       "step": 46710
     },
     {
       "epoch": 0.7590453445110559,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4067,
       "step": 46720
     },
     {
       "epoch": 0.7592078114084255,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.412,
       "step": 46730
     },
     {
       "epoch": 0.7593702783057952,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.411,
+      "loss": 1.3871,
       "step": 46740
     },
     {
       "epoch": 0.7595327452031648,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4177,
       "step": 46750
     },
     {
       "epoch": 0.7596952121005345,
-      "grad_norm": 1.421875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.3495,
       "step": 46760
     },
     {
       "epoch": 0.7598576789979041,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4423,
       "step": 46770
     },
     {
       "epoch": 0.7600201458952739,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4515,
       "step": 46780
     },
     {
       "epoch": 0.7601826127926435,
-      "grad_norm": 2.125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4274,
       "step": 46790
     },
     {
       "epoch": 0.7603450796900132,
-      "grad_norm": 1.8125,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.4024,
       "step": 46800
     },
     {
       "epoch": 0.7605075465873828,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.4287,
       "step": 46810
     },
     {
       "epoch": 0.7606700134847525,
-      "grad_norm": 2.078125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.397,
       "step": 46820
     },
     {
       "epoch": 0.7608324803821221,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3765,
+      "loss": 1.4529,
       "step": 46830
     },
     {
       "epoch": 0.7609949472794918,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3763,
+      "loss": 1.3596,
       "step": 46840
     },
     {
       "epoch": 0.7611574141768614,
-      "grad_norm": 1.953125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.4623,
       "step": 46850
     },
     {
       "epoch": 0.7613198810742311,
-      "grad_norm": 2.875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.3741,
       "step": 46860
     },
     {
       "epoch": 0.7614823479716007,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.3943,
       "step": 46870
     },
     {
       "epoch": 0.7616448148689704,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.3727,
       "step": 46880
     },
     {
       "epoch": 0.7618072817663402,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.393,
       "step": 46890
     },
     {
       "epoch": 0.7619697486637098,
-      "grad_norm": 3.609375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.393,
       "step": 46900
     },
     {
       "epoch": 0.7621322155610795,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4729,
       "step": 46910
     },
     {
       "epoch": 0.7622946824584491,
-      "grad_norm": 3.625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.3771,
       "step": 46920
     },
     {
       "epoch": 0.7624571493558188,
-      "grad_norm": 2.125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3877,
+      "loss": 1.4067,
       "step": 46930
     },
     {
       "epoch": 0.7626196162531884,
-      "grad_norm": 2.03125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.379,
+      "loss": 1.3753,
       "step": 46940
     },
     {
       "epoch": 0.7627820831505581,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.4155,
       "step": 46950
     },
     {
       "epoch": 0.7629445500479277,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4347,
       "step": 46960
     },
     {
       "epoch": 0.7631070169452974,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.4311,
       "step": 46970
     },
     {
       "epoch": 0.763269483842667,
-      "grad_norm": 2.0,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3828,
+      "loss": 1.4416,
       "step": 46980
     },
     {
       "epoch": 0.7634319507400367,
-      "grad_norm": 2.296875,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3653,
       "step": 46990
     },
     {
       "epoch": 0.7635944176374064,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3686,
+      "loss": 1.3958,
       "step": 47000
     },
     {
       "epoch": 0.7637568845347761,
-      "grad_norm": 2.015625,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.3499,
       "step": 47010
     },
     {
       "epoch": 0.7639193514321457,
-      "grad_norm": 1.5546875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4036,
       "step": 47020
     },
     {
       "epoch": 0.7640818183295154,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4091,
       "step": 47030
     },
     {
       "epoch": 0.764244285226885,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.3833,
       "step": 47040
     },
     {
       "epoch": 0.7644067521242547,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4336,
       "step": 47050
     },
     {
       "epoch": 0.7645692190216243,
-      "grad_norm": 1.3515625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.4363,
       "step": 47060
     },
     {
       "epoch": 0.764731685918994,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4675,
       "step": 47070
     },
     {
       "epoch": 0.7648941528163636,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.332,
       "step": 47080
     },
     {
       "epoch": 0.7650566197137333,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3845,
+      "loss": 1.449,
       "step": 47090
     },
     {
       "epoch": 0.7652190866111029,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.4106,
       "step": 47100
     },
     {
       "epoch": 0.7653815535084727,
-      "grad_norm": 1.7578125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4169,
       "step": 47110
     },
     {
       "epoch": 0.7655440204058424,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.3962,
       "step": 47120
     },
     {
       "epoch": 0.765706487303212,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.4137,
       "step": 47130
     },
     {
       "epoch": 0.7658689542005817,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.4117,
       "step": 47140
     },
     {
       "epoch": 0.7660314210979513,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3712,
+      "loss": 1.3982,
       "step": 47150
     },
     {
       "epoch": 0.766193887995321,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.3381,
       "step": 47160
     },
     {
       "epoch": 0.7663563548926906,
-      "grad_norm": 2.5,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.3638,
       "step": 47170
     },
     {
       "epoch": 0.7665188217900603,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3822,
+      "loss": 1.3971,
       "step": 47180
     },
     {
       "epoch": 0.7666812886874299,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.3993,
       "step": 47190
     },
     {
       "epoch": 0.7668437555847996,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4318,
       "step": 47200
     },
     {
       "epoch": 0.7670062224821692,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3776,
+      "loss": 1.4284,
       "step": 47210
     },
     {
       "epoch": 0.767168689379539,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4206,
       "step": 47220
     },
     {
       "epoch": 0.7673311562769086,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3877,
+      "loss": 1.3933,
       "step": 47230
     },
     {
       "epoch": 0.7674936231742783,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.3979,
       "step": 47240
     },
     {
       "epoch": 0.7676560900716479,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.3541,
       "step": 47250
     },
     {
       "epoch": 0.7678185569690176,
-      "grad_norm": 4.46875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4331,
+      "loss": 1.3766,
       "step": 47260
     },
     {
       "epoch": 0.7679810238663872,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.3884,
       "step": 47270
     },
     {
       "epoch": 0.7681434907637569,
-      "grad_norm": 1.65625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.3939,
       "step": 47280
     },
     {
       "epoch": 0.7683059576611265,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.3919,
       "step": 47290
     },
     {
       "epoch": 0.7684684245584962,
-      "grad_norm": 2.0625,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.3356,
       "step": 47300
     },
     {
       "epoch": 0.7686308914558658,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4184,
+      "loss": 1.3829,
       "step": 47310
     },
     {
       "epoch": 0.7687933583532355,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3819,
+      "loss": 1.4075,
       "step": 47320
     },
     {
       "epoch": 0.7689558252506052,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4068,
+      "loss": 1.4263,
       "step": 47330
     },
     {
       "epoch": 0.7691182921479749,
-      "grad_norm": 2.328125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.4002,
+      "loss": 1.4469,
       "step": 47340
     },
     {
       "epoch": 0.7692807590453445,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.4178,
       "step": 47350
     },
     {
       "epoch": 0.7694432259427142,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.4292,
       "step": 47360
     },
     {
       "epoch": 0.7696056928400838,
-      "grad_norm": 5.5625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4222,
       "step": 47370
     },
     {
       "epoch": 0.7697681597374535,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4194,
       "step": 47380
     },
     {
       "epoch": 0.7699306266348231,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.4084,
       "step": 47390
     },
     {
       "epoch": 0.7700930935321928,
-      "grad_norm": 2.875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4096,
       "step": 47400
     },
     {
       "epoch": 0.7702555604295624,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4166,
       "step": 47410
     },
     {
       "epoch": 0.7704180273269321,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.4357,
       "step": 47420
     },
     {
       "epoch": 0.7705804942243017,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3835,
       "step": 47430
     },
     {
       "epoch": 0.7707429611216715,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4292,
       "step": 47440
     },
     {
       "epoch": 0.7709054280190412,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.3614,
       "step": 47450
     },
     {
       "epoch": 0.7710678949164108,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.385,
       "step": 47460
     },
     {
       "epoch": 0.7712303618137805,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4053,
       "step": 47470
     },
     {
       "epoch": 0.7713928287111501,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.3574,
       "step": 47480
     },
     {
       "epoch": 0.7715552956085198,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3746,
+      "loss": 1.414,
       "step": 47490
     },
     {
       "epoch": 0.7717177625058894,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4201,
+      "loss": 1.4257,
       "step": 47500
     },
     {
       "epoch": 0.7718802294032591,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4092,
       "step": 47510
     },
     {
       "epoch": 0.7720426963006287,
-      "grad_norm": 1.7421875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.4059,
       "step": 47520
     },
     {
       "epoch": 0.7722051631979984,
-      "grad_norm": 2.5,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.381,
+      "loss": 1.3866,
       "step": 47530
     },
     {
       "epoch": 0.772367630095368,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4103,
+      "loss": 1.3986,
       "step": 47540
     },
     {
       "epoch": 0.7725300969927378,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4063,
+      "loss": 1.4069,
       "step": 47550
     },
     {
       "epoch": 0.7726925638901074,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.4528,
       "step": 47560
     },
     {
       "epoch": 0.7728550307874771,
-      "grad_norm": 1.9296875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4147,
+      "loss": 1.4253,
       "step": 47570
     },
     {
       "epoch": 0.7730174976848467,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.3638,
       "step": 47580
     },
     {
       "epoch": 0.7731799645822164,
-      "grad_norm": 2.0625,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.4202,
       "step": 47590
     },
     {
       "epoch": 0.773342431479586,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3804,
+      "loss": 1.4031,
       "step": 47600
     },
     {
       "epoch": 0.7735048983769557,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4336,
       "step": 47610
     },
     {
       "epoch": 0.7736673652743253,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.393,
+      "loss": 1.4004,
       "step": 47620
     },
     {
       "epoch": 0.773829832171695,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3872,
+      "loss": 1.3425,
       "step": 47630
     },
     {
       "epoch": 0.7739922990690646,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4135,
+      "loss": 1.4317,
       "step": 47640
     },
     {
       "epoch": 0.7741547659664343,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.3917,
       "step": 47650
     },
     {
       "epoch": 0.774317232863804,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4344,
       "step": 47660
     },
     {
       "epoch": 0.7744796997611737,
-      "grad_norm": 1.578125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3883,
+      "loss": 1.4032,
       "step": 47670
     },
     {
       "epoch": 0.7746421666585434,
-      "grad_norm": 2.5,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.4008,
       "step": 47680
     },
     {
       "epoch": 0.774804633555913,
-      "grad_norm": 2.140625,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.4227,
       "step": 47690
     },
     {
       "epoch": 0.7749671004532827,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.3669,
       "step": 47700
     },
     {
       "epoch": 0.7751295673506523,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.3682,
       "step": 47710
     },
     {
       "epoch": 0.775292034248022,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.3981,
       "step": 47720
     },
     {
       "epoch": 0.7754545011453916,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.3829,
       "step": 47730
     },
     {
       "epoch": 0.7756169680427613,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4036,
       "step": 47740
     },
     {
       "epoch": 0.7757794349401309,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3718,
       "step": 47750
     },
     {
       "epoch": 0.7759419018375006,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4351,
       "step": 47760
     },
     {
       "epoch": 0.7761043687348703,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.3791,
       "step": 47770
     },
     {
       "epoch": 0.77626683563224,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3883,
+      "loss": 1.4386,
       "step": 47780
     },
     {
       "epoch": 0.7764293025296096,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3787,
+      "loss": 1.4549,
       "step": 47790
     },
     {
       "epoch": 0.7765917694269793,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.3806,
       "step": 47800
     },
     {
       "epoch": 0.7767542363243489,
-      "grad_norm": 2.5,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.417,
       "step": 47810
     },
     {
       "epoch": 0.7769167032217186,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.3833,
       "step": 47820
     },
     {
       "epoch": 0.7770791701190882,
-      "grad_norm": 2.0,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4004,
+      "loss": 1.3718,
       "step": 47830
     },
     {
       "epoch": 0.7772416370164579,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3714,
+      "loss": 1.3899,
       "step": 47840
     },
     {
       "epoch": 0.7774041039138275,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3925,
+      "loss": 1.3771,
       "step": 47850
     },
     {
       "epoch": 0.7775665708111972,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.3989,
       "step": 47860
     },
     {
       "epoch": 0.7777290377085668,
-      "grad_norm": 3.75,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3795,
+      "loss": 1.3866,
       "step": 47870
     },
     {
       "epoch": 0.7778915046059366,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.3658,
       "step": 47880
     },
     {
       "epoch": 0.7780539715033062,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.4034,
       "step": 47890
     },
     {
       "epoch": 0.7782164384006759,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3889,
+      "loss": 1.3762,
       "step": 47900
     },
     {
       "epoch": 0.7783789052980455,
-      "grad_norm": 2.078125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.3828,
+      "loss": 1.4224,
       "step": 47910
     },
     {
       "epoch": 0.7785413721954152,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4172,
       "step": 47920
     },
     {
       "epoch": 0.7787038390927848,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3744,
+      "loss": 1.3801,
       "step": 47930
     },
     {
       "epoch": 0.7788663059901545,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4106,
       "step": 47940
     },
     {
       "epoch": 0.7790287728875241,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3795,
+      "loss": 1.3726,
       "step": 47950
     },
     {
       "epoch": 0.7791912397848938,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.3939,
       "step": 47960
     },
     {
       "epoch": 0.7793537066822634,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3975,
       "step": 47970
     },
     {
       "epoch": 0.7795161735796331,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4163,
       "step": 47980
     },
     {
       "epoch": 0.7796786404770029,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4174,
+      "loss": 1.4136,
       "step": 47990
     },
     {
       "epoch": 0.7798411073743725,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3779,
+      "loss": 1.3621,
       "step": 48000
     },
     {
       "epoch": 0.7800035742717422,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3906,
       "step": 48010
     },
     {
       "epoch": 0.7801660411691118,
-      "grad_norm": 1.6328125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.3536,
       "step": 48020
     },
     {
       "epoch": 0.7803285080664815,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.3771,
       "step": 48030
     },
     {
       "epoch": 0.7804909749638511,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4003,
       "step": 48040
     },
     {
       "epoch": 0.7806534418612208,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3822,
+      "loss": 1.3793,
       "step": 48050
     },
     {
       "epoch": 0.7808159087585904,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3703,
+      "loss": 1.3703,
       "step": 48060
     },
     {
       "epoch": 0.7809783756559601,
-      "grad_norm": 2.25,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.4044,
       "step": 48070
     },
     {
       "epoch": 0.7811408425533297,
-      "grad_norm": 3.4375,
+      "grad_norm": 4.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.3996,
       "step": 48080
     },
     {
       "epoch": 0.7813033094506994,
-      "grad_norm": 1.9140625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3925,
       "step": 48090
     },
     {
       "epoch": 0.7814657763480691,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.3764,
       "step": 48100
     },
     {
       "epoch": 0.7816282432454388,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4044,
       "step": 48110
     },
     {
       "epoch": 0.7817907101428084,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4072,
       "step": 48120
     },
     {
       "epoch": 0.7819531770401781,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.4557,
       "step": 48130
     },
     {
       "epoch": 0.7821156439375477,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.3824,
       "step": 48140
     },
     {
       "epoch": 0.7822781108349174,
-      "grad_norm": 1.4140625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.385,
+      "loss": 1.4436,
       "step": 48150
     },
     {
       "epoch": 0.782440577732287,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.381,
+      "loss": 1.3873,
       "step": 48160
     },
     {
       "epoch": 0.7826030446296567,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.4497,
       "step": 48170
     },
     {
       "epoch": 0.7827655115270263,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.3889,
       "step": 48180
     },
     {
       "epoch": 0.782927978424396,
-      "grad_norm": 1.8046875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.379,
+      "loss": 1.363,
       "step": 48190
     },
     {
       "epoch": 0.7830904453217656,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.3935,
       "step": 48200
     },
     {
       "epoch": 0.7832529122191354,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4076,
       "step": 48210
     },
     {
       "epoch": 0.783415379116505,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.4024,
       "step": 48220
     },
     {
       "epoch": 0.7835778460138747,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3957,
+      "loss": 1.3842,
       "step": 48230
     },
     {
       "epoch": 0.7837403129112444,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.4472,
       "step": 48240
     },
     {
       "epoch": 0.783902779808614,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3853,
+      "loss": 1.3769,
       "step": 48250
     },
     {
       "epoch": 0.7840652467059837,
-      "grad_norm": 1.7890625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.3507,
       "step": 48260
     },
     {
       "epoch": 0.7842277136033533,
-      "grad_norm": 1.640625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4527,
       "step": 48270
     },
     {
       "epoch": 0.784390180500723,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.3566,
       "step": 48280
     },
     {
       "epoch": 0.7845526473980926,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4476,
       "step": 48290
     },
     {
       "epoch": 0.7847151142954623,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4022,
       "step": 48300
     },
     {
       "epoch": 0.7848775811928319,
-      "grad_norm": 1.7109375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.3975,
       "step": 48310
     },
     {
       "epoch": 0.7850400480902017,
-      "grad_norm": 2.125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.3976,
       "step": 48320
     },
     {
       "epoch": 0.7852025149875713,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.4332,
       "step": 48330
     },
     {
       "epoch": 0.785364981884941,
-      "grad_norm": 2.875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.3843,
       "step": 48340
     },
     {
       "epoch": 0.7855274487823106,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4149,
       "step": 48350
     },
     {
       "epoch": 0.7856899156796803,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.39,
       "step": 48360
     },
     {
       "epoch": 0.7858523825770499,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.4111,
       "step": 48370
     },
     {
       "epoch": 0.7860148494744196,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3729,
+      "loss": 1.386,
       "step": 48380
     },
     {
       "epoch": 0.7861773163717892,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.4519,
       "step": 48390
     },
     {
       "epoch": 0.7863397832691589,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.4083,
       "step": 48400
     },
     {
       "epoch": 0.7865022501665285,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3804,
+      "loss": 1.4216,
       "step": 48410
     },
     {
       "epoch": 0.7866647170638982,
-      "grad_norm": 1.6875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.4612,
       "step": 48420
     },
     {
       "epoch": 0.7868271839612679,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4275,
       "step": 48430
     },
     {
       "epoch": 0.7869896508586376,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4235,
+      "loss": 1.4084,
       "step": 48440
     },
     {
       "epoch": 0.7871521177560072,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.4344,
       "step": 48450
     },
     {
       "epoch": 0.7873145846533769,
-      "grad_norm": 2.375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.4497,
       "step": 48460
     },
     {
       "epoch": 0.7874770515507465,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.4138,
       "step": 48470
     },
     {
       "epoch": 0.7876395184481162,
-      "grad_norm": 1.875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4926,
       "step": 48480
     },
     {
       "epoch": 0.7878019853454858,
-      "grad_norm": 2.125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.3927,
       "step": 48490
     },
     {
       "epoch": 0.7879644522428555,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3756,
+      "loss": 1.4681,
       "step": 48500
     },
     {
       "epoch": 0.7881269191402251,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.4519,
       "step": 48510
     },
     {
       "epoch": 0.7882893860375948,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4416,
       "step": 48520
     },
     {
       "epoch": 0.7884518529349644,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.3983,
       "step": 48530
     },
     {
       "epoch": 0.7886143198323342,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4277,
       "step": 48540
     },
     {
       "epoch": 0.7887767867297039,
-      "grad_norm": 5.4375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4274,
       "step": 48550
     },
     {
       "epoch": 0.7889392536270735,
-      "grad_norm": 1.7109375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.3844,
       "step": 48560
     },
     {
       "epoch": 0.7891017205244432,
-      "grad_norm": 4.0625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3779,
+      "loss": 1.3793,
       "step": 48570
     },
     {
       "epoch": 0.7892641874218128,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4179,
+      "loss": 1.4009,
       "step": 48580
     },
     {
       "epoch": 0.7894266543191825,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.4058,
       "step": 48590
     },
     {
       "epoch": 0.7895891212165521,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.4203,
       "step": 48600
     },
     {
       "epoch": 0.7897515881139218,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3883,
+      "loss": 1.4328,
       "step": 48610
     },
     {
       "epoch": 0.7899140550112914,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.3951,
       "step": 48620
     },
     {
       "epoch": 0.7900765219086611,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4294,
+      "loss": 1.3818,
       "step": 48630
     },
     {
       "epoch": 0.7902389888060307,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.398,
       "step": 48640
     },
     {
       "epoch": 0.7904014557034005,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4028,
+      "loss": 1.3454,
       "step": 48650
     },
     {
       "epoch": 0.7905639226007701,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3721,
+      "loss": 1.3979,
       "step": 48660
     },
     {
       "epoch": 0.7907263894981398,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.3825,
       "step": 48670
     },
     {
       "epoch": 0.7908888563955094,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3804,
+      "loss": 1.3771,
       "step": 48680
     },
     {
       "epoch": 0.7910513232928791,
-      "grad_norm": 1.6953125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4266,
       "step": 48690
     },
     {
       "epoch": 0.7912137901902487,
-      "grad_norm": 1.734375,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3792,
+      "loss": 1.4011,
       "step": 48700
     },
     {
       "epoch": 0.7913762570876184,
-      "grad_norm": 1.5859375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3696,
+      "loss": 1.4091,
       "step": 48710
     },
     {
       "epoch": 0.791538723984988,
-      "grad_norm": 2.703125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.469,
       "step": 48720
     },
     {
       "epoch": 0.7917011908823577,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.4246,
       "step": 48730
     },
     {
       "epoch": 0.7918636577797273,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4232,
       "step": 48740
     },
     {
       "epoch": 0.792026124677097,
-      "grad_norm": 2.25,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.379,
       "step": 48750
     },
     {
       "epoch": 0.7921885915744667,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4067,
+      "loss": 1.3691,
       "step": 48760
     },
     {
       "epoch": 0.7923510584718364,
-      "grad_norm": 2.671875,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4316,
       "step": 48770
     },
     {
       "epoch": 0.792513525369206,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.43,
       "step": 48780
     },
     {
       "epoch": 0.7926759922665757,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.3698,
       "step": 48790
     },
     {
       "epoch": 0.7928384591639454,
-      "grad_norm": 2.0,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4043,
       "step": 48800
     },
     {
       "epoch": 0.793000926061315,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.44,
       "step": 48810
     },
     {
       "epoch": 0.7931633929586847,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4642,
       "step": 48820
     },
     {
       "epoch": 0.7933258598560543,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4148,
       "step": 48830
     },
     {
       "epoch": 0.793488326753424,
-      "grad_norm": 2.34375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.3878,
       "step": 48840
     },
     {
       "epoch": 0.7936507936507936,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.3999,
       "step": 48850
     },
     {
       "epoch": 0.7938132605481633,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.3925,
       "step": 48860
     },
     {
       "epoch": 0.793975727445533,
-      "grad_norm": 2.3125,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.356,
       "step": 48870
     },
     {
       "epoch": 0.7941381943429027,
-      "grad_norm": 3.296875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4159,
       "step": 48880
     },
     {
       "epoch": 0.7943006612402723,
-      "grad_norm": 2.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.424,
       "step": 48890
     },
     {
       "epoch": 0.794463128137642,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.3949,
       "step": 48900
     },
     {
       "epoch": 0.7946255950350116,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.4166,
       "step": 48910
     },
     {
       "epoch": 0.7947880619323813,
-      "grad_norm": 1.71875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.4808,
       "step": 48920
     },
     {
       "epoch": 0.7949505288297509,
-      "grad_norm": 4.46875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3815,
+      "loss": 1.4189,
       "step": 48930
     },
     {
       "epoch": 0.7951129957271206,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.409,
       "step": 48940
     },
     {
       "epoch": 0.7952754626244902,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4134,
       "step": 48950
     },
     {
       "epoch": 0.7954379295218599,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4257,
+      "loss": 1.3996,
       "step": 48960
     },
     {
       "epoch": 0.7956003964192295,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4092,
+      "loss": 1.4041,
       "step": 48970
     },
     {
       "epoch": 0.7957628633165993,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.3297,
       "step": 48980
     },
     {
       "epoch": 0.7959253302139689,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4359,
       "step": 48990
     },
     {
       "epoch": 0.7960877971113386,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.4567,
       "step": 49000
     },
     {
       "epoch": 0.7962502640087082,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3644,
+      "loss": 1.4224,
       "step": 49010
     },
     {
       "epoch": 0.7964127309060779,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4124,
+      "loss": 1.3566,
       "step": 49020
     },
     {
       "epoch": 0.7965751978034475,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4226,
+      "loss": 1.4108,
       "step": 49030
     },
     {
       "epoch": 0.7967376647008172,
-      "grad_norm": 2.0,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4518,
       "step": 49040
     },
     {
       "epoch": 0.7969001315981868,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.3694,
       "step": 49050
     },
     {
       "epoch": 0.7970625984955565,
-      "grad_norm": 3.5,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.3899,
       "step": 49060
     },
     {
       "epoch": 0.7972250653929261,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.3466,
       "step": 49070
     },
     {
       "epoch": 0.7973875322902958,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.3902,
       "step": 49080
     },
     {
       "epoch": 0.7975499991876656,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.3656,
       "step": 49090
     },
     {
       "epoch": 0.7977124660850352,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4114,
       "step": 49100
     },
     {
       "epoch": 0.7978749329824049,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.4152,
       "step": 49110
     },
     {
       "epoch": 0.7980373998797745,
-      "grad_norm": 4.3125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.425,
       "step": 49120
     },
     {
       "epoch": 0.7981998667771442,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4162,
+      "loss": 1.4082,
       "step": 49130
     },
     {
       "epoch": 0.7983623336745138,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4066,
       "step": 49140
     },
     {
       "epoch": 0.7985248005718835,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.4314,
       "step": 49150
     },
     {
       "epoch": 0.7986872674692531,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.4072,
       "step": 49160
     },
     {
       "epoch": 0.7988497343666228,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3833,
+      "loss": 1.4315,
       "step": 49170
     },
     {
       "epoch": 0.7990122012639924,
-      "grad_norm": 3.65625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.411,
       "step": 49180
     },
     {
       "epoch": 0.7991746681613621,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4196,
       "step": 49190
     },
     {
       "epoch": 0.7993371350587318,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.4589,
       "step": 49200
     },
     {
       "epoch": 0.7994996019561015,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3699,
+      "loss": 1.3909,
       "step": 49210
     },
     {
       "epoch": 0.7996620688534711,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4093,
       "step": 49220
     },
     {
       "epoch": 0.7998245357508408,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4048,
+      "loss": 1.4048,
       "step": 49230
     },
     {
       "epoch": 0.7999870026482104,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3708,
+      "loss": 1.3855,
       "step": 49240
     },
     {
       "epoch": 0.8001494695455801,
-      "grad_norm": 2.125,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3924,
+      "loss": 1.4,
       "step": 49250
     },
     {
       "epoch": 0.8003119364429497,
-      "grad_norm": 3.703125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3904,
+      "loss": 1.4306,
       "step": 49260
     },
     {
       "epoch": 0.8004744033403194,
-      "grad_norm": 1.609375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.3733,
       "step": 49270
     },
     {
       "epoch": 0.800636870237689,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.3894,
       "step": 49280
     },
     {
       "epoch": 0.8007993371350587,
-      "grad_norm": 1.53125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.441,
       "step": 49290
     },
     {
       "epoch": 0.8009618040324283,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.3716,
       "step": 49300
     },
     {
       "epoch": 0.8011242709297981,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3715,
+      "loss": 1.3772,
       "step": 49310
     },
     {
       "epoch": 0.8012867378271677,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.4026,
       "step": 49320
     },
     {
       "epoch": 0.8014492047245374,
-      "grad_norm": 1.71875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.4167,
       "step": 49330
     },
     {
       "epoch": 0.801611671621907,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.379,
       "step": 49340
     },
     {
       "epoch": 0.8017741385192767,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4193,
       "step": 49350
     },
     {
       "epoch": 0.8019366054166464,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3807,
+      "loss": 1.429,
       "step": 49360
     },
     {
       "epoch": 0.802099072314016,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.4,
       "step": 49370
     },
     {
       "epoch": 0.8022615392113857,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.3967,
       "step": 49380
     },
     {
       "epoch": 0.8024240061087553,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.4111,
       "step": 49390
     },
     {
       "epoch": 0.802586473006125,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3898,
+      "loss": 1.3859,
       "step": 49400
     },
     {
       "epoch": 0.8027489399034947,
-      "grad_norm": 2.171875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3749,
+      "loss": 1.4386,
       "step": 49410
     },
     {
       "epoch": 0.8029114068008644,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.3947,
       "step": 49420
     },
     {
       "epoch": 0.803073873698234,
-      "grad_norm": 2.9375,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.3578,
       "step": 49430
     },
     {
       "epoch": 0.8032363405956037,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.4364,
       "step": 49440
     },
     {
       "epoch": 0.8033988074929733,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4178,
+      "loss": 1.3913,
       "step": 49450
     },
     {
       "epoch": 0.803561274390343,
-      "grad_norm": 2.25,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4139,
+      "loss": 1.4222,
       "step": 49460
     },
     {
       "epoch": 0.8037237412877126,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3786,
+      "loss": 1.3832,
       "step": 49470
     },
     {
       "epoch": 0.8038862081850823,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.3965,
       "step": 49480
     },
     {
       "epoch": 0.8040486750824519,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.3422,
       "step": 49490
     },
     {
       "epoch": 0.8042111419798216,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.4047,
       "step": 49500
     },
     {
       "epoch": 0.8043736088771912,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3775,
+      "loss": 1.4591,
       "step": 49510
     },
     {
       "epoch": 0.804536075774561,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.3855,
       "step": 49520
     },
     {
       "epoch": 0.8046985426719306,
-      "grad_norm": 2.328125,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4072,
       "step": 49530
     },
     {
       "epoch": 0.8048610095693003,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4023,
+      "loss": 1.3584,
       "step": 49540
     },
     {
       "epoch": 0.8050234764666699,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4062,
       "step": 49550
     },
     {
       "epoch": 0.8051859433640396,
-      "grad_norm": 2.625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.4454,
       "step": 49560
     },
     {
       "epoch": 0.8053484102614092,
-      "grad_norm": 1.875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.4053,
       "step": 49570
     },
     {
       "epoch": 0.8055108771587789,
-      "grad_norm": 2.375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3819,
+      "loss": 1.362,
       "step": 49580
     },
     {
       "epoch": 0.8056733440561485,
-      "grad_norm": 2.75,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3807,
+      "loss": 1.3837,
       "step": 49590
     },
     {
       "epoch": 0.8058358109535182,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4026,
+      "loss": 1.3691,
       "step": 49600
     },
     {
       "epoch": 0.8059982778508878,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3709,
+      "loss": 1.423,
       "step": 49610
     },
     {
       "epoch": 0.8061607447482575,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.3781,
       "step": 49620
     },
     {
       "epoch": 0.8063232116456273,
-      "grad_norm": 4.125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.3948,
       "step": 49630
     },
     {
       "epoch": 0.8064856785429969,
-      "grad_norm": 1.8671875,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3809,
+      "loss": 1.4345,
       "step": 49640
     },
     {
       "epoch": 0.8066481454403666,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.3807,
       "step": 49650
     },
     {
       "epoch": 0.8068106123377362,
-      "grad_norm": 1.6171875,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3815,
+      "loss": 1.3573,
       "step": 49660
     },
     {
       "epoch": 0.8069730792351059,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4116,
       "step": 49670
     },
     {
       "epoch": 0.8071355461324755,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.4056,
       "step": 49680
     },
     {
       "epoch": 0.8072980130298452,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.3755,
       "step": 49690
     },
     {
       "epoch": 0.8074604799272148,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.3866,
       "step": 49700
     },
     {
       "epoch": 0.8076229468245845,
-      "grad_norm": 4.5,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.3729,
       "step": 49710
     },
     {
       "epoch": 0.8077854137219541,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.3989,
       "step": 49720
     },
     {
       "epoch": 0.8079478806193238,
-      "grad_norm": 1.7109375,
+      "grad_norm": 4.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3717,
+      "loss": 1.4251,
       "step": 49730
     },
     {
       "epoch": 0.8081103475166935,
-      "grad_norm": 2.625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.3884,
       "step": 49740
     },
     {
       "epoch": 0.8082728144140632,
-      "grad_norm": 2.5,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3722,
+      "loss": 1.4031,
       "step": 49750
     },
     {
       "epoch": 0.8084352813114328,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.377,
       "step": 49760
     },
     {
       "epoch": 0.8085977482088025,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3726,
+      "loss": 1.4148,
       "step": 49770
     },
     {
       "epoch": 0.8087602151061721,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.3826,
       "step": 49780
     },
     {
       "epoch": 0.8089226820035418,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.431,
       "step": 49790
     },
     {
       "epoch": 0.8090851489009114,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4078,
+      "loss": 1.423,
       "step": 49800
     },
     {
       "epoch": 0.8092476157982811,
-      "grad_norm": 2.375,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3887,
       "step": 49810
     },
     {
       "epoch": 0.8094100826956507,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3613,
+      "loss": 1.4346,
       "step": 49820
     },
     {
       "epoch": 0.8095725495930204,
-      "grad_norm": 1.6640625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.4116,
       "step": 49830
     },
     {
       "epoch": 0.80973501649039,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4075,
       "step": 49840
     },
     {
       "epoch": 0.8098974833877598,
-      "grad_norm": 3.0,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.4086,
       "step": 49850
     },
     {
       "epoch": 0.8100599502851294,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.3663,
       "step": 49860
     },
     {
       "epoch": 0.8102224171824991,
-      "grad_norm": 3.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.4215,
       "step": 49870
     },
     {
       "epoch": 0.8103848840798688,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.3874,
       "step": 49880
     },
     {
       "epoch": 0.8105473509772384,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4073,
       "step": 49890
     },
     {
       "epoch": 0.810709817874608,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4282,
       "step": 49900
     },
     {
       "epoch": 0.8108722847719777,
-      "grad_norm": 1.6171875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3825,
+      "loss": 1.4291,
       "step": 49910
     },
     {
       "epoch": 0.8110347516693474,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.353,
       "step": 49920
     },
     {
       "epoch": 0.811197218566717,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3752,
+      "loss": 1.3963,
       "step": 49930
     },
     {
       "epoch": 0.8113596854640867,
-      "grad_norm": 1.6171875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.3878,
       "step": 49940
     },
     {
       "epoch": 0.8115221523614563,
-      "grad_norm": 3.75,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4099,
       "step": 49950
     },
     {
       "epoch": 0.8116846192588261,
-      "grad_norm": 2.125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.412,
+      "loss": 1.4499,
       "step": 49960
     },
     {
       "epoch": 0.8118470861561957,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3962,
       "step": 49970
     },
     {
       "epoch": 0.8120095530535654,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.4086,
       "step": 49980
     },
     {
       "epoch": 0.812172019950935,
-      "grad_norm": 2.75,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4051,
+      "loss": 1.4055,
       "step": 49990
     },
     {
       "epoch": 0.8123344868483047,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4118,
+      "loss": 1.4144,
       "step": 50000
     },
     {
       "epoch": 0.8124969537456743,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4148,
+      "loss": 1.372,
       "step": 50010
     },
     {
       "epoch": 0.812659420643044,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.368,
       "step": 50020
     },
     {
       "epoch": 0.8128218875404136,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3703,
+      "loss": 1.4055,
       "step": 50030
     },
     {
       "epoch": 0.8129843544377833,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.4078,
       "step": 50040
     },
     {
       "epoch": 0.8131468213351529,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.4085,
       "step": 50050
     },
     {
       "epoch": 0.8133092882325226,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.4017,
       "step": 50060
     },
     {
       "epoch": 0.8134717551298923,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.4135,
       "step": 50070
     },
     {
       "epoch": 0.813634222027262,
-      "grad_norm": 1.5546875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3738,
+      "loss": 1.3801,
       "step": 50080
     },
     {
       "epoch": 0.8137966889246316,
-      "grad_norm": 2.25,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.3649,
       "step": 50090
     },
     {
       "epoch": 0.8139591558220013,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4263,
       "step": 50100
     },
     {
       "epoch": 0.8141216227193709,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.3734,
       "step": 50110
     },
     {
       "epoch": 0.8142840896167406,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.4405,
       "step": 50120
     },
     {
       "epoch": 0.8144465565141102,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4051,
       "step": 50130
     },
     {
       "epoch": 0.8146090234114799,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.3998,
       "step": 50140
     },
     {
       "epoch": 0.8147714903088495,
-      "grad_norm": 2.75,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.4057,
       "step": 50150
     },
     {
       "epoch": 0.8149339572062192,
-      "grad_norm": 2.375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.4092,
       "step": 50160
     },
     {
       "epoch": 0.8150964241035888,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4035,
       "step": 50170
     },
     {
       "epoch": 0.8152588910009586,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4473,
       "step": 50180
     },
     {
       "epoch": 0.8154213578983283,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3841,
+      "loss": 1.3875,
       "step": 50190
     },
     {
       "epoch": 0.8155838247956979,
-      "grad_norm": 1.796875,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.3942,
       "step": 50200
     },
     {
       "epoch": 0.8157462916930676,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4358,
       "step": 50210
     },
     {
       "epoch": 0.8159087585904372,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4438,
       "step": 50220
     },
     {
       "epoch": 0.8160712254878069,
-      "grad_norm": 2.015625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.384,
+      "loss": 1.4358,
       "step": 50230
     },
     {
       "epoch": 0.8162336923851765,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.4084,
+      "loss": 1.3851,
       "step": 50240
     },
     {
       "epoch": 0.8163961592825462,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4129,
       "step": 50250
     },
     {
       "epoch": 0.8165586261799158,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4296,
       "step": 50260
     },
     {
       "epoch": 0.8167210930772855,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.4198,
       "step": 50270
     },
     {
       "epoch": 0.8168835599746551,
-      "grad_norm": 2.0,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.3675,
       "step": 50280
     },
     {
       "epoch": 0.8170460268720249,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3802,
+      "loss": 1.3723,
       "step": 50290
     },
     {
       "epoch": 0.8172084937693945,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.3987,
       "step": 50300
     },
     {
       "epoch": 0.8173709606667642,
-      "grad_norm": 2.125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4444,
       "step": 50310
     },
     {
       "epoch": 0.8175334275641338,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3816,
+      "loss": 1.433,
       "step": 50320
     },
     {
       "epoch": 0.8176958944615035,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.4101,
       "step": 50330
     },
     {
       "epoch": 0.8178583613588731,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.4196,
       "step": 50340
     },
     {
       "epoch": 0.8180208282562428,
-      "grad_norm": 4.65625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4245,
+      "loss": 1.4297,
       "step": 50350
     },
     {
       "epoch": 0.8181832951536124,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3811,
+      "loss": 1.3917,
       "step": 50360
     },
     {
       "epoch": 0.8183457620509821,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.3883,
       "step": 50370
     },
     {
       "epoch": 0.8185082289483517,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3706,
+      "loss": 1.4442,
       "step": 50380
     },
     {
       "epoch": 0.8186706958457214,
-      "grad_norm": 2.125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.3806,
       "step": 50390
     },
     {
       "epoch": 0.8188331627430911,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.3637,
       "step": 50400
     },
     {
       "epoch": 0.8189956296404608,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.3915,
       "step": 50410
     },
     {
       "epoch": 0.8191580965378304,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.3808,
       "step": 50420
     },
     {
       "epoch": 0.8193205634352001,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4029,
+      "loss": 1.3954,
       "step": 50430
     },
     {
       "epoch": 0.8194830303325698,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.3638,
       "step": 50440
     },
     {
       "epoch": 0.8196454972299394,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.4192,
       "step": 50450
     },
     {
       "epoch": 0.819807964127309,
-      "grad_norm": 1.7578125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4083,
       "step": 50460
     },
     {
       "epoch": 0.8199704310246787,
-      "grad_norm": 2.375,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3822,
+      "loss": 1.3989,
       "step": 50470
     },
     {
       "epoch": 0.8201328979220484,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.4197,
       "step": 50480
     },
     {
       "epoch": 0.820295364819418,
-      "grad_norm": 1.640625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4173,
       "step": 50490
     },
     {
       "epoch": 0.8204578317167877,
-      "grad_norm": 2.25,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.382,
       "step": 50500
     },
     {
       "epoch": 0.8206202986141574,
-      "grad_norm": 2.5,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4335,
       "step": 50510
     },
     {
       "epoch": 0.8207827655115271,
-      "grad_norm": 2.625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4148,
       "step": 50520
     },
     {
       "epoch": 0.8209452324088967,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.3953,
       "step": 50530
     },
     {
       "epoch": 0.8211076993062664,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.359,
       "step": 50540
     },
     {
       "epoch": 0.821270166203636,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.3768,
       "step": 50550
     },
     {
       "epoch": 0.8214326331010057,
-      "grad_norm": 1.9609375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.4185,
       "step": 50560
     },
     {
       "epoch": 0.8215950999983753,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.4409,
       "step": 50570
     },
     {
       "epoch": 0.821757566895745,
-      "grad_norm": 2.25,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.3313,
       "step": 50580
     },
     {
       "epoch": 0.8219200337931146,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3775,
+      "loss": 1.3703,
       "step": 50590
     },
     {
       "epoch": 0.8220825006904843,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.4657,
       "step": 50600
     },
     {
       "epoch": 0.8222449675878539,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3842,
+      "loss": 1.4106,
       "step": 50610
     },
     {
       "epoch": 0.8224074344852237,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.3933,
       "step": 50620
     },
     {
       "epoch": 0.8225699013825933,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.461,
       "step": 50630
     },
     {
       "epoch": 0.822732368279963,
-      "grad_norm": 2.125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.4055,
       "step": 50640
     },
     {
       "epoch": 0.8228948351773326,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3902,
+      "loss": 1.3914,
       "step": 50650
     },
     {
       "epoch": 0.8230573020747023,
-      "grad_norm": 2.296875,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.3857,
       "step": 50660
     },
     {
       "epoch": 0.8232197689720719,
-      "grad_norm": 3.21875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4076,
+      "loss": 1.4035,
       "step": 50670
     },
     {
       "epoch": 0.8233822358694416,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3819,
+      "loss": 1.3573,
       "step": 50680
     },
     {
       "epoch": 0.8235447027668112,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3791,
+      "loss": 1.3729,
       "step": 50690
     },
     {
       "epoch": 0.8237071696641809,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.3375,
       "step": 50700
     },
     {
       "epoch": 0.8238696365615505,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4057,
       "step": 50710
     },
     {
       "epoch": 0.8240321034589202,
-      "grad_norm": 1.640625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4079,
+      "loss": 1.4091,
       "step": 50720
     },
     {
       "epoch": 0.82419457035629,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4227,
       "step": 50730
     },
     {
       "epoch": 0.8243570372536596,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.4095,
       "step": 50740
     },
     {
       "epoch": 0.8245195041510293,
-      "grad_norm": 1.734375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4085,
+      "loss": 1.4173,
       "step": 50750
     },
     {
       "epoch": 0.8246819710483989,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4125,
+      "loss": 1.3872,
       "step": 50760
     },
     {
       "epoch": 0.8248444379457686,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3841,
       "step": 50770
     },
     {
       "epoch": 0.8250069048431382,
-      "grad_norm": 1.703125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.3846,
       "step": 50780
     },
     {
       "epoch": 0.8251693717405079,
-      "grad_norm": 2.5,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4216,
+      "loss": 1.4151,
       "step": 50790
     },
     {
       "epoch": 0.8253318386378775,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.4509,
       "step": 50800
     },
     {
       "epoch": 0.8254943055352472,
-      "grad_norm": 1.75,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.407,
       "step": 50810
     },
     {
       "epoch": 0.8256567724326168,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4141,
       "step": 50820
     },
     {
       "epoch": 0.8258192393299865,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3655,
+      "loss": 1.4371,
       "step": 50830
     },
     {
       "epoch": 0.8259817062273562,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4142,
+      "loss": 1.4135,
       "step": 50840
     },
     {
       "epoch": 0.8261441731247259,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3766,
+      "loss": 1.3741,
       "step": 50850
     },
     {
       "epoch": 0.8263066400220955,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3893,
       "step": 50860
     },
     {
       "epoch": 0.8264691069194652,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3804,
+      "loss": 1.4014,
       "step": 50870
     },
     {
       "epoch": 0.8266315738168348,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.395,
       "step": 50880
     },
     {
       "epoch": 0.8267940407142045,
-      "grad_norm": 3.25,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4091,
       "step": 50890
     },
     {
       "epoch": 0.8269565076115741,
-      "grad_norm": 3.375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.4414,
       "step": 50900
     },
     {
       "epoch": 0.8271189745089438,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.383,
       "step": 50910
     },
     {
       "epoch": 0.8272814414063134,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.3702,
       "step": 50920
     },
     {
       "epoch": 0.8274439083036831,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4011,
       "step": 50930
     },
     {
       "epoch": 0.8276063752010527,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.423,
       "step": 50940
     },
     {
       "epoch": 0.8277688420984225,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.3772,
       "step": 50950
     },
     {
       "epoch": 0.8279313089957921,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.4102,
       "step": 50960
     },
     {
       "epoch": 0.8280937758931618,
-      "grad_norm": 3.0,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.3982,
       "step": 50970
     },
     {
       "epoch": 0.8282562427905314,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.3897,
       "step": 50980
     },
     {
       "epoch": 0.8284187096879011,
-      "grad_norm": 4.75,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3691,
+      "loss": 1.4135,
       "step": 50990
     },
     {
       "epoch": 0.8285811765852708,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3811,
+      "loss": 1.3964,
       "step": 51000
     },
     {
       "epoch": 0.8287436434826404,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4173,
+      "loss": 1.4594,
       "step": 51010
     },
     {
       "epoch": 0.82890611038001,
-      "grad_norm": 3.484375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3798,
+      "loss": 1.4232,
       "step": 51020
     },
     {
       "epoch": 0.8290685772773797,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4384,
       "step": 51030
     },
     {
       "epoch": 0.8292310441747494,
-      "grad_norm": 2.25,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.4004,
       "step": 51040
     },
     {
       "epoch": 0.829393511072119,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4606,
       "step": 51050
     },
     {
       "epoch": 0.8295559779694888,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3908,
+      "loss": 1.3782,
       "step": 51060
     },
     {
       "epoch": 0.8297184448668584,
-      "grad_norm": 2.125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4103,
       "step": 51070
     },
     {
       "epoch": 0.8298809117642281,
-      "grad_norm": 2.25,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.3735,
       "step": 51080
     },
     {
       "epoch": 0.8300433786615977,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.392,
       "step": 51090
     },
     {
       "epoch": 0.8302058455589674,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.4134,
       "step": 51100
     },
     {
       "epoch": 0.830368312456337,
-      "grad_norm": 1.734375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3664,
+      "loss": 1.4217,
       "step": 51110
     },
     {
       "epoch": 0.8305307793537067,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.3589,
       "step": 51120
     },
     {
       "epoch": 0.8306932462510763,
-      "grad_norm": 2.265625,
+      "grad_norm": 12.5,
       "learning_rate": 5e-05,
-      "loss": 0.4027,
+      "loss": 1.4237,
       "step": 51130
     },
     {
       "epoch": 0.830855713148446,
-      "grad_norm": 2.34375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.4323,
       "step": 51140
     },
     {
       "epoch": 0.8310181800458156,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.4286,
       "step": 51150
     },
     {
       "epoch": 0.8311806469431853,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.4149,
       "step": 51160
     },
     {
       "epoch": 0.831343113840555,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4245,
       "step": 51170
     },
     {
       "epoch": 0.8315055807379247,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.406,
       "step": 51180
     },
     {
       "epoch": 0.8316680476352943,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4039,
       "step": 51190
     },
     {
       "epoch": 0.831830514532664,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3958,
+      "loss": 1.4164,
       "step": 51200
     },
     {
       "epoch": 0.8319929814300336,
-      "grad_norm": 2.421875,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.395,
       "step": 51210
     },
     {
       "epoch": 0.8321554483274033,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3813,
       "step": 51220
     },
     {
       "epoch": 0.8323179152247729,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3684,
+      "loss": 1.4225,
       "step": 51230
     },
     {
       "epoch": 0.8324803821221426,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4091,
+      "loss": 1.4678,
       "step": 51240
     },
     {
       "epoch": 0.8326428490195122,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.3536,
       "step": 51250
     },
     {
       "epoch": 0.8328053159168819,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3692,
+      "loss": 1.4133,
       "step": 51260
     },
     {
       "epoch": 0.8329677828142515,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4148,
+      "loss": 1.4109,
       "step": 51270
     },
     {
       "epoch": 0.8331302497116213,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.3857,
       "step": 51280
     },
     {
       "epoch": 0.833292716608991,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4363,
       "step": 51290
     },
     {
       "epoch": 0.8334551835063606,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4165,
+      "loss": 1.4416,
       "step": 51300
     },
     {
       "epoch": 0.8336176504037303,
-      "grad_norm": 1.7421875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3983,
+      "loss": 1.4101,
       "step": 51310
     },
     {
       "epoch": 0.8337801173010999,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3705,
+      "loss": 1.4589,
       "step": 51320
     },
     {
       "epoch": 0.8339425841984696,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.3586,
       "step": 51330
     },
     {
       "epoch": 0.8341050510958392,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.3741,
       "step": 51340
     },
     {
       "epoch": 0.8342675179932089,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.376,
+      "loss": 1.3421,
       "step": 51350
     },
     {
       "epoch": 0.8344299848905785,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.3824,
       "step": 51360
     },
     {
       "epoch": 0.8345924517879482,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4071,
       "step": 51370
     },
     {
       "epoch": 0.8347549186853178,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3916,
+      "loss": 1.3962,
       "step": 51380
     },
     {
       "epoch": 0.8349173855826876,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.3831,
       "step": 51390
     },
     {
       "epoch": 0.8350798524800572,
-      "grad_norm": 1.8203125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.3982,
       "step": 51400
     },
     {
       "epoch": 0.8352423193774269,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.3755,
       "step": 51410
     },
     {
       "epoch": 0.8354047862747965,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4131,
       "step": 51420
     },
     {
       "epoch": 0.8355672531721662,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.4363,
       "step": 51430
     },
     {
       "epoch": 0.8357297200695358,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3808,
+      "loss": 1.3823,
       "step": 51440
     },
     {
       "epoch": 0.8358921869669055,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3996,
+      "loss": 1.3635,
       "step": 51450
     },
     {
       "epoch": 0.8360546538642751,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4061,
       "step": 51460
     },
     {
       "epoch": 0.8362171207616448,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.4115,
       "step": 51470
     },
     {
       "epoch": 0.8363795876590144,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.3734,
       "step": 51480
     },
     {
       "epoch": 0.8365420545563841,
-      "grad_norm": 1.75,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4042,
+      "loss": 1.4131,
       "step": 51490
     },
     {
       "epoch": 0.8367045214537538,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3702,
+      "loss": 1.3647,
       "step": 51500
     },
     {
       "epoch": 0.8368669883511235,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3842,
+      "loss": 1.4213,
       "step": 51510
     },
     {
       "epoch": 0.8370294552484931,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.3716,
       "step": 51520
     },
     {
       "epoch": 0.8371919221458628,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4083,
+      "loss": 1.4048,
       "step": 51530
     },
     {
       "epoch": 0.8373543890432324,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3982,
+      "loss": 1.3874,
       "step": 51540
     },
     {
       "epoch": 0.8375168559406021,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4291,
+      "loss": 1.3747,
       "step": 51550
     },
     {
       "epoch": 0.8376793228379718,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.4027,
       "step": 51560
     },
     {
       "epoch": 0.8378417897353414,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4186,
+      "loss": 1.41,
       "step": 51570
     },
     {
       "epoch": 0.838004256632711,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4461,
       "step": 51580
     },
     {
       "epoch": 0.8381667235300807,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.4089,
       "step": 51590
     },
     {
       "epoch": 0.8383291904274504,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.4188,
       "step": 51600
     },
     {
       "epoch": 0.8384916573248201,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3931,
+      "loss": 1.4216,
       "step": 51610
     },
     {
       "epoch": 0.8386541242221898,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3692,
+      "loss": 1.4078,
       "step": 51620
     },
     {
       "epoch": 0.8388165911195594,
-      "grad_norm": 1.7734375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3737,
+      "loss": 1.385,
       "step": 51630
     },
     {
       "epoch": 0.8389790580169291,
-      "grad_norm": 4.1875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.3762,
       "step": 51640
     },
     {
       "epoch": 0.8391415249142987,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.3901,
       "step": 51650
     },
     {
       "epoch": 0.8393039918116684,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.364,
       "step": 51660
     },
     {
       "epoch": 0.839466458709038,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.4396,
       "step": 51670
     },
     {
       "epoch": 0.8396289256064077,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.4192,
       "step": 51680
     },
     {
       "epoch": 0.8397913925037773,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4113,
       "step": 51690
     },
     {
       "epoch": 0.839953859401147,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.4317,
       "step": 51700
     },
     {
       "epoch": 0.8401163262985166,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.3803,
       "step": 51710
     },
     {
       "epoch": 0.8402787931958864,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.3892,
       "step": 51720
     },
     {
       "epoch": 0.840441260093256,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.393,
       "step": 51730
     },
     {
       "epoch": 0.8406037269906257,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.38,
+      "loss": 1.3912,
       "step": 51740
     },
     {
       "epoch": 0.8407661938879953,
-      "grad_norm": 1.796875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.3311,
       "step": 51750
     },
     {
       "epoch": 0.840928660785365,
-      "grad_norm": 3.125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3872,
+      "loss": 1.3855,
       "step": 51760
     },
     {
       "epoch": 0.8410911276827346,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.3781,
       "step": 51770
     },
     {
       "epoch": 0.8412535945801043,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4056,
+      "loss": 1.4129,
       "step": 51780
     },
     {
       "epoch": 0.8414160614774739,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4007,
       "step": 51790
     },
     {
       "epoch": 0.8415785283748436,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3766,
+      "loss": 1.4365,
       "step": 51800
     },
     {
       "epoch": 0.8417409952722132,
-      "grad_norm": 2.75,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.424,
       "step": 51810
     },
     {
       "epoch": 0.8419034621695829,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.4338,
       "step": 51820
     },
     {
       "epoch": 0.8420659290669527,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.4253,
       "step": 51830
     },
     {
       "epoch": 0.8422283959643223,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4333,
       "step": 51840
     },
     {
       "epoch": 0.842390862861692,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.3622,
       "step": 51850
     },
     {
       "epoch": 0.8425533297590616,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3978,
+      "loss": 1.4203,
       "step": 51860
     },
     {
       "epoch": 0.8427157966564313,
-      "grad_norm": 2.0,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.3947,
       "step": 51870
     },
     {
       "epoch": 0.8428782635538009,
-      "grad_norm": 4.71875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.4057,
       "step": 51880
     },
     {
       "epoch": 0.8430407304511706,
-      "grad_norm": 2.03125,
+      "grad_norm": 4.25,
       "learning_rate": 5e-05,
-      "loss": 0.3804,
+      "loss": 1.3751,
       "step": 51890
     },
     {
       "epoch": 0.8432031973485402,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3767,
+      "loss": 1.4488,
       "step": 51900
     },
     {
       "epoch": 0.8433656642459099,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4087,
+      "loss": 1.3987,
       "step": 51910
     },
     {
       "epoch": 0.8435281311432795,
-      "grad_norm": 1.625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.383,
       "step": 51920
     },
     {
       "epoch": 0.8436905980406492,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.4325,
       "step": 51930
     },
     {
       "epoch": 0.8438530649380189,
-      "grad_norm": 1.5234375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3735,
+      "loss": 1.4216,
       "step": 51940
     },
     {
       "epoch": 0.8440155318353886,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3808,
+      "loss": 1.3698,
       "step": 51950
     },
     {
       "epoch": 0.8441779987327582,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4128,
       "step": 51960
     },
     {
       "epoch": 0.8443404656301279,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.3941,
       "step": 51970
     },
     {
       "epoch": 0.8445029325274975,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4446,
       "step": 51980
     },
     {
       "epoch": 0.8446653994248672,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.3913,
       "step": 51990
     },
     {
       "epoch": 0.8448278663222368,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.4372,
       "step": 52000
     },
     {
       "epoch": 0.8449903332196065,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.3562,
       "step": 52010
     },
     {
       "epoch": 0.8451528001169761,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.3897,
       "step": 52020
     },
     {
       "epoch": 0.8453152670143458,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3834,
+      "loss": 1.3444,
       "step": 52030
     },
     {
       "epoch": 0.8454777339117154,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.3924,
       "step": 52040
     },
     {
       "epoch": 0.8456402008090852,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.3818,
       "step": 52050
     },
     {
       "epoch": 0.8458026677064548,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.4221,
       "step": 52060
     },
     {
       "epoch": 0.8459651346038245,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.3803,
       "step": 52070
     },
     {
       "epoch": 0.8461276015011941,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.3934,
       "step": 52080
     },
     {
       "epoch": 0.8462900683985638,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4104,
+      "loss": 1.3096,
       "step": 52090
     },
     {
       "epoch": 0.8464525352959335,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.4196,
       "step": 52100
     },
     {
       "epoch": 0.8466150021933031,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.414,
+      "loss": 1.3985,
       "step": 52110
     },
     {
       "epoch": 0.8467774690906728,
-      "grad_norm": 1.796875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.3717,
       "step": 52120
     },
     {
       "epoch": 0.8469399359880424,
-      "grad_norm": 2.125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.438,
       "step": 52130
     },
     {
       "epoch": 0.847102402885412,
-      "grad_norm": 1.765625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4177,
       "step": 52140
     },
     {
       "epoch": 0.8472648697827817,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4171,
+      "loss": 1.4323,
       "step": 52150
     },
     {
       "epoch": 0.8474273366801515,
-      "grad_norm": 1.6953125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.3644,
       "step": 52160
     },
     {
       "epoch": 0.8475898035775211,
-      "grad_norm": 3.625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3805,
+      "loss": 1.3864,
       "step": 52170
     },
     {
       "epoch": 0.8477522704748908,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.3949,
       "step": 52180
     },
     {
       "epoch": 0.8479147373722604,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.4475,
       "step": 52190
     },
     {
       "epoch": 0.8480772042696301,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.3726,
       "step": 52200
     },
     {
       "epoch": 0.8482396711669997,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.4238,
       "step": 52210
     },
     {
       "epoch": 0.8484021380643694,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.3832,
       "step": 52220
     },
     {
       "epoch": 0.848564604961739,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3687,
       "step": 52230
     },
     {
       "epoch": 0.8487270718591087,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.3921,
       "step": 52240
     },
     {
       "epoch": 0.8488895387564783,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4006,
       "step": 52250
     },
     {
       "epoch": 0.849052005653848,
-      "grad_norm": 2.0,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.3699,
       "step": 52260
     },
     {
       "epoch": 0.8492144725512177,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.3769,
       "step": 52270
     },
     {
       "epoch": 0.8493769394485874,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.3551,
       "step": 52280
     },
     {
       "epoch": 0.849539406345957,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3825,
+      "loss": 1.3873,
       "step": 52290
     },
     {
       "epoch": 0.8497018732433267,
-      "grad_norm": 1.8203125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.43,
       "step": 52300
     },
     {
       "epoch": 0.8498643401406963,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.389,
       "step": 52310
     },
     {
       "epoch": 0.850026807038066,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3767,
+      "loss": 1.4294,
       "step": 52320
     },
     {
       "epoch": 0.8501892739354356,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4075,
       "step": 52330
     },
     {
       "epoch": 0.8503517408328053,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3627,
+      "loss": 1.3581,
       "step": 52340
     },
     {
       "epoch": 0.8505142077301749,
-      "grad_norm": 1.8828125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3788,
+      "loss": 1.4197,
       "step": 52350
     },
     {
       "epoch": 0.8506766746275446,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3845,
+      "loss": 1.4147,
       "step": 52360
     },
     {
       "epoch": 0.8508391415249142,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.3563,
       "step": 52370
     },
     {
       "epoch": 0.851001608422284,
-      "grad_norm": 1.7109375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4433,
       "step": 52380
     },
     {
       "epoch": 0.8511640753196537,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4014,
       "step": 52390
     },
     {
       "epoch": 0.8513265422170233,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.4327,
       "step": 52400
     },
     {
       "epoch": 0.851489009114393,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.3782,
       "step": 52410
     },
     {
       "epoch": 0.8516514760117626,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.4187,
       "step": 52420
     },
     {
       "epoch": 0.8518139429091323,
-      "grad_norm": 1.734375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.384,
+      "loss": 1.415,
       "step": 52430
     },
     {
       "epoch": 0.8519764098065019,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.3709,
       "step": 52440
     },
     {
       "epoch": 0.8521388767038716,
-      "grad_norm": 2.5,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.3832,
       "step": 52450
     },
     {
       "epoch": 0.8523013436012412,
-      "grad_norm": 3.109375,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3697,
+      "loss": 1.4758,
       "step": 52460
     },
     {
       "epoch": 0.8524638104986109,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.3929,
       "step": 52470
     },
     {
       "epoch": 0.8526262773959805,
-      "grad_norm": 2.234375,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.3937,
       "step": 52480
     },
     {
       "epoch": 0.8527887442933503,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.4087,
       "step": 52490
     },
     {
       "epoch": 0.8529512111907199,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3815,
+      "loss": 1.4064,
       "step": 52500
     },
     {
       "epoch": 0.8531136780880896,
-      "grad_norm": 1.8984375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.4005,
       "step": 52510
     },
     {
       "epoch": 0.8532761449854592,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.3905,
       "step": 52520
     },
     {
       "epoch": 0.8534386118828289,
-      "grad_norm": 2.0,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.3909,
       "step": 52530
     },
     {
       "epoch": 0.8536010787801985,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.3546,
       "step": 52540
     },
     {
       "epoch": 0.8537635456775682,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.3921,
       "step": 52550
     },
     {
       "epoch": 0.8539260125749378,
-      "grad_norm": 1.765625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.3799,
       "step": 52560
     },
     {
       "epoch": 0.8540884794723075,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.3956,
       "step": 52570
     },
     {
       "epoch": 0.8542509463696771,
-      "grad_norm": 2.078125,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.3953,
       "step": 52580
     },
     {
       "epoch": 0.8544134132670468,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4241,
+      "loss": 1.3878,
       "step": 52590
     },
     {
       "epoch": 0.8545758801644165,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.3714,
       "step": 52600
     },
     {
       "epoch": 0.8547383470617862,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4089,
       "step": 52610
     },
     {
       "epoch": 0.8549008139591558,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.375,
+      "loss": 1.3582,
       "step": 52620
     },
     {
       "epoch": 0.8550632808565255,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4204,
+      "loss": 1.3998,
       "step": 52630
     },
     {
       "epoch": 0.8552257477538951,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4101,
+      "loss": 1.4739,
       "step": 52640
     },
     {
       "epoch": 0.8553882146512648,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.4002,
       "step": 52650
     },
     {
       "epoch": 0.8555506815486345,
-      "grad_norm": 1.6328125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.3902,
       "step": 52660
     },
     {
       "epoch": 0.8557131484460041,
-      "grad_norm": 2.375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4223,
+      "loss": 1.4443,
       "step": 52670
     },
     {
       "epoch": 0.8558756153433738,
-      "grad_norm": 2.75,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.444,
       "step": 52680
     },
     {
       "epoch": 0.8560380822407434,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3967,
+      "loss": 1.4147,
       "step": 52690
     },
     {
       "epoch": 0.856200549138113,
-      "grad_norm": 2.625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.4177,
       "step": 52700
     },
     {
       "epoch": 0.8563630160354828,
-      "grad_norm": 3.0,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.3879,
       "step": 52710
     },
     {
       "epoch": 0.8565254829328525,
-      "grad_norm": 2.265625,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.4001,
       "step": 52720
     },
     {
       "epoch": 0.8566879498302221,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3753,
+      "loss": 1.4169,
       "step": 52730
     },
     {
       "epoch": 0.8568504167275918,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3886,
+      "loss": 1.3905,
       "step": 52740
     },
     {
       "epoch": 0.8570128836249614,
-      "grad_norm": 2.015625,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4086,
       "step": 52750
     },
     {
       "epoch": 0.8571753505223311,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.37,
+      "loss": 1.412,
       "step": 52760
     },
     {
       "epoch": 0.8573378174197007,
-      "grad_norm": 1.8203125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.4107,
       "step": 52770
     },
     {
       "epoch": 0.8575002843170704,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4127,
       "step": 52780
     },
     {
       "epoch": 0.85766275121444,
-      "grad_norm": 1.640625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.4131,
       "step": 52790
     },
     {
       "epoch": 0.8578252181118097,
-      "grad_norm": 2.5,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4042,
       "step": 52800
     },
     {
       "epoch": 0.8579876850091794,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.4273,
       "step": 52810
     },
     {
       "epoch": 0.8581501519065491,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.4326,
       "step": 52820
     },
     {
       "epoch": 0.8583126188039187,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.395,
       "step": 52830
     },
     {
       "epoch": 0.8584750857012884,
-      "grad_norm": 3.25,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4208,
+      "loss": 1.392,
       "step": 52840
     },
     {
       "epoch": 0.858637552598658,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.3702,
       "step": 52850
     },
     {
       "epoch": 0.8588000194960277,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.3966,
       "step": 52860
     },
     {
       "epoch": 0.8589624863933973,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4087,
       "step": 52870
     },
     {
       "epoch": 0.859124953290767,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3678,
+      "loss": 1.3992,
       "step": 52880
     },
     {
       "epoch": 0.8592874201881366,
-      "grad_norm": 1.421875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.404,
+      "loss": 1.4146,
       "step": 52890
     },
     {
       "epoch": 0.8594498870855063,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4121,
+      "loss": 1.389,
       "step": 52900
     },
     {
       "epoch": 0.8596123539828759,
-      "grad_norm": 4.21875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4152,
       "step": 52910
     },
     {
       "epoch": 0.8597748208802457,
-      "grad_norm": 3.0,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.3899,
       "step": 52920
     },
     {
       "epoch": 0.8599372877776154,
-      "grad_norm": 2.25,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4058,
       "step": 52930
     },
     {
       "epoch": 0.860099754674985,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.4318,
       "step": 52940
     },
     {
       "epoch": 0.8602622215723547,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.3533,
       "step": 52950
     },
     {
       "epoch": 0.8604246884697243,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.3626,
       "step": 52960
     },
     {
       "epoch": 0.860587155367094,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.4063,
       "step": 52970
     },
     {
       "epoch": 0.8607496222644636,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.4227,
       "step": 52980
     },
     {
       "epoch": 0.8609120891618333,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.398,
       "step": 52990
     },
     {
       "epoch": 0.8610745560592029,
-      "grad_norm": 1.7265625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3889,
+      "loss": 1.4037,
       "step": 53000
     },
     {
       "epoch": 0.8612370229565726,
-      "grad_norm": 2.953125,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4256,
       "step": 53010
     },
     {
       "epoch": 0.8613994898539422,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.3393,
       "step": 53020
     },
     {
       "epoch": 0.861561956751312,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3805,
+      "loss": 1.4379,
       "step": 53030
     },
     {
       "epoch": 0.8617244236486816,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3769,
+      "loss": 1.4905,
       "step": 53040
     },
     {
       "epoch": 0.8618868905460513,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3856,
       "step": 53050
     },
     {
       "epoch": 0.8620493574434209,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.3665,
       "step": 53060
     },
     {
       "epoch": 0.8622118243407906,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.4006,
+      "loss": 1.3825,
       "step": 53070
     },
     {
       "epoch": 0.8623742912381602,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.3846,
       "step": 53080
     },
     {
       "epoch": 0.8625367581355299,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3839,
       "step": 53090
     },
     {
       "epoch": 0.8626992250328995,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.4422,
       "step": 53100
     },
     {
       "epoch": 0.8628616919302692,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.369,
       "step": 53110
     },
     {
       "epoch": 0.8630241588276388,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3841,
+      "loss": 1.3417,
       "step": 53120
     },
     {
       "epoch": 0.8631866257250085,
-      "grad_norm": 2.5,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4109,
       "step": 53130
     },
     {
       "epoch": 0.8633490926223782,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.403,
       "step": 53140
     },
     {
       "epoch": 0.8635115595197479,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4029,
+      "loss": 1.371,
       "step": 53150
     },
     {
       "epoch": 0.8636740264171175,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3869,
+      "loss": 1.397,
       "step": 53160
     },
     {
       "epoch": 0.8638364933144872,
-      "grad_norm": 2.375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4109,
+      "loss": 1.3932,
       "step": 53170
     },
     {
       "epoch": 0.8639989602118568,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.4162,
       "step": 53180
     },
     {
       "epoch": 0.8641614271092265,
-      "grad_norm": 2.515625,
+      "grad_norm": 4.375,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3988,
       "step": 53190
     },
     {
       "epoch": 0.8643238940065961,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.3275,
       "step": 53200
     },
     {
       "epoch": 0.8644863609039658,
-      "grad_norm": 2.25,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.3876,
       "step": 53210
     },
     {
       "epoch": 0.8646488278013355,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.3723,
       "step": 53220
     },
     {
       "epoch": 0.8648112946987051,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.3927,
       "step": 53230
     },
     {
       "epoch": 0.8649737615960748,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.3862,
       "step": 53240
     },
     {
       "epoch": 0.8651362284934445,
-      "grad_norm": 4.9375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3845,
+      "loss": 1.3748,
       "step": 53250
     },
     {
       "epoch": 0.8652986953908142,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.402,
+      "loss": 1.4029,
       "step": 53260
     },
     {
       "epoch": 0.8654611622881838,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.3685,
       "step": 53270
     },
     {
       "epoch": 0.8656236291855535,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.3678,
       "step": 53280
     },
     {
       "epoch": 0.8657860960829231,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3972,
+      "loss": 1.375,
       "step": 53290
     },
     {
       "epoch": 0.8659485629802928,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4005,
+      "loss": 1.4093,
       "step": 53300
     },
     {
       "epoch": 0.8661110298776624,
-      "grad_norm": 2.125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4112,
+      "loss": 1.392,
       "step": 53310
     },
     {
       "epoch": 0.8662734967750321,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.4628,
       "step": 53320
     },
     {
       "epoch": 0.8664359636724017,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4122,
+      "loss": 1.4231,
       "step": 53330
     },
     {
       "epoch": 0.8665984305697714,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.4175,
       "step": 53340
     },
     {
       "epoch": 0.866760897467141,
-      "grad_norm": 3.21875,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.3515,
       "step": 53350
     },
     {
       "epoch": 0.8669233643645108,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.4028,
       "step": 53360
     },
     {
       "epoch": 0.8670858312618804,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4347,
       "step": 53370
     },
     {
       "epoch": 0.8672482981592501,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.4081,
+      "loss": 1.371,
       "step": 53380
     },
     {
       "epoch": 0.8674107650566197,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.3778,
       "step": 53390
     },
     {
       "epoch": 0.8675732319539894,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4205,
+      "loss": 1.4308,
       "step": 53400
     },
     {
       "epoch": 0.867735698851359,
-      "grad_norm": 3.21875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.4211,
       "step": 53410
     },
     {
       "epoch": 0.8678981657487287,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.3935,
       "step": 53420
     },
     {
       "epoch": 0.8680606326460983,
-      "grad_norm": 2.375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.4219,
       "step": 53430
     },
     {
       "epoch": 0.868223099543468,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4102,
+      "loss": 1.413,
       "step": 53440
     },
     {
       "epoch": 0.8683855664408376,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3643,
+      "loss": 1.3886,
       "step": 53450
     },
     {
       "epoch": 0.8685480333382073,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3828,
+      "loss": 1.4689,
       "step": 53460
     },
     {
       "epoch": 0.868710500235577,
-      "grad_norm": 2.5,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.4085,
       "step": 53470
     },
     {
       "epoch": 0.8688729671329467,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3743,
+      "loss": 1.4037,
       "step": 53480
     },
     {
       "epoch": 0.8690354340303164,
-      "grad_norm": 2.625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3743,
+      "loss": 1.3438,
       "step": 53490
     },
     {
       "epoch": 0.869197900927686,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.3804,
       "step": 53500
     },
     {
       "epoch": 0.8693603678250557,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3853,
+      "loss": 1.3858,
       "step": 53510
     },
     {
       "epoch": 0.8695228347224253,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4389,
       "step": 53520
     },
     {
       "epoch": 0.869685301619795,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.4674,
       "step": 53530
     },
     {
       "epoch": 0.8698477685171646,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3995,
+      "loss": 1.4392,
       "step": 53540
     },
     {
       "epoch": 0.8700102354145343,
-      "grad_norm": 1.9765625,
+      "grad_norm": 4.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.4085,
       "step": 53550
     },
     {
       "epoch": 0.8701727023119039,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.3825,
       "step": 53560
     },
     {
       "epoch": 0.8703351692092736,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3792,
+      "loss": 1.4492,
       "step": 53570
     },
     {
       "epoch": 0.8704976361066433,
-      "grad_norm": 3.875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4426,
       "step": 53580
     },
     {
       "epoch": 0.870660103004013,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4148,
+      "loss": 1.3547,
       "step": 53590
     },
     {
       "epoch": 0.8708225699013826,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4,
+      "loss": 1.4116,
       "step": 53600
     },
     {
       "epoch": 0.8709850367987523,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.4122,
       "step": 53610
     },
     {
       "epoch": 0.8711475036961219,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.3699,
       "step": 53620
     },
     {
       "epoch": 0.8713099705934916,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3986,
       "step": 53630
     },
     {
       "epoch": 0.8714724374908612,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3837,
+      "loss": 1.424,
       "step": 53640
     },
     {
       "epoch": 0.8716349043882309,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4437,
       "step": 53650
     },
     {
       "epoch": 0.8717973712856005,
-      "grad_norm": 2.234375,
+      "grad_norm": 4.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3691,
+      "loss": 1.407,
       "step": 53660
     },
     {
       "epoch": 0.8719598381829702,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.3958,
       "step": 53670
     },
     {
       "epoch": 0.8721223050803398,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.3477,
       "step": 53680
     },
     {
       "epoch": 0.8722847719777096,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.3805,
       "step": 53690
     },
     {
       "epoch": 0.8724472388750792,
-      "grad_norm": 1.7265625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3687,
+      "loss": 1.4358,
       "step": 53700
     },
     {
       "epoch": 0.8726097057724489,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3322,
       "step": 53710
     },
     {
       "epoch": 0.8727721726698185,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4102,
       "step": 53720
     },
     {
       "epoch": 0.8729346395671882,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4213,
+      "loss": 1.3967,
       "step": 53730
     },
     {
       "epoch": 0.8730971064645578,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4041,
       "step": 53740
     },
     {
       "epoch": 0.8732595733619275,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.3534,
       "step": 53750
     },
     {
       "epoch": 0.8734220402592971,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.4063,
       "step": 53760
     },
     {
       "epoch": 0.8735845071566668,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.3862,
       "step": 53770
     },
     {
       "epoch": 0.8737469740540365,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.4129,
       "step": 53780
     },
     {
       "epoch": 0.8739094409514061,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3881,
+      "loss": 1.4208,
       "step": 53790
     },
     {
       "epoch": 0.8740719078487759,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.384,
+      "loss": 1.4064,
       "step": 53800
     },
     {
       "epoch": 0.8742343747461455,
-      "grad_norm": 3.25,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3976,
+      "loss": 1.4081,
       "step": 53810
     },
     {
       "epoch": 0.8743968416435152,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.4097,
       "step": 53820
     },
     {
       "epoch": 0.8745593085408848,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.3997,
       "step": 53830
     },
     {
       "epoch": 0.8747217754382545,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3779,
+      "loss": 1.3782,
       "step": 53840
     },
     {
       "epoch": 0.8748842423356241,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4098,
+      "loss": 1.4546,
       "step": 53850
     },
     {
       "epoch": 0.8750467092329938,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.3823,
       "step": 53860
     },
     {
       "epoch": 0.8752091761303634,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.3849,
       "step": 53870
     },
     {
       "epoch": 0.8753716430277331,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.3974,
       "step": 53880
     },
     {
       "epoch": 0.8755341099251027,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3873,
+      "loss": 1.3924,
       "step": 53890
     },
     {
       "epoch": 0.8756965768224724,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4048,
+      "loss": 1.4114,
       "step": 53900
     },
     {
       "epoch": 0.8758590437198421,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3812,
+      "loss": 1.3759,
       "step": 53910
     },
     {
       "epoch": 0.8760215106172118,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.4105,
       "step": 53920
     },
     {
       "epoch": 0.8761839775145814,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.427,
       "step": 53930
     },
     {
       "epoch": 0.8763464444119511,
-      "grad_norm": 2.578125,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4029,
+      "loss": 1.3937,
       "step": 53940
     },
     {
       "epoch": 0.8765089113093207,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3797,
+      "loss": 1.4649,
       "step": 53950
     },
     {
       "epoch": 0.8766713782066904,
-      "grad_norm": 1.734375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.3925,
       "step": 53960
     },
     {
       "epoch": 0.87683384510406,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.3669,
       "step": 53970
     },
     {
       "epoch": 0.8769963120014297,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3841,
+      "loss": 1.424,
       "step": 53980
     },
     {
       "epoch": 0.8771587788987993,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.4033,
       "step": 53990
     },
     {
       "epoch": 0.877321245796169,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3883,
+      "loss": 1.3599,
       "step": 54000
     },
     {
       "epoch": 0.8774837126935386,
-      "grad_norm": 1.890625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3735,
+      "loss": 1.3989,
       "step": 54010
     },
     {
       "epoch": 0.8776461795909084,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3604,
       "step": 54020
     },
     {
       "epoch": 0.877808646488278,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.3741,
       "step": 54030
     },
     {
       "epoch": 0.8779711133856477,
-      "grad_norm": 1.84375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.4197,
       "step": 54040
     },
     {
       "epoch": 0.8781335802830174,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3985,
+      "loss": 1.4312,
       "step": 54050
     },
     {
       "epoch": 0.878296047180387,
-      "grad_norm": 1.515625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4002,
+      "loss": 1.3703,
       "step": 54060
     },
     {
       "epoch": 0.8784585140777567,
-      "grad_norm": 2.75,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.3859,
       "step": 54070
     },
     {
       "epoch": 0.8786209809751263,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.3874,
       "step": 54080
     },
     {
       "epoch": 0.878783447872496,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3805,
       "step": 54090
     },
     {
       "epoch": 0.8789459147698656,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4235,
+      "loss": 1.4033,
       "step": 54100
     },
     {
       "epoch": 0.8791083816672353,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4053,
+      "loss": 1.4193,
       "step": 54110
     },
     {
       "epoch": 0.8792708485646049,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4052,
+      "loss": 1.3896,
       "step": 54120
     },
     {
       "epoch": 0.8794333154619747,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3965,
+      "loss": 1.3625,
       "step": 54130
     },
     {
       "epoch": 0.8795957823593443,
-      "grad_norm": 3.125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.3899,
       "step": 54140
     },
     {
       "epoch": 0.879758249256714,
-      "grad_norm": 1.609375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3964,
+      "loss": 1.4073,
       "step": 54150
     },
     {
       "epoch": 0.8799207161540836,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3678,
+      "loss": 1.3758,
       "step": 54160
     },
     {
       "epoch": 0.8800831830514533,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.3802,
       "step": 54170
     },
     {
       "epoch": 0.8802456499488229,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.3653,
       "step": 54180
     },
     {
       "epoch": 0.8804081168461926,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.3807,
       "step": 54190
     },
     {
       "epoch": 0.8805705837435622,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3896,
+      "loss": 1.3297,
       "step": 54200
     },
     {
       "epoch": 0.8807330506409319,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.4129,
+      "loss": 1.385,
       "step": 54210
     },
     {
       "epoch": 0.8808955175383015,
-      "grad_norm": 1.671875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3791,
+      "loss": 1.3503,
       "step": 54220
     },
     {
       "epoch": 0.8810579844356712,
-      "grad_norm": 2.5,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.3872,
       "step": 54230
     },
     {
       "epoch": 0.8812204513330409,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3759,
+      "loss": 1.394,
       "step": 54240
     },
     {
       "epoch": 0.8813829182304106,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.3582,
       "step": 54250
     },
     {
       "epoch": 0.8815453851277802,
-      "grad_norm": 2.046875,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4089,
+      "loss": 1.3574,
       "step": 54260
     },
     {
       "epoch": 0.8817078520251499,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3777,
+      "loss": 1.3408,
       "step": 54270
     },
     {
       "epoch": 0.8818703189225195,
-      "grad_norm": 2.578125,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.3357,
       "step": 54280
     },
     {
       "epoch": 0.8820327858198892,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.378,
       "step": 54290
     },
     {
       "epoch": 0.8821952527172588,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.4337,
       "step": 54300
     },
     {
       "epoch": 0.8823577196146285,
-      "grad_norm": 1.8046875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.4297,
       "step": 54310
     },
     {
       "epoch": 0.8825201865119982,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.4279,
       "step": 54320
     },
     {
       "epoch": 0.8826826534093678,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.4159,
       "step": 54330
     },
     {
       "epoch": 0.8828451203067375,
-      "grad_norm": 1.640625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3915,
       "step": 54340
     },
     {
       "epoch": 0.8830075872041072,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3823,
+      "loss": 1.4163,
       "step": 54350
     },
     {
       "epoch": 0.8831700541014769,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3858,
+      "loss": 1.3973,
       "step": 54360
     },
     {
       "epoch": 0.8833325209988465,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3785,
+      "loss": 1.3786,
       "step": 54370
     },
     {
       "epoch": 0.8834949878962162,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.3816,
       "step": 54380
     },
     {
       "epoch": 0.8836574547935858,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.4145,
       "step": 54390
     },
     {
       "epoch": 0.8838199216909555,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3904,
+      "loss": 1.3765,
       "step": 54400
     },
     {
       "epoch": 0.8839823885883251,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4008,
+      "loss": 1.3391,
       "step": 54410
     },
     {
       "epoch": 0.8841448554856948,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3728,
+      "loss": 1.42,
       "step": 54420
     },
     {
       "epoch": 0.8843073223830644,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.3736,
       "step": 54430
     },
     {
       "epoch": 0.8844697892804341,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3883,
+      "loss": 1.3749,
       "step": 54440
     },
     {
       "epoch": 0.8846322561778037,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.3755,
       "step": 54450
     },
     {
       "epoch": 0.8847947230751735,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.392,
+      "loss": 1.4016,
       "step": 54460
     },
     {
       "epoch": 0.8849571899725431,
-      "grad_norm": 3.25,
+      "grad_norm": 4.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4222,
       "step": 54470
     },
     {
       "epoch": 0.8851196568699128,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3745,
+      "loss": 1.4194,
       "step": 54480
     },
     {
       "epoch": 0.8852821237672824,
-      "grad_norm": 1.875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.3936,
       "step": 54490
     },
     {
       "epoch": 0.8854445906646521,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.4198,
       "step": 54500
     },
     {
       "epoch": 0.8856070575620217,
-      "grad_norm": 1.890625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.3904,
       "step": 54510
     },
     {
       "epoch": 0.8857695244593914,
-      "grad_norm": 1.9609375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.3986,
       "step": 54520
     },
     {
       "epoch": 0.885931991356761,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.384,
       "step": 54530
     },
     {
       "epoch": 0.8860944582541307,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.3794,
       "step": 54540
     },
     {
       "epoch": 0.8862569251515003,
-      "grad_norm": 1.8359375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3841,
+      "loss": 1.4366,
       "step": 54550
     },
     {
       "epoch": 0.88641939204887,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4283,
       "step": 54560
     },
     {
       "epoch": 0.8865818589462398,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4168,
       "step": 54570
     },
     {
       "epoch": 0.8867443258436094,
-      "grad_norm": 1.6171875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.381,
+      "loss": 1.3679,
       "step": 54580
     },
     {
       "epoch": 0.886906792740979,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3734,
+      "loss": 1.3932,
       "step": 54590
     },
     {
       "epoch": 0.8870692596383487,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3929,
       "step": 54600
     },
     {
       "epoch": 0.8872317265357184,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.3796,
       "step": 54610
     },
     {
       "epoch": 0.887394193433088,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4044,
       "step": 54620
     },
     {
       "epoch": 0.8875566603304577,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3776,
+      "loss": 1.3896,
       "step": 54630
     },
     {
       "epoch": 0.8877191272278273,
-      "grad_norm": 2.75,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.3678,
       "step": 54640
     },
     {
       "epoch": 0.887881594125197,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4193,
       "step": 54650
     },
     {
       "epoch": 0.8880440610225666,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3787,
+      "loss": 1.3564,
       "step": 54660
     },
     {
       "epoch": 0.8882065279199363,
-      "grad_norm": 1.7734375,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.4151,
+      "loss": 1.4286,
       "step": 54670
     },
     {
       "epoch": 0.888368994817306,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.4003,
       "step": 54680
     },
     {
       "epoch": 0.8885314617146757,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3717,
+      "loss": 1.3951,
       "step": 54690
     },
     {
       "epoch": 0.8886939286120453,
-      "grad_norm": 2.375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3778,
+      "loss": 1.3916,
       "step": 54700
     },
     {
       "epoch": 0.888856395509415,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.3304,
       "step": 54710
     },
     {
       "epoch": 0.8890188624067846,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.3918,
       "step": 54720
     },
     {
       "epoch": 0.8891813293041543,
-      "grad_norm": 2.09375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3686,
       "step": 54730
     },
     {
       "epoch": 0.8893437962015239,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.3732,
       "step": 54740
     },
     {
       "epoch": 0.8895062630988936,
-      "grad_norm": 2.125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.3647,
       "step": 54750
     },
     {
       "epoch": 0.8896687299962632,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.35,
       "step": 54760
     },
     {
       "epoch": 0.8898311968936329,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.374,
+      "loss": 1.3899,
       "step": 54770
     },
     {
       "epoch": 0.8899936637910025,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4051,
+      "loss": 1.3841,
       "step": 54780
     },
     {
       "epoch": 0.8901561306883723,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3788,
+      "loss": 1.4066,
       "step": 54790
     },
     {
       "epoch": 0.8903185975857419,
-      "grad_norm": 2.25,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3957,
       "step": 54800
     },
     {
       "epoch": 0.8904810644831116,
-      "grad_norm": 2.15625,
+      "grad_norm": 4.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3739,
+      "loss": 1.3704,
       "step": 54810
     },
     {
       "epoch": 0.8906435313804812,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.379,
       "step": 54820
     },
     {
       "epoch": 0.8908059982778509,
-      "grad_norm": 3.78125,
+      "grad_norm": 4.59375,
       "learning_rate": 5e-05,
-      "loss": 0.403,
+      "loss": 1.3903,
       "step": 54830
     },
     {
       "epoch": 0.8909684651752205,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4199,
       "step": 54840
     },
     {
       "epoch": 0.8911309320725902,
-      "grad_norm": 1.59375,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3757,
+      "loss": 1.3841,
       "step": 54850
     },
     {
       "epoch": 0.8912933989699598,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4221,
       "step": 54860
     },
     {
       "epoch": 0.8914558658673295,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4147,
       "step": 54870
     },
     {
       "epoch": 0.8916183327646992,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.4049,
       "step": 54880
     },
     {
       "epoch": 0.8917807996620688,
-      "grad_norm": 2.375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.3842,
       "step": 54890
     },
     {
       "epoch": 0.8919432665594386,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.4239,
       "step": 54900
     },
     {
       "epoch": 0.8921057334568082,
-      "grad_norm": 2.546875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.3884,
       "step": 54910
     },
     {
       "epoch": 0.8922682003541779,
-      "grad_norm": 2.234375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3596,
       "step": 54920
     },
     {
       "epoch": 0.8924306672515475,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3768,
+      "loss": 1.3948,
       "step": 54930
     },
     {
       "epoch": 0.8925931341489172,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4145,
+      "loss": 1.4,
       "step": 54940
     },
     {
       "epoch": 0.8927556010462868,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.416,
+      "loss": 1.4174,
       "step": 54950
     },
     {
       "epoch": 0.8929180679436565,
-      "grad_norm": 1.875,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3742,
+      "loss": 1.3806,
       "step": 54960
     },
     {
       "epoch": 0.8930805348410261,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.3886,
       "step": 54970
     },
     {
       "epoch": 0.8932430017383958,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.417,
       "step": 54980
     },
     {
       "epoch": 0.8934054686357654,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4119,
+      "loss": 1.4462,
       "step": 54990
     },
     {
       "epoch": 0.8935679355331351,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3811,
+      "loss": 1.3564,
       "step": 55000
     },
     {
       "epoch": 0.8937304024305048,
-      "grad_norm": 4.03125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4094,
+      "loss": 1.3819,
       "step": 55010
     },
     {
       "epoch": 0.8938928693278745,
-      "grad_norm": 1.40625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.3755,
       "step": 55020
     },
     {
       "epoch": 0.8940553362252441,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4288,
       "step": 55030
     },
     {
       "epoch": 0.8942178031226138,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4102,
       "step": 55040
     },
     {
       "epoch": 0.8943802700199834,
-      "grad_norm": 2.21875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.3819,
       "step": 55050
     },
     {
       "epoch": 0.8945427369173531,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.379,
+      "loss": 1.3937,
       "step": 55060
     },
     {
       "epoch": 0.8947052038147227,
-      "grad_norm": 1.6640625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3898,
+      "loss": 1.4232,
       "step": 55070
     },
     {
       "epoch": 0.8948676707120924,
-      "grad_norm": 2.84375,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4144,
       "step": 55080
     },
     {
       "epoch": 0.895030137609462,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3742,
+      "loss": 1.4191,
       "step": 55090
     },
     {
       "epoch": 0.8951926045068317,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.4113,
       "step": 55100
     },
     {
       "epoch": 0.8953550714042013,
-      "grad_norm": 1.6484375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3761,
+      "loss": 1.4197,
       "step": 55110
     },
     {
       "epoch": 0.8955175383015711,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.4293,
       "step": 55120
     },
     {
       "epoch": 0.8956800051989408,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.4346,
       "step": 55130
     },
     {
       "epoch": 0.8958424720963104,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.3755,
       "step": 55140
     },
     {
       "epoch": 0.8960049389936801,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3894,
+      "loss": 1.3508,
       "step": 55150
     },
     {
       "epoch": 0.8961674058910497,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.3923,
       "step": 55160
     },
     {
       "epoch": 0.8963298727884194,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3709,
+      "loss": 1.4128,
       "step": 55170
     },
     {
       "epoch": 0.896492339685789,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.4311,
       "step": 55180
     },
     {
       "epoch": 0.8966548065831587,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4457,
       "step": 55190
     },
     {
       "epoch": 0.8968172734805283,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.3752,
       "step": 55200
     },
     {
       "epoch": 0.896979740377898,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3876,
+      "loss": 1.3684,
       "step": 55210
     },
     {
       "epoch": 0.8971422072752676,
-      "grad_norm": 1.796875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.4024,
       "step": 55220
     },
     {
       "epoch": 0.8973046741726374,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.365,
       "step": 55230
     },
     {
       "epoch": 0.897467141070007,
-      "grad_norm": 2.5,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3771,
+      "loss": 1.4122,
       "step": 55240
     },
     {
       "epoch": 0.8976296079673767,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.3919,
       "step": 55250
     },
     {
       "epoch": 0.8977920748647463,
-      "grad_norm": 2.015625,
+      "grad_norm": 4.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.3924,
       "step": 55260
     },
     {
       "epoch": 0.897954541762116,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.3934,
       "step": 55270
     },
     {
       "epoch": 0.8981170086594856,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3716,
+      "loss": 1.3899,
       "step": 55280
     },
     {
       "epoch": 0.8982794755568553,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3733,
+      "loss": 1.3859,
       "step": 55290
     },
     {
       "epoch": 0.8984419424542249,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.382,
       "step": 55300
     },
     {
       "epoch": 0.8986044093515946,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.3981,
       "step": 55310
     },
     {
       "epoch": 0.8987668762489642,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3637,
+      "loss": 1.3774,
       "step": 55320
     },
     {
       "epoch": 0.8989293431463339,
-      "grad_norm": 1.5625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.4001,
       "step": 55330
     },
     {
       "epoch": 0.8990918100437036,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.3684,
       "step": 55340
     },
     {
       "epoch": 0.8992542769410733,
-      "grad_norm": 1.5078125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.3899,
       "step": 55350
     },
     {
       "epoch": 0.899416743838443,
-      "grad_norm": 2.25,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.4011,
       "step": 55360
     },
     {
       "epoch": 0.8995792107358126,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.3893,
       "step": 55370
     },
     {
       "epoch": 0.8997416776331822,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.4106,
       "step": 55380
     },
     {
       "epoch": 0.8999041445305519,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3829,
+      "loss": 1.4402,
       "step": 55390
     },
     {
       "epoch": 0.9000666114279215,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3781,
+      "loss": 1.4062,
       "step": 55400
     },
     {
       "epoch": 0.9002290783252912,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.3756,
       "step": 55410
     },
     {
       "epoch": 0.9003915452226608,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3634,
+      "loss": 1.405,
       "step": 55420
     },
     {
       "epoch": 0.9005540121200305,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3858,
+      "loss": 1.4238,
       "step": 55430
     },
     {
       "epoch": 0.9007164790174002,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4126,
+      "loss": 1.3973,
       "step": 55440
     },
     {
       "epoch": 0.9008789459147699,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.3707,
       "step": 55450
     },
     {
       "epoch": 0.9010414128121396,
-      "grad_norm": 2.0,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.4106,
       "step": 55460
     },
     {
       "epoch": 0.9012038797095092,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3844,
       "step": 55470
     },
     {
       "epoch": 0.9013663466068789,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.3819,
       "step": 55480
     },
     {
       "epoch": 0.9015288135042485,
-      "grad_norm": 2.15625,
+      "grad_norm": 4.25,
       "learning_rate": 5e-05,
-      "loss": 0.3865,
+      "loss": 1.4178,
       "step": 55490
     },
     {
       "epoch": 0.9016912804016182,
-      "grad_norm": 2.59375,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.377,
       "step": 55500
     },
     {
       "epoch": 0.9018537472989878,
-      "grad_norm": 2.296875,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3757,
+      "loss": 1.4126,
       "step": 55510
     },
     {
       "epoch": 0.9020162141963575,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4202,
+      "loss": 1.3763,
       "step": 55520
     },
     {
       "epoch": 0.9021786810937271,
-      "grad_norm": 1.6875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3731,
+      "loss": 1.3811,
       "step": 55530
     },
     {
       "epoch": 0.9023411479910968,
-      "grad_norm": 2.0,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3853,
+      "loss": 1.372,
       "step": 55540
     },
     {
       "epoch": 0.9025036148884664,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.3916,
       "step": 55550
     },
     {
       "epoch": 0.9026660817858362,
-      "grad_norm": 4.59375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.3713,
       "step": 55560
     },
     {
       "epoch": 0.9028285486832058,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4117,
+      "loss": 1.3632,
       "step": 55570
     },
     {
       "epoch": 0.9029910155805755,
-      "grad_norm": 2.625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.4065,
       "step": 55580
     },
     {
       "epoch": 0.9031534824779451,
-      "grad_norm": 1.8515625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.4099,
       "step": 55590
     },
     {
       "epoch": 0.9033159493753148,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.4007,
       "step": 55600
     },
     {
       "epoch": 0.9034784162726844,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.3113,
       "step": 55610
     },
     {
       "epoch": 0.9036408831700541,
       "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3851,
+      "loss": 1.3858,
       "step": 55620
     },
     {
       "epoch": 0.9038033500674237,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.3761,
       "step": 55630
     },
     {
       "epoch": 0.9039658169647934,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.3687,
       "step": 55640
     },
     {
       "epoch": 0.904128283862163,
-      "grad_norm": 2.609375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.3469,
       "step": 55650
     },
     {
       "epoch": 0.9042907507595327,
-      "grad_norm": 2.625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.38,
+      "loss": 1.3972,
       "step": 55660
     },
     {
       "epoch": 0.9044532176569025,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3843,
+      "loss": 1.3843,
       "step": 55670
     },
     {
       "epoch": 0.9046156845542721,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.3758,
       "step": 55680
     },
     {
       "epoch": 0.9047781514516418,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.4055,
       "step": 55690
     },
     {
       "epoch": 0.9049406183490114,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4096,
+      "loss": 1.415,
       "step": 55700
     },
     {
       "epoch": 0.9051030852463811,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4127,
+      "loss": 1.4151,
       "step": 55710
     },
     {
       "epoch": 0.9052655521437507,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.4325,
       "step": 55720
     },
     {
       "epoch": 0.9054280190411204,
-      "grad_norm": 2.671875,
+      "grad_norm": 4.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3824,
+      "loss": 1.4237,
       "step": 55730
     },
     {
       "epoch": 0.90559048593849,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.4086,
+      "loss": 1.3793,
       "step": 55740
     },
     {
       "epoch": 0.9057529528358597,
-      "grad_norm": 1.546875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.3762,
       "step": 55750
     },
     {
       "epoch": 0.9059154197332293,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.407,
       "step": 55760
     },
     {
       "epoch": 0.906077886630599,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.3594,
       "step": 55770
     },
     {
       "epoch": 0.9062403535279687,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.4013,
+      "loss": 1.4186,
       "step": 55780
     },
     {
       "epoch": 0.9064028204253384,
-      "grad_norm": 1.7578125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3812,
+      "loss": 1.3975,
       "step": 55790
     },
     {
       "epoch": 0.906565287322708,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.4463,
       "step": 55800
     },
     {
       "epoch": 0.9067277542200777,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4008,
       "step": 55810
     },
     {
       "epoch": 0.9068902211174473,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.3881,
       "step": 55820
     },
     {
       "epoch": 0.907052688014817,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3719,
+      "loss": 1.3618,
       "step": 55830
     },
     {
       "epoch": 0.9072151549121866,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3994,
+      "loss": 1.3621,
       "step": 55840
     },
     {
       "epoch": 0.9073776218095563,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.4361,
       "step": 55850
     },
     {
       "epoch": 0.9075400887069259,
-      "grad_norm": 1.7265625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4234,
       "step": 55860
     },
     {
       "epoch": 0.9077025556042956,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.3992,
       "step": 55870
     },
     {
       "epoch": 0.9078650225016652,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.385,
+      "loss": 1.388,
       "step": 55880
     },
     {
       "epoch": 0.908027489399035,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4035,
       "step": 55890
     },
     {
       "epoch": 0.9081899562964046,
-      "grad_norm": 2.125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4044,
+      "loss": 1.4023,
       "step": 55900
     },
     {
       "epoch": 0.9083524231937743,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.4215,
       "step": 55910
     },
     {
       "epoch": 0.908514890091144,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4059,
+      "loss": 1.4283,
       "step": 55920
     },
     {
       "epoch": 0.9086773569885136,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.3602,
       "step": 55930
     },
     {
       "epoch": 0.9088398238858832,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4032,
+      "loss": 1.3884,
       "step": 55940
     },
     {
       "epoch": 0.9090022907832529,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3802,
+      "loss": 1.356,
       "step": 55950
     },
     {
       "epoch": 0.9091647576806225,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3794,
+      "loss": 1.4013,
       "step": 55960
     },
     {
       "epoch": 0.9093272245779922,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4194,
+      "loss": 1.3932,
       "step": 55970
     },
     {
       "epoch": 0.9094896914753618,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.3593,
       "step": 55980
     },
     {
       "epoch": 0.9096521583727315,
-      "grad_norm": 2.34375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3987,
+      "loss": 1.4283,
       "step": 55990
     },
     {
       "epoch": 0.9098146252701013,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.3827,
       "step": 56000
     },
     {
       "epoch": 0.9099770921674709,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.3883,
       "step": 56010
     },
     {
       "epoch": 0.9101395590648406,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.3732,
       "step": 56020
     },
     {
       "epoch": 0.9103020259622102,
-      "grad_norm": 2.140625,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4082,
+      "loss": 1.3825,
       "step": 56030
     },
     {
       "epoch": 0.9104644928595799,
-      "grad_norm": 1.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.3954,
       "step": 56040
     },
     {
       "epoch": 0.9106269597569495,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.365,
+      "loss": 1.3914,
       "step": 56050
     },
     {
       "epoch": 0.9107894266543192,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.373,
+      "loss": 1.3792,
       "step": 56060
     },
     {
       "epoch": 0.9109518935516888,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3937,
+      "loss": 1.3755,
       "step": 56070
     },
     {
       "epoch": 0.9111143604490585,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4175,
+      "loss": 1.4063,
       "step": 56080
     },
     {
       "epoch": 0.9112768273464281,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3732,
+      "loss": 1.3888,
       "step": 56090
     },
     {
       "epoch": 0.9114392942437978,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4123,
+      "loss": 1.3967,
       "step": 56100
     },
     {
       "epoch": 0.9116017611411675,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3781,
+      "loss": 1.3636,
       "step": 56110
     },
     {
       "epoch": 0.9117642280385372,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3869,
+      "loss": 1.41,
       "step": 56120
     },
     {
       "epoch": 0.9119266949359068,
-      "grad_norm": 1.6953125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.4344,
       "step": 56130
     },
     {
       "epoch": 0.9120891618332765,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.398,
+      "loss": 1.402,
       "step": 56140
     },
     {
       "epoch": 0.9122516287306461,
-      "grad_norm": 1.875,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.4048,
+      "loss": 1.3526,
       "step": 56150
     },
     {
       "epoch": 0.9124140956280158,
-      "grad_norm": 2.515625,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.4061,
+      "loss": 1.3961,
       "step": 56160
     },
     {
       "epoch": 0.9125765625253854,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.4033,
       "step": 56170
     },
     {
       "epoch": 0.9127390294227551,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.4027,
       "step": 56180
     },
     {
       "epoch": 0.9129014963201247,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.4175,
       "step": 56190
     },
     {
       "epoch": 0.9130639632174944,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.399,
       "step": 56200
     },
     {
       "epoch": 0.9132264301148642,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4161,
+      "loss": 1.3944,
       "step": 56210
     },
     {
       "epoch": 0.9133888970122338,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.3783,
       "step": 56220
     },
     {
       "epoch": 0.9135513639096035,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.3801,
       "step": 56230
     },
     {
       "epoch": 0.9137138308069731,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3822,
+      "loss": 1.4255,
       "step": 56240
     },
     {
       "epoch": 0.9138762977043428,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.3813,
       "step": 56250
     },
     {
       "epoch": 0.9140387646017124,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.3885,
       "step": 56260
     },
     {
       "epoch": 0.9142012314990821,
-      "grad_norm": 1.8046875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.3591,
       "step": 56270
     },
     {
       "epoch": 0.9143636983964517,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3979,
+      "loss": 1.3378,
       "step": 56280
     },
     {
       "epoch": 0.9145261652938214,
-      "grad_norm": 1.6875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3809,
+      "loss": 1.3824,
       "step": 56290
     },
     {
       "epoch": 0.914688632191191,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3829,
+      "loss": 1.3688,
       "step": 56300
     },
     {
       "epoch": 0.9148510990885607,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3767,
+      "loss": 1.3986,
       "step": 56310
     },
     {
       "epoch": 0.9150135659859304,
-      "grad_norm": 1.921875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.422,
       "step": 56320
     },
     {
       "epoch": 0.9151760328833001,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.3587,
       "step": 56330
     },
     {
       "epoch": 0.9153384997806697,
-      "grad_norm": 2.5,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.3992,
       "step": 56340
     },
     {
       "epoch": 0.9155009666780394,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3908,
+      "loss": 1.4097,
       "step": 56350
     },
     {
       "epoch": 0.915663433575409,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3662,
+      "loss": 1.398,
       "step": 56360
     },
     {
       "epoch": 0.9158259004727787,
-      "grad_norm": 2.421875,
+      "grad_norm": 4.375,
       "learning_rate": 5e-05,
-      "loss": 0.4025,
+      "loss": 1.3896,
       "step": 56370
     },
     {
       "epoch": 0.9159883673701483,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4153,
+      "loss": 1.3837,
       "step": 56380
     },
     {
       "epoch": 0.916150834267518,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3874,
+      "loss": 1.3646,
       "step": 56390
     },
     {
       "epoch": 0.9163133011648876,
-      "grad_norm": 2.46875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3803,
       "step": 56400
     },
     {
       "epoch": 0.9164757680622573,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.4085,
       "step": 56410
     },
     {
       "epoch": 0.9166382349596269,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3828,
+      "loss": 1.365,
       "step": 56420
     },
     {
       "epoch": 0.9168007018569967,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.3604,
       "step": 56430
     },
     {
       "epoch": 0.9169631687543663,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3924,
+      "loss": 1.3575,
       "step": 56440
     },
     {
       "epoch": 0.917125635651736,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.3918,
       "step": 56450
     },
     {
       "epoch": 0.9172881025491056,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.3979,
       "step": 56460
     },
     {
       "epoch": 0.9174505694464753,
-      "grad_norm": 1.5546875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3972,
+      "loss": 1.3781,
       "step": 56470
     },
     {
       "epoch": 0.917613036343845,
-      "grad_norm": 1.8515625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3638,
+      "loss": 1.4015,
       "step": 56480
     },
     {
       "epoch": 0.9177755032412146,
-      "grad_norm": 1.6328125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4173,
       "step": 56490
     },
     {
       "epoch": 0.9179379701385842,
-      "grad_norm": 2.25,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4111,
+      "loss": 1.4361,
       "step": 56500
     },
     {
       "epoch": 0.9181004370359539,
-      "grad_norm": 2.375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.3865,
       "step": 56510
     },
     {
       "epoch": 0.9182629039333235,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.3721,
       "step": 56520
     },
     {
       "epoch": 0.9184253708306932,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.375,
+      "loss": 1.4106,
       "step": 56530
     },
     {
       "epoch": 0.918587837728063,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.3797,
       "step": 56540
     },
     {
       "epoch": 0.9187503046254326,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3792,
+      "loss": 1.4239,
       "step": 56550
     },
     {
       "epoch": 0.9189127715228023,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.3742,
       "step": 56560
     },
     {
       "epoch": 0.9190752384201719,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3746,
+      "loss": 1.3761,
       "step": 56570
     },
     {
       "epoch": 0.9192377053175416,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.376,
+      "loss": 1.4052,
       "step": 56580
     },
     {
       "epoch": 0.9194001722149112,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.3929,
       "step": 56590
     },
     {
       "epoch": 0.9195626391122809,
-      "grad_norm": 1.5703125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3684,
+      "loss": 1.3833,
       "step": 56600
     },
     {
       "epoch": 0.9197251060096505,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3986,
+      "loss": 1.4049,
       "step": 56610
     },
     {
       "epoch": 0.9198875729070202,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4115,
+      "loss": 1.3856,
       "step": 56620
     },
     {
       "epoch": 0.9200500398043898,
-      "grad_norm": 1.9921875,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3802,
+      "loss": 1.3713,
       "step": 56630
     },
     {
       "epoch": 0.9202125067017595,
-      "grad_norm": 2.5,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4143,
+      "loss": 1.4126,
       "step": 56640
     },
     {
       "epoch": 0.9203749735991292,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.3994,
       "step": 56650
     },
     {
       "epoch": 0.9205374404964989,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.4057,
       "step": 56660
     },
     {
       "epoch": 0.9206999073938685,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.4178,
       "step": 56670
     },
     {
       "epoch": 0.9208623742912382,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.344,
       "step": 56680
     },
     {
       "epoch": 0.9210248411886078,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3926,
+      "loss": 1.4054,
       "step": 56690
     },
     {
       "epoch": 0.9211873080859775,
-      "grad_norm": 3.59375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3667,
+      "loss": 1.3999,
       "step": 56700
     },
     {
       "epoch": 0.9213497749833471,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.397,
       "step": 56710
     },
     {
       "epoch": 0.9215122418807168,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3766,
+      "loss": 1.3559,
       "step": 56720
     },
     {
       "epoch": 0.9216747087780864,
-      "grad_norm": 2.0,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.4093,
       "step": 56730
     },
     {
       "epoch": 0.9218371756754561,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.3426,
       "step": 56740
     },
     {
       "epoch": 0.9219996425728257,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.3904,
       "step": 56750
     },
     {
       "epoch": 0.9221621094701955,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3795,
+      "loss": 1.4289,
       "step": 56760
     },
     {
       "epoch": 0.9223245763675652,
-      "grad_norm": 2.46875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.3857,
       "step": 56770
     },
     {
       "epoch": 0.9224870432649348,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.4032,
       "step": 56780
     },
     {
       "epoch": 0.9226495101623045,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.3936,
       "step": 56790
     },
     {
       "epoch": 0.9228119770596741,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.4043,
       "step": 56800
     },
     {
       "epoch": 0.9229744439570438,
-      "grad_norm": 1.640625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.3971,
       "step": 56810
     },
     {
       "epoch": 0.9231369108544134,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.3458,
       "step": 56820
     },
     {
       "epoch": 0.9232993777517831,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4182,
+      "loss": 1.3546,
       "step": 56830
     },
     {
       "epoch": 0.9234618446491527,
-      "grad_norm": 2.625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3802,
+      "loss": 1.4037,
       "step": 56840
     },
     {
       "epoch": 0.9236243115465224,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.4305,
       "step": 56850
     },
     {
       "epoch": 0.923786778443892,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3981,
+      "loss": 1.4041,
       "step": 56860
     },
     {
       "epoch": 0.9239492453412618,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.4113,
       "step": 56870
     },
     {
       "epoch": 0.9241117122386314,
-      "grad_norm": 2.25,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.3927,
       "step": 56880
     },
     {
       "epoch": 0.9242741791360011,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.3934,
       "step": 56890
     },
     {
       "epoch": 0.9244366460333707,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3905,
+      "loss": 1.3926,
       "step": 56900
     },
     {
       "epoch": 0.9245991129307404,
-      "grad_norm": 2.375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.3508,
       "step": 56910
     },
     {
       "epoch": 0.92476157982811,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.3926,
       "step": 56920
     },
     {
       "epoch": 0.9249240467254797,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.3546,
       "step": 56930
     },
     {
       "epoch": 0.9250865136228493,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.409,
+      "loss": 1.4058,
       "step": 56940
     },
     {
       "epoch": 0.925248980520219,
-      "grad_norm": 1.5078125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3863,
+      "loss": 1.3832,
       "step": 56950
     },
     {
       "epoch": 0.9254114474175886,
-      "grad_norm": 1.703125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.3802,
       "step": 56960
     },
     {
       "epoch": 0.9255739143149583,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3875,
+      "loss": 1.3553,
       "step": 56970
     },
     {
       "epoch": 0.925736381212328,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.3912,
       "step": 56980
     },
     {
       "epoch": 0.9258988481096977,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.4242,
       "step": 56990
     },
     {
       "epoch": 0.9260613150070673,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4285,
+      "loss": 1.4209,
       "step": 57000
     },
     {
       "epoch": 0.926223781904437,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.3868,
       "step": 57010
     },
     {
       "epoch": 0.9263862488018066,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.426,
       "step": 57020
     },
     {
       "epoch": 0.9265487156991763,
-      "grad_norm": 2.0,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3645,
+      "loss": 1.3897,
       "step": 57030
     },
     {
       "epoch": 0.926711182596546,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4103,
+      "loss": 1.4046,
       "step": 57040
     },
     {
       "epoch": 0.9268736494939156,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4164,
+      "loss": 1.2915,
       "step": 57050
     },
     {
       "epoch": 0.9270361163912852,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3728,
+      "loss": 1.3904,
       "step": 57060
     },
     {
       "epoch": 0.9271985832886549,
-      "grad_norm": 1.703125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3741,
+      "loss": 1.3796,
       "step": 57070
     },
     {
       "epoch": 0.9273610501860245,
-      "grad_norm": 1.7578125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3734,
+      "loss": 1.3951,
       "step": 57080
     },
     {
       "epoch": 0.9275235170833943,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.3909,
       "step": 57090
     },
     {
       "epoch": 0.927685983980764,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.3558,
       "step": 57100
     },
     {
       "epoch": 0.9278484508781336,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.3505,
       "step": 57110
     },
     {
       "epoch": 0.9280109177755033,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4251,
+      "loss": 1.3745,
       "step": 57120
     },
     {
       "epoch": 0.9281733846728729,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3746,
+      "loss": 1.3888,
       "step": 57130
     },
     {
       "epoch": 0.9283358515702426,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.3608,
       "step": 57140
     },
     {
       "epoch": 0.9284983184676122,
-      "grad_norm": 2.265625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3772,
+      "loss": 1.3865,
       "step": 57150
     },
     {
       "epoch": 0.9286607853649819,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.4136,
+      "loss": 1.4065,
       "step": 57160
     },
     {
       "epoch": 0.9288232522623515,
-      "grad_norm": 1.578125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.39,
+      "loss": 1.4145,
       "step": 57170
     },
     {
       "epoch": 0.9289857191597212,
-      "grad_norm": 2.25,
+      "grad_norm": 4.625,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.3963,
       "step": 57180
     },
     {
       "epoch": 0.9291481860570908,
-      "grad_norm": 2.0,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.4038,
       "step": 57190
     },
     {
       "epoch": 0.9293106529544606,
-      "grad_norm": 1.8203125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.4424,
       "step": 57200
     },
     {
       "epoch": 0.9294731198518302,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4107,
       "step": 57210
     },
     {
       "epoch": 0.9296355867491999,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.3308,
       "step": 57220
     },
     {
       "epoch": 0.9297980536465695,
-      "grad_norm": 4.78125,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.4004,
       "step": 57230
     },
     {
       "epoch": 0.9299605205439392,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.4259,
       "step": 57240
     },
     {
       "epoch": 0.9301229874413088,
-      "grad_norm": 2.21875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.3624,
       "step": 57250
     },
     {
       "epoch": 0.9302854543386785,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.384,
       "step": 57260
     },
     {
       "epoch": 0.9304479212360481,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.3832,
       "step": 57270
     },
     {
       "epoch": 0.9306103881334178,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.4122,
       "step": 57280
     },
     {
       "epoch": 0.9307728550307874,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3984,
+      "loss": 1.3575,
       "step": 57290
     },
     {
       "epoch": 0.9309353219281571,
-      "grad_norm": 1.734375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.4105,
       "step": 57300
     },
     {
       "epoch": 0.9310977888255269,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3762,
+      "loss": 1.3656,
       "step": 57310
     },
     {
       "epoch": 0.9312602557228965,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.406,
+      "loss": 1.4092,
       "step": 57320
     },
     {
       "epoch": 0.9314227226202662,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4001,
+      "loss": 1.3682,
       "step": 57330
     },
     {
       "epoch": 0.9315851895176358,
-      "grad_norm": 6.03125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3796,
+      "loss": 1.3665,
       "step": 57340
     },
     {
       "epoch": 0.9317476564150055,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.4003,
+      "loss": 1.4276,
       "step": 57350
     },
     {
       "epoch": 0.9319101233123751,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.3917,
       "step": 57360
     },
     {
       "epoch": 0.9320725902097448,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4022,
+      "loss": 1.3886,
       "step": 57370
     },
     {
       "epoch": 0.9322350571071144,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.3968,
       "step": 57380
     },
     {
       "epoch": 0.9323975240044841,
-      "grad_norm": 1.4609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.3753,
       "step": 57390
     },
     {
       "epoch": 0.9325599909018537,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.3451,
       "step": 57400
     },
     {
       "epoch": 0.9327224577992234,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4048,
+      "loss": 1.4121,
       "step": 57410
     },
     {
       "epoch": 0.9328849246965931,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3904,
+      "loss": 1.3775,
       "step": 57420
     },
     {
       "epoch": 0.9330473915939628,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.3582,
       "step": 57430
     },
     {
       "epoch": 0.9332098584913324,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3798,
+      "loss": 1.3605,
       "step": 57440
     },
     {
       "epoch": 0.9333723253887021,
-      "grad_norm": 3.96875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.3732,
       "step": 57450
     },
     {
       "epoch": 0.9335347922860717,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.3481,
       "step": 57460
     },
     {
       "epoch": 0.9336972591834414,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3795,
+      "loss": 1.4121,
       "step": 57470
     },
     {
       "epoch": 0.933859726080811,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3785,
+      "loss": 1.402,
       "step": 57480
     },
     {
       "epoch": 0.9340221929781807,
-      "grad_norm": 4.40625,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.4173,
       "step": 57490
     },
     {
       "epoch": 0.9341846598755503,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.3494,
       "step": 57500
     },
     {
       "epoch": 0.93434712677292,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.3862,
       "step": 57510
     },
     {
       "epoch": 0.9345095936702896,
-      "grad_norm": 2.6875,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.389,
+      "loss": 1.4255,
       "step": 57520
     },
     {
       "epoch": 0.9346720605676594,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.3538,
       "step": 57530
     },
     {
       "epoch": 0.934834527465029,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.369,
+      "loss": 1.3443,
       "step": 57540
     },
     {
       "epoch": 0.9349969943623987,
-      "grad_norm": 2.40625,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.4172,
       "step": 57550
     },
     {
       "epoch": 0.9351594612597683,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.3661,
       "step": 57560
     },
     {
       "epoch": 0.935321928157138,
-      "grad_norm": 2.203125,
+      "grad_norm": 15.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.366,
       "step": 57570
     },
     {
       "epoch": 0.9354843950545076,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3802,
+      "loss": 1.3833,
       "step": 57580
     },
     {
       "epoch": 0.9356468619518773,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.39,
       "step": 57590
     },
     {
       "epoch": 0.935809328849247,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.3715,
       "step": 57600
     },
     {
       "epoch": 0.9359717957466166,
-      "grad_norm": 2.0,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3696,
+      "loss": 1.412,
       "step": 57610
     },
     {
       "epoch": 0.9361342626439862,
-      "grad_norm": 2.25,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.4354,
       "step": 57620
     },
     {
       "epoch": 0.9362967295413559,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3761,
+      "loss": 1.3988,
       "step": 57630
     },
     {
       "epoch": 0.9364591964387257,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3769,
+      "loss": 1.3853,
       "step": 57640
     },
     {
       "epoch": 0.9366216633360953,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3867,
+      "loss": 1.3824,
       "step": 57650
     },
     {
       "epoch": 0.936784130233465,
-      "grad_norm": 3.5,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3766,
+      "loss": 1.4116,
       "step": 57660
     },
     {
       "epoch": 0.9369465971308346,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4131,
+      "loss": 1.4226,
       "step": 57670
     },
     {
       "epoch": 0.9371090640282043,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3647,
       "step": 57680
     },
     {
       "epoch": 0.9372715309255739,
-      "grad_norm": 2.25,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.3703,
       "step": 57690
     },
     {
       "epoch": 0.9374339978229436,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3739,
+      "loss": 1.3833,
       "step": 57700
     },
     {
       "epoch": 0.9375964647203132,
-      "grad_norm": 1.609375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3869,
+      "loss": 1.4092,
       "step": 57710
     },
     {
       "epoch": 0.9377589316176829,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.4037,
+      "loss": 1.4075,
       "step": 57720
     },
     {
       "epoch": 0.9379213985150525,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4036,
+      "loss": 1.3802,
       "step": 57730
     },
     {
       "epoch": 0.9380838654124222,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.3814,
       "step": 57740
     },
     {
       "epoch": 0.9382463323097919,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4106,
+      "loss": 1.4086,
       "step": 57750
     },
     {
       "epoch": 0.9384087992071616,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3746,
+      "loss": 1.3992,
       "step": 57760
     },
     {
       "epoch": 0.9385712661045312,
-      "grad_norm": 1.9765625,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3766,
+      "loss": 1.3588,
       "step": 57770
     },
     {
       "epoch": 0.9387337330019009,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3735,
+      "loss": 1.4151,
       "step": 57780
     },
     {
       "epoch": 0.9388961998992705,
-      "grad_norm": 2.609375,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.403,
       "step": 57790
     },
     {
       "epoch": 0.9390586667966402,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3904,
+      "loss": 1.3782,
       "step": 57800
     },
     {
       "epoch": 0.9392211336940098,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4263,
       "step": 57810
     },
     {
       "epoch": 0.9393836005913795,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.3887,
       "step": 57820
     },
     {
       "epoch": 0.9395460674887491,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.3653,
       "step": 57830
     },
     {
       "epoch": 0.9397085343861188,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.3519,
       "step": 57840
     },
     {
       "epoch": 0.9398710012834884,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3751,
+      "loss": 1.3826,
       "step": 57850
     },
     {
       "epoch": 0.9400334681808582,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.3918,
       "step": 57860
     },
     {
       "epoch": 0.9401959350782279,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3731,
+      "loss": 1.3715,
       "step": 57870
     },
     {
       "epoch": 0.9403584019755975,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3646,
+      "loss": 1.3879,
       "step": 57880
     },
     {
       "epoch": 0.9405208688729672,
-      "grad_norm": 4.1875,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.3867,
       "step": 57890
     },
     {
       "epoch": 0.9406833357703368,
-      "grad_norm": 2.0,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.3672,
       "step": 57900
     },
     {
       "epoch": 0.9408458026677065,
-      "grad_norm": 1.5,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.3814,
       "step": 57910
     },
     {
       "epoch": 0.9410082695650761,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4074,
+      "loss": 1.3719,
       "step": 57920
     },
     {
       "epoch": 0.9411707364624458,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4482,
       "step": 57930
     },
     {
       "epoch": 0.9413332033598154,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.4095,
       "step": 57940
     },
     {
       "epoch": 0.9414956702571851,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3974,
+      "loss": 1.3807,
       "step": 57950
     },
     {
       "epoch": 0.9416581371545547,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.355,
       "step": 57960
     },
     {
       "epoch": 0.9418206040519245,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3929,
+      "loss": 1.4125,
       "step": 57970
     },
     {
       "epoch": 0.9419830709492941,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.4099,
+      "loss": 1.3902,
       "step": 57980
     },
     {
       "epoch": 0.9421455378466638,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.4037,
       "step": 57990
     },
     {
       "epoch": 0.9423080047440334,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.359,
       "step": 58000
     },
     {
       "epoch": 0.9424704716414031,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3795,
+      "loss": 1.4373,
       "step": 58010
     },
     {
       "epoch": 0.9426329385387727,
-      "grad_norm": 2.8125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.3739,
       "step": 58020
     },
     {
       "epoch": 0.9427954054361424,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3903,
+      "loss": 1.4272,
       "step": 58030
     },
     {
       "epoch": 0.942957872333512,
-      "grad_norm": 5.65625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.43,
       "step": 58040
     },
     {
       "epoch": 0.9431203392308817,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.397,
       "step": 58050
     },
     {
       "epoch": 0.9432828061282513,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.4263,
       "step": 58060
     },
     {
       "epoch": 0.943445273025621,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.4269,
       "step": 58070
     },
     {
       "epoch": 0.9436077399229907,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4123,
       "step": 58080
     },
     {
       "epoch": 0.9437702068203604,
-      "grad_norm": 4.03125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.423,
+      "loss": 1.405,
       "step": 58090
     },
     {
       "epoch": 0.94393267371773,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.3767,
       "step": 58100
     },
     {
       "epoch": 0.9440951406150997,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3947,
+      "loss": 1.389,
       "step": 58110
     },
     {
       "epoch": 0.9442576075124693,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3817,
+      "loss": 1.3932,
       "step": 58120
     },
     {
       "epoch": 0.944420074409839,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.3734,
       "step": 58130
     },
     {
       "epoch": 0.9445825413072086,
-      "grad_norm": 1.34375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4326,
       "step": 58140
     },
     {
       "epoch": 0.9447450082045783,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.4173,
       "step": 58150
     },
     {
       "epoch": 0.944907475101948,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3676,
+      "loss": 1.3812,
       "step": 58160
     },
     {
       "epoch": 0.9450699419993176,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.4156,
+      "loss": 1.4245,
       "step": 58170
     },
     {
       "epoch": 0.9452324088966872,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.374,
       "step": 58180
     },
     {
       "epoch": 0.945394875794057,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.3967,
       "step": 58190
     },
     {
       "epoch": 0.9455573426914267,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3932,
+      "loss": 1.424,
       "step": 58200
     },
     {
       "epoch": 0.9457198095887963,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3812,
+      "loss": 1.3714,
       "step": 58210
     },
     {
       "epoch": 0.945882276486166,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.3833,
       "step": 58220
     },
     {
       "epoch": 0.9460447433835356,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.3722,
       "step": 58230
     },
     {
       "epoch": 0.9462072102809053,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.4092,
       "step": 58240
     },
     {
       "epoch": 0.9463696771782749,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4046,
+      "loss": 1.3878,
       "step": 58250
     },
     {
       "epoch": 0.9465321440756446,
-      "grad_norm": 2.484375,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3828,
+      "loss": 1.3471,
       "step": 58260
     },
     {
       "epoch": 0.9466946109730142,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3783,
+      "loss": 1.3873,
       "step": 58270
     },
     {
       "epoch": 0.9468570778703839,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3711,
+      "loss": 1.3993,
       "step": 58280
     },
     {
       "epoch": 0.9470195447677535,
-      "grad_norm": 1.5703125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3763,
+      "loss": 1.379,
       "step": 58290
     },
     {
       "epoch": 0.9471820116651233,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.374,
+      "loss": 1.3802,
       "step": 58300
     },
     {
       "epoch": 0.9473444785624929,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.3796,
       "step": 58310
     },
     {
       "epoch": 0.9475069454598626,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3764,
+      "loss": 1.4383,
       "step": 58320
     },
     {
       "epoch": 0.9476694123572322,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.3203,
       "step": 58330
     },
     {
       "epoch": 0.9478318792546019,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.4211,
       "step": 58340
     },
     {
       "epoch": 0.9479943461519715,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.3656,
       "step": 58350
     },
     {
       "epoch": 0.9481568130493412,
-      "grad_norm": 1.6953125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.3995,
       "step": 58360
     },
     {
       "epoch": 0.9483192799467108,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4149,
       "step": 58370
     },
     {
       "epoch": 0.9484817468440805,
-      "grad_norm": 2.8125,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4041,
       "step": 58380
     },
     {
       "epoch": 0.9486442137414501,
-      "grad_norm": 2.5,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.4201,
       "step": 58390
     },
     {
       "epoch": 0.9488066806388198,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3607,
+      "loss": 1.3429,
       "step": 58400
     },
     {
       "epoch": 0.9489691475361896,
-      "grad_norm": 3.875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.3687,
       "step": 58410
     },
     {
       "epoch": 0.9491316144335592,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3919,
+      "loss": 1.3616,
       "step": 58420
     },
     {
       "epoch": 0.9492940813309289,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3778,
+      "loss": 1.3969,
       "step": 58430
     },
     {
       "epoch": 0.9494565482282985,
-      "grad_norm": 2.625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4045,
+      "loss": 1.4052,
       "step": 58440
     },
     {
       "epoch": 0.9496190151256682,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.4328,
       "step": 58450
     },
     {
       "epoch": 0.9497814820230378,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4047,
+      "loss": 1.4065,
       "step": 58460
     },
     {
       "epoch": 0.9499439489204075,
-      "grad_norm": 2.625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.4079,
       "step": 58470
     },
     {
       "epoch": 0.9501064158177771,
-      "grad_norm": 2.609375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4019,
+      "loss": 1.3813,
       "step": 58480
     },
     {
       "epoch": 0.9502688827151468,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3991,
+      "loss": 1.4,
       "step": 58490
     },
     {
       "epoch": 0.9504313496125164,
-      "grad_norm": 3.125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3882,
+      "loss": 1.4168,
       "step": 58500
     },
     {
       "epoch": 0.9505938165098861,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.403,
       "step": 58510
     },
     {
       "epoch": 0.9507562834072558,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.4014,
       "step": 58520
     },
     {
       "epoch": 0.9509187503046255,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3716,
+      "loss": 1.4012,
       "step": 58530
     },
     {
       "epoch": 0.9510812172019951,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.3995,
       "step": 58540
     },
     {
       "epoch": 0.9512436840993648,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3961,
+      "loss": 1.421,
       "step": 58550
     },
     {
       "epoch": 0.9514061509967344,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3941,
+      "loss": 1.4086,
       "step": 58560
     },
     {
       "epoch": 0.9515686178941041,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4273,
+      "loss": 1.3918,
       "step": 58570
     },
     {
       "epoch": 0.9517310847914737,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.3523,
       "step": 58580
     },
     {
       "epoch": 0.9518935516888434,
-      "grad_norm": 2.125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3791,
+      "loss": 1.397,
       "step": 58590
     },
     {
       "epoch": 0.952056018586213,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3948,
       "step": 58600
     },
     {
       "epoch": 0.9522184854835827,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.3744,
       "step": 58610
     },
     {
       "epoch": 0.9523809523809523,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.3744,
       "step": 58620
     },
     {
       "epoch": 0.9525434192783221,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.377,
       "step": 58630
     },
     {
       "epoch": 0.9527058861756917,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3988,
+      "loss": 1.4103,
       "step": 58640
     },
     {
       "epoch": 0.9528683530730614,
-      "grad_norm": 2.296875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.3823,
+      "loss": 1.3962,
       "step": 58650
     },
     {
       "epoch": 0.953030819970431,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.4067,
       "step": 58660
     },
     {
       "epoch": 0.9531932868678007,
-      "grad_norm": 1.7734375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.4081,
       "step": 58670
     },
     {
       "epoch": 0.9533557537651703,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.3784,
       "step": 58680
     },
     {
       "epoch": 0.95351822066254,
-      "grad_norm": 2.796875,
+      "grad_norm": 4.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3679,
+      "loss": 1.4037,
       "step": 58690
     },
     {
       "epoch": 0.9536806875599096,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3765,
+      "loss": 1.3703,
       "step": 58700
     },
     {
       "epoch": 0.9538431544572793,
-      "grad_norm": 1.5234375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.3886,
       "step": 58710
     },
     {
       "epoch": 0.954005621354649,
-      "grad_norm": 1.84375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3724,
+      "loss": 1.4163,
       "step": 58720
     },
     {
       "epoch": 0.9541680882520186,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.3653,
       "step": 58730
     },
     {
       "epoch": 0.9543305551493884,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.387,
+      "loss": 1.3937,
       "step": 58740
     },
     {
       "epoch": 0.954493022046758,
-      "grad_norm": 1.8046875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.3822,
       "step": 58750
     },
     {
       "epoch": 0.9546554889441277,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.4007,
+      "loss": 1.392,
       "step": 58760
     },
     {
       "epoch": 0.9548179558414973,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3855,
+      "loss": 1.4093,
       "step": 58770
     },
     {
       "epoch": 0.954980422738867,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3928,
+      "loss": 1.393,
       "step": 58780
     },
     {
       "epoch": 0.9551428896362366,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3864,
+      "loss": 1.3799,
       "step": 58790
     },
     {
       "epoch": 0.9553053565336063,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4095,
+      "loss": 1.3566,
       "step": 58800
     },
     {
       "epoch": 0.9554678234309759,
-      "grad_norm": 2.625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3752,
+      "loss": 1.4288,
       "step": 58810
     },
     {
       "epoch": 0.9556302903283456,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3972,
+      "loss": 1.4325,
       "step": 58820
     },
     {
       "epoch": 0.9557927572257152,
-      "grad_norm": 2.875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.4017,
       "step": 58830
     },
     {
       "epoch": 0.9559552241230849,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3852,
       "step": 58840
     },
     {
       "epoch": 0.9561176910204546,
-      "grad_norm": 3.453125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3936,
+      "loss": 1.405,
       "step": 58850
     },
     {
       "epoch": 0.9562801579178243,
-      "grad_norm": 3.375,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.4098,
       "step": 58860
     },
     {
       "epoch": 0.9564426248151939,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.3964,
       "step": 58870
     },
     {
       "epoch": 0.9566050917125636,
-      "grad_norm": 2.875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3931,
+      "loss": 1.3728,
       "step": 58880
     },
     {
       "epoch": 0.9567675586099332,
-      "grad_norm": 3.0,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.3739,
       "step": 58890
     },
     {
       "epoch": 0.9569300255073029,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.382,
       "step": 58900
     },
     {
       "epoch": 0.9570924924046725,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3847,
+      "loss": 1.422,
       "step": 58910
     },
     {
       "epoch": 0.9572549593020422,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3914,
+      "loss": 1.3992,
       "step": 58920
     },
     {
       "epoch": 0.9574174261994118,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.4088,
+      "loss": 1.3936,
       "step": 58930
     },
     {
       "epoch": 0.9575798930967815,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.3714,
       "step": 58940
     },
     {
       "epoch": 0.9577423599941511,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3899,
+      "loss": 1.4167,
       "step": 58950
     },
     {
       "epoch": 0.9579048268915209,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.4074,
       "step": 58960
     },
     {
       "epoch": 0.9580672937888906,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4011,
+      "loss": 1.3556,
       "step": 58970
     },
     {
       "epoch": 0.9582297606862602,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3767,
+      "loss": 1.3604,
       "step": 58980
     },
     {
       "epoch": 0.9583922275836299,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3778,
+      "loss": 1.3902,
       "step": 58990
     },
     {
       "epoch": 0.9585546944809995,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.4243,
       "step": 59000
     },
     {
       "epoch": 0.9587171613783692,
-      "grad_norm": 1.7265625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.42,
       "step": 59010
     },
     {
       "epoch": 0.9588796282757388,
-      "grad_norm": 3.0,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3805,
+      "loss": 1.3635,
       "step": 59020
     },
     {
       "epoch": 0.9590420951731085,
-      "grad_norm": 2.375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.352,
       "step": 59030
     },
     {
       "epoch": 0.9592045620704781,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.4402,
       "step": 59040
     },
     {
       "epoch": 0.9593670289678478,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3889,
+      "loss": 1.3891,
       "step": 59050
     },
     {
       "epoch": 0.9595294958652174,
-      "grad_norm": 2.125,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.3741,
       "step": 59060
     },
     {
       "epoch": 0.9596919627625872,
-      "grad_norm": 3.65625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3969,
+      "loss": 1.3752,
       "step": 59070
     },
     {
       "epoch": 0.9598544296599568,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.3755,
       "step": 59080
     },
     {
       "epoch": 0.9600168965573265,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3752,
+      "loss": 1.3821,
       "step": 59090
     },
     {
       "epoch": 0.9601793634546961,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3712,
       "step": 59100
     },
     {
       "epoch": 0.9603418303520658,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3823,
+      "loss": 1.4197,
       "step": 59110
     },
     {
       "epoch": 0.9605042972494354,
-      "grad_norm": 3.921875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3921,
+      "loss": 1.353,
       "step": 59120
     },
     {
       "epoch": 0.9606667641468051,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3962,
+      "loss": 1.3639,
       "step": 59130
     },
     {
       "epoch": 0.9608292310441747,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4024,
+      "loss": 1.3685,
       "step": 59140
     },
     {
       "epoch": 0.9609916979415444,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.3611,
       "step": 59150
     },
     {
       "epoch": 0.961154164838914,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3913,
+      "loss": 1.3982,
       "step": 59160
     },
     {
       "epoch": 0.9613166317362837,
-      "grad_norm": 3.453125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.3642,
       "step": 59170
     },
     {
       "epoch": 0.9614790986336534,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3796,
+      "loss": 1.3715,
       "step": 59180
     },
     {
       "epoch": 0.9616415655310231,
-      "grad_norm": 2.5,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.3654,
       "step": 59190
     },
     {
       "epoch": 0.9618040324283927,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4054,
+      "loss": 1.3872,
       "step": 59200
     },
     {
       "epoch": 0.9619664993257624,
-      "grad_norm": 2.375,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3917,
+      "loss": 1.3995,
       "step": 59210
     },
     {
       "epoch": 0.962128966223132,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.393,
+      "loss": 1.3638,
       "step": 59220
     },
     {
       "epoch": 0.9622914331205017,
-      "grad_norm": 1.78125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3831,
+      "loss": 1.3591,
       "step": 59230
     },
     {
       "epoch": 0.9624539000178713,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.4105,
       "step": 59240
     },
     {
       "epoch": 0.962616366915241,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3788,
+      "loss": 1.3699,
       "step": 59250
     },
     {
       "epoch": 0.9627788338126106,
-      "grad_norm": 1.6796875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3909,
+      "loss": 1.4349,
       "step": 59260
     },
     {
       "epoch": 0.9629413007099803,
-      "grad_norm": 1.7890625,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3698,
+      "loss": 1.3823,
       "step": 59270
     },
     {
       "epoch": 0.96310376760735,
-      "grad_norm": 3.265625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.4014,
+      "loss": 1.3798,
       "step": 59280
     },
     {
       "epoch": 0.9632662345047197,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3892,
+      "loss": 1.4143,
       "step": 59290
     },
     {
       "epoch": 0.9634287014020894,
-      "grad_norm": 1.7734375,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.3819,
+      "loss": 1.4028,
       "step": 59300
     },
     {
       "epoch": 0.963591168299459,
-      "grad_norm": 1.8359375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3862,
+      "loss": 1.4282,
       "step": 59310
     },
     {
       "epoch": 0.9637536351968287,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3779,
+      "loss": 1.4113,
       "step": 59320
     },
     {
       "epoch": 0.9639161020941983,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3912,
+      "loss": 1.4188,
       "step": 59330
     },
     {
       "epoch": 0.964078568991568,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.374,
+      "loss": 1.3304,
       "step": 59340
     },
     {
       "epoch": 0.9642410358889376,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3889,
+      "loss": 1.4183,
       "step": 59350
     },
     {
       "epoch": 0.9644035027863073,
-      "grad_norm": 3.375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.4192,
+      "loss": 1.4175,
       "step": 59360
     },
     {
       "epoch": 0.9645659696836769,
-      "grad_norm": 1.6171875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.3803,
       "step": 59370
     },
     {
       "epoch": 0.9647284365810466,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.417,
+      "loss": 1.375,
       "step": 59380
     },
     {
       "epoch": 0.9648909034784162,
-      "grad_norm": 1.921875,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3856,
+      "loss": 1.3495,
       "step": 59390
     },
     {
       "epoch": 0.965053370375786,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.3681,
       "step": 59400
     },
     {
       "epoch": 0.9652158372731556,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3659,
+      "loss": 1.3567,
       "step": 59410
     },
     {
       "epoch": 0.9653783041705253,
-      "grad_norm": 1.8125,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.3769,
       "step": 59420
     },
     {
       "epoch": 0.9655407710678949,
-      "grad_norm": 2.109375,
+      "grad_norm": 4.125,
       "learning_rate": 5e-05,
-      "loss": 0.3999,
+      "loss": 1.4462,
       "step": 59430
     },
     {
       "epoch": 0.9657032379652646,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3885,
+      "loss": 1.3823,
       "step": 59440
     },
     {
       "epoch": 0.9658657048626342,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.391,
+      "loss": 1.3693,
       "step": 59450
     },
     {
       "epoch": 0.9660281717600039,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3665,
+      "loss": 1.3623,
       "step": 59460
     },
     {
       "epoch": 0.9661906386573735,
-      "grad_norm": 2.125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3769,
+      "loss": 1.3843,
       "step": 59470
     },
     {
       "epoch": 0.9663531055547432,
-      "grad_norm": 1.6484375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3759,
+      "loss": 1.3977,
       "step": 59480
     },
     {
       "epoch": 0.9665155724521128,
-      "grad_norm": 3.984375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.401,
+      "loss": 1.3105,
       "step": 59490
     },
     {
       "epoch": 0.9666780393494825,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3673,
+      "loss": 1.4211,
       "step": 59500
     },
     {
       "epoch": 0.9668405062468522,
-      "grad_norm": 2.546875,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.3567,
       "step": 59510
     },
     {
       "epoch": 0.9670029731442219,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.3731,
       "step": 59520
     },
     {
       "epoch": 0.9671654400415916,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3959,
+      "loss": 1.4364,
       "step": 59530
     },
     {
       "epoch": 0.9673279069389612,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.4455,
       "step": 59540
     },
     {
       "epoch": 0.9674903738363309,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.4039,
+      "loss": 1.3627,
       "step": 59550
     },
     {
       "epoch": 0.9676528407337005,
-      "grad_norm": 1.78125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3944,
+      "loss": 1.3757,
       "step": 59560
     },
     {
       "epoch": 0.9678153076310702,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3694,
+      "loss": 1.412,
       "step": 59570
     },
     {
       "epoch": 0.9679777745284398,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3842,
+      "loss": 1.3704,
       "step": 59580
     },
     {
       "epoch": 0.9681402414258095,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.3981,
       "step": 59590
     },
     {
       "epoch": 0.9683027083231791,
-      "grad_norm": 2.0,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3629,
+      "loss": 1.4234,
       "step": 59600
     },
     {
       "epoch": 0.9684651752205489,
-      "grad_norm": 1.8046875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.3912,
       "step": 59610
     },
     {
       "epoch": 0.9686276421179185,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3737,
+      "loss": 1.3877,
       "step": 59620
     },
     {
       "epoch": 0.9687901090152882,
-      "grad_norm": 2.640625,
+      "grad_norm": 4.375,
       "learning_rate": 5e-05,
-      "loss": 0.3911,
+      "loss": 1.4004,
       "step": 59630
     },
     {
       "epoch": 0.9689525759126578,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3772,
+      "loss": 1.4109,
       "step": 59640
     },
     {
       "epoch": 0.9691150428100275,
-      "grad_norm": 1.59375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.3865,
       "step": 59650
     },
     {
       "epoch": 0.9692775097073971,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.4379,
       "step": 59660
     },
     {
       "epoch": 0.9694399766047668,
-      "grad_norm": 2.625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3399,
       "step": 59670
     },
     {
       "epoch": 0.9696024435021364,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3977,
+      "loss": 1.3788,
       "step": 59680
     },
     {
       "epoch": 0.9697649103995061,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3839,
+      "loss": 1.4023,
       "step": 59690
     },
     {
       "epoch": 0.9699273772968757,
-      "grad_norm": 1.8203125,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.4018,
+      "loss": 1.3609,
       "step": 59700
     },
     {
       "epoch": 0.9700898441942454,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3731,
+      "loss": 1.412,
       "step": 59710
     },
     {
       "epoch": 0.9702523110916151,
-      "grad_norm": 2.875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.3867,
       "step": 59720
     },
     {
       "epoch": 0.9704147779889848,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3746,
+      "loss": 1.3849,
       "step": 59730
     },
     {
       "epoch": 0.9705772448863544,
-      "grad_norm": 2.75,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.3778,
       "step": 59740
     },
     {
       "epoch": 0.9707397117837241,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4069,
+      "loss": 1.4031,
       "step": 59750
     },
     {
       "epoch": 0.9709021786810937,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.3896,
       "step": 59760
     },
     {
       "epoch": 0.9710646455784634,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.3754,
       "step": 59770
     },
     {
       "epoch": 0.971227112475833,
-      "grad_norm": 3.3125,
+      "grad_norm": 4.875,
       "learning_rate": 5e-05,
-      "loss": 0.387,
+      "loss": 1.4108,
       "step": 59780
     },
     {
       "epoch": 0.9713895793732027,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.4044,
       "step": 59790
     },
     {
       "epoch": 0.9715520462705723,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.4015,
+      "loss": 1.4191,
       "step": 59800
     },
     {
       "epoch": 0.971714513167942,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.34,
       "step": 59810
     },
     {
       "epoch": 0.9718769800653116,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.4366,
       "step": 59820
     },
     {
       "epoch": 0.9720394469626814,
-      "grad_norm": 2.375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4049,
+      "loss": 1.4105,
       "step": 59830
     },
     {
       "epoch": 0.9722019138600511,
-      "grad_norm": 2.1875,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.378,
+      "loss": 1.3336,
       "step": 59840
     },
     {
       "epoch": 0.9723643807574207,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3729,
+      "loss": 1.3539,
       "step": 59850
     },
     {
       "epoch": 0.9725268476547904,
-      "grad_norm": 1.7578125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3884,
+      "loss": 1.3571,
       "step": 59860
     },
     {
       "epoch": 0.97268931455216,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3942,
+      "loss": 1.394,
       "step": 59870
     },
     {
       "epoch": 0.9728517814495297,
-      "grad_norm": 1.6953125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.3903,
       "step": 59880
     },
     {
       "epoch": 0.9730142483468993,
-      "grad_norm": 1.890625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3774,
+      "loss": 1.4018,
       "step": 59890
     },
     {
       "epoch": 0.973176715244269,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3946,
+      "loss": 1.3764,
       "step": 59900
     },
     {
       "epoch": 0.9733391821416386,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.38,
       "step": 59910
     },
     {
       "epoch": 0.9735016490390083,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3878,
+      "loss": 1.3654,
       "step": 59920
     },
     {
       "epoch": 0.9736641159363779,
-      "grad_norm": 1.8203125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3675,
+      "loss": 1.3986,
       "step": 59930
     },
     {
       "epoch": 0.9738265828337477,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3777,
+      "loss": 1.3872,
       "step": 59940
     },
     {
       "epoch": 0.9739890497311173,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.382,
+      "loss": 1.4312,
       "step": 59950
     },
     {
       "epoch": 0.974151516628487,
-      "grad_norm": 1.4453125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.4492,
       "step": 59960
     },
     {
       "epoch": 0.9743139835258566,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.3805,
       "step": 59970
     },
     {
       "epoch": 0.9744764504232263,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3773,
+      "loss": 1.3703,
       "step": 59980
     },
     {
       "epoch": 0.9746389173205959,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.3582,
       "step": 59990
     },
     {
       "epoch": 0.9748013842179656,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4097,
+      "loss": 1.3599,
       "step": 60000
     },
     {
       "epoch": 0.9749638511153352,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.3665,
       "step": 60010
     },
     {
       "epoch": 0.9751263180127049,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.4034,
+      "loss": 1.3887,
       "step": 60020
     },
     {
       "epoch": 0.9752887849100745,
-      "grad_norm": 3.09375,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.382,
+      "loss": 1.3735,
       "step": 60030
     },
     {
       "epoch": 0.9754512518074442,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3967,
+      "loss": 1.3594,
       "step": 60040
     },
     {
       "epoch": 0.975613718704814,
-      "grad_norm": 1.6328125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.3889,
       "step": 60050
     },
     {
       "epoch": 0.9757761856021836,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3784,
+      "loss": 1.4081,
       "step": 60060
     },
     {
       "epoch": 0.9759386524995532,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.3915,
       "step": 60070
     },
     {
       "epoch": 0.9761011193969229,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3963,
+      "loss": 1.3843,
       "step": 60080
     },
     {
       "epoch": 0.9762635862942926,
-      "grad_norm": 2.1875,
+      "grad_norm": 4.71875,
       "learning_rate": 5e-05,
-      "loss": 0.4016,
+      "loss": 1.3331,
       "step": 60090
     },
     {
       "epoch": 0.9764260531916622,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3814,
+      "loss": 1.3408,
       "step": 60100
     },
     {
       "epoch": 0.9765885200890319,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.388,
+      "loss": 1.3502,
       "step": 60110
     },
     {
       "epoch": 0.9767509869864015,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.4057,
+      "loss": 1.393,
       "step": 60120
     },
     {
       "epoch": 0.9769134538837712,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3866,
+      "loss": 1.4349,
       "step": 60130
     },
     {
       "epoch": 0.9770759207811408,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.3356,
       "step": 60140
     },
     {
       "epoch": 0.9772383876785105,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.1875,
       "learning_rate": 5e-05,
-      "loss": 0.4041,
+      "loss": 1.4611,
       "step": 60150
     },
     {
       "epoch": 0.9774008545758802,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3939,
+      "loss": 1.4237,
       "step": 60160
     },
     {
       "epoch": 0.9775633214732499,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3769,
+      "loss": 1.4672,
       "step": 60170
     },
     {
       "epoch": 0.9777257883706195,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.4009,
+      "loss": 1.4428,
       "step": 60180
     },
     {
       "epoch": 0.9778882552679892,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3895,
+      "loss": 1.427,
       "step": 60190
     },
     {
       "epoch": 0.9780507221653588,
-      "grad_norm": 1.75,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3992,
+      "loss": 1.4274,
       "step": 60200
     },
     {
       "epoch": 0.9782131890627285,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3786,
+      "loss": 1.3933,
       "step": 60210
     },
     {
       "epoch": 0.9783756559600981,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3953,
+      "loss": 1.4162,
       "step": 60220
     },
     {
       "epoch": 0.9785381228574678,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3938,
+      "loss": 1.3772,
       "step": 60230
     },
     {
       "epoch": 0.9787005897548374,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3662,
+      "loss": 1.3726,
       "step": 60240
     },
     {
       "epoch": 0.9788630566522071,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3955,
+      "loss": 1.3574,
       "step": 60250
     },
     {
       "epoch": 0.9790255235495767,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.358,
       "step": 60260
     },
     {
       "epoch": 0.9791879904469465,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3713,
+      "loss": 1.3825,
       "step": 60270
     },
     {
       "epoch": 0.9793504573443161,
-      "grad_norm": 2.5,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3806,
+      "loss": 1.3918,
       "step": 60280
     },
     {
       "epoch": 0.9795129242416858,
-      "grad_norm": 2.078125,
+      "grad_norm": 4.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3763,
+      "loss": 1.328,
       "step": 60290
     },
     {
       "epoch": 0.9796753911390554,
-      "grad_norm": 2.5,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3927,
+      "loss": 1.377,
       "step": 60300
     },
     {
       "epoch": 0.9798378580364251,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.4149,
+      "loss": 1.3701,
       "step": 60310
     },
     {
       "epoch": 0.9800003249337947,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.3698,
       "step": 60320
     },
     {
       "epoch": 0.9801627918311644,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3858,
+      "loss": 1.3692,
       "step": 60330
     },
     {
       "epoch": 0.980325258728534,
-      "grad_norm": 3.015625,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3854,
+      "loss": 1.397,
       "step": 60340
     },
     {
       "epoch": 0.9804877256259037,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.4058,
+      "loss": 1.4341,
       "step": 60350
     },
     {
       "epoch": 0.9806501925232733,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.4072,
+      "loss": 1.4174,
       "step": 60360
     },
     {
       "epoch": 0.980812659420643,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3897,
+      "loss": 1.4116,
       "step": 60370
     },
     {
       "epoch": 0.9809751263180128,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.4035,
+      "loss": 1.4107,
       "step": 60380
     },
     {
       "epoch": 0.9811375932153824,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.4065,
+      "loss": 1.3843,
       "step": 60390
     },
     {
       "epoch": 0.9813000601127521,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.3858,
       "step": 60400
     },
     {
       "epoch": 0.9814625270101217,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3624,
+      "loss": 1.4226,
       "step": 60410
     },
     {
       "epoch": 0.9816249939074914,
-      "grad_norm": 2.25,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3688,
+      "loss": 1.3572,
       "step": 60420
     },
     {
       "epoch": 0.981787460804861,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3745,
+      "loss": 1.3905,
       "step": 60430
     },
     {
       "epoch": 0.9819499277022307,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3789,
+      "loss": 1.429,
       "step": 60440
     },
     {
       "epoch": 0.9821123945996003,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.3806,
       "step": 60450
     },
     {
       "epoch": 0.98227486149697,
-      "grad_norm": 3.25,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.394,
       "step": 60460
     },
     {
       "epoch": 0.9824373283943396,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3848,
+      "loss": 1.3753,
       "step": 60470
     },
     {
       "epoch": 0.9825997952917093,
-      "grad_norm": 1.796875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3877,
+      "loss": 1.3756,
       "step": 60480
     },
     {
       "epoch": 0.982762262189079,
-      "grad_norm": 1.8828125,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.4151,
       "step": 60490
     },
     {
       "epoch": 0.9829247290864487,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3668,
+      "loss": 1.3585,
       "step": 60500
     },
     {
       "epoch": 0.9830871959838183,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3859,
+      "loss": 1.3851,
       "step": 60510
     },
     {
       "epoch": 0.983249662881188,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.3957,
       "step": 60520
     },
     {
       "epoch": 0.9834121297785576,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.394,
+      "loss": 1.3829,
       "step": 60530
     },
     {
       "epoch": 0.9835745966759273,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.4053,
       "step": 60540
     },
     {
       "epoch": 0.9837370635732969,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3791,
+      "loss": 1.4004,
       "step": 60550
     },
     {
       "epoch": 0.9838995304706666,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.3774,
       "step": 60560
     },
     {
       "epoch": 0.9840619973680362,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.3798,
       "step": 60570
     },
     {
       "epoch": 0.9842244642654059,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.4115,
       "step": 60580
     },
     {
       "epoch": 0.9843869311627755,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3968,
+      "loss": 1.3391,
       "step": 60590
     },
     {
       "epoch": 0.9845493980601453,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3901,
+      "loss": 1.3649,
       "step": 60600
     },
     {
       "epoch": 0.984711864957515,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.4077,
+      "loss": 1.3741,
       "step": 60610
     },
     {
       "epoch": 0.9848743318548846,
-      "grad_norm": 1.6953125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3739,
+      "loss": 1.3684,
       "step": 60620
     },
     {
       "epoch": 0.9850367987522543,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3845,
+      "loss": 1.4427,
       "step": 60630
     },
     {
       "epoch": 0.9851992656496239,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3752,
+      "loss": 1.3873,
       "step": 60640
     },
     {
       "epoch": 0.9853617325469936,
-      "grad_norm": 1.984375,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3857,
+      "loss": 1.3821,
       "step": 60650
     },
     {
       "epoch": 0.9855241994443632,
-      "grad_norm": 2.375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3775,
+      "loss": 1.4366,
       "step": 60660
     },
     {
       "epoch": 0.9856866663417329,
-      "grad_norm": 2.234375,
+      "grad_norm": 4.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3844,
+      "loss": 1.3914,
       "step": 60670
     },
     {
       "epoch": 0.9858491332391025,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.4085,
       "step": 60680
     },
     {
       "epoch": 0.9860116001364722,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.41,
+      "loss": 1.3137,
       "step": 60690
     },
     {
       "epoch": 0.9861740670338418,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.3481,
       "step": 60700
     },
     {
       "epoch": 0.9863365339312116,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.3762,
       "step": 60710
     },
     {
       "epoch": 0.9864990008285812,
-      "grad_norm": 1.625,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3873,
+      "loss": 1.4034,
       "step": 60720
     },
     {
       "epoch": 0.9866614677259509,
-      "grad_norm": 1.5703125,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3907,
+      "loss": 1.4232,
       "step": 60730
     },
     {
       "epoch": 0.9868239346233205,
-      "grad_norm": 3.5625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3693,
+      "loss": 1.4051,
       "step": 60740
     },
     {
       "epoch": 0.9869864015206902,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3887,
+      "loss": 1.393,
       "step": 60750
     },
     {
       "epoch": 0.9871488684180598,
-      "grad_norm": 2.75,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3852,
+      "loss": 1.4041,
       "step": 60760
     },
     {
       "epoch": 0.9873113353154295,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3813,
+      "loss": 1.3875,
       "step": 60770
     },
     {
       "epoch": 0.9874738022127991,
-      "grad_norm": 1.703125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3792,
+      "loss": 1.3932,
       "step": 60780
     },
     {
       "epoch": 0.9876362691101688,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3835,
+      "loss": 1.3927,
       "step": 60790
     },
     {
       "epoch": 0.9877987360075384,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.3451,
       "step": 60800
     },
     {
       "epoch": 0.9879612029049081,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.397,
+      "loss": 1.4297,
       "step": 60810
     },
     {
       "epoch": 0.9881236698022778,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3691,
+      "loss": 1.429,
       "step": 60820
     },
     {
       "epoch": 0.9882861366996475,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3975,
+      "loss": 1.369,
       "step": 60830
     },
     {
       "epoch": 0.9884486035970171,
-      "grad_norm": 3.40625,
+      "grad_norm": 4.75,
       "learning_rate": 5e-05,
-      "loss": 0.3846,
+      "loss": 1.3503,
       "step": 60840
     },
     {
       "epoch": 0.9886110704943868,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3934,
+      "loss": 1.3846,
       "step": 60850
     },
     {
       "epoch": 0.9887735373917564,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.4075,
+      "loss": 1.3791,
       "step": 60860
     },
     {
       "epoch": 0.9889360042891261,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3933,
+      "loss": 1.4187,
       "step": 60870
     },
     {
       "epoch": 0.9890984711864957,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3893,
+      "loss": 1.421,
       "step": 60880
     },
     {
       "epoch": 0.9892609380838654,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.395,
+      "loss": 1.3915,
       "step": 60890
     },
     {
       "epoch": 0.989423404981235,
-      "grad_norm": 2.953125,
+      "grad_norm": 4.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.3864,
       "step": 60900
     },
     {
       "epoch": 0.9895858718786047,
-      "grad_norm": 2.125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3858,
+      "loss": 1.3805,
       "step": 60910
     },
     {
       "epoch": 0.9897483387759743,
-      "grad_norm": 3.03125,
+      "grad_norm": 4.5625,
       "learning_rate": 5e-05,
-      "loss": 0.405,
+      "loss": 1.3807,
       "step": 60920
     },
     {
       "epoch": 0.9899108056733441,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3731,
+      "loss": 1.4079,
       "step": 60930
     },
     {
       "epoch": 0.9900732725707138,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3915,
+      "loss": 1.3701,
       "step": 60940
     },
     {
       "epoch": 0.9902357394680834,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3812,
+      "loss": 1.3513,
       "step": 60950
     },
     {
       "epoch": 0.9903982063654531,
-      "grad_norm": 1.953125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3971,
+      "loss": 1.4388,
       "step": 60960
     },
     {
       "epoch": 0.9905606732628227,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.4114,
+      "loss": 1.4059,
       "step": 60970
     },
     {
       "epoch": 0.9907231401601924,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3723,
+      "loss": 1.3695,
       "step": 60980
     },
     {
       "epoch": 0.990885607057562,
-      "grad_norm": 3.796875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.3995,
       "step": 60990
     },
     {
       "epoch": 0.9910480739549317,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3797,
+      "loss": 1.3768,
       "step": 61000
     },
     {
       "epoch": 0.9912105408523013,
-      "grad_norm": 2.375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3935,
+      "loss": 1.3964,
       "step": 61010
     },
     {
       "epoch": 0.991373007749671,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3827,
+      "loss": 1.418,
       "step": 61020
     },
     {
       "epoch": 0.9915354746470406,
-      "grad_norm": 1.7109375,
+      "grad_norm": 5.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3954,
+      "loss": 1.4002,
       "step": 61030
     },
     {
       "epoch": 0.9916979415444104,
-      "grad_norm": 2.125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.385,
+      "loss": 1.4431,
       "step": 61040
     },
     {
       "epoch": 0.99186040844178,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.4021,
+      "loss": 1.4076,
       "step": 61050
     },
     {
       "epoch": 0.9920228753391497,
-      "grad_norm": 1.9140625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.4043,
+      "loss": 1.404,
       "step": 61060
     },
     {
       "epoch": 0.9921853422365193,
-      "grad_norm": 3.46875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3703,
+      "loss": 1.384,
       "step": 61070
     },
     {
       "epoch": 0.992347809133889,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3956,
+      "loss": 1.4062,
       "step": 61080
     },
     {
       "epoch": 0.9925102760312586,
-      "grad_norm": 1.8515625,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3888,
+      "loss": 1.4197,
       "step": 61090
     },
     {
       "epoch": 0.9926727429286283,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.3538,
       "step": 61100
     },
     {
       "epoch": 0.9928352098259979,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3952,
+      "loss": 1.4014,
       "step": 61110
     },
     {
       "epoch": 0.9929976767233676,
-      "grad_norm": 2.359375,
+      "grad_norm": 4.6875,
       "learning_rate": 5e-05,
-      "loss": 0.377,
+      "loss": 1.4156,
       "step": 61120
     },
     {
       "epoch": 0.9931601436207372,
-      "grad_norm": 1.6328125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3826,
+      "loss": 1.4127,
       "step": 61130
     },
     {
       "epoch": 0.9933226105181069,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3989,
+      "loss": 1.3881,
       "step": 61140
     },
     {
       "epoch": 0.9934850774154766,
-      "grad_norm": 2.234375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.399,
+      "loss": 1.384,
       "step": 61150
     },
     {
       "epoch": 0.9936475443128463,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3832,
+      "loss": 1.3987,
       "step": 61160
     },
     {
       "epoch": 0.993810011210216,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3918,
+      "loss": 1.3668,
       "step": 61170
     },
     {
       "epoch": 0.9939724781075856,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.383,
+      "loss": 1.3511,
       "step": 61180
     },
     {
       "epoch": 0.9941349450049553,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.4167,
+      "loss": 1.3619,
       "step": 61190
     },
     {
       "epoch": 0.9942974119023249,
-      "grad_norm": 2.4375,
+      "grad_norm": 4.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3945,
+      "loss": 1.3654,
       "step": 61200
     },
     {
       "epoch": 0.9944598787996946,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3691,
+      "loss": 1.4037,
       "step": 61210
     },
     {
       "epoch": 0.9946223456970642,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3659,
+      "loss": 1.4276,
       "step": 61220
     },
     {
       "epoch": 0.9947848125944339,
-      "grad_norm": 1.703125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.4012,
+      "loss": 1.3563,
       "step": 61230
     },
     {
       "epoch": 0.9949472794918035,
-      "grad_norm": 2.625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3997,
+      "loss": 1.3999,
       "step": 61240
     },
     {
       "epoch": 0.9951097463891732,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.372,
       "step": 61250
     },
     {
       "epoch": 0.9952722132865429,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3993,
+      "loss": 1.4162,
       "step": 61260
     },
     {
       "epoch": 0.9954346801839126,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3729,
+      "loss": 1.403,
       "step": 61270
     },
     {
       "epoch": 0.9955971470812822,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3833,
+      "loss": 1.4033,
       "step": 61280
     },
     {
       "epoch": 0.9957596139786519,
-      "grad_norm": 4.03125,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.4033,
+      "loss": 1.3688,
       "step": 61290
     },
     {
       "epoch": 0.9959220808760215,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.3962,
       "step": 61300
     },
     {
       "epoch": 0.9960845477733912,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.125,
       "learning_rate": 5e-05,
-      "loss": 0.3891,
+      "loss": 1.3694,
       "step": 61310
     },
     {
       "epoch": 0.9962470146707608,
-      "grad_norm": 3.109375,
+      "grad_norm": 5.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3735,
+      "loss": 1.4161,
       "step": 61320
     },
     {
       "epoch": 0.9964094815681305,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.4073,
+      "loss": 1.3891,
       "step": 61330
     },
     {
       "epoch": 0.9965719484655001,
-      "grad_norm": 1.8203125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3713,
+      "loss": 1.3847,
       "step": 61340
     },
     {
       "epoch": 0.9967344153628698,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3923,
+      "loss": 1.3967,
       "step": 61350
     },
     {
       "epoch": 0.9968968822602394,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3998,
+      "loss": 1.4093,
       "step": 61360
     },
     {
       "epoch": 0.9970593491576092,
-      "grad_norm": 4.90625,
+      "grad_norm": 5.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3973,
+      "loss": 1.347,
       "step": 61370
     },
     {
       "epoch": 0.9972218160549788,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3948,
+      "loss": 1.3492,
       "step": 61380
     },
     {
       "epoch": 0.9973842829523485,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.413,
+      "loss": 1.3706,
       "step": 61390
     },
     {
       "epoch": 0.9975467498497181,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3943,
+      "loss": 1.4111,
       "step": 61400
     },
     {
       "epoch": 0.9977092167470878,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3922,
+      "loss": 1.3395,
       "step": 61410
     },
     {
       "epoch": 0.9978716836444574,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3966,
+      "loss": 1.3695,
       "step": 61420
     },
     {
       "epoch": 0.9980341505418271,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3818,
+      "loss": 1.3588,
       "step": 61430
     },
     {
       "epoch": 0.9981966174391967,
-      "grad_norm": 1.8046875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3906,
+      "loss": 1.4165,
       "step": 61440
     },
     {
       "epoch": 0.9983590843365664,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3858,
+      "loss": 1.3826,
       "step": 61450
     },
     {
       "epoch": 0.998521551233936,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3951,
+      "loss": 1.3058,
       "step": 61460
     },
     {
       "epoch": 0.9986840181313057,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3816,
+      "loss": 1.3805,
       "step": 61470
     },
     {
       "epoch": 0.9988464850286755,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3861,
+      "loss": 1.3548,
       "step": 61480
     },
     {
       "epoch": 0.9990089519260451,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.371,
+      "loss": 1.3384,
       "step": 61490
     },
     {
       "epoch": 0.9991714188234148,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3849,
+      "loss": 1.369,
       "step": 61500
     },
     {
       "epoch": 0.9993338857207844,
-      "grad_norm": 1.9375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3821,
+      "loss": 1.3852,
       "step": 61510
     },
     {
       "epoch": 0.9994963526181541,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3634,
+      "loss": 1.3838,
       "step": 61520
     },
     {
       "epoch": 0.9996588195155237,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.03125,
       "learning_rate": 5e-05,
-      "loss": 0.386,
+      "loss": 1.3795,
       "step": 61530
     },
     {
       "epoch": 0.9998212864128934,
-      "grad_norm": 2.125,
+      "grad_norm": 5.0,
       "learning_rate": 5e-05,
-      "loss": 0.3868,
+      "loss": 1.3781,
       "step": 61540
     },
     {
       "epoch": 0.999983753310263,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3949,
+      "loss": 1.3969,
       "step": 61550
     },
     {
       "epoch": 1.0001462202076328,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3714,
+      "loss": 1.2278,
       "step": 61560
     },
     {
       "epoch": 1.0003086871050024,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3737,
+      "loss": 1.2029,
       "step": 61570
     },
     {
       "epoch": 1.000471154002372,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3763,
+      "loss": 1.1922,
       "step": 61580
     },
     {
       "epoch": 1.0006336208997417,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.1943,
       "step": 61590
     },
     {
       "epoch": 1.0007960877971114,
-      "grad_norm": 2.015625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3643,
+      "loss": 1.2199,
       "step": 61600
     },
     {
       "epoch": 1.000958554694481,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3871,
+      "loss": 1.2117,
       "step": 61610
     },
     {
       "epoch": 1.0011210215918507,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3678,
+      "loss": 1.2091,
       "step": 61620
     },
     {
       "epoch": 1.0012834884892203,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2011,
       "step": 61630
     },
     {
       "epoch": 1.00144595538659,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.1871,
       "step": 61640
     },
     {
       "epoch": 1.0016084222839596,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3772,
+      "loss": 1.2293,
       "step": 61650
     },
     {
       "epoch": 1.0017708891813293,
-      "grad_norm": 1.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.1513,
       "step": 61660
     },
     {
       "epoch": 1.001933356078699,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3702,
+      "loss": 1.214,
       "step": 61670
     },
     {
       "epoch": 1.0020958229760686,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2189,
       "step": 61680
     },
     {
       "epoch": 1.0022582898734382,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3706,
+      "loss": 1.217,
       "step": 61690
     },
     {
       "epoch": 1.0024207567708079,
-      "grad_norm": 2.125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3703,
+      "loss": 1.219,
       "step": 61700
     },
     {
       "epoch": 1.0025832236681775,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.1911,
       "step": 61710
     },
     {
       "epoch": 1.0027456905655472,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.1718,
       "step": 61720
     },
     {
       "epoch": 1.0029081574629168,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3752,
+      "loss": 1.2012,
       "step": 61730
     },
     {
       "epoch": 1.0030706243602865,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1915,
       "step": 61740
     },
     {
       "epoch": 1.0032330912576564,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3761,
+      "loss": 1.1876,
       "step": 61750
     },
     {
       "epoch": 1.003395558155026,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3641,
+      "loss": 1.2135,
       "step": 61760
     },
     {
       "epoch": 1.0035580250523957,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2014,
       "step": 61770
     },
     {
       "epoch": 1.0037204919497653,
-      "grad_norm": 1.875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3592,
+      "loss": 1.2054,
       "step": 61780
     },
     {
       "epoch": 1.003882958847135,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3674,
+      "loss": 1.1835,
       "step": 61790
     },
     {
       "epoch": 1.0040454257445046,
-      "grad_norm": 1.6875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3708,
+      "loss": 1.1748,
       "step": 61800
     },
     {
       "epoch": 1.0042078926418743,
-      "grad_norm": 1.765625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3879,
+      "loss": 1.2187,
       "step": 61810
     },
     {
       "epoch": 1.004370359539244,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.2015,
       "step": 61820
     },
     {
       "epoch": 1.0045328264366136,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3741,
+      "loss": 1.1648,
       "step": 61830
     },
     {
       "epoch": 1.0046952933339832,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.1731,
       "step": 61840
     },
     {
       "epoch": 1.0048577602313529,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3644,
+      "loss": 1.1395,
       "step": 61850
     },
     {
       "epoch": 1.0050202271287225,
-      "grad_norm": 2.34375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3681,
+      "loss": 1.1846,
       "step": 61860
     },
     {
       "epoch": 1.0051826940260922,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3794,
+      "loss": 1.1958,
       "step": 61870
     },
     {
       "epoch": 1.0053451609234618,
-      "grad_norm": 1.6875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.199,
       "step": 61880
     },
     {
       "epoch": 1.0055076278208315,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3744,
+      "loss": 1.2076,
       "step": 61890
     },
     {
       "epoch": 1.0056700947182011,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.2138,
       "step": 61900
     },
     {
       "epoch": 1.0058325616155708,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3627,
+      "loss": 1.1961,
       "step": 61910
     },
     {
       "epoch": 1.0059950285129404,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3527,
+      "loss": 1.1882,
       "step": 61920
     },
     {
       "epoch": 1.00615749541031,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3577,
+      "loss": 1.1865,
       "step": 61930
     },
     {
       "epoch": 1.0063199623076797,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3635,
+      "loss": 1.1614,
       "step": 61940
     },
     {
       "epoch": 1.0064824292050494,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.1997,
       "step": 61950
     },
     {
       "epoch": 1.006644896102419,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3763,
+      "loss": 1.1621,
       "step": 61960
     },
     {
       "epoch": 1.006807362999789,
-      "grad_norm": 1.65625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2298,
       "step": 61970
     },
     {
       "epoch": 1.0069698298971586,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.1896,
       "step": 61980
     },
     {
       "epoch": 1.0071322967945282,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3697,
+      "loss": 1.1943,
       "step": 61990
     },
     {
       "epoch": 1.0072947636918979,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.1889,
       "step": 62000
     },
     {
       "epoch": 1.0074572305892675,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.1806,
       "step": 62010
     },
     {
       "epoch": 1.0076196974866372,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.1887,
       "step": 62020
     },
     {
       "epoch": 1.0077821643840068,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3639,
+      "loss": 1.1687,
       "step": 62030
     },
     {
       "epoch": 1.0079446312813765,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3726,
+      "loss": 1.1752,
       "step": 62040
     },
     {
       "epoch": 1.008107098178746,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.1571,
       "step": 62050
     },
     {
       "epoch": 1.0082695650761158,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.1832,
       "step": 62060
     },
     {
       "epoch": 1.0084320319734854,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.228,
       "step": 62070
     },
     {
       "epoch": 1.008594498870855,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3672,
+      "loss": 1.1921,
       "step": 62080
     },
     {
       "epoch": 1.0087569657682247,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.1621,
       "step": 62090
     },
     {
       "epoch": 1.0089194326655944,
-      "grad_norm": 2.375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.1707,
       "step": 62100
     },
     {
       "epoch": 1.009081899562964,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.1689,
       "step": 62110
     },
     {
       "epoch": 1.0092443664603337,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.2105,
       "step": 62120
     },
     {
       "epoch": 1.0094068333577033,
-      "grad_norm": 2.125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.1817,
       "step": 62130
     },
     {
       "epoch": 1.009569300255073,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1692,
       "step": 62140
     },
     {
       "epoch": 1.0097317671524426,
-      "grad_norm": 2.625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2034,
       "step": 62150
     },
     {
       "epoch": 1.0098942340498123,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3743,
+      "loss": 1.1998,
       "step": 62160
     },
     {
       "epoch": 1.010056700947182,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3628,
+      "loss": 1.1921,
       "step": 62170
     },
     {
       "epoch": 1.0102191678445516,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2233,
       "step": 62180
     },
     {
       "epoch": 1.0103816347419214,
-      "grad_norm": 2.875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1626,
       "step": 62190
     },
     {
       "epoch": 1.010544101639291,
-      "grad_norm": 2.875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.1823,
       "step": 62200
     },
     {
       "epoch": 1.0107065685366607,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1991,
       "step": 62210
     },
     {
       "epoch": 1.0108690354340304,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1506,
       "step": 62220
     },
     {
       "epoch": 1.0110315023314,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3665,
+      "loss": 1.1798,
       "step": 62230
     },
     {
       "epoch": 1.0111939692287697,
-      "grad_norm": 1.765625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1813,
       "step": 62240
     },
     {
       "epoch": 1.0113564361261393,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3682,
+      "loss": 1.1405,
       "step": 62250
     },
     {
       "epoch": 1.011518903023509,
-      "grad_norm": 3.125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1787,
       "step": 62260
     },
     {
       "epoch": 1.0116813699208786,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.1389,
       "step": 62270
     },
     {
       "epoch": 1.0118438368182483,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.208,
       "step": 62280
     },
     {
       "epoch": 1.012006303715618,
-      "grad_norm": 1.71875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.2157,
       "step": 62290
     },
     {
       "epoch": 1.0121687706129876,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1922,
       "step": 62300
     },
     {
       "epoch": 1.0123312375103573,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.1722,
       "step": 62310
     },
     {
       "epoch": 1.012493704407727,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.1606,
       "step": 62320
     },
     {
       "epoch": 1.0126561713050966,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.1752,
       "step": 62330
     },
     {
       "epoch": 1.0128186382024662,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.1811,
       "step": 62340
     },
     {
       "epoch": 1.0129811050998359,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1505,
       "step": 62350
     },
     {
       "epoch": 1.0131435719972055,
-      "grad_norm": 1.796875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2128,
       "step": 62360
     },
     {
       "epoch": 1.0133060388945752,
-      "grad_norm": 2.625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1846,
       "step": 62370
     },
     {
       "epoch": 1.0134685057919448,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.1987,
       "step": 62380
     },
     {
       "epoch": 1.0136309726893145,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.1887,
       "step": 62390
     },
     {
       "epoch": 1.013793439586684,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2178,
       "step": 62400
     },
     {
       "epoch": 1.013955906484054,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.1116,
       "step": 62410
     },
     {
       "epoch": 1.0141183733814236,
-      "grad_norm": 2.171875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.1899,
       "step": 62420
     },
     {
       "epoch": 1.0142808402787933,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3728,
+      "loss": 1.2185,
       "step": 62430
     },
     {
       "epoch": 1.014443307176163,
-      "grad_norm": 2.875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.1724,
       "step": 62440
     },
     {
       "epoch": 1.0146057740735326,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2147,
       "step": 62450
     },
     {
       "epoch": 1.0147682409709022,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.1946,
       "step": 62460
     },
     {
       "epoch": 1.0149307078682719,
-      "grad_norm": 1.71875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2086,
       "step": 62470
     },
     {
       "epoch": 1.0150931747656415,
-      "grad_norm": 1.8046875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.1819,
       "step": 62480
     },
     {
       "epoch": 1.0152556416630112,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2172,
       "step": 62490
     },
     {
       "epoch": 1.0154181085603808,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.1612,
       "step": 62500
     },
     {
       "epoch": 1.0155805754577505,
-      "grad_norm": 1.6484375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2379,
       "step": 62510
     },
     {
       "epoch": 1.0157430423551201,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1998,
       "step": 62520
     },
     {
       "epoch": 1.0159055092524898,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2161,
       "step": 62530
     },
     {
       "epoch": 1.0160679761498594,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.1867,
       "step": 62540
     },
     {
       "epoch": 1.016230443047229,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2192,
       "step": 62550
     },
     {
       "epoch": 1.0163929099445987,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.1655,
       "step": 62560
     },
     {
       "epoch": 1.0165553768419684,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1533,
       "step": 62570
     },
     {
       "epoch": 1.016717843739338,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.1936,
       "step": 62580
     },
     {
       "epoch": 1.0168803106367077,
-      "grad_norm": 2.234375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3789,
+      "loss": 1.1643,
       "step": 62590
     },
     {
       "epoch": 1.0170427775340773,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.224,
       "step": 62600
     },
     {
       "epoch": 1.017205244431447,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2241,
       "step": 62610
     },
     {
       "epoch": 1.0173677113288166,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.2179,
       "step": 62620
     },
     {
       "epoch": 1.0175301782261865,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.1602,
       "step": 62630
     },
     {
       "epoch": 1.0176926451235562,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1754,
       "step": 62640
     },
     {
       "epoch": 1.0178551120209258,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.1663,
       "step": 62650
     },
     {
       "epoch": 1.0180175789182955,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2041,
       "step": 62660
     },
     {
       "epoch": 1.0181800458156651,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2224,
       "step": 62670
     },
     {
       "epoch": 1.0183425127130348,
-      "grad_norm": 1.7734375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2053,
       "step": 62680
     },
     {
       "epoch": 1.0185049796104044,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.1906,
       "step": 62690
     },
     {
       "epoch": 1.018667446507774,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2272,
       "step": 62700
     },
     {
       "epoch": 1.0188299134051437,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.1412,
       "step": 62710
     },
     {
       "epoch": 1.0189923803025134,
-      "grad_norm": 1.9140625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2045,
       "step": 62720
     },
     {
       "epoch": 1.019154847199883,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.1977,
       "step": 62730
     },
     {
       "epoch": 1.0193173140972527,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2099,
       "step": 62740
     },
     {
       "epoch": 1.0194797809946223,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.171,
       "step": 62750
     },
     {
       "epoch": 1.019642247891992,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1901,
       "step": 62760
     },
     {
       "epoch": 1.0198047147893616,
-      "grad_norm": 3.734375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3652,
+      "loss": 1.1862,
       "step": 62770
     },
     {
       "epoch": 1.0199671816867313,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2037,
       "step": 62780
     },
     {
       "epoch": 1.020129648584101,
-      "grad_norm": 1.8203125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1813,
       "step": 62790
     },
     {
       "epoch": 1.0202921154814706,
-      "grad_norm": 1.53125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2358,
       "step": 62800
     },
     {
       "epoch": 1.0204545823788402,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2132,
       "step": 62810
     },
     {
       "epoch": 1.0206170492762099,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2029,
       "step": 62820
     },
     {
       "epoch": 1.0207795161735795,
-      "grad_norm": 2.5,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2101,
       "step": 62830
     },
     {
       "epoch": 1.0209419830709492,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1785,
       "step": 62840
     },
     {
       "epoch": 1.021104449968319,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.1915,
       "step": 62850
     },
     {
       "epoch": 1.0212669168656887,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2677,
       "step": 62860
     },
     {
       "epoch": 1.0214293837630584,
-      "grad_norm": 2.75,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.1539,
       "step": 62870
     },
     {
       "epoch": 1.021591850660428,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1698,
       "step": 62880
     },
     {
       "epoch": 1.0217543175577977,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.1612,
       "step": 62890
     },
     {
       "epoch": 1.0219167844551673,
-      "grad_norm": 1.7578125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.1778,
       "step": 62900
     },
     {
       "epoch": 1.022079251352537,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2098,
       "step": 62910
     },
     {
       "epoch": 1.0222417182499066,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2328,
       "step": 62920
     },
     {
       "epoch": 1.0224041851472763,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.2155,
       "step": 62930
     },
     {
       "epoch": 1.022566652044646,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1776,
       "step": 62940
     },
     {
       "epoch": 1.0227291189420156,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2207,
       "step": 62950
     },
     {
       "epoch": 1.0228915858393852,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.1528,
       "step": 62960
     },
     {
       "epoch": 1.0230540527367549,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2196,
       "step": 62970
     },
     {
       "epoch": 1.0232165196341245,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3721,
+      "loss": 1.1813,
       "step": 62980
     },
     {
       "epoch": 1.0233789865314942,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.1409,
       "step": 62990
     },
     {
       "epoch": 1.0235414534288638,
-      "grad_norm": 4.125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.1889,
       "step": 63000
     },
     {
       "epoch": 1.0237039203262335,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2012,
       "step": 63010
     },
     {
       "epoch": 1.0238663872236031,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.1859,
       "step": 63020
     },
     {
       "epoch": 1.0240288541209728,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.1754,
       "step": 63030
     },
     {
       "epoch": 1.0241913210183424,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1932,
       "step": 63040
     },
     {
       "epoch": 1.024353787915712,
-      "grad_norm": 2.375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1974,
       "step": 63050
     },
     {
       "epoch": 1.0245162548130817,
-      "grad_norm": 1.859375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.1686,
       "step": 63060
     },
     {
       "epoch": 1.0246787217104516,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2093,
       "step": 63070
     },
     {
       "epoch": 1.0248411886078213,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.1729,
       "step": 63080
     },
     {
       "epoch": 1.025003655505191,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1641,
       "step": 63090
     },
     {
       "epoch": 1.0251661224025606,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2128,
       "step": 63100
     },
     {
       "epoch": 1.0253285892999302,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2231,
       "step": 63110
     },
     {
       "epoch": 1.0254910561972999,
-      "grad_norm": 2.25,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3566,
+      "loss": 1.1647,
       "step": 63120
     },
     {
       "epoch": 1.0256535230946695,
-      "grad_norm": 2.625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1605,
       "step": 63130
     },
     {
       "epoch": 1.0258159899920392,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1514,
       "step": 63140
     },
     {
       "epoch": 1.0259784568894088,
-      "grad_norm": 2.5,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2058,
       "step": 63150
     },
     {
       "epoch": 1.0261409237867785,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2194,
       "step": 63160
     },
     {
       "epoch": 1.026303390684148,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.1872,
       "step": 63170
     },
     {
       "epoch": 1.0264658575815178,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1624,
       "step": 63180
     },
     {
       "epoch": 1.0266283244788874,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2174,
       "step": 63190
     },
     {
       "epoch": 1.026790791376257,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1994,
       "step": 63200
     },
     {
       "epoch": 1.0269532582736267,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1642,
       "step": 63210
     },
     {
       "epoch": 1.0271157251709964,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2077,
       "step": 63220
     },
     {
       "epoch": 1.027278192068366,
-      "grad_norm": 1.921875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2101,
       "step": 63230
     },
     {
       "epoch": 1.0274406589657357,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2005,
       "step": 63240
     },
     {
       "epoch": 1.0276031258631053,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1715,
       "step": 63250
     },
     {
       "epoch": 1.027765592760475,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2012,
       "step": 63260
     },
     {
       "epoch": 1.0279280596578446,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3527,
+      "loss": 1.2116,
       "step": 63270
     },
     {
       "epoch": 1.0280905265552143,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2016,
       "step": 63280
     },
     {
       "epoch": 1.0282529934525841,
-      "grad_norm": 1.859375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3185,
+      "loss": 1.1347,
       "step": 63290
     },
     {
       "epoch": 1.0284154603499538,
-      "grad_norm": 1.8203125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2153,
       "step": 63300
     },
     {
       "epoch": 1.0285779272473234,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.177,
       "step": 63310
     },
     {
       "epoch": 1.028740394144693,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2003,
       "step": 63320
     },
     {
       "epoch": 1.0289028610420627,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2054,
       "step": 63330
     },
     {
       "epoch": 1.0290653279394324,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3153,
+      "loss": 1.2152,
       "step": 63340
     },
     {
       "epoch": 1.029227794836802,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.1932,
       "step": 63350
     },
     {
       "epoch": 1.0293902617341717,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2224,
       "step": 63360
     },
     {
       "epoch": 1.0295527286315413,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2326,
       "step": 63370
     },
     {
       "epoch": 1.029715195528911,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2051,
       "step": 63380
     },
     {
       "epoch": 1.0298776624262806,
-      "grad_norm": 1.90625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.1544,
       "step": 63390
     },
     {
       "epoch": 1.0300401293236503,
-      "grad_norm": 1.8203125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2113,
       "step": 63400
     },
     {
       "epoch": 1.03020259622102,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1594,
       "step": 63410
     },
     {
       "epoch": 1.0303650631183896,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.202,
       "step": 63420
     },
     {
       "epoch": 1.0305275300157593,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2421,
       "step": 63430
     },
     {
       "epoch": 1.030689996913129,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2149,
       "step": 63440
     },
     {
       "epoch": 1.0308524638104986,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2048,
       "step": 63450
     },
     {
       "epoch": 1.0310149307078682,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.1792,
       "step": 63460
     },
     {
       "epoch": 1.0311773976052379,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.2242,
       "step": 63470
     },
     {
       "epoch": 1.0313398645026075,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1883,
       "step": 63480
     },
     {
       "epoch": 1.0315023313999772,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.1719,
       "step": 63490
     },
     {
       "epoch": 1.031664798297347,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.2331,
       "step": 63500
     },
     {
       "epoch": 1.0318272651947167,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3622,
+      "loss": 1.1997,
       "step": 63510
     },
     {
       "epoch": 1.0319897320920863,
-      "grad_norm": 1.6640625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.1836,
       "step": 63520
     },
     {
       "epoch": 1.032152198989456,
-      "grad_norm": 1.9140625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.169,
       "step": 63530
     },
     {
       "epoch": 1.0323146658868256,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2134,
       "step": 63540
     },
     {
       "epoch": 1.0324771327841953,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3757,
+      "loss": 1.1993,
       "step": 63550
     },
     {
       "epoch": 1.032639599681565,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2129,
       "step": 63560
     },
     {
       "epoch": 1.0328020665789346,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.1246,
       "step": 63570
     },
     {
       "epoch": 1.0329645334763042,
-      "grad_norm": 2.109375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1771,
       "step": 63580
     },
     {
       "epoch": 1.0331270003736739,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.1674,
       "step": 63590
     },
     {
       "epoch": 1.0332894672710435,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2165,
       "step": 63600
     },
     {
       "epoch": 1.0334519341684132,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2009,
       "step": 63610
     },
     {
       "epoch": 1.0336144010657828,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2077,
       "step": 63620
     },
     {
       "epoch": 1.0337768679631525,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2277,
       "step": 63630
     },
     {
       "epoch": 1.0339393348605221,
-      "grad_norm": 1.8828125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.1949,
       "step": 63640
     },
     {
       "epoch": 1.0341018017578918,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.2312,
       "step": 63650
     },
     {
       "epoch": 1.0342642686552614,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.1784,
       "step": 63660
     },
     {
       "epoch": 1.034426735552631,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.1996,
       "step": 63670
     },
     {
       "epoch": 1.0345892024500007,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2149,
       "step": 63680
     },
     {
       "epoch": 1.0347516693473704,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2293,
       "step": 63690
     },
     {
       "epoch": 1.03491413624474,
-      "grad_norm": 2.25,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.1722,
       "step": 63700
     },
     {
       "epoch": 1.0350766031421097,
-      "grad_norm": 1.765625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2103,
       "step": 63710
     },
     {
       "epoch": 1.0352390700394793,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2045,
       "step": 63720
     },
     {
       "epoch": 1.0354015369368492,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.183,
       "step": 63730
     },
     {
       "epoch": 1.0355640038342189,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.1816,
       "step": 63740
     },
     {
       "epoch": 1.0357264707315885,
-      "grad_norm": 2.078125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.1746,
       "step": 63750
     },
     {
       "epoch": 1.0358889376289582,
-      "grad_norm": 2.875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2104,
       "step": 63760
     },
     {
       "epoch": 1.0360514045263278,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2002,
       "step": 63770
     },
     {
       "epoch": 1.0362138714236975,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.1688,
       "step": 63780
     },
     {
       "epoch": 1.0363763383210671,
-      "grad_norm": 2.75,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.1638,
       "step": 63790
     },
     {
       "epoch": 1.0365388052184368,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2015,
       "step": 63800
     },
     {
       "epoch": 1.0367012721158064,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.1656,
       "step": 63810
     },
     {
       "epoch": 1.036863739013176,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.1849,
       "step": 63820
     },
     {
       "epoch": 1.0370262059105457,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.1716,
       "step": 63830
     },
     {
       "epoch": 1.0371886728079154,
-      "grad_norm": 2.875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.1789,
       "step": 63840
     },
     {
       "epoch": 1.037351139705285,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.196,
       "step": 63850
     },
     {
       "epoch": 1.0375136066026547,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.1776,
       "step": 63860
     },
     {
       "epoch": 1.0376760735000243,
-      "grad_norm": 2.125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1905,
       "step": 63870
     },
     {
       "epoch": 1.037838540397394,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.1819,
       "step": 63880
     },
     {
       "epoch": 1.0380010072947636,
-      "grad_norm": 2.75,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2332,
       "step": 63890
     },
     {
       "epoch": 1.0381634741921333,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.215,
       "step": 63900
     },
     {
       "epoch": 1.038325941089503,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.2069,
       "step": 63910
     },
     {
       "epoch": 1.0384884079868726,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3176,
+      "loss": 1.1559,
       "step": 63920
     },
     {
       "epoch": 1.0386508748842422,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3646,
+      "loss": 1.1591,
       "step": 63930
     },
     {
       "epoch": 1.038813341781612,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2062,
       "step": 63940
     },
     {
       "epoch": 1.0389758086789818,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.1825,
       "step": 63950
     },
     {
       "epoch": 1.0391382755763514,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.235,
       "step": 63960
     },
     {
       "epoch": 1.039300742473721,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.1746,
       "step": 63970
     },
     {
       "epoch": 1.0394632093710907,
-      "grad_norm": 2.125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.1661,
       "step": 63980
     },
     {
       "epoch": 1.0396256762684604,
-      "grad_norm": 1.734375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.203,
       "step": 63990
     },
     {
       "epoch": 1.03978814316583,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.1928,
       "step": 64000
     },
     {
       "epoch": 1.0399506100631997,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1851,
       "step": 64010
     },
     {
       "epoch": 1.0401130769605693,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.324,
+      "loss": 1.1632,
       "step": 64020
     },
     {
       "epoch": 1.040275543857939,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.1665,
       "step": 64030
     },
     {
       "epoch": 1.0404380107553086,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3188,
+      "loss": 1.2179,
       "step": 64040
     },
     {
       "epoch": 1.0406004776526783,
-      "grad_norm": 2.25,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.1944,
       "step": 64050
     },
     {
       "epoch": 1.040762944550048,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2031,
       "step": 64060
     },
     {
       "epoch": 1.0409254114474176,
-      "grad_norm": 2.25,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2351,
       "step": 64070
     },
     {
       "epoch": 1.0410878783447872,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.1762,
       "step": 64080
     },
     {
       "epoch": 1.0412503452421569,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1806,
       "step": 64090
     },
     {
       "epoch": 1.0414128121395265,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2337,
       "step": 64100
     },
     {
       "epoch": 1.0415752790368962,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2234,
       "step": 64110
     },
     {
       "epoch": 1.0417377459342658,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.1756,
       "step": 64120
     },
     {
       "epoch": 1.0419002128316355,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3208,
+      "loss": 1.1504,
       "step": 64130
     },
     {
       "epoch": 1.0420626797290051,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.1906,
       "step": 64140
     },
     {
       "epoch": 1.0422251466263748,
-      "grad_norm": 2.0,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1635,
       "step": 64150
     },
     {
       "epoch": 1.0423876135237444,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.202,
       "step": 64160
     },
     {
       "epoch": 1.0425500804211143,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.1756,
       "step": 64170
     },
     {
       "epoch": 1.042712547318484,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1792,
       "step": 64180
     },
     {
       "epoch": 1.0428750142158536,
-      "grad_norm": 1.875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.1404,
       "step": 64190
     },
     {
       "epoch": 1.0430374811132233,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.233,
       "step": 64200
     },
     {
       "epoch": 1.043199948010593,
-      "grad_norm": 1.90625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2174,
       "step": 64210
     },
     {
       "epoch": 1.0433624149079626,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.1625,
       "step": 64220
     },
     {
       "epoch": 1.0435248818053322,
-      "grad_norm": 2.625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.1956,
       "step": 64230
     },
     {
       "epoch": 1.0436873487027019,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.1594,
       "step": 64240
     },
     {
       "epoch": 1.0438498156000715,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.1981,
       "step": 64250
     },
     {
       "epoch": 1.0440122824974412,
-      "grad_norm": 2.375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.158,
       "step": 64260
     },
     {
       "epoch": 1.0441747493948108,
-      "grad_norm": 1.8359375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1497,
       "step": 64270
     },
     {
       "epoch": 1.0443372162921805,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.1802,
       "step": 64280
     },
     {
       "epoch": 1.04449968318955,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2069,
       "step": 64290
     },
     {
       "epoch": 1.0446621500869198,
-      "grad_norm": 2.25,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2272,
       "step": 64300
     },
     {
       "epoch": 1.0448246169842894,
-      "grad_norm": 1.90625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1776,
       "step": 64310
     },
     {
       "epoch": 1.044987083881659,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.1939,
       "step": 64320
     },
     {
       "epoch": 1.0451495507790287,
-      "grad_norm": 1.640625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.1989,
       "step": 64330
     },
     {
       "epoch": 1.0453120176763984,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3619,
+      "loss": 1.1754,
       "step": 64340
     },
     {
       "epoch": 1.045474484573768,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2027,
       "step": 64350
     },
     {
       "epoch": 1.0456369514711377,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1884,
       "step": 64360
     },
     {
       "epoch": 1.0457994183685073,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2151,
       "step": 64370
     },
     {
       "epoch": 1.0459618852658772,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3751,
+      "loss": 1.1785,
       "step": 64380
     },
     {
       "epoch": 1.0461243521632468,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1474,
       "step": 64390
     },
     {
       "epoch": 1.0462868190606165,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.313,
+      "loss": 1.1823,
       "step": 64400
     },
     {
       "epoch": 1.0464492859579861,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3239,
+      "loss": 1.2238,
       "step": 64410
     },
     {
       "epoch": 1.0466117528553558,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2361,
       "step": 64420
     },
     {
       "epoch": 1.0467742197527254,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2354,
       "step": 64430
     },
     {
       "epoch": 1.046936686650095,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.1948,
       "step": 64440
     },
     {
       "epoch": 1.0470991535474647,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3193,
+      "loss": 1.1919,
       "step": 64450
     },
     {
       "epoch": 1.0472616204448344,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.1718,
       "step": 64460
     },
     {
       "epoch": 1.047424087342204,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2082,
       "step": 64470
     },
     {
       "epoch": 1.0475865542395737,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3104,
+      "loss": 1.1891,
       "step": 64480
     },
     {
       "epoch": 1.0477490211369433,
-      "grad_norm": 2.125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2075,
       "step": 64490
     },
     {
       "epoch": 1.047911488034313,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2424,
       "step": 64500
     },
     {
       "epoch": 1.0480739549316826,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.191,
       "step": 64510
     },
     {
       "epoch": 1.0482364218290523,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2176,
       "step": 64520
     },
     {
       "epoch": 1.048398888726422,
-      "grad_norm": 1.671875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.1841,
       "step": 64530
     },
     {
       "epoch": 1.0485613556237916,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.1796,
       "step": 64540
     },
     {
       "epoch": 1.0487238225211613,
-      "grad_norm": 1.6953125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.1974,
       "step": 64550
     },
     {
       "epoch": 1.048886289418531,
-      "grad_norm": 2.0,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.1954,
       "step": 64560
     },
     {
       "epoch": 1.0490487563159006,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.1779,
       "step": 64570
     },
     {
       "epoch": 1.0492112232132702,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2226,
       "step": 64580
     },
     {
       "epoch": 1.0493736901106399,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.1989,
       "step": 64590
     },
     {
       "epoch": 1.0495361570080095,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3115,
+      "loss": 1.2077,
       "step": 64600
     },
     {
       "epoch": 1.0496986239053794,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.1766,
       "step": 64610
     },
     {
       "epoch": 1.049861090802749,
-      "grad_norm": 1.90625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.1821,
       "step": 64620
     },
     {
       "epoch": 1.0500235577001187,
-      "grad_norm": 1.7265625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.1988,
       "step": 64630
     },
     {
       "epoch": 1.0501860245974883,
-      "grad_norm": 1.5625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.2067,
       "step": 64640
     },
     {
       "epoch": 1.050348491494858,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.168,
       "step": 64650
     },
     {
       "epoch": 1.0505109583922276,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2019,
       "step": 64660
     },
     {
       "epoch": 1.0506734252895973,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2292,
       "step": 64670
     },
     {
       "epoch": 1.050835892186967,
-      "grad_norm": 1.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2012,
       "step": 64680
     },
     {
       "epoch": 1.0509983590843366,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2155,
       "step": 64690
     },
     {
       "epoch": 1.0511608259817062,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3184,
+      "loss": 1.1816,
       "step": 64700
     },
     {
       "epoch": 1.0513232928790759,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.1773,
       "step": 64710
     },
     {
       "epoch": 1.0514857597764455,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.1902,
       "step": 64720
     },
     {
       "epoch": 1.0516482266738152,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1963,
       "step": 64730
     },
     {
       "epoch": 1.0518106935711848,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2117,
       "step": 64740
     },
     {
       "epoch": 1.0519731604685545,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.1857,
       "step": 64750
     },
     {
       "epoch": 1.0521356273659241,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2063,
       "step": 64760
     },
     {
       "epoch": 1.0522980942632938,
-      "grad_norm": 2.015625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2148,
       "step": 64770
     },
     {
       "epoch": 1.0524605611606634,
-      "grad_norm": 1.625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2216,
       "step": 64780
     },
     {
       "epoch": 1.052623028058033,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3632,
+      "loss": 1.1729,
       "step": 64790
     },
     {
       "epoch": 1.0527854949554027,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2233,
       "step": 64800
     },
     {
       "epoch": 1.0529479618527724,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.1283,
       "step": 64810
     },
     {
       "epoch": 1.0531104287501423,
-      "grad_norm": 2.25,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3188,
+      "loss": 1.1725,
       "step": 64820
     },
     {
       "epoch": 1.053272895647512,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 1.1634,
       "step": 64830
     },
     {
       "epoch": 1.0534353625448816,
-      "grad_norm": 2.125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2178,
       "step": 64840
     },
     {
       "epoch": 1.0535978294422512,
-      "grad_norm": 1.8125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2094,
       "step": 64850
     },
     {
       "epoch": 1.0537602963396209,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.1469,
       "step": 64860
     },
     {
       "epoch": 1.0539227632369905,
-      "grad_norm": 2.625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.169,
       "step": 64870
     },
     {
       "epoch": 1.0540852301343602,
-      "grad_norm": 1.9453125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.1871,
       "step": 64880
     },
     {
       "epoch": 1.0542476970317298,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2205,
       "step": 64890
     },
     {
       "epoch": 1.0544101639290995,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.1915,
       "step": 64900
     },
     {
       "epoch": 1.0545726308264691,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2128,
       "step": 64910
     },
     {
       "epoch": 1.0547350977238388,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3139,
+      "loss": 1.2509,
       "step": 64920
     },
     {
       "epoch": 1.0548975646212084,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2107,
       "step": 64930
     },
     {
       "epoch": 1.055060031518578,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.1828,
       "step": 64940
     },
     {
       "epoch": 1.0552224984159477,
-      "grad_norm": 4.3125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2229,
       "step": 64950
     },
     {
       "epoch": 1.0553849653133174,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2197,
       "step": 64960
     },
     {
       "epoch": 1.055547432210687,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2088,
       "step": 64970
     },
     {
       "epoch": 1.0557098991080567,
-      "grad_norm": 2.0,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.1769,
       "step": 64980
     },
     {
       "epoch": 1.0558723660054263,
-      "grad_norm": 1.6875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.1913,
       "step": 64990
     },
     {
       "epoch": 1.056034832902796,
-      "grad_norm": 1.921875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.2077,
       "step": 65000
     },
     {
       "epoch": 1.0561972998001656,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.2128,
       "step": 65010
     },
     {
       "epoch": 1.0563597666975353,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1567,
       "step": 65020
     },
     {
       "epoch": 1.056522233594905,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2273,
       "step": 65030
     },
     {
       "epoch": 1.0566847004922746,
-      "grad_norm": 2.5,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.1789,
       "step": 65040
     },
     {
       "epoch": 1.0568471673896445,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.1962,
       "step": 65050
     },
     {
       "epoch": 1.057009634287014,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.1588,
       "step": 65060
     },
     {
       "epoch": 1.0571721011843838,
-      "grad_norm": 2.0,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2559,
       "step": 65070
     },
     {
       "epoch": 1.0573345680817534,
-      "grad_norm": 1.796875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.1924,
       "step": 65080
     },
     {
       "epoch": 1.057497034979123,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2158,
       "step": 65090
     },
     {
       "epoch": 1.0576595018764927,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.1608,
       "step": 65100
     },
     {
       "epoch": 1.0578219687738624,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3171,
+      "loss": 1.2599,
       "step": 65110
     },
     {
       "epoch": 1.057984435671232,
-      "grad_norm": 2.0,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2335,
       "step": 65120
     },
     {
       "epoch": 1.0581469025686017,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2099,
       "step": 65130
     },
     {
       "epoch": 1.0583093694659713,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2273,
       "step": 65140
     },
     {
       "epoch": 1.058471836363341,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.1621,
       "step": 65150
     },
     {
       "epoch": 1.0586343032607106,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2114,
       "step": 65160
     },
     {
       "epoch": 1.0587967701580803,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1561,
       "step": 65170
     },
     {
       "epoch": 1.05895923705545,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.1622,
       "step": 65180
     },
     {
       "epoch": 1.0591217039528196,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.1633,
       "step": 65190
     },
     {
       "epoch": 1.0592841708501892,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.1506,
       "step": 65200
     },
     {
       "epoch": 1.0594466377475589,
-      "grad_norm": 2.5,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3224,
+      "loss": 1.1947,
       "step": 65210
     },
     {
       "epoch": 1.0596091046449285,
-      "grad_norm": 3.25,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.179,
       "step": 65220
     },
     {
       "epoch": 1.0597715715422982,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.1852,
       "step": 65230
     },
     {
       "epoch": 1.0599340384396678,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.323,
+      "loss": 1.2037,
       "step": 65240
     },
     {
       "epoch": 1.0600965053370375,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.206,
       "step": 65250
     },
     {
       "epoch": 1.0602589722344073,
-      "grad_norm": 2.28125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.188,
       "step": 65260
     },
     {
       "epoch": 1.060421439131777,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.1773,
       "step": 65270
     },
     {
       "epoch": 1.0605839060291467,
-      "grad_norm": 1.8984375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.1466,
       "step": 65280
     },
     {
       "epoch": 1.0607463729265163,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3217,
+      "loss": 1.2079,
       "step": 65290
     },
     {
       "epoch": 1.060908839823886,
-      "grad_norm": 2.125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.1774,
       "step": 65300
     },
     {
       "epoch": 1.0610713067212556,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.2133,
       "step": 65310
     },
     {
       "epoch": 1.0612337736186253,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2442,
       "step": 65320
     },
     {
       "epoch": 1.061396240515995,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2044,
       "step": 65330
     },
     {
       "epoch": 1.0615587074133646,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.1924,
       "step": 65340
     },
     {
       "epoch": 1.0617211743107342,
-      "grad_norm": 2.25,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.1787,
       "step": 65350
     },
     {
       "epoch": 1.0618836412081039,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.1627,
       "step": 65360
     },
     {
       "epoch": 1.0620461081054735,
-      "grad_norm": 3.375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2131,
       "step": 65370
     },
     {
       "epoch": 1.0622085750028432,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.1715,
       "step": 65380
     },
     {
       "epoch": 1.0623710419002128,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.1678,
       "step": 65390
     },
     {
       "epoch": 1.0625335087975825,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2278,
       "step": 65400
     },
     {
       "epoch": 1.062695975694952,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.1947,
       "step": 65410
     },
     {
       "epoch": 1.0628584425923218,
-      "grad_norm": 2.375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.1624,
       "step": 65420
     },
     {
       "epoch": 1.0630209094896914,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2035,
       "step": 65430
     },
     {
       "epoch": 1.063183376387061,
-      "grad_norm": 2.125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3204,
+      "loss": 1.1506,
       "step": 65440
     },
     {
       "epoch": 1.0633458432844307,
-      "grad_norm": 1.9609375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2338,
       "step": 65450
     },
     {
       "epoch": 1.0635083101818004,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.16,
       "step": 65460
     },
     {
       "epoch": 1.06367077707917,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.1951,
       "step": 65470
     },
     {
       "epoch": 1.0638332439765397,
-      "grad_norm": 3.375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2221,
       "step": 65480
     },
     {
       "epoch": 1.0639957108739095,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.1995,
       "step": 65490
     },
     {
       "epoch": 1.0641581777712792,
-      "grad_norm": 3.0,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.1528,
       "step": 65500
     },
     {
       "epoch": 1.0643206446686488,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.1724,
       "step": 65510
     },
     {
       "epoch": 1.0644831115660185,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.1811,
       "step": 65520
     },
     {
       "epoch": 1.0646455784633881,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.1886,
       "step": 65530
     },
     {
       "epoch": 1.0648080453607578,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1768,
       "step": 65540
     },
     {
       "epoch": 1.0649705122581274,
-      "grad_norm": 1.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3118,
+      "loss": 1.1863,
       "step": 65550
     },
     {
       "epoch": 1.065132979155497,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2003,
       "step": 65560
     },
     {
       "epoch": 1.0652954460528667,
-      "grad_norm": 2.25,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.1788,
       "step": 65570
     },
     {
       "epoch": 1.0654579129502364,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3247,
+      "loss": 1.2162,
       "step": 65580
     },
     {
       "epoch": 1.065620379847606,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.174,
       "step": 65590
     },
     {
       "epoch": 1.0657828467449757,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2119,
       "step": 65600
     },
     {
       "epoch": 1.0659453136423453,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.174,
       "step": 65610
     },
     {
       "epoch": 1.066107780539715,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.1749,
       "step": 65620
     },
     {
       "epoch": 1.0662702474370847,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.1975,
       "step": 65630
     },
     {
       "epoch": 1.0664327143344543,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.1949,
       "step": 65640
     },
     {
       "epoch": 1.066595181231824,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.1685,
       "step": 65650
     },
     {
       "epoch": 1.0667576481291936,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3115,
+      "loss": 1.2134,
       "step": 65660
     },
     {
       "epoch": 1.0669201150265633,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2078,
       "step": 65670
     },
     {
       "epoch": 1.067082581923933,
-      "grad_norm": 3.5,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.324,
+      "loss": 1.1699,
       "step": 65680
     },
     {
       "epoch": 1.0672450488213028,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3683,
+      "loss": 1.2082,
       "step": 65690
     },
     {
       "epoch": 1.0674075157186724,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.164,
       "step": 65700
     },
     {
       "epoch": 1.067569982616042,
-      "grad_norm": 3.125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.1556,
       "step": 65710
     },
     {
       "epoch": 1.0677324495134117,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.1512,
       "step": 65720
     },
     {
       "epoch": 1.0678949164107814,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2238,
       "step": 65730
     },
     {
       "epoch": 1.068057383308151,
-      "grad_norm": 2.375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.1532,
       "step": 65740
     },
     {
       "epoch": 1.0682198502055207,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.1963,
       "step": 65750
     },
     {
       "epoch": 1.0683823171028903,
-      "grad_norm": 2.625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.1907,
       "step": 65760
     },
     {
       "epoch": 1.06854478400026,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2055,
       "step": 65770
     },
     {
       "epoch": 1.0687072508976296,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2204,
       "step": 65780
     },
     {
       "epoch": 1.0688697177949993,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.1617,
       "step": 65790
     },
     {
       "epoch": 1.069032184692369,
-      "grad_norm": 1.5625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.153,
       "step": 65800
     },
     {
       "epoch": 1.0691946515897386,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2035,
       "step": 65810
     },
     {
       "epoch": 1.0693571184871082,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1766,
       "step": 65820
     },
     {
       "epoch": 1.0695195853844779,
-      "grad_norm": 3.828125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.1824,
       "step": 65830
     },
     {
       "epoch": 1.0696820522818475,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.1901,
       "step": 65840
     },
     {
       "epoch": 1.0698445191792172,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.1884,
       "step": 65850
     },
     {
       "epoch": 1.0700069860765868,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3206,
+      "loss": 1.2206,
       "step": 65860
     },
     {
       "epoch": 1.0701694529739565,
-      "grad_norm": 1.6796875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.196,
       "step": 65870
     },
     {
       "epoch": 1.0703319198713261,
-      "grad_norm": 2.25,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.1833,
       "step": 65880
     },
     {
       "epoch": 1.0704943867686958,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.1819,
       "step": 65890
     },
     {
       "epoch": 1.0706568536660654,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.1766,
       "step": 65900
     },
     {
       "epoch": 1.070819320563435,
-      "grad_norm": 1.7421875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.1953,
       "step": 65910
     },
     {
       "epoch": 1.0709817874608047,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.1665,
       "step": 65920
     },
     {
       "epoch": 1.0711442543581746,
-      "grad_norm": 2.25,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.1946,
       "step": 65930
     },
     {
       "epoch": 1.0713067212555443,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.2351,
       "step": 65940
     },
     {
       "epoch": 1.071469188152914,
-      "grad_norm": 1.9296875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1857,
       "step": 65950
     },
     {
       "epoch": 1.0716316550502836,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3239,
+      "loss": 1.1649,
       "step": 65960
     },
     {
       "epoch": 1.0717941219476532,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.319,
+      "loss": 1.2112,
       "step": 65970
     },
     {
       "epoch": 1.0719565888450229,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.1995,
       "step": 65980
     },
     {
       "epoch": 1.0721190557423925,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2242,
       "step": 65990
     },
     {
       "epoch": 1.0722815226397622,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2222,
       "step": 66000
     },
     {
       "epoch": 1.0724439895371318,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.1773,
       "step": 66010
     },
     {
       "epoch": 1.0726064564345015,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.26,
       "step": 66020
     },
     {
       "epoch": 1.0727689233318711,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1839,
       "step": 66030
     },
     {
       "epoch": 1.0729313902292408,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.1991,
       "step": 66040
     },
     {
       "epoch": 1.0730938571266104,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.2027,
       "step": 66050
     },
     {
       "epoch": 1.07325632402398,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1998,
       "step": 66060
     },
     {
       "epoch": 1.0734187909213497,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2053,
       "step": 66070
     },
     {
       "epoch": 1.0735812578187194,
-      "grad_norm": 2.5,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2484,
       "step": 66080
     },
     {
       "epoch": 1.073743724716089,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.2087,
       "step": 66090
     },
     {
       "epoch": 1.0739061916134587,
-      "grad_norm": 2.5,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2166,
       "step": 66100
     },
     {
       "epoch": 1.0740686585108283,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.1872,
       "step": 66110
     },
     {
       "epoch": 1.074231125408198,
-      "grad_norm": 2.25,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2163,
       "step": 66120
     },
     {
       "epoch": 1.0743935923055679,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.204,
       "step": 66130
     },
     {
       "epoch": 1.0745560592029375,
-      "grad_norm": 2.0,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2127,
       "step": 66140
     },
     {
       "epoch": 1.0747185261003072,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2377,
       "step": 66150
     },
     {
       "epoch": 1.0748809929976768,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.1913,
       "step": 66160
     },
     {
       "epoch": 1.0750434598950465,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.1811,
       "step": 66170
     },
     {
       "epoch": 1.0752059267924161,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.1747,
       "step": 66180
     },
     {
       "epoch": 1.0753683936897858,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.1949,
       "step": 66190
     },
     {
       "epoch": 1.0755308605871554,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.1821,
       "step": 66200
     },
     {
       "epoch": 1.075693327484525,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2032,
       "step": 66210
     },
     {
       "epoch": 1.0758557943818947,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.221,
       "step": 66220
     },
     {
       "epoch": 1.0760182612792644,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.1801,
       "step": 66230
     },
     {
       "epoch": 1.076180728176634,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3622,
+      "loss": 1.1831,
       "step": 66240
     },
     {
       "epoch": 1.0763431950740037,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.223,
       "step": 66250
     },
     {
       "epoch": 1.0765056619713733,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3151,
+      "loss": 1.2418,
       "step": 66260
     },
     {
       "epoch": 1.076668128868743,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2334,
       "step": 66270
     },
     {
       "epoch": 1.0768305957661126,
-      "grad_norm": 2.125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.1833,
       "step": 66280
     },
     {
       "epoch": 1.0769930626634823,
-      "grad_norm": 2.71875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2225,
       "step": 66290
     },
     {
       "epoch": 1.077155529560852,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2271,
       "step": 66300
     },
     {
       "epoch": 1.0773179964582216,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.1739,
       "step": 66310
     },
     {
       "epoch": 1.0774804633555912,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3237,
+      "loss": 1.1824,
       "step": 66320
     },
     {
       "epoch": 1.0776429302529609,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2219,
       "step": 66330
     },
     {
       "epoch": 1.0778053971503305,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3238,
+      "loss": 1.1885,
       "step": 66340
     },
     {
       "epoch": 1.0779678640477002,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.1949,
       "step": 66350
     },
     {
       "epoch": 1.0781303309450698,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1943,
       "step": 66360
     },
     {
       "epoch": 1.0782927978424397,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2096,
       "step": 66370
     },
     {
       "epoch": 1.0784552647398093,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.194,
       "step": 66380
     },
     {
       "epoch": 1.078617731637179,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.2048,
       "step": 66390
     },
     {
       "epoch": 1.0787801985345487,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1994,
       "step": 66400
     },
     {
       "epoch": 1.0789426654319183,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2297,
       "step": 66410
     },
     {
       "epoch": 1.079105132329288,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.1908,
       "step": 66420
     },
     {
       "epoch": 1.0792675992266576,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2015,
       "step": 66430
     },
     {
       "epoch": 1.0794300661240273,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.1961,
       "step": 66440
     },
     {
       "epoch": 1.079592533021397,
-      "grad_norm": 1.734375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2049,
       "step": 66450
     },
     {
       "epoch": 1.0797549999187666,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2266,
       "step": 66460
     },
     {
       "epoch": 1.0799174668161362,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2023,
       "step": 66470
     },
     {
       "epoch": 1.0800799337135059,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1893,
       "step": 66480
     },
     {
       "epoch": 1.0802424006108755,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.321,
+      "loss": 1.219,
       "step": 66490
     },
     {
       "epoch": 1.0804048675082452,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2171,
       "step": 66500
     },
     {
       "epoch": 1.0805673344056148,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2267,
       "step": 66510
     },
     {
       "epoch": 1.0807298013029845,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2064,
       "step": 66520
     },
     {
       "epoch": 1.0808922682003541,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.1578,
       "step": 66530
     },
     {
       "epoch": 1.0810547350977238,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3128,
+      "loss": 1.1761,
       "step": 66540
     },
     {
       "epoch": 1.0812172019950934,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2346,
       "step": 66550
     },
     {
       "epoch": 1.081379668892463,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3144,
+      "loss": 1.2401,
       "step": 66560
     },
     {
       "epoch": 1.081542135789833,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3111,
+      "loss": 1.2598,
       "step": 66570
     },
     {
       "epoch": 1.0817046026872026,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.1765,
       "step": 66580
     },
     {
       "epoch": 1.0818670695845722,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2086,
       "step": 66590
     },
     {
       "epoch": 1.082029536481942,
-      "grad_norm": 1.875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2377,
       "step": 66600
     },
     {
       "epoch": 1.0821920033793115,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.1825,
       "step": 66610
     },
     {
       "epoch": 1.0823544702766812,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.1918,
       "step": 66620
     },
     {
       "epoch": 1.0825169371740508,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2213,
       "step": 66630
     },
     {
       "epoch": 1.0826794040714205,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2024,
       "step": 66640
     },
     {
       "epoch": 1.0828418709687901,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.193,
       "step": 66650
     },
     {
       "epoch": 1.0830043378661598,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.1873,
       "step": 66660
     },
     {
       "epoch": 1.0831668047635294,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3178,
+      "loss": 1.1662,
       "step": 66670
     },
     {
       "epoch": 1.083329271660899,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.1697,
       "step": 66680
     },
     {
       "epoch": 1.0834917385582687,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2112,
       "step": 66690
     },
     {
       "epoch": 1.0836542054556384,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.1617,
       "step": 66700
     },
     {
       "epoch": 1.083816672353008,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1832,
       "step": 66710
     },
     {
       "epoch": 1.0839791392503777,
-      "grad_norm": 2.109375,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.1909,
       "step": 66720
     },
     {
       "epoch": 1.0841416061477473,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2413,
       "step": 66730
     },
     {
       "epoch": 1.084304073045117,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.302,
+      "loss": 1.2249,
       "step": 66740
     },
     {
       "epoch": 1.0844665399424867,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3232,
+      "loss": 1.185,
       "step": 66750
     },
     {
       "epoch": 1.0846290068398563,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2196,
       "step": 66760
     },
     {
       "epoch": 1.084791473737226,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.365,
+      "loss": 1.2306,
       "step": 66770
     },
     {
       "epoch": 1.0849539406345956,
-      "grad_norm": 2.875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.1664,
       "step": 66780
     },
     {
       "epoch": 1.0851164075319653,
-      "grad_norm": 3.0,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2011,
       "step": 66790
     },
     {
       "epoch": 1.085278874429335,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.1942,
       "step": 66800
     },
     {
       "epoch": 1.0854413413267048,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.1947,
       "step": 66810
     },
     {
       "epoch": 1.0856038082240744,
-      "grad_norm": 2.5,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.1823,
       "step": 66820
     },
     {
       "epoch": 1.085766275121444,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2104,
       "step": 66830
     },
     {
       "epoch": 1.0859287420188137,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2197,
       "step": 66840
     },
     {
       "epoch": 1.0860912089161834,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2346,
       "step": 66850
     },
     {
       "epoch": 1.086253675813553,
-      "grad_norm": 1.9921875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.1788,
       "step": 66860
     },
     {
       "epoch": 1.0864161427109227,
-      "grad_norm": 2.28125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2187,
       "step": 66870
     },
     {
       "epoch": 1.0865786096082923,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3635,
+      "loss": 1.1759,
       "step": 66880
     },
     {
       "epoch": 1.086741076505662,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2224,
       "step": 66890
     },
     {
       "epoch": 1.0869035434030316,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2048,
       "step": 66900
     },
     {
       "epoch": 1.0870660103004013,
-      "grad_norm": 1.875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2071,
       "step": 66910
     },
     {
       "epoch": 1.087228477197771,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2036,
       "step": 66920
     },
     {
       "epoch": 1.0873909440951406,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2201,
       "step": 66930
     },
     {
       "epoch": 1.0875534109925102,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.252,
       "step": 66940
     },
     {
       "epoch": 1.08771587788988,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3127,
+      "loss": 1.1493,
       "step": 66950
     },
     {
       "epoch": 1.0878783447872495,
-      "grad_norm": 3.0,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3258,
+      "loss": 1.2058,
       "step": 66960
     },
     {
       "epoch": 1.0880408116846192,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.223,
       "step": 66970
     },
     {
       "epoch": 1.0882032785819888,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2207,
       "step": 66980
     },
     {
       "epoch": 1.0883657454793585,
-      "grad_norm": 1.6875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.238,
       "step": 66990
     },
     {
       "epoch": 1.0885282123767281,
-      "grad_norm": 2.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2233,
       "step": 67000
     },
     {
       "epoch": 1.088690679274098,
-      "grad_norm": 2.359375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2038,
       "step": 67010
     },
     {
       "epoch": 1.0888531461714677,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.1737,
       "step": 67020
     },
     {
       "epoch": 1.0890156130688373,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3268,
+      "loss": 1.1905,
       "step": 67030
     },
     {
       "epoch": 1.089178079966207,
-      "grad_norm": 2.625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3651,
+      "loss": 1.2122,
       "step": 67040
     },
     {
       "epoch": 1.0893405468635766,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.2024,
       "step": 67050
     },
     {
       "epoch": 1.0895030137609463,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2166,
       "step": 67060
     },
     {
       "epoch": 1.089665480658316,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.324,
+      "loss": 1.1988,
       "step": 67070
     },
     {
       "epoch": 1.0898279475556856,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.2009,
       "step": 67080
     },
     {
       "epoch": 1.0899904144530552,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.1762,
       "step": 67090
     },
     {
       "epoch": 1.0901528813504249,
-      "grad_norm": 1.890625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2316,
       "step": 67100
     },
     {
       "epoch": 1.0903153482477945,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2338,
       "step": 67110
     },
     {
       "epoch": 1.0904778151451642,
-      "grad_norm": 1.8046875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.1675,
       "step": 67120
     },
     {
       "epoch": 1.0906402820425338,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2101,
       "step": 67130
     },
     {
       "epoch": 1.0908027489399035,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.2128,
       "step": 67140
     },
     {
       "epoch": 1.0909652158372731,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2041,
       "step": 67150
     },
     {
       "epoch": 1.0911276827346428,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.1977,
       "step": 67160
     },
     {
       "epoch": 1.0912901496320124,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1963,
       "step": 67170
     },
     {
       "epoch": 1.091452616529382,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.1951,
       "step": 67180
     },
     {
       "epoch": 1.0916150834267517,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2076,
       "step": 67190
     },
     {
       "epoch": 1.0917775503241214,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.147,
       "step": 67200
     },
     {
       "epoch": 1.091940017221491,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2048,
       "step": 67210
     },
     {
       "epoch": 1.0921024841188607,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2398,
       "step": 67220
     },
     {
       "epoch": 1.0922649510162303,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2011,
       "step": 67230
     },
     {
       "epoch": 1.0924274179136002,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.1891,
       "step": 67240
     },
     {
       "epoch": 1.0925898848109699,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.1986,
       "step": 67250
     },
     {
       "epoch": 1.0927523517083395,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2396,
       "step": 67260
     },
     {
       "epoch": 1.0929148186057092,
-      "grad_norm": 1.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3095,
+      "loss": 1.1683,
       "step": 67270
     },
     {
       "epoch": 1.0930772855030788,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2256,
       "step": 67280
     },
     {
       "epoch": 1.0932397524004485,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.1666,
       "step": 67290
     },
     {
       "epoch": 1.0934022192978181,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2087,
       "step": 67300
     },
     {
       "epoch": 1.0935646861951878,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3118,
+      "loss": 1.2194,
       "step": 67310
     },
     {
       "epoch": 1.0937271530925574,
-      "grad_norm": 2.0625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2215,
       "step": 67320
     },
     {
       "epoch": 1.093889619989927,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1986,
       "step": 67330
     },
     {
       "epoch": 1.0940520868872967,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2232,
       "step": 67340
     },
     {
       "epoch": 1.0942145537846664,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2337,
       "step": 67350
     },
     {
       "epoch": 1.094377020682036,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1856,
       "step": 67360
     },
     {
       "epoch": 1.0945394875794057,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.1668,
       "step": 67370
     },
     {
       "epoch": 1.0947019544767753,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2365,
       "step": 67380
     },
     {
       "epoch": 1.094864421374145,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.1998,
       "step": 67390
     },
     {
       "epoch": 1.0950268882715146,
-      "grad_norm": 2.25,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.1853,
       "step": 67400
     },
     {
       "epoch": 1.0951893551688843,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.1559,
       "step": 67410
     },
     {
       "epoch": 1.095351822066254,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.1769,
       "step": 67420
     },
     {
       "epoch": 1.0955142889636236,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.1609,
       "step": 67430
     },
     {
       "epoch": 1.0956767558609932,
-      "grad_norm": 3.0,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2214,
       "step": 67440
     },
     {
       "epoch": 1.095839222758363,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3194,
+      "loss": 1.2705,
       "step": 67450
     },
     {
       "epoch": 1.0960016896557327,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.177,
       "step": 67460
     },
     {
       "epoch": 1.0961641565531024,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.1749,
       "step": 67470
     },
     {
       "epoch": 1.096326623450472,
-      "grad_norm": 2.625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2038,
       "step": 67480
     },
     {
       "epoch": 1.0964890903478417,
-      "grad_norm": 3.640625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.1088,
       "step": 67490
     },
     {
       "epoch": 1.0966515572452114,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2431,
       "step": 67500
     },
     {
       "epoch": 1.096814024142581,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.2112,
       "step": 67510
     },
     {
       "epoch": 1.0969764910399507,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3091,
+      "loss": 1.1922,
       "step": 67520
     },
     {
       "epoch": 1.0971389579373203,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.1467,
       "step": 67530
     },
     {
       "epoch": 1.09730142483469,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.1782,
       "step": 67540
     },
     {
       "epoch": 1.0974638917320596,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2088,
       "step": 67550
     },
     {
       "epoch": 1.0976263586294293,
-      "grad_norm": 2.734375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3587,
+      "loss": 1.2221,
       "step": 67560
     },
     {
       "epoch": 1.097788825526799,
-      "grad_norm": 2.625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3223,
+      "loss": 1.2085,
       "step": 67570
     },
     {
       "epoch": 1.0979512924241686,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2186,
       "step": 67580
     },
     {
       "epoch": 1.0981137593215382,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.1822,
       "step": 67590
     },
     {
       "epoch": 1.0982762262189079,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.188,
       "step": 67600
     },
     {
       "epoch": 1.0984386931162775,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.1967,
       "step": 67610
     },
     {
       "epoch": 1.0986011600136472,
-      "grad_norm": 1.828125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.2074,
       "step": 67620
     },
     {
       "epoch": 1.0987636269110168,
-      "grad_norm": 1.96875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.1753,
       "step": 67630
     },
     {
       "epoch": 1.0989260938083865,
-      "grad_norm": 1.875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3175,
+      "loss": 1.2192,
       "step": 67640
     },
     {
       "epoch": 1.0990885607057561,
-      "grad_norm": 2.875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.207,
       "step": 67650
     },
     {
       "epoch": 1.0992510276031258,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2479,
       "step": 67660
     },
     {
       "epoch": 1.0994134945004954,
-      "grad_norm": 2.375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2068,
       "step": 67670
     },
     {
       "epoch": 1.0995759613978653,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2344,
       "step": 67680
     },
     {
       "epoch": 1.099738428295235,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.2025,
       "step": 67690
     },
     {
       "epoch": 1.0999008951926046,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.222,
       "step": 67700
     },
     {
       "epoch": 1.1000633620899742,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2338,
       "step": 67710
     },
     {
       "epoch": 1.100225828987344,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.1971,
       "step": 67720
     },
     {
       "epoch": 1.1003882958847135,
-      "grad_norm": 2.5,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2276,
       "step": 67730
     },
     {
       "epoch": 1.1005507627820832,
-      "grad_norm": 2.125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1974,
       "step": 67740
     },
     {
       "epoch": 1.1007132296794528,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2318,
       "step": 67750
     },
     {
       "epoch": 1.1008756965768225,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.2231,
       "step": 67760
     },
     {
       "epoch": 1.1010381634741921,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2492,
       "step": 67770
     },
     {
       "epoch": 1.1012006303715618,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3149,
+      "loss": 1.196,
       "step": 67780
     },
     {
       "epoch": 1.1013630972689314,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2309,
       "step": 67790
     },
     {
       "epoch": 1.101525564166301,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.1939,
       "step": 67800
     },
     {
       "epoch": 1.1016880310636707,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.1772,
       "step": 67810
     },
     {
       "epoch": 1.1018504979610404,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.2196,
       "step": 67820
     },
     {
       "epoch": 1.10201296485841,
-      "grad_norm": 2.375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3266,
+      "loss": 1.2035,
       "step": 67830
     },
     {
       "epoch": 1.1021754317557797,
-      "grad_norm": 1.65625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.1591,
       "step": 67840
     },
     {
       "epoch": 1.1023378986531494,
-      "grad_norm": 1.84375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3147,
+      "loss": 1.2212,
       "step": 67850
     },
     {
       "epoch": 1.102500365550519,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.189,
       "step": 67860
     },
     {
       "epoch": 1.1026628324478887,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1775,
       "step": 67870
     },
     {
       "epoch": 1.1028252993452583,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2187,
       "step": 67880
     },
     {
       "epoch": 1.1029877662426282,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2082,
       "step": 67890
     },
     {
       "epoch": 1.1031502331399978,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3055,
+      "loss": 1.2479,
       "step": 67900
     },
     {
       "epoch": 1.1033127000373675,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2453,
       "step": 67910
     },
     {
       "epoch": 1.1034751669347371,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2103,
       "step": 67920
     },
     {
       "epoch": 1.1036376338321068,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.19,
       "step": 67930
     },
     {
       "epoch": 1.1038001007294764,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2146,
       "step": 67940
     },
     {
       "epoch": 1.103962567626846,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.1765,
       "step": 67950
     },
     {
       "epoch": 1.1041250345242157,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1776,
       "step": 67960
     },
     {
       "epoch": 1.1042875014215854,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2103,
       "step": 67970
     },
     {
       "epoch": 1.104449968318955,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3117,
+      "loss": 1.204,
       "step": 67980
     },
     {
       "epoch": 1.1046124352163247,
-      "grad_norm": 2.25,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.1903,
       "step": 67990
     },
     {
       "epoch": 1.1047749021136943,
-      "grad_norm": 1.90625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.1522,
       "step": 68000
     },
     {
       "epoch": 1.104937369011064,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.1793,
       "step": 68010
     },
     {
       "epoch": 1.1050998359084336,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.197,
       "step": 68020
     },
     {
       "epoch": 1.1052623028058033,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.1992,
       "step": 68030
     },
     {
       "epoch": 1.105424769703173,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.1759,
       "step": 68040
     },
     {
       "epoch": 1.1055872366005426,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2477,
       "step": 68050
     },
     {
       "epoch": 1.1057497034979122,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2561,
       "step": 68060
     },
     {
       "epoch": 1.105912170395282,
-      "grad_norm": 2.375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.1936,
       "step": 68070
     },
     {
       "epoch": 1.1060746372926515,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2064,
       "step": 68080
     },
     {
       "epoch": 1.1062371041900212,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.1967,
       "step": 68090
     },
     {
       "epoch": 1.1063995710873908,
-      "grad_norm": 1.78125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.198,
       "step": 68100
     },
     {
       "epoch": 1.1065620379847605,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2226,
       "step": 68110
     },
     {
       "epoch": 1.1067245048821304,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.1863,
       "step": 68120
     },
     {
       "epoch": 1.1068869717795,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2044,
       "step": 68130
     },
     {
       "epoch": 1.1070494386768697,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2011,
       "step": 68140
     },
     {
       "epoch": 1.1072119055742393,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.188,
       "step": 68150
     },
     {
       "epoch": 1.107374372471609,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.1846,
       "step": 68160
     },
     {
       "epoch": 1.1075368393689786,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2049,
       "step": 68170
     },
     {
       "epoch": 1.1076993062663483,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.219,
       "step": 68180
     },
     {
       "epoch": 1.107861773163718,
-      "grad_norm": 3.25,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.1966,
       "step": 68190
     },
     {
       "epoch": 1.1080242400610876,
-      "grad_norm": 2.0,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3223,
+      "loss": 1.1975,
       "step": 68200
     },
     {
       "epoch": 1.1081867069584572,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2002,
       "step": 68210
     },
     {
       "epoch": 1.1083491738558269,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.1938,
       "step": 68220
     },
     {
       "epoch": 1.1085116407531965,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2284,
       "step": 68230
     },
     {
       "epoch": 1.1086741076505662,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3194,
+      "loss": 1.1935,
       "step": 68240
     },
     {
       "epoch": 1.1088365745479358,
-      "grad_norm": 3.203125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.1811,
       "step": 68250
     },
     {
       "epoch": 1.1089990414453055,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.1956,
       "step": 68260
     },
     {
       "epoch": 1.1091615083426751,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.2267,
       "step": 68270
     },
     {
       "epoch": 1.1093239752400448,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.234,
       "step": 68280
     },
     {
       "epoch": 1.1094864421374144,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1846,
       "step": 68290
     },
     {
       "epoch": 1.109648909034784,
-      "grad_norm": 3.0,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2509,
       "step": 68300
     },
     {
       "epoch": 1.1098113759321537,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.2007,
       "step": 68310
     },
     {
       "epoch": 1.1099738428295234,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.1941,
       "step": 68320
     },
     {
       "epoch": 1.1101363097268933,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.2107,
       "step": 68330
     },
     {
       "epoch": 1.110298776624263,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2326,
       "step": 68340
     },
     {
       "epoch": 1.1104612435216326,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2003,
       "step": 68350
     },
     {
       "epoch": 1.1106237104190022,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.237,
       "step": 68360
     },
     {
       "epoch": 1.1107861773163719,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3704,
+      "loss": 1.2279,
       "step": 68370
     },
     {
       "epoch": 1.1109486442137415,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.1708,
       "step": 68380
     },
     {
       "epoch": 1.1111111111111112,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3264,
+      "loss": 1.2186,
       "step": 68390
     },
     {
       "epoch": 1.1112735780084808,
-      "grad_norm": 2.75,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.207,
       "step": 68400
     },
     {
       "epoch": 1.1114360449058505,
-      "grad_norm": 1.875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.2031,
       "step": 68410
     },
     {
       "epoch": 1.1115985118032201,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2005,
       "step": 68420
     },
     {
       "epoch": 1.1117609787005898,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.2071,
       "step": 68430
     },
     {
       "epoch": 1.1119234455979594,
-      "grad_norm": 3.5,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.209,
       "step": 68440
     },
     {
       "epoch": 1.112085912495329,
-      "grad_norm": 3.5,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2023,
       "step": 68450
     },
     {
       "epoch": 1.1122483793926987,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 1.2416,
       "step": 68460
     },
     {
       "epoch": 1.1124108462900684,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2132,
       "step": 68470
     },
     {
       "epoch": 1.112573313187438,
-      "grad_norm": 1.9609375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.1602,
       "step": 68480
     },
     {
       "epoch": 1.1127357800848077,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3624,
+      "loss": 1.1991,
       "step": 68490
     },
     {
       "epoch": 1.1128982469821773,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.203,
       "step": 68500
     },
     {
       "epoch": 1.113060713879547,
-      "grad_norm": 2.0,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2006,
       "step": 68510
     },
     {
       "epoch": 1.1132231807769166,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.362,
+      "loss": 1.1764,
       "step": 68520
     },
     {
       "epoch": 1.1133856476742863,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3609,
+      "loss": 1.1703,
       "step": 68530
     },
     {
       "epoch": 1.113548114571656,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2222,
       "step": 68540
     },
     {
       "epoch": 1.1137105814690256,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.196,
       "step": 68550
     },
     {
       "epoch": 1.1138730483663954,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.1992,
       "step": 68560
     },
     {
       "epoch": 1.114035515263765,
-      "grad_norm": 3.3125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.1812,
       "step": 68570
     },
     {
       "epoch": 1.1141979821611347,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1894,
       "step": 68580
     },
     {
       "epoch": 1.1143604490585044,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1757,
       "step": 68590
     },
     {
       "epoch": 1.114522915955874,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2007,
       "step": 68600
     },
     {
       "epoch": 1.1146853828532437,
-      "grad_norm": 2.625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3202,
+      "loss": 1.2156,
       "step": 68610
     },
     {
       "epoch": 1.1148478497506134,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.1678,
       "step": 68620
     },
     {
       "epoch": 1.115010316647983,
-      "grad_norm": 2.875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1785,
       "step": 68630
     },
     {
       "epoch": 1.1151727835453527,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.164,
       "step": 68640
     },
     {
       "epoch": 1.1153352504427223,
-      "grad_norm": 1.7578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3161,
+      "loss": 1.1982,
       "step": 68650
     },
     {
       "epoch": 1.115497717340092,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2005,
       "step": 68660
     },
     {
       "epoch": 1.1156601842374616,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.1758,
       "step": 68670
     },
     {
       "epoch": 1.1158226511348313,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1979,
       "step": 68680
     },
     {
       "epoch": 1.115985118032201,
-      "grad_norm": 3.125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1565,
       "step": 68690
     },
     {
       "epoch": 1.1161475849295706,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.1784,
       "step": 68700
     },
     {
       "epoch": 1.1163100518269402,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2266,
       "step": 68710
     },
     {
       "epoch": 1.1164725187243099,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.1734,
       "step": 68720
     },
     {
       "epoch": 1.1166349856216795,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2188,
       "step": 68730
     },
     {
       "epoch": 1.1167974525190492,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2152,
       "step": 68740
     },
     {
       "epoch": 1.1169599194164188,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.239,
       "step": 68750
     },
     {
       "epoch": 1.1171223863137885,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2084,
       "step": 68760
     },
     {
       "epoch": 1.1172848532111583,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2183,
       "step": 68770
     },
     {
       "epoch": 1.117447320108528,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.1896,
       "step": 68780
     },
     {
       "epoch": 1.1176097870058976,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2249,
       "step": 68790
     },
     {
       "epoch": 1.1177722539032673,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1826,
       "step": 68800
     },
     {
       "epoch": 1.117934720800637,
-      "grad_norm": 2.5,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.1852,
       "step": 68810
     },
     {
       "epoch": 1.1180971876980066,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1863,
       "step": 68820
     },
     {
       "epoch": 1.1182596545953762,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.1998,
       "step": 68830
     },
     {
       "epoch": 1.118422121492746,
-      "grad_norm": 2.40625,
+      "grad_norm": 12.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.2164,
       "step": 68840
     },
     {
       "epoch": 1.1185845883901155,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2207,
       "step": 68850
     },
     {
       "epoch": 1.1187470552874852,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.1972,
       "step": 68860
     },
     {
       "epoch": 1.1189095221848548,
-      "grad_norm": 2.75,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2264,
       "step": 68870
     },
     {
       "epoch": 1.1190719890822245,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3603,
+      "loss": 1.1745,
       "step": 68880
     },
     {
       "epoch": 1.1192344559795941,
-      "grad_norm": 1.578125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.187,
       "step": 68890
     },
     {
       "epoch": 1.1193969228769638,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.2361,
       "step": 68900
     },
     {
       "epoch": 1.1195593897743334,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.1928,
       "step": 68910
     },
     {
       "epoch": 1.119721856671703,
-      "grad_norm": 1.8828125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.1842,
       "step": 68920
     },
     {
       "epoch": 1.1198843235690727,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2388,
       "step": 68930
     },
     {
       "epoch": 1.1200467904664424,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.1935,
       "step": 68940
     },
     {
       "epoch": 1.120209257363812,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.1503,
       "step": 68950
     },
     {
       "epoch": 1.1203717242611817,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3258,
+      "loss": 1.2114,
       "step": 68960
     },
     {
       "epoch": 1.1205341911585514,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2186,
       "step": 68970
     },
     {
       "epoch": 1.120696658055921,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1858,
       "step": 68980
     },
     {
       "epoch": 1.1208591249532907,
-      "grad_norm": 3.0,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.22,
       "step": 68990
     },
     {
       "epoch": 1.1210215918506605,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.1825,
       "step": 69000
     },
     {
       "epoch": 1.1211840587480302,
-      "grad_norm": 1.9453125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2516,
       "step": 69010
     },
     {
       "epoch": 1.1213465256453998,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.1853,
       "step": 69020
     },
     {
       "epoch": 1.1215089925427695,
-      "grad_norm": 1.9921875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.242,
       "step": 69030
     },
     {
       "epoch": 1.1216714594401391,
-      "grad_norm": 2.125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2472,
       "step": 69040
     },
     {
       "epoch": 1.1218339263375088,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2155,
       "step": 69050
     },
     {
       "epoch": 1.1219963932348784,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2083,
       "step": 69060
     },
     {
       "epoch": 1.122158860132248,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2731,
       "step": 69070
     },
     {
       "epoch": 1.1223213270296177,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2106,
       "step": 69080
     },
     {
       "epoch": 1.1224837939269874,
-      "grad_norm": 2.140625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.2561,
       "step": 69090
     },
     {
       "epoch": 1.122646260824357,
-      "grad_norm": 2.375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.2083,
       "step": 69100
     },
     {
       "epoch": 1.1228087277217267,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.203,
       "step": 69110
     },
     {
       "epoch": 1.1229711946190963,
-      "grad_norm": 2.625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.1844,
       "step": 69120
     },
     {
       "epoch": 1.123133661516466,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.1983,
       "step": 69130
     },
     {
       "epoch": 1.1232961284138356,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1947,
       "step": 69140
     },
     {
       "epoch": 1.1234585953112053,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2399,
       "step": 69150
     },
     {
       "epoch": 1.123621062208575,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.1757,
       "step": 69160
     },
     {
       "epoch": 1.1237835291059446,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2018,
       "step": 69170
     },
     {
       "epoch": 1.1239459960033142,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2043,
       "step": 69180
     },
     {
       "epoch": 1.124108462900684,
-      "grad_norm": 2.5,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2197,
       "step": 69190
     },
     {
       "epoch": 1.1242709297980538,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.1725,
       "step": 69200
     },
     {
       "epoch": 1.1244333966954234,
-      "grad_norm": 2.125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.153,
       "step": 69210
     },
     {
       "epoch": 1.124595863592793,
-      "grad_norm": 2.75,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.1754,
       "step": 69220
     },
     {
       "epoch": 1.1247583304901627,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.229,
       "step": 69230
     },
     {
       "epoch": 1.1249207973875324,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2089,
       "step": 69240
     },
     {
       "epoch": 1.125083264284902,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.1624,
       "step": 69250
     },
     {
       "epoch": 1.1252457311822717,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2528,
       "step": 69260
     },
     {
       "epoch": 1.1254081980796413,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.1934,
       "step": 69270
     },
     {
       "epoch": 1.125570664977011,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3614,
+      "loss": 1.2202,
       "step": 69280
     },
     {
       "epoch": 1.1257331318743806,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.1912,
       "step": 69290
     },
     {
       "epoch": 1.1258955987717503,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2049,
       "step": 69300
     },
     {
       "epoch": 1.12605806566912,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.1765,
       "step": 69310
     },
     {
       "epoch": 1.1262205325664896,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.1934,
       "step": 69320
     },
     {
       "epoch": 1.1263829994638592,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2015,
       "step": 69330
     },
     {
       "epoch": 1.1265454663612289,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.167,
       "step": 69340
     },
     {
       "epoch": 1.1267079332585985,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.1952,
       "step": 69350
     },
     {
       "epoch": 1.1268704001559682,
-      "grad_norm": 1.6328125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3086,
+      "loss": 1.1546,
       "step": 69360
     },
     {
       "epoch": 1.1270328670533378,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.1793,
       "step": 69370
     },
     {
       "epoch": 1.1271953339507075,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.1849,
       "step": 69380
     },
     {
       "epoch": 1.1273578008480771,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2179,
       "step": 69390
     },
     {
       "epoch": 1.1275202677454468,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2056,
       "step": 69400
     },
     {
       "epoch": 1.1276827346428164,
-      "grad_norm": 2.75,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2066,
       "step": 69410
     },
     {
       "epoch": 1.127845201540186,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.1967,
       "step": 69420
     },
     {
       "epoch": 1.1280076684375557,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.1837,
       "step": 69430
     },
     {
       "epoch": 1.1281701353349256,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2215,
       "step": 69440
     },
     {
       "epoch": 1.1283326022322953,
-      "grad_norm": 2.5,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1841,
       "step": 69450
     },
     {
       "epoch": 1.128495069129665,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.1809,
       "step": 69460
     },
     {
       "epoch": 1.1286575360270346,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3579,
+      "loss": 1.175,
       "step": 69470
     },
     {
       "epoch": 1.1288200029244042,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2223,
       "step": 69480
     },
     {
       "epoch": 1.1289824698217739,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1993,
       "step": 69490
     },
     {
       "epoch": 1.1291449367191435,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2052,
       "step": 69500
     },
     {
       "epoch": 1.1293074036165132,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.1911,
       "step": 69510
     },
     {
       "epoch": 1.1294698705138828,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.1674,
       "step": 69520
     },
     {
       "epoch": 1.1296323374112525,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.1801,
       "step": 69530
     },
     {
       "epoch": 1.1297948043086221,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.1833,
       "step": 69540
     },
     {
       "epoch": 1.1299572712059918,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2215,
       "step": 69550
     },
     {
       "epoch": 1.1301197381033614,
-      "grad_norm": 2.75,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.206,
       "step": 69560
     },
     {
       "epoch": 1.130282205000731,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2156,
       "step": 69570
     },
     {
       "epoch": 1.1304446718981007,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2584,
       "step": 69580
     },
     {
       "epoch": 1.1306071387954704,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2156,
       "step": 69590
     },
     {
       "epoch": 1.13076960569284,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.2299,
       "step": 69600
     },
     {
       "epoch": 1.1309320725902097,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.1955,
       "step": 69610
     },
     {
       "epoch": 1.1310945394875793,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.1874,
       "step": 69620
     },
     {
       "epoch": 1.131257006384949,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3208,
+      "loss": 1.2102,
       "step": 69630
     },
     {
       "epoch": 1.1314194732823188,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.1906,
       "step": 69640
     },
     {
       "epoch": 1.1315819401796885,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.1787,
       "step": 69650
     },
     {
       "epoch": 1.1317444070770581,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.2078,
       "step": 69660
     },
     {
       "epoch": 1.1319068739744278,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.1808,
       "step": 69670
     },
     {
       "epoch": 1.1320693408717974,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2121,
       "step": 69680
     },
     {
       "epoch": 1.132231807769167,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.1952,
       "step": 69690
     },
     {
       "epoch": 1.1323942746665367,
-      "grad_norm": 3.890625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1812,
       "step": 69700
     },
     {
       "epoch": 1.1325567415639064,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2057,
       "step": 69710
     },
     {
       "epoch": 1.132719208461276,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2337,
       "step": 69720
     },
     {
       "epoch": 1.1328816753586457,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2217,
       "step": 69730
     },
     {
       "epoch": 1.1330441422560154,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2386,
       "step": 69740
     },
     {
       "epoch": 1.133206609153385,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3674,
+      "loss": 1.1947,
       "step": 69750
     },
     {
       "epoch": 1.1333690760507547,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.1869,
       "step": 69760
     },
     {
       "epoch": 1.1335315429481243,
-      "grad_norm": 2.5,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.1939,
       "step": 69770
     },
     {
       "epoch": 1.133694009845494,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.1987,
       "step": 69780
     },
     {
       "epoch": 1.1338564767428636,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2072,
       "step": 69790
     },
     {
       "epoch": 1.1340189436402333,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2138,
       "step": 69800
     },
     {
       "epoch": 1.134181410537603,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.182,
       "step": 69810
     },
     {
       "epoch": 1.1343438774349726,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2259,
       "step": 69820
     },
     {
       "epoch": 1.1345063443323422,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.1827,
       "step": 69830
     },
     {
       "epoch": 1.1346688112297119,
-      "grad_norm": 2.25,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.1962,
       "step": 69840
     },
     {
       "epoch": 1.1348312781270815,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.1787,
       "step": 69850
     },
     {
       "epoch": 1.1349937450244512,
-      "grad_norm": 2.875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1643,
       "step": 69860
     },
     {
       "epoch": 1.1351562119218208,
-      "grad_norm": 2.875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.1884,
       "step": 69870
     },
     {
       "epoch": 1.1353186788191907,
-      "grad_norm": 2.171875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.32,
+      "loss": 1.2071,
       "step": 69880
     },
     {
       "epoch": 1.1354811457165603,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2237,
       "step": 69890
     },
     {
       "epoch": 1.13564361261393,
-      "grad_norm": 2.375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.1978,
       "step": 69900
     },
     {
       "epoch": 1.1358060795112996,
-      "grad_norm": 3.25,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2182,
       "step": 69910
     },
     {
       "epoch": 1.1359685464086693,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2077,
       "step": 69920
     },
     {
       "epoch": 1.136131013306039,
-      "grad_norm": 3.546875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.1828,
       "step": 69930
     },
     {
       "epoch": 1.1362934802034086,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.204,
       "step": 69940
     },
     {
       "epoch": 1.1364559471007782,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2252,
       "step": 69950
     },
     {
       "epoch": 1.136618413998148,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2364,
       "step": 69960
     },
     {
       "epoch": 1.1367808808955175,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.2084,
       "step": 69970
     },
     {
       "epoch": 1.1369433477928872,
-      "grad_norm": 2.75,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1955,
       "step": 69980
     },
     {
       "epoch": 1.1371058146902568,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.1962,
       "step": 69990
     },
     {
       "epoch": 1.1372682815876265,
-      "grad_norm": 3.375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3666,
+      "loss": 1.1995,
       "step": 70000
     },
     {
       "epoch": 1.1374307484849961,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2008,
       "step": 70010
     },
     {
       "epoch": 1.1375932153823658,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3565,
+      "loss": 1.2037,
       "step": 70020
     },
     {
       "epoch": 1.1377556822797354,
-      "grad_norm": 3.0,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.1809,
       "step": 70030
     },
     {
       "epoch": 1.137918149177105,
-      "grad_norm": 2.5,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.198,
       "step": 70040
     },
     {
       "epoch": 1.1380806160744747,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.1998,
       "step": 70050
     },
     {
       "epoch": 1.1382430829718444,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2084,
       "step": 70060
     },
     {
       "epoch": 1.138405549869214,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.1877,
       "step": 70070
     },
     {
       "epoch": 1.138568016766584,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3235,
+      "loss": 1.1612,
       "step": 70080
     },
     {
       "epoch": 1.1387304836639536,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2028,
       "step": 70090
     },
     {
       "epoch": 1.1388929505613232,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.188,
       "step": 70100
     },
     {
       "epoch": 1.1390554174586929,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.1984,
       "step": 70110
     },
     {
       "epoch": 1.1392178843560625,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2239,
       "step": 70120
     },
     {
       "epoch": 1.1393803512534322,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.2066,
       "step": 70130
     },
     {
       "epoch": 1.1395428181508018,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.1886,
       "step": 70140
     },
     {
       "epoch": 1.1397052850481715,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.1802,
       "step": 70150
     },
     {
       "epoch": 1.1398677519455411,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.379,
+      "loss": 1.2082,
       "step": 70160
     },
     {
       "epoch": 1.1400302188429108,
-      "grad_norm": 3.75,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2392,
       "step": 70170
     },
     {
       "epoch": 1.1401926857402804,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.1955,
       "step": 70180
     },
     {
       "epoch": 1.14035515263765,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.1968,
       "step": 70190
     },
     {
       "epoch": 1.1405176195350197,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2339,
       "step": 70200
     },
     {
       "epoch": 1.1406800864323894,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.1946,
       "step": 70210
     },
     {
       "epoch": 1.140842553329759,
-      "grad_norm": 2.75,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.1733,
       "step": 70220
     },
     {
       "epoch": 1.1410050202271287,
-      "grad_norm": 2.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2119,
       "step": 70230
     },
     {
       "epoch": 1.1411674871244983,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.1921,
       "step": 70240
     },
     {
       "epoch": 1.141329954021868,
-      "grad_norm": 2.265625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2355,
       "step": 70250
     },
     {
       "epoch": 1.1414924209192376,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.1747,
       "step": 70260
     },
     {
       "epoch": 1.1416548878166073,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2368,
       "step": 70270
     },
     {
       "epoch": 1.141817354713977,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2228,
       "step": 70280
     },
     {
       "epoch": 1.1419798216113466,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.212,
       "step": 70290
     },
     {
       "epoch": 1.1421422885087162,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.1892,
       "step": 70300
     },
     {
       "epoch": 1.142304755406086,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.21,
       "step": 70310
     },
     {
       "epoch": 1.1424672223034558,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2478,
       "step": 70320
     },
     {
       "epoch": 1.1426296892008254,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3098,
+      "loss": 1.1724,
       "step": 70330
     },
     {
       "epoch": 1.142792156098195,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.2023,
       "step": 70340
     },
     {
       "epoch": 1.1429546229955647,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2147,
       "step": 70350
     },
     {
       "epoch": 1.1431170898929344,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.1762,
       "step": 70360
     },
     {
       "epoch": 1.143279556790304,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3163,
+      "loss": 1.2031,
       "step": 70370
     },
     {
       "epoch": 1.1434420236876737,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.2352,
       "step": 70380
     },
     {
       "epoch": 1.1436044905850433,
-      "grad_norm": 3.640625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2059,
       "step": 70390
     },
     {
       "epoch": 1.143766957482413,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.226,
       "step": 70400
     },
     {
       "epoch": 1.1439294243797826,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2296,
       "step": 70410
     },
     {
       "epoch": 1.1440918912771523,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.1668,
       "step": 70420
     },
     {
       "epoch": 1.144254358174522,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2552,
       "step": 70430
     },
     {
       "epoch": 1.1444168250718916,
-      "grad_norm": 4.8125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.1969,
       "step": 70440
     },
     {
       "epoch": 1.1445792919692612,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.2084,
       "step": 70450
     },
     {
       "epoch": 1.1447417588666309,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3587,
+      "loss": 1.2273,
       "step": 70460
     },
     {
       "epoch": 1.1449042257640005,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2274,
       "step": 70470
     },
     {
       "epoch": 1.1450666926613702,
-      "grad_norm": 2.140625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3603,
+      "loss": 1.1674,
       "step": 70480
     },
     {
       "epoch": 1.1452291595587398,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.1904,
       "step": 70490
     },
     {
       "epoch": 1.1453916264561095,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.1903,
       "step": 70500
     },
     {
       "epoch": 1.1455540933534794,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.1929,
       "step": 70510
     },
     {
       "epoch": 1.145716560250849,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.2191,
       "step": 70520
     },
     {
       "epoch": 1.1458790271482187,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2394,
       "step": 70530
     },
     {
       "epoch": 1.1460414940455883,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2305,
       "step": 70540
     },
     {
       "epoch": 1.146203960942958,
-      "grad_norm": 2.75,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.1778,
       "step": 70550
     },
     {
       "epoch": 1.1463664278403276,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2035,
       "step": 70560
     },
     {
       "epoch": 1.1465288947376973,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.1694,
       "step": 70570
     },
     {
       "epoch": 1.146691361635067,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.2051,
       "step": 70580
     },
     {
       "epoch": 1.1468538285324366,
-      "grad_norm": 2.5,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2225,
       "step": 70590
     },
     {
       "epoch": 1.1470162954298062,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2124,
       "step": 70600
     },
     {
       "epoch": 1.1471787623271759,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3688,
+      "loss": 1.2018,
       "step": 70610
     },
     {
       "epoch": 1.1473412292245455,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.2422,
       "step": 70620
     },
     {
       "epoch": 1.1475036961219152,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3102,
+      "loss": 1.2351,
       "step": 70630
     },
     {
       "epoch": 1.1476661630192848,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2255,
       "step": 70640
     },
     {
       "epoch": 1.1478286299166545,
-      "grad_norm": 1.875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3149,
+      "loss": 1.1919,
       "step": 70650
     },
     {
       "epoch": 1.1479910968140241,
-      "grad_norm": 1.4765625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2036,
       "step": 70660
     },
     {
       "epoch": 1.1481535637113938,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.197,
       "step": 70670
     },
     {
       "epoch": 1.1483160306087634,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2193,
       "step": 70680
     },
     {
       "epoch": 1.148478497506133,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.1699,
       "step": 70690
     },
     {
       "epoch": 1.1486409644035027,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.1958,
       "step": 70700
     },
     {
       "epoch": 1.1488034313008724,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.2574,
       "step": 70710
     },
     {
       "epoch": 1.148965898198242,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2258,
       "step": 70720
     },
     {
       "epoch": 1.1491283650956117,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2186,
       "step": 70730
     },
     {
       "epoch": 1.1492908319929813,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2038,
       "step": 70740
     },
     {
       "epoch": 1.149453298890351,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.1524,
       "step": 70750
     },
     {
       "epoch": 1.1496157657877208,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2149,
       "step": 70760
     },
     {
       "epoch": 1.1497782326850905,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2063,
       "step": 70770
     },
     {
       "epoch": 1.1499406995824601,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.1972,
       "step": 70780
     },
     {
       "epoch": 1.1501031664798298,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.1704,
       "step": 70790
     },
     {
       "epoch": 1.1502656333771994,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3108,
+      "loss": 1.1843,
       "step": 70800
     },
     {
       "epoch": 1.150428100274569,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2334,
       "step": 70810
     },
     {
       "epoch": 1.1505905671719387,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.1958,
       "step": 70820
     },
     {
       "epoch": 1.1507530340693084,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2412,
       "step": 70830
     },
     {
       "epoch": 1.150915500966678,
-      "grad_norm": 1.9140625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2107,
       "step": 70840
     },
     {
       "epoch": 1.1510779678640477,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2213,
       "step": 70850
     },
     {
       "epoch": 1.1512404347614174,
-      "grad_norm": 1.96875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.2218,
       "step": 70860
     },
     {
       "epoch": 1.151402901658787,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.1966,
       "step": 70870
     },
     {
       "epoch": 1.1515653685561567,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1742,
       "step": 70880
     },
     {
       "epoch": 1.1517278354535263,
-      "grad_norm": 2.1875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2368,
       "step": 70890
     },
     {
       "epoch": 1.151890302350896,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2092,
       "step": 70900
     },
     {
       "epoch": 1.1520527692482656,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.1601,
       "step": 70910
     },
     {
       "epoch": 1.1522152361456353,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2049,
       "step": 70920
     },
     {
       "epoch": 1.152377703043005,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.2436,
       "step": 70930
     },
     {
       "epoch": 1.1525401699403746,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2133,
       "step": 70940
     },
     {
       "epoch": 1.1527026368377444,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.1685,
       "step": 70950
     },
     {
       "epoch": 1.152865103735114,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.218,
       "step": 70960
     },
     {
       "epoch": 1.1530275706324837,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3131,
+      "loss": 1.2048,
       "step": 70970
     },
     {
       "epoch": 1.1531900375298534,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2452,
       "step": 70980
     },
     {
       "epoch": 1.153352504427223,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2196,
       "step": 70990
     },
     {
       "epoch": 1.1535149713245927,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3165,
+      "loss": 1.197,
       "step": 71000
     },
     {
       "epoch": 1.1536774382219623,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3072,
+      "loss": 1.2132,
       "step": 71010
     },
     {
       "epoch": 1.153839905119332,
-      "grad_norm": 4.0,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3611,
+      "loss": 1.2068,
       "step": 71020
     },
     {
       "epoch": 1.1540023720167016,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2348,
       "step": 71030
     },
     {
       "epoch": 1.1541648389140713,
-      "grad_norm": 2.375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.1608,
       "step": 71040
     },
     {
       "epoch": 1.154327305811441,
-      "grad_norm": 2.625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.1688,
       "step": 71050
     },
     {
       "epoch": 1.1544897727088106,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2013,
       "step": 71060
     },
     {
       "epoch": 1.1546522396061802,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.1751,
       "step": 71070
     },
     {
       "epoch": 1.15481470650355,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.1801,
       "step": 71080
     },
     {
       "epoch": 1.1549771734009195,
-      "grad_norm": 2.640625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3238,
+      "loss": 1.2221,
       "step": 71090
     },
     {
       "epoch": 1.1551396402982892,
-      "grad_norm": 2.75,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.209,
       "step": 71100
     },
     {
       "epoch": 1.1553021071956588,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.1944,
       "step": 71110
     },
     {
       "epoch": 1.1554645740930285,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2321,
       "step": 71120
     },
     {
       "epoch": 1.1556270409903981,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2528,
       "step": 71130
     },
     {
       "epoch": 1.1557895078877678,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.1966,
       "step": 71140
     },
     {
       "epoch": 1.1559519747851374,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2154,
       "step": 71150
     },
     {
       "epoch": 1.156114441682507,
-      "grad_norm": 2.625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2446,
       "step": 71160
     },
     {
       "epoch": 1.1562769085798767,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.1936,
       "step": 71170
     },
     {
       "epoch": 1.1564393754772464,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.2397,
       "step": 71180
     },
     {
       "epoch": 1.156601842374616,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.1603,
       "step": 71190
     },
     {
       "epoch": 1.156764309271986,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2099,
       "step": 71200
     },
     {
       "epoch": 1.1569267761693556,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2314,
       "step": 71210
     },
     {
       "epoch": 1.1570892430667252,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2047,
       "step": 71220
     },
     {
       "epoch": 1.1572517099640949,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3719,
+      "loss": 1.19,
       "step": 71230
     },
     {
       "epoch": 1.1574141768614645,
-      "grad_norm": 2.5,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3178,
+      "loss": 1.2022,
       "step": 71240
     },
     {
       "epoch": 1.1575766437588342,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2037,
       "step": 71250
     },
     {
       "epoch": 1.1577391106562038,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.2328,
       "step": 71260
     },
     {
       "epoch": 1.1579015775535735,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2083,
       "step": 71270
     },
     {
       "epoch": 1.1580640444509431,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3636,
+      "loss": 1.2243,
       "step": 71280
     },
     {
       "epoch": 1.1582265113483128,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3616,
+      "loss": 1.1689,
       "step": 71290
     },
     {
       "epoch": 1.1583889782456824,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.234,
       "step": 71300
     },
     {
       "epoch": 1.158551445143052,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2329,
       "step": 71310
     },
     {
       "epoch": 1.1587139120404217,
-      "grad_norm": 2.21875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3223,
+      "loss": 1.2128,
       "step": 71320
     },
     {
       "epoch": 1.1588763789377914,
-      "grad_norm": 3.421875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.2027,
       "step": 71330
     },
     {
       "epoch": 1.159038845835161,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.1956,
       "step": 71340
     },
     {
       "epoch": 1.1592013127325307,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2348,
       "step": 71350
     },
     {
       "epoch": 1.1593637796299003,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.1973,
       "step": 71360
     },
     {
       "epoch": 1.15952624652727,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2292,
       "step": 71370
     },
     {
       "epoch": 1.1596887134246396,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.1939,
       "step": 71380
     },
     {
       "epoch": 1.1598511803220095,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1994,
       "step": 71390
     },
     {
       "epoch": 1.1600136472193792,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2193,
       "step": 71400
     },
     {
       "epoch": 1.1601761141167488,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2175,
       "step": 71410
     },
     {
       "epoch": 1.1603385810141185,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2101,
       "step": 71420
     },
     {
       "epoch": 1.1605010479114881,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3666,
+      "loss": 1.1809,
       "step": 71430
     },
     {
       "epoch": 1.1606635148088578,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3254,
+      "loss": 1.2153,
       "step": 71440
     },
     {
       "epoch": 1.1608259817062274,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2163,
       "step": 71450
     },
     {
       "epoch": 1.160988448603597,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2219,
       "step": 71460
     },
     {
       "epoch": 1.1611509155009667,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.317,
+      "loss": 1.242,
       "step": 71470
     },
     {
       "epoch": 1.1613133823983364,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.1775,
       "step": 71480
     },
     {
       "epoch": 1.161475849295706,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.257,
       "step": 71490
     },
     {
       "epoch": 1.1616383161930757,
-      "grad_norm": 2.5,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.194,
       "step": 71500
     },
     {
       "epoch": 1.1618007830904453,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.1952,
       "step": 71510
     },
     {
       "epoch": 1.161963249987815,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.1703,
       "step": 71520
     },
     {
       "epoch": 1.1621257168851846,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.2383,
       "step": 71530
     },
     {
       "epoch": 1.1622881837825543,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1889,
       "step": 71540
     },
     {
       "epoch": 1.162450650679924,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.1687,
       "step": 71550
     },
     {
       "epoch": 1.1626131175772936,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2182,
       "step": 71560
     },
     {
       "epoch": 1.1627755844746632,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1975,
       "step": 71570
     },
     {
       "epoch": 1.1629380513720329,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.1909,
       "step": 71580
     },
     {
       "epoch": 1.1631005182694025,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.2386,
       "step": 71590
     },
     {
       "epoch": 1.1632629851667722,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3173,
+      "loss": 1.1764,
       "step": 71600
     },
     {
       "epoch": 1.1634254520641418,
-      "grad_norm": 2.0,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.1914,
       "step": 71610
     },
     {
       "epoch": 1.1635879189615115,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3136,
+      "loss": 1.2349,
       "step": 71620
     },
     {
       "epoch": 1.1637503858588811,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2339,
       "step": 71630
     },
     {
       "epoch": 1.163912852756251,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2072,
       "step": 71640
     },
     {
       "epoch": 1.1640753196536207,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2084,
       "step": 71650
     },
     {
       "epoch": 1.1642377865509903,
-      "grad_norm": 2.75,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.1884,
       "step": 71660
     },
     {
       "epoch": 1.16440025344836,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.1961,
       "step": 71670
     },
     {
       "epoch": 1.1645627203457296,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2232,
       "step": 71680
     },
     {
       "epoch": 1.1647251872430993,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2075,
       "step": 71690
     },
     {
       "epoch": 1.164887654140469,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3144,
+      "loss": 1.1946,
       "step": 71700
     },
     {
       "epoch": 1.1650501210378386,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2331,
       "step": 71710
     },
     {
       "epoch": 1.1652125879352082,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2265,
       "step": 71720
     },
     {
       "epoch": 1.1653750548325779,
-      "grad_norm": 4.25,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2055,
       "step": 71730
     },
     {
       "epoch": 1.1655375217299475,
-      "grad_norm": 2.625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.1758,
       "step": 71740
     },
     {
       "epoch": 1.1656999886273172,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.224,
       "step": 71750
     },
     {
       "epoch": 1.1658624555246868,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.197,
       "step": 71760
     },
     {
       "epoch": 1.1660249224220565,
-      "grad_norm": 2.5,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.1994,
       "step": 71770
     },
     {
       "epoch": 1.1661873893194261,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2367,
       "step": 71780
     },
     {
       "epoch": 1.1663498562167958,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.2072,
       "step": 71790
     },
     {
       "epoch": 1.1665123231141654,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2136,
       "step": 71800
     },
     {
       "epoch": 1.166674790011535,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.2394,
       "step": 71810
     },
     {
       "epoch": 1.1668372569089047,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2453,
       "step": 71820
     },
     {
       "epoch": 1.1669997238062746,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2501,
       "step": 71830
     },
     {
       "epoch": 1.1671621907036442,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1799,
       "step": 71840
     },
     {
       "epoch": 1.167324657601014,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.2233,
       "step": 71850
     },
     {
       "epoch": 1.1674871244983835,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1933,
       "step": 71860
     },
     {
       "epoch": 1.1676495913957532,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2208,
       "step": 71870
     },
     {
       "epoch": 1.1678120582931228,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2342,
       "step": 71880
     },
     {
       "epoch": 1.1679745251904925,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3668,
+      "loss": 1.1874,
       "step": 71890
     },
     {
       "epoch": 1.1681369920878621,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2181,
       "step": 71900
     },
     {
       "epoch": 1.1682994589852318,
-      "grad_norm": 1.7890625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2399,
       "step": 71910
     },
     {
       "epoch": 1.1684619258826014,
-      "grad_norm": 3.25,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.1855,
       "step": 71920
     },
     {
       "epoch": 1.168624392779971,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2311,
       "step": 71930
     },
     {
       "epoch": 1.1687868596773408,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2119,
       "step": 71940
     },
     {
       "epoch": 1.1689493265747104,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.2359,
       "step": 71950
     },
     {
       "epoch": 1.16911179347208,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.2234,
       "step": 71960
     },
     {
       "epoch": 1.1692742603694497,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3139,
+      "loss": 1.2179,
       "step": 71970
     },
     {
       "epoch": 1.1694367272668194,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2316,
       "step": 71980
     },
     {
       "epoch": 1.169599194164189,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2221,
       "step": 71990
     },
     {
       "epoch": 1.1697616610615587,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.1941,
       "step": 72000
     },
     {
       "epoch": 1.1699241279589283,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1989,
       "step": 72010
     },
     {
       "epoch": 1.170086594856298,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2081,
       "step": 72020
     },
     {
       "epoch": 1.1702490617536676,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.1906,
       "step": 72030
     },
     {
       "epoch": 1.1704115286510373,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2041,
       "step": 72040
     },
     {
       "epoch": 1.170573995548407,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1921,
       "step": 72050
     },
     {
       "epoch": 1.1707364624457766,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2404,
       "step": 72060
     },
     {
       "epoch": 1.1708989293431462,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.1997,
       "step": 72070
     },
     {
       "epoch": 1.171061396240516,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2256,
       "step": 72080
     },
     {
       "epoch": 1.1712238631378857,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3254,
+      "loss": 1.198,
       "step": 72090
     },
     {
       "epoch": 1.1713863300352554,
-      "grad_norm": 2.875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.2071,
       "step": 72100
     },
     {
       "epoch": 1.171548796932625,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2178,
       "step": 72110
     },
     {
       "epoch": 1.1717112638299947,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.1877,
       "step": 72120
     },
     {
       "epoch": 1.1718737307273643,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2319,
       "step": 72130
     },
     {
       "epoch": 1.172036197624734,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2071,
       "step": 72140
     },
     {
       "epoch": 1.1721986645221036,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2138,
       "step": 72150
     },
     {
       "epoch": 1.1723611314194733,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.2215,
       "step": 72160
     },
     {
       "epoch": 1.172523598316843,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2657,
       "step": 72170
     },
     {
       "epoch": 1.1726860652142126,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.1423,
       "step": 72180
     },
     {
       "epoch": 1.1728485321115822,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3206,
+      "loss": 1.2004,
       "step": 72190
     },
     {
       "epoch": 1.173010999008952,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2113,
       "step": 72200
     },
     {
       "epoch": 1.1731734659063215,
-      "grad_norm": 1.9453125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3266,
+      "loss": 1.1807,
       "step": 72210
     },
     {
       "epoch": 1.1733359328036912,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2038,
       "step": 72220
     },
     {
       "epoch": 1.1734983997010608,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3656,
+      "loss": 1.1819,
       "step": 72230
     },
     {
       "epoch": 1.1736608665984305,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.2254,
       "step": 72240
     },
     {
       "epoch": 1.1738233334958001,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.1876,
       "step": 72250
     },
     {
       "epoch": 1.1739858003931698,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.2107,
       "step": 72260
     },
     {
       "epoch": 1.1741482672905397,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.1728,
       "step": 72270
     },
     {
       "epoch": 1.1743107341879093,
-      "grad_norm": 1.734375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2105,
       "step": 72280
     },
     {
       "epoch": 1.174473201085279,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2328,
       "step": 72290
     },
     {
       "epoch": 1.1746356679826486,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2162,
       "step": 72300
     },
     {
       "epoch": 1.1747981348800183,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2075,
       "step": 72310
     },
     {
       "epoch": 1.174960601777388,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2295,
       "step": 72320
     },
     {
       "epoch": 1.1751230686747576,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2539,
       "step": 72330
     },
     {
       "epoch": 1.1752855355721272,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2151,
       "step": 72340
     },
     {
       "epoch": 1.1754480024694969,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2385,
       "step": 72350
     },
     {
       "epoch": 1.1756104693668665,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.1697,
       "step": 72360
     },
     {
       "epoch": 1.1757729362642362,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3564,
+      "loss": 1.2032,
       "step": 72370
     },
     {
       "epoch": 1.1759354031616058,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.1847,
       "step": 72380
     },
     {
       "epoch": 1.1760978700589755,
-      "grad_norm": 1.90625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2047,
       "step": 72390
     },
     {
       "epoch": 1.1762603369563451,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2062,
       "step": 72400
     },
     {
       "epoch": 1.1764228038537148,
-      "grad_norm": 3.625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2055,
       "step": 72410
     },
     {
       "epoch": 1.1765852707510844,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2099,
       "step": 72420
     },
     {
       "epoch": 1.176747737648454,
-      "grad_norm": 1.7109375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2054,
       "step": 72430
     },
     {
       "epoch": 1.1769102045458237,
-      "grad_norm": 1.6640625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2313,
       "step": 72440
     },
     {
       "epoch": 1.1770726714431934,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3107,
+      "loss": 1.2171,
       "step": 72450
     },
     {
       "epoch": 1.177235138340563,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.1947,
       "step": 72460
     },
     {
       "epoch": 1.1773976052379327,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.1923,
       "step": 72470
     },
     {
       "epoch": 1.1775600721353023,
-      "grad_norm": 1.9296875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.2366,
       "step": 72480
     },
     {
       "epoch": 1.177722539032672,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2255,
       "step": 72490
     },
     {
       "epoch": 1.1778850059300416,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3117,
+      "loss": 1.2043,
       "step": 72500
     },
     {
       "epoch": 1.1780474728274113,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2526,
       "step": 72510
     },
     {
       "epoch": 1.1782099397247812,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1748,
       "step": 72520
     },
     {
       "epoch": 1.1783724066221508,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.1924,
       "step": 72530
     },
     {
       "epoch": 1.1785348735195205,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.1895,
       "step": 72540
     },
     {
       "epoch": 1.1786973404168901,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2673,
       "step": 72550
     },
     {
       "epoch": 1.1788598073142598,
-      "grad_norm": 1.9296875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.23,
       "step": 72560
     },
     {
       "epoch": 1.1790222742116294,
-      "grad_norm": 2.625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.189,
       "step": 72570
     },
     {
       "epoch": 1.179184741108999,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2215,
       "step": 72580
     },
     {
       "epoch": 1.1793472080063687,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2187,
       "step": 72590
     },
     {
       "epoch": 1.1795096749037384,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3611,
+      "loss": 1.1992,
       "step": 72600
     },
     {
       "epoch": 1.179672141801108,
-      "grad_norm": 2.375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.216,
       "step": 72610
     },
     {
       "epoch": 1.1798346086984777,
-      "grad_norm": 2.203125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.1852,
       "step": 72620
     },
     {
       "epoch": 1.1799970755958473,
-      "grad_norm": 4.40625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2361,
       "step": 72630
     },
     {
       "epoch": 1.180159542493217,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2227,
       "step": 72640
     },
     {
       "epoch": 1.1803220093905866,
-      "grad_norm": 2.140625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2453,
       "step": 72650
     },
     {
       "epoch": 1.1804844762879563,
-      "grad_norm": 1.75,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2182,
       "step": 72660
     },
     {
       "epoch": 1.180646943185326,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.1982,
       "step": 72670
     },
     {
       "epoch": 1.1808094100826956,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.1932,
       "step": 72680
     },
     {
       "epoch": 1.1809718769800652,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2145,
       "step": 72690
     },
     {
       "epoch": 1.1811343438774349,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2379,
       "step": 72700
     },
     {
       "epoch": 1.1812968107748048,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2317,
       "step": 72710
     },
     {
       "epoch": 1.1814592776721744,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1793,
       "step": 72720
     },
     {
       "epoch": 1.181621744569544,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.2327,
       "step": 72730
     },
     {
       "epoch": 1.1817842114669137,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2262,
       "step": 72740
     },
     {
       "epoch": 1.1819466783642834,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2589,
       "step": 72750
     },
     {
       "epoch": 1.182109145261653,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2059,
       "step": 72760
     },
     {
       "epoch": 1.1822716121590227,
-      "grad_norm": 2.25,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.1937,
       "step": 72770
     },
     {
       "epoch": 1.1824340790563923,
-      "grad_norm": 2.125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.234,
       "step": 72780
     },
     {
       "epoch": 1.182596545953762,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.1919,
       "step": 72790
     },
     {
       "epoch": 1.1827590128511316,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.1925,
       "step": 72800
     },
     {
       "epoch": 1.1829214797485013,
-      "grad_norm": 2.75,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2228,
       "step": 72810
     },
     {
       "epoch": 1.183083946645871,
-      "grad_norm": 1.953125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.1764,
       "step": 72820
     },
     {
       "epoch": 1.1832464135432406,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3264,
+      "loss": 1.2305,
       "step": 72830
     },
     {
       "epoch": 1.1834088804406102,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1653,
       "step": 72840
     },
     {
       "epoch": 1.1835713473379799,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2194,
       "step": 72850
     },
     {
       "epoch": 1.1837338142353495,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2279,
       "step": 72860
     },
     {
       "epoch": 1.1838962811327192,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.38,
+      "loss": 1.205,
       "step": 72870
     },
     {
       "epoch": 1.1840587480300888,
-      "grad_norm": 1.9609375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.177,
       "step": 72880
     },
     {
       "epoch": 1.1842212149274585,
-      "grad_norm": 1.6875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.1797,
       "step": 72890
     },
     {
       "epoch": 1.1843836818248281,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3655,
+      "loss": 1.2128,
       "step": 72900
     },
     {
       "epoch": 1.1845461487221978,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.2089,
       "step": 72910
     },
     {
       "epoch": 1.1847086156195674,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2045,
       "step": 72920
     },
     {
       "epoch": 1.184871082516937,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.2228,
       "step": 72930
     },
     {
       "epoch": 1.1850335494143067,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.1977,
       "step": 72940
     },
     {
       "epoch": 1.1851960163116766,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1998,
       "step": 72950
     },
     {
       "epoch": 1.1853584832090462,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.2488,
       "step": 72960
     },
     {
       "epoch": 1.185520950106416,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.1676,
       "step": 72970
     },
     {
       "epoch": 1.1856834170037855,
-      "grad_norm": 2.828125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2225,
       "step": 72980
     },
     {
       "epoch": 1.1858458839011552,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.1625,
       "step": 72990
     },
     {
       "epoch": 1.1860083507985248,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2071,
       "step": 73000
     },
     {
       "epoch": 1.1861708176958945,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2367,
       "step": 73010
     },
     {
       "epoch": 1.1863332845932641,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.1896,
       "step": 73020
     },
     {
       "epoch": 1.1864957514906338,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3195,
+      "loss": 1.216,
       "step": 73030
     },
     {
       "epoch": 1.1866582183880034,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.2522,
       "step": 73040
     },
     {
       "epoch": 1.186820685285373,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2057,
       "step": 73050
     },
     {
       "epoch": 1.1869831521827428,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.249,
       "step": 73060
     },
     {
       "epoch": 1.1871456190801124,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3177,
+      "loss": 1.2045,
       "step": 73070
     },
     {
       "epoch": 1.187308085977482,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2129,
       "step": 73080
     },
     {
       "epoch": 1.1874705528748517,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2466,
       "step": 73090
     },
     {
       "epoch": 1.1876330197722214,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2242,
       "step": 73100
     },
     {
       "epoch": 1.187795486669591,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.1993,
       "step": 73110
     },
     {
       "epoch": 1.1879579535669607,
-      "grad_norm": 2.25,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3178,
+      "loss": 1.2408,
       "step": 73120
     },
     {
       "epoch": 1.1881204204643303,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.1682,
       "step": 73130
     },
     {
       "epoch": 1.1882828873617,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.1921,
       "step": 73140
     },
     {
       "epoch": 1.1884453542590698,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2144,
       "step": 73150
     },
     {
       "epoch": 1.1886078211564395,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2058,
       "step": 73160
     },
     {
       "epoch": 1.1887702880538091,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.1877,
       "step": 73170
     },
     {
       "epoch": 1.1889327549511788,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2217,
       "step": 73180
     },
     {
       "epoch": 1.1890952218485484,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.2202,
       "step": 73190
     },
     {
       "epoch": 1.189257688745918,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.1996,
       "step": 73200
     },
     {
       "epoch": 1.1894201556432877,
-      "grad_norm": 3.0,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.2353,
       "step": 73210
     },
     {
       "epoch": 1.1895826225406574,
-      "grad_norm": 2.625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.1663,
       "step": 73220
     },
     {
       "epoch": 1.189745089438027,
-      "grad_norm": 2.625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.1904,
       "step": 73230
     },
     {
       "epoch": 1.1899075563353967,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2086,
       "step": 73240
     },
     {
       "epoch": 1.1900700232327663,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3145,
+      "loss": 1.199,
       "step": 73250
     },
     {
       "epoch": 1.190232490130136,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2212,
       "step": 73260
     },
     {
       "epoch": 1.1903949570275056,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2219,
       "step": 73270
     },
     {
       "epoch": 1.1905574239248753,
-      "grad_norm": 2.875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3577,
+      "loss": 1.2014,
       "step": 73280
     },
     {
       "epoch": 1.190719890822245,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2339,
       "step": 73290
     },
     {
       "epoch": 1.1908823577196146,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3252,
+      "loss": 1.2379,
       "step": 73300
     },
     {
       "epoch": 1.1910448246169842,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2108,
       "step": 73310
     },
     {
       "epoch": 1.191207291514354,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.207,
       "step": 73320
     },
     {
       "epoch": 1.1913697584117235,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3232,
+      "loss": 1.2045,
       "step": 73330
     },
     {
       "epoch": 1.1915322253090932,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2439,
       "step": 73340
     },
     {
       "epoch": 1.1916946922064628,
-      "grad_norm": 3.375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.1719,
       "step": 73350
     },
     {
       "epoch": 1.1918571591038325,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.194,
       "step": 73360
     },
     {
       "epoch": 1.1920196260012021,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2444,
       "step": 73370
     },
     {
       "epoch": 1.1921820928985718,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2302,
       "step": 73380
     },
     {
       "epoch": 1.1923445597959417,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.2226,
       "step": 73390
     },
     {
       "epoch": 1.1925070266933113,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.1931,
       "step": 73400
     },
     {
       "epoch": 1.192669493590681,
-      "grad_norm": 3.0,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.1762,
       "step": 73410
     },
     {
       "epoch": 1.1928319604880506,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.2038,
       "step": 73420
     },
     {
       "epoch": 1.1929944273854203,
-      "grad_norm": 2.375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2209,
       "step": 73430
     },
     {
       "epoch": 1.19315689428279,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3266,
+      "loss": 1.1886,
       "step": 73440
     },
     {
       "epoch": 1.1933193611801596,
-      "grad_norm": 3.375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2316,
       "step": 73450
     },
     {
       "epoch": 1.1934818280775292,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3574,
+      "loss": 1.1842,
       "step": 73460
     },
     {
       "epoch": 1.1936442949748989,
-      "grad_norm": 1.9453125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2328,
       "step": 73470
     },
     {
       "epoch": 1.1938067618722685,
-      "grad_norm": 2.578125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2281,
       "step": 73480
     },
     {
       "epoch": 1.1939692287696382,
-      "grad_norm": 2.625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2549,
       "step": 73490
     },
     {
       "epoch": 1.1941316956670078,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1958,
       "step": 73500
     },
     {
       "epoch": 1.1942941625643775,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1805,
       "step": 73510
     },
     {
       "epoch": 1.1944566294617471,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.2569,
       "step": 73520
     },
     {
       "epoch": 1.1946190963591168,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.1718,
       "step": 73530
     },
     {
       "epoch": 1.1947815632564864,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2626,
       "step": 73540
     },
     {
       "epoch": 1.194944030153856,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.2224,
       "step": 73550
     },
     {
       "epoch": 1.1951064970512257,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.1781,
       "step": 73560
     },
     {
       "epoch": 1.1952689639485954,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2602,
       "step": 73570
     },
     {
       "epoch": 1.1954314308459653,
-      "grad_norm": 2.03125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.227,
       "step": 73580
     },
     {
       "epoch": 1.195593897743335,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2049,
       "step": 73590
     },
     {
       "epoch": 1.1957563646407046,
-      "grad_norm": 2.25,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3684,
+      "loss": 1.1807,
       "step": 73600
     },
     {
       "epoch": 1.1959188315380742,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2124,
       "step": 73610
     },
     {
       "epoch": 1.1960812984354439,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2529,
       "step": 73620
     },
     {
       "epoch": 1.1962437653328135,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3664,
+      "loss": 1.1634,
       "step": 73630
     },
     {
       "epoch": 1.1964062322301832,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2059,
       "step": 73640
     },
     {
       "epoch": 1.1965686991275528,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2241,
       "step": 73650
     },
     {
       "epoch": 1.1967311660249225,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.1964,
       "step": 73660
     },
     {
       "epoch": 1.1968936329222921,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.1873,
       "step": 73670
     },
     {
       "epoch": 1.1970560998196618,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3095,
+      "loss": 1.2196,
       "step": 73680
     },
     {
       "epoch": 1.1972185667170314,
-      "grad_norm": 1.6953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2245,
       "step": 73690
     },
     {
       "epoch": 1.197381033614401,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2313,
       "step": 73700
     },
     {
       "epoch": 1.1975435005117707,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.2806,
       "step": 73710
     },
     {
       "epoch": 1.1977059674091404,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2298,
       "step": 73720
     },
     {
       "epoch": 1.19786843430651,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3629,
+      "loss": 1.2281,
       "step": 73730
     },
     {
       "epoch": 1.1980309012038797,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.1925,
       "step": 73740
     },
     {
       "epoch": 1.1981933681012493,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.1765,
       "step": 73750
     },
     {
       "epoch": 1.198355834998619,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3593,
+      "loss": 1.1995,
       "step": 73760
     },
     {
       "epoch": 1.1985183018959886,
-      "grad_norm": 2.625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3281,
+      "loss": 1.1976,
       "step": 73770
     },
     {
       "epoch": 1.1986807687933583,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2137,
       "step": 73780
     },
     {
       "epoch": 1.198843235690728,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2532,
       "step": 73790
     },
     {
       "epoch": 1.1990057025880976,
-      "grad_norm": 3.125,
+      "grad_norm": 4.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.1995,
       "step": 73800
     },
     {
       "epoch": 1.1991681694854672,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2214,
       "step": 73810
     },
     {
       "epoch": 1.1993306363828369,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2039,
       "step": 73820
     },
     {
       "epoch": 1.1994931032802068,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.1996,
       "step": 73830
     },
     {
       "epoch": 1.1996555701775764,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.202,
       "step": 73840
     },
     {
       "epoch": 1.199818037074946,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.2342,
       "step": 73850
     },
     {
       "epoch": 1.1999805039723157,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2173,
       "step": 73860
     },
     {
       "epoch": 1.2001429708696854,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2104,
       "step": 73870
     },
     {
       "epoch": 1.200305437767055,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2499,
       "step": 73880
     },
     {
       "epoch": 1.2004679046644247,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2698,
       "step": 73890
     },
     {
       "epoch": 1.2006303715617943,
-      "grad_norm": 2.375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.1787,
       "step": 73900
     },
     {
       "epoch": 1.200792838459164,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2414,
       "step": 73910
     },
     {
       "epoch": 1.2009553053565336,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2051,
       "step": 73920
     },
     {
       "epoch": 1.2011177722539033,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.2317,
       "step": 73930
     },
     {
       "epoch": 1.201280239151273,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2087,
       "step": 73940
     },
     {
       "epoch": 1.2014427060486426,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.1948,
       "step": 73950
     },
     {
       "epoch": 1.2016051729460122,
-      "grad_norm": 2.46875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3252,
+      "loss": 1.2036,
       "step": 73960
     },
     {
       "epoch": 1.2017676398433819,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2254,
       "step": 73970
     },
     {
       "epoch": 1.2019301067407515,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2167,
       "step": 73980
     },
     {
       "epoch": 1.2020925736381212,
-      "grad_norm": 2.625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1882,
       "step": 73990
     },
     {
       "epoch": 1.2022550405354908,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.1753,
       "step": 74000
     },
     {
       "epoch": 1.2024175074328605,
-      "grad_norm": 2.625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2069,
       "step": 74010
     },
     {
       "epoch": 1.2025799743302303,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.211,
       "step": 74020
     },
     {
       "epoch": 1.2027424412276,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.1652,
       "step": 74030
     },
     {
       "epoch": 1.2029049081249696,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2036,
       "step": 74040
     },
     {
       "epoch": 1.2030673750223393,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2251,
       "step": 74050
     },
     {
       "epoch": 1.203229841919709,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2169,
       "step": 74060
     },
     {
       "epoch": 1.2033923088170786,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2264,
       "step": 74070
     },
     {
       "epoch": 1.2035547757144482,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2025,
       "step": 74080
     },
     {
       "epoch": 1.203717242611818,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.213,
       "step": 74090
     },
     {
       "epoch": 1.2038797095091875,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3198,
+      "loss": 1.204,
       "step": 74100
     },
     {
       "epoch": 1.2040421764065572,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2353,
       "step": 74110
     },
     {
       "epoch": 1.2042046433039268,
-      "grad_norm": 2.203125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2084,
       "step": 74120
     },
     {
       "epoch": 1.2043671102012965,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2023,
       "step": 74130
     },
     {
       "epoch": 1.2045295770986661,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.1952,
       "step": 74140
     },
     {
       "epoch": 1.2046920439960358,
-      "grad_norm": 2.75,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2001,
       "step": 74150
     },
     {
       "epoch": 1.2048545108934055,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.177,
       "step": 74160
     },
     {
       "epoch": 1.205016977790775,
-      "grad_norm": 3.0,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2296,
       "step": 74170
     },
     {
       "epoch": 1.2051794446881448,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2096,
       "step": 74180
     },
     {
       "epoch": 1.2053419115855144,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.187,
       "step": 74190
     },
     {
       "epoch": 1.205504378482884,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.1633,
       "step": 74200
     },
     {
       "epoch": 1.2056668453802537,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.1961,
       "step": 74210
     },
     {
       "epoch": 1.2058293122776234,
-      "grad_norm": 1.671875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3169,
+      "loss": 1.2139,
       "step": 74220
     },
     {
       "epoch": 1.205991779174993,
-      "grad_norm": 2.375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2161,
       "step": 74230
     },
     {
       "epoch": 1.2061542460723627,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2163,
       "step": 74240
     },
     {
       "epoch": 1.2063167129697323,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.1953,
       "step": 74250
     },
     {
       "epoch": 1.206479179867102,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.2227,
       "step": 74260
     },
     {
       "epoch": 1.2066416467644718,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2208,
       "step": 74270
     },
     {
       "epoch": 1.2068041136618415,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2288,
       "step": 74280
     },
     {
       "epoch": 1.2069665805592111,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3643,
+      "loss": 1.1961,
       "step": 74290
     },
     {
       "epoch": 1.2071290474565808,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1969,
       "step": 74300
     },
     {
       "epoch": 1.2072915143539504,
-      "grad_norm": 2.125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2411,
       "step": 74310
     },
     {
       "epoch": 1.20745398125132,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3098,
+      "loss": 1.2837,
       "step": 74320
     },
     {
       "epoch": 1.2076164481486897,
-      "grad_norm": 2.75,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2296,
       "step": 74330
     },
     {
       "epoch": 1.2077789150460594,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3093,
+      "loss": 1.188,
       "step": 74340
     },
     {
       "epoch": 1.207941381943429,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2094,
       "step": 74350
     },
     {
       "epoch": 1.2081038488407987,
-      "grad_norm": 3.375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.1835,
       "step": 74360
     },
     {
       "epoch": 1.2082663157381683,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2535,
       "step": 74370
     },
     {
       "epoch": 1.208428782635538,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.1906,
       "step": 74380
     },
     {
       "epoch": 1.2085912495329076,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.2208,
       "step": 74390
     },
     {
       "epoch": 1.2087537164302773,
-      "grad_norm": 2.5,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.2046,
       "step": 74400
     },
     {
       "epoch": 1.208916183327647,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2467,
       "step": 74410
     },
     {
       "epoch": 1.2090786502250166,
-      "grad_norm": 2.71875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2064,
       "step": 74420
     },
     {
       "epoch": 1.2092411171223862,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.1991,
       "step": 74430
     },
     {
       "epoch": 1.209403584019756,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.1909,
       "step": 74440
     },
     {
       "epoch": 1.2095660509171255,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2034,
       "step": 74450
     },
     {
       "epoch": 1.2097285178144954,
-      "grad_norm": 2.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3204,
+      "loss": 1.1845,
       "step": 74460
     },
     {
       "epoch": 1.209890984711865,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.1799,
       "step": 74470
     },
     {
       "epoch": 1.2100534516092347,
-      "grad_norm": 1.953125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.2327,
       "step": 74480
     },
     {
       "epoch": 1.2102159185066044,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.1809,
       "step": 74490
     },
     {
       "epoch": 1.210378385403974,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.1883,
       "step": 74500
     },
     {
       "epoch": 1.2105408523013437,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.2215,
       "step": 74510
     },
     {
       "epoch": 1.2107033191987133,
-      "grad_norm": 3.375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.22,
       "step": 74520
     },
     {
       "epoch": 1.210865786096083,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.367,
+      "loss": 1.2021,
       "step": 74530
     },
     {
       "epoch": 1.2110282529934526,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3166,
+      "loss": 1.1694,
       "step": 74540
     },
     {
       "epoch": 1.2111907198908223,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.2296,
       "step": 74550
     },
     {
       "epoch": 1.211353186788192,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3141,
+      "loss": 1.2432,
       "step": 74560
     },
     {
       "epoch": 1.2115156536855616,
-      "grad_norm": 2.25,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2157,
       "step": 74570
     },
     {
       "epoch": 1.2116781205829312,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.214,
       "step": 74580
     },
     {
       "epoch": 1.2118405874803009,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2255,
       "step": 74590
     },
     {
       "epoch": 1.2120030543776705,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.2026,
       "step": 74600
     },
     {
       "epoch": 1.2121655212750402,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3026,
+      "loss": 1.1989,
       "step": 74610
     },
     {
       "epoch": 1.2123279881724098,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.201,
       "step": 74620
     },
     {
       "epoch": 1.2124904550697795,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2109,
       "step": 74630
     },
     {
       "epoch": 1.2126529219671491,
-      "grad_norm": 2.5,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.1894,
       "step": 74640
     },
     {
       "epoch": 1.2128153888645188,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.1904,
       "step": 74650
     },
     {
       "epoch": 1.2129778557618884,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2021,
       "step": 74660
     },
     {
       "epoch": 1.213140322659258,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.2558,
       "step": 74670
     },
     {
       "epoch": 1.2133027895566277,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2241,
       "step": 74680
     },
     {
       "epoch": 1.2134652564539974,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1961,
       "step": 74690
     },
     {
       "epoch": 1.213627723351367,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2262,
       "step": 74700
     },
     {
       "epoch": 1.213790190248737,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2302,
       "step": 74710
     },
     {
       "epoch": 1.2139526571461066,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2085,
       "step": 74720
     },
     {
       "epoch": 1.2141151240434762,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1741,
       "step": 74730
     },
     {
       "epoch": 1.2142775909408459,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2484,
       "step": 74740
     },
     {
       "epoch": 1.2144400578382155,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.1838,
       "step": 74750
     },
     {
       "epoch": 1.2146025247355852,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2138,
       "step": 74760
     },
     {
       "epoch": 1.2147649916329548,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1775,
       "step": 74770
     },
     {
       "epoch": 1.2149274585303245,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1753,
       "step": 74780
     },
     {
       "epoch": 1.2150899254276941,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2146,
       "step": 74790
     },
     {
       "epoch": 1.2152523923250638,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.1827,
       "step": 74800
     },
     {
       "epoch": 1.2154148592224334,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2233,
       "step": 74810
     },
     {
       "epoch": 1.215577326119803,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.1894,
       "step": 74820
     },
     {
       "epoch": 1.2157397930171727,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2334,
       "step": 74830
     },
     {
       "epoch": 1.2159022599145424,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2303,
       "step": 74840
     },
     {
       "epoch": 1.216064726811912,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1959,
       "step": 74850
     },
     {
       "epoch": 1.2162271937092817,
-      "grad_norm": 1.90625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2182,
       "step": 74860
     },
     {
       "epoch": 1.2163896606066513,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2419,
       "step": 74870
     },
     {
       "epoch": 1.216552127504021,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.1903,
       "step": 74880
     },
     {
       "epoch": 1.2167145944013906,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.1894,
       "step": 74890
     },
     {
       "epoch": 1.2168770612987605,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3691,
+      "loss": 1.2342,
       "step": 74900
     },
     {
       "epoch": 1.2170395281961301,
-      "grad_norm": 1.84375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2437,
       "step": 74910
     },
     {
       "epoch": 1.2172019950934998,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.236,
       "step": 74920
     },
     {
       "epoch": 1.2173644619908695,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2055,
       "step": 74930
     },
     {
       "epoch": 1.217526928888239,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2183,
       "step": 74940
     },
     {
       "epoch": 1.2176893957856088,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2213,
       "step": 74950
     },
     {
       "epoch": 1.2178518626829784,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3264,
+      "loss": 1.1955,
       "step": 74960
     },
     {
       "epoch": 1.218014329580348,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1765,
       "step": 74970
     },
     {
       "epoch": 1.2181767964777177,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.1488,
       "step": 74980
     },
     {
       "epoch": 1.2183392633750874,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2027,
       "step": 74990
     },
     {
       "epoch": 1.218501730272457,
-      "grad_norm": 2.625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3617,
+      "loss": 1.1894,
       "step": 75000
     },
     {
       "epoch": 1.2186641971698267,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2015,
       "step": 75010
     },
     {
       "epoch": 1.2188266640671963,
-      "grad_norm": 2.875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3669,
+      "loss": 1.2179,
       "step": 75020
     },
     {
       "epoch": 1.218989130964566,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2092,
       "step": 75030
     },
     {
       "epoch": 1.2191515978619356,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.1956,
       "step": 75040
     },
     {
       "epoch": 1.2193140647593053,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.1964,
       "step": 75050
     },
     {
       "epoch": 1.219476531656675,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2324,
       "step": 75060
     },
     {
       "epoch": 1.2196389985540446,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2102,
       "step": 75070
     },
     {
       "epoch": 1.2198014654514142,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.219,
       "step": 75080
     },
     {
       "epoch": 1.2199639323487839,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.1831,
       "step": 75090
     },
     {
       "epoch": 1.2201263992461535,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.1673,
       "step": 75100
     },
     {
       "epoch": 1.2202888661435232,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2023,
       "step": 75110
     },
     {
       "epoch": 1.2204513330408928,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.1779,
       "step": 75120
     },
     {
       "epoch": 1.2206137999382625,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.213,
       "step": 75130
     },
     {
       "epoch": 1.2207762668356321,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.2011,
       "step": 75140
     },
     {
       "epoch": 1.220938733733002,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2267,
       "step": 75150
     },
     {
       "epoch": 1.2211012006303716,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2174,
       "step": 75160
     },
     {
       "epoch": 1.2212636675277413,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3624,
+      "loss": 1.2271,
       "step": 75170
     },
     {
       "epoch": 1.221426134425111,
-      "grad_norm": 2.5,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.1918,
       "step": 75180
     },
     {
       "epoch": 1.2215886013224806,
-      "grad_norm": 2.375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.1946,
       "step": 75190
     },
     {
       "epoch": 1.2217510682198502,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.1762,
       "step": 75200
     },
     {
       "epoch": 1.22191353511722,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3601,
+      "loss": 1.1571,
       "step": 75210
     },
     {
       "epoch": 1.2220760020145895,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.1983,
       "step": 75220
     },
     {
       "epoch": 1.2222384689119592,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2075,
       "step": 75230
     },
     {
       "epoch": 1.2224009358093288,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3722,
+      "loss": 1.2293,
       "step": 75240
     },
     {
       "epoch": 1.2225634027066985,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2187,
       "step": 75250
     },
     {
       "epoch": 1.2227258696040681,
-      "grad_norm": 2.25,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.313,
+      "loss": 1.1974,
       "step": 75260
     },
     {
       "epoch": 1.2228883365014378,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.2396,
       "step": 75270
     },
     {
       "epoch": 1.2230508033988075,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.1975,
       "step": 75280
     },
     {
       "epoch": 1.223213270296177,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2063,
       "step": 75290
     },
     {
       "epoch": 1.2233757371935468,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.2175,
       "step": 75300
     },
     {
       "epoch": 1.2235382040909164,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2124,
       "step": 75310
     },
     {
       "epoch": 1.223700670988286,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.1835,
       "step": 75320
     },
     {
       "epoch": 1.2238631378856557,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.23,
       "step": 75330
     },
     {
       "epoch": 1.2240256047830256,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.1849,
       "step": 75340
     },
     {
       "epoch": 1.2241880716803952,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3112,
+      "loss": 1.2223,
       "step": 75350
     },
     {
       "epoch": 1.2243505385777649,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.1889,
       "step": 75360
     },
     {
       "epoch": 1.2245130054751345,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2478,
       "step": 75370
     },
     {
       "epoch": 1.2246754723725042,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2158,
       "step": 75380
     },
     {
       "epoch": 1.2248379392698738,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2501,
       "step": 75390
     },
     {
       "epoch": 1.2250004061672435,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2152,
       "step": 75400
     },
     {
       "epoch": 1.2251628730646131,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.1936,
       "step": 75410
     },
     {
       "epoch": 1.2253253399619828,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.1898,
       "step": 75420
     },
     {
       "epoch": 1.2254878068593524,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2263,
       "step": 75430
     },
     {
       "epoch": 1.225650273756722,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.1766,
       "step": 75440
     },
     {
       "epoch": 1.2258127406540917,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2313,
       "step": 75450
     },
     {
       "epoch": 1.2259752075514614,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2084,
       "step": 75460
     },
     {
       "epoch": 1.226137674448831,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.2019,
       "step": 75470
     },
     {
       "epoch": 1.2263001413462007,
-      "grad_norm": 2.625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.1691,
       "step": 75480
     },
     {
       "epoch": 1.2264626082435703,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.215,
       "step": 75490
     },
     {
       "epoch": 1.22662507514094,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2313,
       "step": 75500
     },
     {
       "epoch": 1.2267875420383096,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.2728,
       "step": 75510
     },
     {
       "epoch": 1.2269500089356793,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.1774,
       "step": 75520
     },
     {
       "epoch": 1.227112475833049,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2025,
       "step": 75530
     },
     {
       "epoch": 1.2272749427304186,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1927,
       "step": 75540
     },
     {
       "epoch": 1.2274374096277882,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2381,
       "step": 75550
     },
     {
       "epoch": 1.227599876525158,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.1946,
       "step": 75560
     },
     {
       "epoch": 1.2277623434225275,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.205,
       "step": 75570
     },
     {
       "epoch": 1.2279248103198972,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2255,
       "step": 75580
     },
     {
       "epoch": 1.228087277217267,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1877,
       "step": 75590
     },
     {
       "epoch": 1.2282497441146367,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2011,
       "step": 75600
     },
     {
       "epoch": 1.2284122110120064,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.2168,
       "step": 75610
     },
     {
       "epoch": 1.228574677909376,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2687,
       "step": 75620
     },
     {
       "epoch": 1.2287371448067457,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.185,
       "step": 75630
     },
     {
       "epoch": 1.2288996117041153,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3636,
+      "loss": 1.2012,
       "step": 75640
     },
     {
       "epoch": 1.229062078601485,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2043,
       "step": 75650
     },
     {
       "epoch": 1.2292245454988546,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.1561,
       "step": 75660
     },
     {
       "epoch": 1.2293870123962243,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.2431,
       "step": 75670
     },
     {
       "epoch": 1.229549479293594,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3221,
+      "loss": 1.1706,
       "step": 75680
     },
     {
       "epoch": 1.2297119461909636,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.1949,
       "step": 75690
     },
     {
       "epoch": 1.2298744130883332,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.2292,
       "step": 75700
     },
     {
       "epoch": 1.2300368799857029,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.1989,
       "step": 75710
     },
     {
       "epoch": 1.2301993468830725,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2063,
       "step": 75720
     },
     {
       "epoch": 1.2303618137804422,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2132,
       "step": 75730
     },
     {
       "epoch": 1.2305242806778118,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3643,
+      "loss": 1.1839,
       "step": 75740
     },
     {
       "epoch": 1.2306867475751815,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2118,
       "step": 75750
     },
     {
       "epoch": 1.2308492144725511,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.26,
       "step": 75760
     },
     {
       "epoch": 1.2310116813699208,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2083,
       "step": 75770
     },
     {
       "epoch": 1.2311741482672907,
-      "grad_norm": 1.953125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.215,
       "step": 75780
     },
     {
       "epoch": 1.2313366151646603,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.2108,
       "step": 75790
     },
     {
       "epoch": 1.23149908206203,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2597,
       "step": 75800
     },
     {
       "epoch": 1.2316615489593996,
-      "grad_norm": 1.796875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.2592,
       "step": 75810
     },
     {
       "epoch": 1.2318240158567693,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2054,
       "step": 75820
     },
     {
       "epoch": 1.231986482754139,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2031,
       "step": 75830
     },
     {
       "epoch": 1.2321489496515086,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.1785,
       "step": 75840
     },
     {
       "epoch": 1.2323114165488782,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2234,
       "step": 75850
     },
     {
       "epoch": 1.2324738834462479,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2134,
       "step": 75860
     },
     {
       "epoch": 1.2326363503436175,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.1995,
       "step": 75870
     },
     {
       "epoch": 1.2327988172409872,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2256,
       "step": 75880
     },
     {
       "epoch": 1.2329612841383568,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.1997,
       "step": 75890
     },
     {
       "epoch": 1.2331237510357265,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.1907,
       "step": 75900
     },
     {
       "epoch": 1.2332862179330961,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2093,
       "step": 75910
     },
     {
       "epoch": 1.2334486848304658,
-      "grad_norm": 2.5,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.2034,
       "step": 75920
     },
     {
       "epoch": 1.2336111517278354,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2232,
       "step": 75930
     },
     {
       "epoch": 1.233773618625205,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3156,
+      "loss": 1.1836,
       "step": 75940
     },
     {
       "epoch": 1.2339360855225747,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2588,
       "step": 75950
     },
     {
       "epoch": 1.2340985524199444,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.1927,
       "step": 75960
     },
     {
       "epoch": 1.234261019317314,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3527,
+      "loss": 1.199,
       "step": 75970
     },
     {
       "epoch": 1.2344234862146837,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.1965,
       "step": 75980
     },
     {
       "epoch": 1.2345859531120533,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2348,
       "step": 75990
     },
     {
       "epoch": 1.234748420009423,
-      "grad_norm": 3.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2148,
       "step": 76000
     },
     {
       "epoch": 1.2349108869067926,
-      "grad_norm": 2.375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2147,
       "step": 76010
     },
     {
       "epoch": 1.2350733538041623,
-      "grad_norm": 2.25,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2219,
       "step": 76020
     },
     {
       "epoch": 1.2352358207015322,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.1636,
       "step": 76030
     },
     {
       "epoch": 1.2353982875989018,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3679,
+      "loss": 1.2548,
       "step": 76040
     },
     {
       "epoch": 1.2355607544962715,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.258,
       "step": 76050
     },
     {
       "epoch": 1.235723221393641,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.19,
       "step": 76060
     },
     {
       "epoch": 1.2358856882910108,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2098,
       "step": 76070
     },
     {
       "epoch": 1.2360481551883804,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.2809,
       "step": 76080
     },
     {
       "epoch": 1.23621062208575,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1956,
       "step": 76090
     },
     {
       "epoch": 1.2363730889831197,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3122,
+      "loss": 1.2163,
       "step": 76100
     },
     {
       "epoch": 1.2365355558804894,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.1956,
       "step": 76110
     },
     {
       "epoch": 1.236698022777859,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.1715,
       "step": 76120
     },
     {
       "epoch": 1.2368604896752287,
-      "grad_norm": 2.140625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2105,
       "step": 76130
     },
     {
       "epoch": 1.2370229565725983,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2068,
       "step": 76140
     },
     {
       "epoch": 1.237185423469968,
-      "grad_norm": 1.6796875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2039,
       "step": 76150
     },
     {
       "epoch": 1.2373478903673376,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.1939,
       "step": 76160
     },
     {
       "epoch": 1.2375103572647073,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2054,
       "step": 76170
     },
     {
       "epoch": 1.237672824162077,
-      "grad_norm": 2.75,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2468,
       "step": 76180
     },
     {
       "epoch": 1.2378352910594466,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2193,
       "step": 76190
     },
     {
       "epoch": 1.2379977579568162,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.1895,
       "step": 76200
     },
     {
       "epoch": 1.2381602248541859,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2112,
       "step": 76210
     },
     {
       "epoch": 1.2383226917515557,
-      "grad_norm": 1.8046875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.2705,
       "step": 76220
     },
     {
       "epoch": 1.2384851586489254,
-      "grad_norm": 3.0,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.1735,
       "step": 76230
     },
     {
       "epoch": 1.238647625546295,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2331,
       "step": 76240
     },
     {
       "epoch": 1.2388100924436647,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3221,
+      "loss": 1.2085,
       "step": 76250
     },
     {
       "epoch": 1.2389725593410343,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2079,
       "step": 76260
     },
     {
       "epoch": 1.239135026238404,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3112,
+      "loss": 1.1849,
       "step": 76270
     },
     {
       "epoch": 1.2392974931357736,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3194,
+      "loss": 1.2086,
       "step": 76280
     },
     {
       "epoch": 1.2394599600331433,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2272,
       "step": 76290
     },
     {
       "epoch": 1.239622426930513,
-      "grad_norm": 3.125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2102,
       "step": 76300
     },
     {
       "epoch": 1.2397848938278826,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.1797,
       "step": 76310
     },
     {
       "epoch": 1.2399473607252522,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.2152,
       "step": 76320
     },
     {
       "epoch": 1.240109827622622,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.189,
       "step": 76330
     },
     {
       "epoch": 1.2402722945199915,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.2517,
       "step": 76340
     },
     {
       "epoch": 1.2404347614173612,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.1893,
       "step": 76350
     },
     {
       "epoch": 1.2405972283147308,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.1851,
       "step": 76360
     },
     {
       "epoch": 1.2407596952121005,
-      "grad_norm": 2.34375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2174,
       "step": 76370
     },
     {
       "epoch": 1.2409221621094702,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2139,
       "step": 76380
     },
     {
       "epoch": 1.2410846290068398,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.1562,
       "step": 76390
     },
     {
       "epoch": 1.2412470959042095,
-      "grad_norm": 2.3125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3133,
+      "loss": 1.152,
       "step": 76400
     },
     {
       "epoch": 1.241409562801579,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3125,
+      "loss": 1.2235,
       "step": 76410
     },
     {
       "epoch": 1.2415720296989488,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2076,
       "step": 76420
     },
     {
       "epoch": 1.2417344965963184,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2022,
       "step": 76430
     },
     {
       "epoch": 1.241896963493688,
-      "grad_norm": 3.125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2112,
       "step": 76440
     },
     {
       "epoch": 1.2420594303910577,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2142,
       "step": 76450
     },
     {
       "epoch": 1.2422218972884276,
-      "grad_norm": 2.375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2507,
       "step": 76460
     },
     {
       "epoch": 1.2423843641857972,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.1834,
       "step": 76470
     },
     {
       "epoch": 1.2425468310831669,
-      "grad_norm": 3.0,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.1546,
       "step": 76480
     },
     {
       "epoch": 1.2427092979805365,
-      "grad_norm": 1.8828125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2051,
       "step": 76490
     },
     {
       "epoch": 1.2428717648779062,
-      "grad_norm": 2.375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2447,
       "step": 76500
     },
     {
       "epoch": 1.2430342317752758,
-      "grad_norm": 1.8671875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.1737,
       "step": 76510
     },
     {
       "epoch": 1.2431966986726455,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3195,
+      "loss": 1.2258,
       "step": 76520
     },
     {
       "epoch": 1.2433591655700151,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1825,
       "step": 76530
     },
     {
       "epoch": 1.2435216324673848,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2279,
       "step": 76540
     },
     {
       "epoch": 1.2436840993647544,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.215,
       "step": 76550
     },
     {
       "epoch": 1.243846566262124,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.1612,
       "step": 76560
     },
     {
       "epoch": 1.2440090331594937,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.214,
       "step": 76570
     },
     {
       "epoch": 1.2441715000568634,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2146,
       "step": 76580
     },
     {
       "epoch": 1.244333966954233,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2038,
       "step": 76590
     },
     {
       "epoch": 1.2444964338516027,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2133,
       "step": 76600
     },
     {
       "epoch": 1.2446589007489723,
-      "grad_norm": 1.8359375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1873,
       "step": 76610
     },
     {
       "epoch": 1.244821367646342,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.1927,
       "step": 76620
     },
     {
       "epoch": 1.2449838345437116,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2436,
       "step": 76630
     },
     {
       "epoch": 1.2451463014410813,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.243,
       "step": 76640
     },
     {
       "epoch": 1.245308768338451,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2165,
       "step": 76650
     },
     {
       "epoch": 1.2454712352358208,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2359,
       "step": 76660
     },
     {
       "epoch": 1.2456337021331905,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.2486,
       "step": 76670
     },
     {
       "epoch": 1.2457961690305601,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2076,
       "step": 76680
     },
     {
       "epoch": 1.2459586359279298,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2336,
       "step": 76690
     },
     {
       "epoch": 1.2461211028252994,
-      "grad_norm": 1.9921875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2116,
       "step": 76700
     },
     {
       "epoch": 1.246283569722669,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2163,
       "step": 76710
     },
     {
       "epoch": 1.2464460366200387,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2108,
       "step": 76720
     },
     {
       "epoch": 1.2466085035174084,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.244,
       "step": 76730
     },
     {
       "epoch": 1.246770970414778,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.1956,
       "step": 76740
     },
     {
       "epoch": 1.2469334373121477,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2361,
       "step": 76750
     },
     {
       "epoch": 1.2470959042095173,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 1.2175,
       "step": 76760
     },
     {
       "epoch": 1.247258371106887,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3167,
+      "loss": 1.2075,
       "step": 76770
     },
     {
       "epoch": 1.2474208380042566,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2578,
       "step": 76780
     },
     {
       "epoch": 1.2475833049016263,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.229,
       "step": 76790
     },
     {
       "epoch": 1.247745771798996,
-      "grad_norm": 3.3125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3579,
+      "loss": 1.2403,
       "step": 76800
     },
     {
       "epoch": 1.2479082386963656,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2293,
       "step": 76810
     },
     {
       "epoch": 1.2480707055937352,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2037,
       "step": 76820
     },
     {
       "epoch": 1.2482331724911049,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.1941,
       "step": 76830
     },
     {
       "epoch": 1.2483956393884745,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.2326,
       "step": 76840
     },
     {
       "epoch": 1.2485581062858442,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.1797,
       "step": 76850
     },
     {
       "epoch": 1.2487205731832138,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2224,
       "step": 76860
     },
     {
       "epoch": 1.2488830400805835,
-      "grad_norm": 1.9453125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.1977,
       "step": 76870
     },
     {
       "epoch": 1.2490455069779531,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.1636,
       "step": 76880
     },
     {
       "epoch": 1.2492079738753228,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2054,
       "step": 76890
     },
     {
       "epoch": 1.2493704407726927,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3258,
+      "loss": 1.1779,
       "step": 76900
     },
     {
       "epoch": 1.2495329076700623,
-      "grad_norm": 2.8125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.1945,
       "step": 76910
     },
     {
       "epoch": 1.249695374567432,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.1638,
       "step": 76920
     },
     {
       "epoch": 1.2498578414648016,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.187,
       "step": 76930
     },
     {
       "epoch": 1.2500203083621713,
-      "grad_norm": 2.625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2435,
       "step": 76940
     },
     {
       "epoch": 1.250182775259541,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2525,
       "step": 76950
     },
     {
       "epoch": 1.2503452421569106,
-      "grad_norm": 3.125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2271,
       "step": 76960
     },
     {
       "epoch": 1.2505077090542802,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2254,
       "step": 76970
     },
     {
       "epoch": 1.2506701759516499,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2129,
       "step": 76980
     },
     {
       "epoch": 1.2508326428490195,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.239,
       "step": 76990
     },
     {
       "epoch": 1.2509951097463892,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2225,
       "step": 77000
     },
     {
       "epoch": 1.2511575766437588,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3115,
+      "loss": 1.21,
       "step": 77010
     },
     {
       "epoch": 1.2513200435411285,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2441,
       "step": 77020
     },
     {
       "epoch": 1.2514825104384981,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2256,
       "step": 77030
     },
     {
       "epoch": 1.2516449773358678,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.2257,
       "step": 77040
     },
     {
       "epoch": 1.2518074442332374,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3184,
+      "loss": 1.2116,
       "step": 77050
     },
     {
       "epoch": 1.251969911130607,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.2246,
       "step": 77060
     },
     {
       "epoch": 1.2521323780279767,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3724,
+      "loss": 1.166,
       "step": 77070
     },
     {
       "epoch": 1.2522948449253466,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.1725,
       "step": 77080
     },
     {
       "epoch": 1.2524573118227162,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2099,
       "step": 77090
     },
     {
       "epoch": 1.252619778720086,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2232,
       "step": 77100
     },
     {
       "epoch": 1.2527822456174555,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2274,
       "step": 77110
     },
     {
       "epoch": 1.2529447125148252,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2132,
       "step": 77120
     },
     {
       "epoch": 1.2531071794121948,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.2122,
       "step": 77130
     },
     {
       "epoch": 1.2532696463095645,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2076,
       "step": 77140
     },
     {
       "epoch": 1.2534321132069342,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2443,
       "step": 77150
     },
     {
       "epoch": 1.2535945801043038,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2233,
       "step": 77160
     },
     {
       "epoch": 1.2537570470016735,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2396,
       "step": 77170
     },
     {
       "epoch": 1.253919513899043,
-      "grad_norm": 2.25,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.2253,
       "step": 77180
     },
     {
       "epoch": 1.2540819807964128,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.2047,
       "step": 77190
     },
     {
       "epoch": 1.2542444476937824,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.1945,
       "step": 77200
     },
     {
       "epoch": 1.254406914591152,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2378,
       "step": 77210
     },
     {
       "epoch": 1.2545693814885217,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2197,
       "step": 77220
     },
     {
       "epoch": 1.2547318483858914,
-      "grad_norm": 3.265625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2386,
       "step": 77230
     },
     {
       "epoch": 1.254894315283261,
-      "grad_norm": 2.25,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2187,
       "step": 77240
     },
     {
       "epoch": 1.2550567821806307,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.221,
       "step": 77250
     },
     {
       "epoch": 1.2552192490780003,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.2294,
       "step": 77260
     },
     {
       "epoch": 1.25538171597537,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2431,
       "step": 77270
     },
     {
       "epoch": 1.2555441828727396,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2466,
       "step": 77280
     },
     {
       "epoch": 1.2557066497701093,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2582,
       "step": 77290
     },
     {
       "epoch": 1.255869116667479,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2315,
       "step": 77300
     },
     {
       "epoch": 1.2560315835648486,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.1926,
       "step": 77310
     },
     {
       "epoch": 1.2561940504622182,
-      "grad_norm": 2.125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.197,
       "step": 77320
     },
     {
       "epoch": 1.2563565173595879,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.237,
       "step": 77330
     },
     {
       "epoch": 1.2565189842569575,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.2164,
       "step": 77340
     },
     {
       "epoch": 1.2566814511543272,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.1806,
       "step": 77350
     },
     {
       "epoch": 1.256843918051697,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2225,
       "step": 77360
     },
     {
       "epoch": 1.2570063849490667,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2031,
       "step": 77370
     },
     {
       "epoch": 1.2571688518464363,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3202,
+      "loss": 1.2061,
       "step": 77380
     },
     {
       "epoch": 1.257331318743806,
-      "grad_norm": 1.765625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.206,
       "step": 77390
     },
     {
       "epoch": 1.2574937856411756,
-      "grad_norm": 1.859375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2401,
       "step": 77400
     },
     {
       "epoch": 1.2576562525385453,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2197,
       "step": 77410
     },
     {
       "epoch": 1.257818719435915,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2105,
       "step": 77420
     },
     {
       "epoch": 1.2579811863332846,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2006,
       "step": 77430
     },
     {
       "epoch": 1.2581436532306542,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3701,
+      "loss": 1.2349,
       "step": 77440
     },
     {
       "epoch": 1.258306120128024,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.1777,
       "step": 77450
     },
     {
       "epoch": 1.2584685870253935,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.1754,
       "step": 77460
     },
     {
       "epoch": 1.2586310539227632,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3611,
+      "loss": 1.2126,
       "step": 77470
     },
     {
       "epoch": 1.2587935208201328,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2251,
       "step": 77480
     },
     {
       "epoch": 1.2589559877175025,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2227,
       "step": 77490
     },
     {
       "epoch": 1.2591184546148722,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.2513,
       "step": 77500
     },
     {
       "epoch": 1.2592809215122418,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.25,
       "step": 77510
     },
     {
       "epoch": 1.2594433884096117,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3659,
+      "loss": 1.1966,
       "step": 77520
     },
     {
       "epoch": 1.2596058553069813,
-      "grad_norm": 2.25,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.242,
       "step": 77530
     },
     {
       "epoch": 1.259768322204351,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2029,
       "step": 77540
     },
     {
       "epoch": 1.2599307891017206,
-      "grad_norm": 2.296875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2082,
       "step": 77550
     },
     {
       "epoch": 1.2600932559990903,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2337,
       "step": 77560
     },
     {
       "epoch": 1.26025572289646,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.1911,
       "step": 77570
     },
     {
       "epoch": 1.2604181897938296,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2225,
       "step": 77580
     },
     {
       "epoch": 1.2605806566911992,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3156,
+      "loss": 1.2133,
       "step": 77590
     },
     {
       "epoch": 1.2607431235885689,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2001,
       "step": 77600
     },
     {
       "epoch": 1.2609055904859385,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2274,
       "step": 77610
     },
     {
       "epoch": 1.2610680573833082,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2294,
       "step": 77620
     },
     {
       "epoch": 1.2612305242806778,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2096,
       "step": 77630
     },
     {
       "epoch": 1.2613929911780475,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3111,
+      "loss": 1.1704,
       "step": 77640
     },
     {
       "epoch": 1.2615554580754171,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3142,
+      "loss": 1.1953,
       "step": 77650
     },
     {
       "epoch": 1.2617179249727868,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2452,
       "step": 77660
     },
     {
       "epoch": 1.2618803918701564,
-      "grad_norm": 2.5,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2135,
       "step": 77670
     },
     {
       "epoch": 1.262042858767526,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.1975,
       "step": 77680
     },
     {
       "epoch": 1.2622053256648957,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2447,
       "step": 77690
     },
     {
       "epoch": 1.2623677925622654,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.2175,
       "step": 77700
     },
     {
       "epoch": 1.262530259459635,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.1838,
       "step": 77710
     },
     {
       "epoch": 1.2626927263570047,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.2202,
       "step": 77720
     },
     {
       "epoch": 1.2628551932543743,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3706,
+      "loss": 1.215,
       "step": 77730
     },
     {
       "epoch": 1.263017660151744,
-      "grad_norm": 1.625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.1931,
       "step": 77740
     },
     {
       "epoch": 1.2631801270491136,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2199,
       "step": 77750
     },
     {
       "epoch": 1.2633425939464833,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.1793,
       "step": 77760
     },
     {
       "epoch": 1.263505060843853,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.1758,
       "step": 77770
     },
     {
       "epoch": 1.2636675277412226,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.1734,
       "step": 77780
     },
     {
       "epoch": 1.2638299946385925,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2695,
       "step": 77790
     },
     {
       "epoch": 1.2639924615359621,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3138,
+      "loss": 1.2245,
       "step": 77800
     },
     {
       "epoch": 1.2641549284333318,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2088,
       "step": 77810
     },
     {
       "epoch": 1.2643173953307014,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2245,
       "step": 77820
     },
     {
       "epoch": 1.264479862228071,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1987,
       "step": 77830
     },
     {
       "epoch": 1.2646423291254407,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3669,
+      "loss": 1.2554,
       "step": 77840
     },
     {
       "epoch": 1.2648047960228104,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.1981,
       "step": 77850
     },
     {
       "epoch": 1.26496726292018,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.1861,
       "step": 77860
     },
     {
       "epoch": 1.2651297298175497,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2113,
       "step": 77870
     },
     {
       "epoch": 1.2652921967149193,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2142,
       "step": 77880
     },
     {
       "epoch": 1.265454663612289,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.161,
       "step": 77890
     },
     {
       "epoch": 1.2656171305096586,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3558,
+      "loss": 1.2041,
       "step": 77900
     },
     {
       "epoch": 1.2657795974070283,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2039,
       "step": 77910
     },
     {
       "epoch": 1.265942064304398,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.2234,
       "step": 77920
     },
     {
       "epoch": 1.2661045312017676,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3145,
+      "loss": 1.1999,
       "step": 77930
     },
     {
       "epoch": 1.2662669980991372,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.203,
       "step": 77940
     },
     {
       "epoch": 1.2664294649965069,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2066,
       "step": 77950
     },
     {
       "epoch": 1.2665919318938768,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.231,
       "step": 77960
     },
     {
       "epoch": 1.2667543987912464,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2225,
       "step": 77970
     },
     {
       "epoch": 1.266916865688616,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1723,
       "step": 77980
     },
     {
       "epoch": 1.2670793325859857,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.1928,
       "step": 77990
     },
     {
       "epoch": 1.2672417994833554,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.1762,
       "step": 78000
     },
     {
       "epoch": 1.267404266380725,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.243,
       "step": 78010
     },
     {
       "epoch": 1.2675667332780947,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2438,
       "step": 78020
     },
     {
       "epoch": 1.2677292001754643,
-      "grad_norm": 2.375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3603,
+      "loss": 1.2255,
       "step": 78030
     },
     {
       "epoch": 1.267891667072834,
-      "grad_norm": 2.875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.232,
       "step": 78040
     },
     {
       "epoch": 1.2680541339702036,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3217,
+      "loss": 1.2003,
       "step": 78050
     },
     {
       "epoch": 1.2682166008675733,
-      "grad_norm": 1.7109375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.2171,
       "step": 78060
     },
     {
       "epoch": 1.268379067764943,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2408,
       "step": 78070
     },
     {
       "epoch": 1.2685415346623126,
-      "grad_norm": 2.046875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1697,
       "step": 78080
     },
     {
       "epoch": 1.2687040015596822,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.2103,
       "step": 78090
     },
     {
       "epoch": 1.2688664684570519,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2029,
       "step": 78100
     },
     {
       "epoch": 1.2690289353544215,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.324,
+      "loss": 1.2256,
       "step": 78110
     },
     {
       "epoch": 1.2691914022517912,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.1869,
       "step": 78120
     },
     {
       "epoch": 1.2693538691491608,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.1944,
       "step": 78130
     },
     {
       "epoch": 1.2695163360465305,
-      "grad_norm": 2.796875,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.208,
       "step": 78140
     },
     {
       "epoch": 1.2696788029439001,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2583,
       "step": 78150
     },
     {
       "epoch": 1.2698412698412698,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.1896,
       "step": 78160
     },
     {
       "epoch": 1.2700037367386394,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.1679,
       "step": 78170
     },
     {
       "epoch": 1.270166203636009,
-      "grad_norm": 3.0625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.228,
       "step": 78180
     },
     {
       "epoch": 1.2703286705333787,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3271,
+      "loss": 1.1698,
       "step": 78190
     },
     {
       "epoch": 1.2704911374307484,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3081,
+      "loss": 1.2298,
       "step": 78200
     },
     {
       "epoch": 1.270653604328118,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.2203,
       "step": 78210
     },
     {
       "epoch": 1.2708160712254877,
-      "grad_norm": 3.125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2174,
       "step": 78220
     },
     {
       "epoch": 1.2709785381228575,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.1941,
       "step": 78230
     },
     {
       "epoch": 1.2711410050202272,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2661,
       "step": 78240
     },
     {
       "epoch": 1.2713034719175969,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.2406,
       "step": 78250
     },
     {
       "epoch": 1.2714659388149665,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.1936,
       "step": 78260
     },
     {
       "epoch": 1.2716284057123362,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2038,
       "step": 78270
     },
     {
       "epoch": 1.2717908726097058,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2264,
       "step": 78280
     },
     {
       "epoch": 1.2719533395070755,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2145,
       "step": 78290
     },
     {
       "epoch": 1.272115806404445,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2113,
       "step": 78300
     },
     {
       "epoch": 1.2722782733018148,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2421,
       "step": 78310
     },
     {
       "epoch": 1.2724407401991844,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2004,
       "step": 78320
     },
     {
       "epoch": 1.272603207096554,
-      "grad_norm": 2.375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.199,
       "step": 78330
     },
     {
       "epoch": 1.2727656739939237,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2284,
       "step": 78340
     },
     {
       "epoch": 1.2729281408912934,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.1881,
       "step": 78350
     },
     {
       "epoch": 1.273090607788663,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2142,
       "step": 78360
     },
     {
       "epoch": 1.2732530746860327,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2255,
       "step": 78370
     },
     {
       "epoch": 1.2734155415834023,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2563,
       "step": 78380
     },
     {
       "epoch": 1.273578008480772,
-      "grad_norm": 2.359375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3106,
+      "loss": 1.2352,
       "step": 78390
     },
     {
       "epoch": 1.2737404753781418,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2241,
       "step": 78400
     },
     {
       "epoch": 1.2739029422755115,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2288,
       "step": 78410
     },
     {
       "epoch": 1.2740654091728811,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.2224,
       "step": 78420
     },
     {
       "epoch": 1.2742278760702508,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.1713,
       "step": 78430
     },
     {
       "epoch": 1.2743903429676204,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.2275,
       "step": 78440
     },
     {
       "epoch": 1.27455280986499,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.2251,
       "step": 78450
     },
     {
       "epoch": 1.2747152767623597,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2308,
       "step": 78460
     },
     {
       "epoch": 1.2748777436597294,
-      "grad_norm": 3.3125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.2148,
       "step": 78470
     },
     {
       "epoch": 1.275040210557099,
-      "grad_norm": 2.859375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.1736,
       "step": 78480
     },
     {
       "epoch": 1.2752026774544687,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2205,
       "step": 78490
     },
     {
       "epoch": 1.2753651443518383,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2482,
       "step": 78500
     },
     {
       "epoch": 1.275527611249208,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.2178,
       "step": 78510
     },
     {
       "epoch": 1.2756900781465776,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2234,
       "step": 78520
     },
     {
       "epoch": 1.2758525450439473,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.1867,
       "step": 78530
     },
     {
       "epoch": 1.276015011941317,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.2371,
       "step": 78540
     },
     {
       "epoch": 1.2761774788386866,
-      "grad_norm": 3.0,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.193,
       "step": 78550
     },
     {
       "epoch": 1.2763399457360562,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.1887,
       "step": 78560
     },
     {
       "epoch": 1.276502412633426,
-      "grad_norm": 4.9375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.2064,
       "step": 78570
     },
     {
       "epoch": 1.2766648795307955,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2189,
       "step": 78580
     },
     {
       "epoch": 1.2768273464281652,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.1582,
       "step": 78590
     },
     {
       "epoch": 1.2769898133255349,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2281,
       "step": 78600
     },
     {
       "epoch": 1.2771522802229045,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2014,
       "step": 78610
     },
     {
       "epoch": 1.2773147471202742,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.1996,
       "step": 78620
     },
     {
       "epoch": 1.2774772140176438,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.1995,
       "step": 78630
     },
     {
       "epoch": 1.2776396809150135,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2334,
       "step": 78640
     },
     {
       "epoch": 1.277802147812383,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2516,
       "step": 78650
     },
     {
       "epoch": 1.2779646147097528,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.2025,
       "step": 78660
     },
     {
       "epoch": 1.2781270816071226,
-      "grad_norm": 1.5625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.214,
       "step": 78670
     },
     {
       "epoch": 1.2782895485044923,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2262,
       "step": 78680
     },
     {
       "epoch": 1.278452015401862,
-      "grad_norm": 2.640625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2375,
       "step": 78690
     },
     {
       "epoch": 1.2786144822992316,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2261,
       "step": 78700
     },
     {
       "epoch": 1.2787769491966012,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.1815,
       "step": 78710
     },
     {
       "epoch": 1.2789394160939709,
-      "grad_norm": 3.0,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3592,
+      "loss": 1.2197,
       "step": 78720
     },
     {
       "epoch": 1.2791018829913405,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2208,
       "step": 78730
     },
     {
       "epoch": 1.2792643498887102,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.2008,
       "step": 78740
     },
     {
       "epoch": 1.2794268167860798,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2532,
       "step": 78750
     },
     {
       "epoch": 1.2795892836834495,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2042,
       "step": 78760
     },
     {
       "epoch": 1.2797517505808191,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2381,
       "step": 78770
     },
     {
       "epoch": 1.2799142174781888,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.2134,
       "step": 78780
     },
     {
       "epoch": 1.2800766843755584,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2406,
       "step": 78790
     },
     {
       "epoch": 1.280239151272928,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.1902,
       "step": 78800
     },
     {
       "epoch": 1.2804016181702977,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.1934,
       "step": 78810
     },
     {
       "epoch": 1.2805640850676674,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.1811,
       "step": 78820
     },
     {
       "epoch": 1.280726551965037,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.222,
       "step": 78830
     },
     {
       "epoch": 1.280889018862407,
-      "grad_norm": 3.125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1922,
       "step": 78840
     },
     {
       "epoch": 1.2810514857597766,
-      "grad_norm": 2.09375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.199,
       "step": 78850
     },
     {
       "epoch": 1.2812139526571462,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.1981,
       "step": 78860
     },
     {
       "epoch": 1.2813764195545159,
-      "grad_norm": 1.7265625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2028,
       "step": 78870
     },
     {
       "epoch": 1.2815388864518855,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2049,
       "step": 78880
     },
     {
       "epoch": 1.2817013533492552,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2219,
       "step": 78890
     },
     {
       "epoch": 1.2818638202466248,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2034,
       "step": 78900
     },
     {
       "epoch": 1.2820262871439945,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.185,
       "step": 78910
     },
     {
       "epoch": 1.2821887540413641,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1966,
       "step": 78920
     },
     {
       "epoch": 1.2823512209387338,
-      "grad_norm": 2.25,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.1745,
       "step": 78930
     },
     {
       "epoch": 1.2825136878361034,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2654,
       "step": 78940
     },
     {
       "epoch": 1.282676154733473,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.1883,
       "step": 78950
     },
     {
       "epoch": 1.2828386216308427,
-      "grad_norm": 2.125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.246,
       "step": 78960
     },
     {
       "epoch": 1.2830010885282124,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2235,
       "step": 78970
     },
     {
       "epoch": 1.283163555425582,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.214,
       "step": 78980
     },
     {
       "epoch": 1.2833260223229517,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2087,
       "step": 78990
     },
     {
       "epoch": 1.2834884892203213,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2531,
       "step": 79000
     },
     {
       "epoch": 1.283650956117691,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3247,
+      "loss": 1.2004,
       "step": 79010
     },
     {
       "epoch": 1.2838134230150606,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3644,
+      "loss": 1.2343,
       "step": 79020
     },
     {
       "epoch": 1.2839758899124303,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.2163,
       "step": 79030
     },
     {
       "epoch": 1.2841383568098,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.194,
       "step": 79040
     },
     {
       "epoch": 1.2843008237071696,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.2341,
       "step": 79050
     },
     {
       "epoch": 1.2844632906045392,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2411,
       "step": 79060
     },
     {
       "epoch": 1.2846257575019089,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.2258,
       "step": 79070
     },
     {
       "epoch": 1.2847882243992785,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2244,
       "step": 79080
     },
     {
       "epoch": 1.2849506912966482,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.17,
       "step": 79090
     },
     {
       "epoch": 1.2851131581940178,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2339,
       "step": 79100
     },
     {
       "epoch": 1.2852756250913877,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.1724,
       "step": 79110
     },
     {
       "epoch": 1.2854380919887574,
-      "grad_norm": 5.1875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2432,
       "step": 79120
     },
     {
       "epoch": 1.285600558886127,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2224,
       "step": 79130
     },
     {
       "epoch": 1.2857630257834967,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.2317,
       "step": 79140
     },
     {
       "epoch": 1.2859254926808663,
-      "grad_norm": 2.15625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1974,
       "step": 79150
     },
     {
       "epoch": 1.286087959578236,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2192,
       "step": 79160
     },
     {
       "epoch": 1.2862504264756056,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2231,
       "step": 79170
     },
     {
       "epoch": 1.2864128933729753,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2182,
       "step": 79180
     },
     {
       "epoch": 1.286575360270345,
-      "grad_norm": 3.125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.1862,
       "step": 79190
     },
     {
       "epoch": 1.2867378271677146,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2239,
       "step": 79200
     },
     {
       "epoch": 1.2869002940650842,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.1974,
       "step": 79210
     },
     {
       "epoch": 1.2870627609624539,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2267,
       "step": 79220
     },
     {
       "epoch": 1.2872252278598235,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2235,
       "step": 79230
     },
     {
       "epoch": 1.2873876947571932,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2755,
       "step": 79240
     },
     {
       "epoch": 1.2875501616545628,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2037,
       "step": 79250
     },
     {
       "epoch": 1.2877126285519325,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.229,
       "step": 79260
     },
     {
       "epoch": 1.2878750954493023,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.1802,
       "step": 79270
     },
     {
       "epoch": 1.288037562346672,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3637,
+      "loss": 1.1849,
       "step": 79280
     },
     {
       "epoch": 1.2882000292440416,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2085,
       "step": 79290
     },
     {
       "epoch": 1.2883624961414113,
-      "grad_norm": 3.125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2223,
       "step": 79300
     },
     {
       "epoch": 1.288524963038781,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2394,
       "step": 79310
     },
     {
       "epoch": 1.2886874299361506,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.1838,
       "step": 79320
     },
     {
       "epoch": 1.2888498968335202,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2122,
       "step": 79330
     },
     {
       "epoch": 1.28901236373089,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2445,
       "step": 79340
     },
     {
       "epoch": 1.2891748306282595,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3566,
+      "loss": 1.2237,
       "step": 79350
     },
     {
       "epoch": 1.2893372975256292,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.2334,
       "step": 79360
     },
     {
       "epoch": 1.2894997644229989,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3258,
+      "loss": 1.2262,
       "step": 79370
     },
     {
       "epoch": 1.2896622313203685,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2222,
       "step": 79380
     },
     {
       "epoch": 1.2898246982177382,
-      "grad_norm": 1.9375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2235,
       "step": 79390
     },
     {
       "epoch": 1.2899871651151078,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3628,
+      "loss": 1.2176,
       "step": 79400
     },
     {
       "epoch": 1.2901496320124775,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.183,
       "step": 79410
     },
     {
       "epoch": 1.290312098909847,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2127,
       "step": 79420
     },
     {
       "epoch": 1.2904745658072168,
-      "grad_norm": 3.421875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.236,
       "step": 79430
     },
     {
       "epoch": 1.2906370327045864,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2178,
       "step": 79440
     },
     {
       "epoch": 1.290799499601956,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2421,
       "step": 79450
     },
     {
       "epoch": 1.2909619664993257,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3731,
+      "loss": 1.2254,
       "step": 79460
     },
     {
       "epoch": 1.2911244333966954,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3232,
+      "loss": 1.2327,
       "step": 79470
     },
     {
       "epoch": 1.291286900294065,
-      "grad_norm": 2.125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.221,
       "step": 79480
     },
     {
       "epoch": 1.2914493671914347,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1877,
       "step": 79490
     },
     {
       "epoch": 1.2916118340888043,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2434,
       "step": 79500
     },
     {
       "epoch": 1.291774300986174,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.2393,
       "step": 79510
     },
     {
       "epoch": 1.2919367678835436,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.2649,
       "step": 79520
     },
     {
       "epoch": 1.2920992347809133,
-      "grad_norm": 2.15625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2276,
       "step": 79530
     },
     {
       "epoch": 1.292261701678283,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2432,
       "step": 79540
     },
     {
       "epoch": 1.2924241685756528,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2222,
       "step": 79550
     },
     {
       "epoch": 1.2925866354730224,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2202,
       "step": 79560
     },
     {
       "epoch": 1.292749102370392,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3685,
+      "loss": 1.2144,
       "step": 79570
     },
     {
       "epoch": 1.2929115692677617,
-      "grad_norm": 2.015625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.317,
+      "loss": 1.204,
       "step": 79580
     },
     {
       "epoch": 1.2930740361651314,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2731,
       "step": 79590
     },
     {
       "epoch": 1.293236503062501,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2378,
       "step": 79600
     },
     {
       "epoch": 1.2933989699598707,
-      "grad_norm": 3.828125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.177,
       "step": 79610
     },
     {
       "epoch": 1.2935614368572403,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2271,
       "step": 79620
     },
     {
       "epoch": 1.29372390375461,
-      "grad_norm": 2.875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2349,
       "step": 79630
     },
     {
       "epoch": 1.2938863706519796,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.1899,
       "step": 79640
     },
     {
       "epoch": 1.2940488375493493,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2183,
       "step": 79650
     },
     {
       "epoch": 1.294211304446719,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2471,
       "step": 79660
     },
     {
       "epoch": 1.2943737713440886,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2278,
       "step": 79670
     },
     {
       "epoch": 1.2945362382414582,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.1735,
       "step": 79680
     },
     {
       "epoch": 1.294698705138828,
-      "grad_norm": 2.125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2529,
       "step": 79690
     },
     {
       "epoch": 1.2948611720361975,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2066,
       "step": 79700
     },
     {
       "epoch": 1.2950236389335674,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.1883,
       "step": 79710
     },
     {
       "epoch": 1.295186105830937,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.1837,
       "step": 79720
     },
     {
       "epoch": 1.2953485727283067,
-      "grad_norm": 3.125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2162,
       "step": 79730
     },
     {
       "epoch": 1.2955110396256764,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.1983,
       "step": 79740
     },
     {
       "epoch": 1.295673506523046,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.1616,
       "step": 79750
     },
     {
       "epoch": 1.2958359734204157,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2802,
       "step": 79760
     },
     {
       "epoch": 1.2959984403177853,
-      "grad_norm": 2.5,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2084,
       "step": 79770
     },
     {
       "epoch": 1.296160907215155,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.1704,
       "step": 79780
     },
     {
       "epoch": 1.2963233741125246,
-      "grad_norm": 3.921875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.231,
       "step": 79790
     },
     {
       "epoch": 1.2964858410098943,
-      "grad_norm": 2.25,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2027,
       "step": 79800
     },
     {
       "epoch": 1.296648307907264,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1735,
       "step": 79810
     },
     {
       "epoch": 1.2968107748046336,
-      "grad_norm": 2.25,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2145,
       "step": 79820
     },
     {
       "epoch": 1.2969732417020032,
-      "grad_norm": 2.59375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2188,
       "step": 79830
     },
     {
       "epoch": 1.2971357085993729,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2264,
       "step": 79840
     },
     {
       "epoch": 1.2972981754967425,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2281,
       "step": 79850
     },
     {
       "epoch": 1.2974606423941122,
-      "grad_norm": 2.125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2747,
       "step": 79860
     },
     {
       "epoch": 1.2976231092914818,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3281,
+      "loss": 1.2094,
       "step": 79870
     },
     {
       "epoch": 1.2977855761888515,
-      "grad_norm": 1.8515625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.1983,
       "step": 79880
     },
     {
       "epoch": 1.2979480430862211,
-      "grad_norm": 2.71875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2004,
       "step": 79890
     },
     {
       "epoch": 1.2981105099835908,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.218,
       "step": 79900
     },
     {
       "epoch": 1.2982729768809604,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.208,
       "step": 79910
     },
     {
       "epoch": 1.29843544377833,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2353,
       "step": 79920
     },
     {
       "epoch": 1.2985979106756997,
-      "grad_norm": 2.75,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2732,
       "step": 79930
     },
     {
       "epoch": 1.2987603775730694,
-      "grad_norm": 2.125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2256,
       "step": 79940
     },
     {
       "epoch": 1.298922844470439,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.2286,
       "step": 79950
     },
     {
       "epoch": 1.2990853113678087,
-      "grad_norm": 2.328125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1909,
       "step": 79960
     },
     {
       "epoch": 1.2992477782651783,
-      "grad_norm": 3.375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3598,
+      "loss": 1.1821,
       "step": 79970
     },
     {
       "epoch": 1.299410245162548,
-      "grad_norm": 3.875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.2358,
       "step": 79980
     },
     {
       "epoch": 1.2995727120599179,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3678,
+      "loss": 1.1786,
       "step": 79990
     },
     {
       "epoch": 1.2997351789572875,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.1773,
       "step": 80000
     },
     {
       "epoch": 1.2998976458546572,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2054,
       "step": 80010
     },
     {
       "epoch": 1.3000601127520268,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2131,
       "step": 80020
     },
     {
       "epoch": 1.3002225796493965,
-      "grad_norm": 2.625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.1825,
       "step": 80030
     },
     {
       "epoch": 1.3003850465467661,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2009,
       "step": 80040
     },
     {
       "epoch": 1.3005475134441358,
-      "grad_norm": 3.375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3085,
+      "loss": 1.2115,
       "step": 80050
     },
     {
       "epoch": 1.3007099803415054,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2078,
       "step": 80060
     },
     {
       "epoch": 1.300872447238875,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2266,
       "step": 80070
     },
     {
       "epoch": 1.3010349141362447,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2123,
       "step": 80080
     },
     {
       "epoch": 1.3011973810336144,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.2123,
       "step": 80090
     },
     {
       "epoch": 1.301359847930984,
-      "grad_norm": 1.7421875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2182,
       "step": 80100
     },
     {
       "epoch": 1.3015223148283537,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2533,
       "step": 80110
     },
     {
       "epoch": 1.3016847817257233,
-      "grad_norm": 3.0,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2101,
       "step": 80120
     },
     {
       "epoch": 1.301847248623093,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.1885,
       "step": 80130
     },
     {
       "epoch": 1.3020097155204626,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2366,
       "step": 80140
     },
     {
       "epoch": 1.3021721824178325,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2195,
       "step": 80150
     },
     {
       "epoch": 1.3023346493152022,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2397,
       "step": 80160
     },
     {
       "epoch": 1.3024971162125718,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2175,
       "step": 80170
     },
     {
       "epoch": 1.3026595831099415,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2207,
       "step": 80180
     },
     {
       "epoch": 1.302822050007311,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.1891,
       "step": 80190
     },
     {
       "epoch": 1.3029845169046808,
-      "grad_norm": 2.25,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2071,
       "step": 80200
     },
     {
       "epoch": 1.3031469838020504,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.1682,
       "step": 80210
     },
     {
       "epoch": 1.30330945069942,
-      "grad_norm": 2.875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2011,
       "step": 80220
     },
     {
       "epoch": 1.3034719175967897,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2509,
       "step": 80230
     },
     {
       "epoch": 1.3036343844941594,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2341,
       "step": 80240
     },
     {
       "epoch": 1.303796851391529,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.247,
       "step": 80250
     },
     {
       "epoch": 1.3039593182888987,
-      "grad_norm": 3.5,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2133,
       "step": 80260
     },
     {
       "epoch": 1.3041217851862683,
-      "grad_norm": 3.25,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3271,
+      "loss": 1.2616,
       "step": 80270
     },
     {
       "epoch": 1.304284252083638,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.237,
       "step": 80280
     },
     {
       "epoch": 1.3044467189810076,
-      "grad_norm": 2.375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2341,
       "step": 80290
     },
     {
       "epoch": 1.3046091858783773,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2017,
       "step": 80300
     },
     {
       "epoch": 1.304771652775747,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3579,
+      "loss": 1.1769,
       "step": 80310
     },
     {
       "epoch": 1.3049341196731166,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.1757,
       "step": 80320
     },
     {
       "epoch": 1.3050965865704862,
-      "grad_norm": 3.421875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3565,
+      "loss": 1.2144,
       "step": 80330
     },
     {
       "epoch": 1.3052590534678559,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3616,
+      "loss": 1.2018,
       "step": 80340
     },
     {
       "epoch": 1.3054215203652255,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.1745,
       "step": 80350
     },
     {
       "epoch": 1.3055839872625952,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.216,
       "step": 80360
     },
     {
       "epoch": 1.3057464541599648,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.1878,
       "step": 80370
     },
     {
       "epoch": 1.3059089210573345,
-      "grad_norm": 2.25,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3266,
+      "loss": 1.2155,
       "step": 80380
     },
     {
       "epoch": 1.3060713879547041,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3141,
+      "loss": 1.226,
       "step": 80390
     },
     {
       "epoch": 1.3062338548520738,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.2075,
       "step": 80400
     },
     {
       "epoch": 1.3063963217494434,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3109,
+      "loss": 1.2371,
       "step": 80410
     },
     {
       "epoch": 1.306558788646813,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2283,
       "step": 80420
     },
     {
       "epoch": 1.306721255544183,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.2421,
       "step": 80430
     },
     {
       "epoch": 1.3068837224415526,
-      "grad_norm": 1.875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2008,
       "step": 80440
     },
     {
       "epoch": 1.3070461893389222,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.2169,
       "step": 80450
     },
     {
       "epoch": 1.307208656236292,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2102,
       "step": 80460
     },
     {
       "epoch": 1.3073711231336616,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3607,
+      "loss": 1.2261,
       "step": 80470
     },
     {
       "epoch": 1.3075335900310312,
-      "grad_norm": 2.0,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.1931,
       "step": 80480
     },
     {
       "epoch": 1.3076960569284009,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2214,
       "step": 80490
     },
     {
       "epoch": 1.3078585238257705,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1926,
       "step": 80500
     },
     {
       "epoch": 1.3080209907231402,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2349,
       "step": 80510
     },
     {
       "epoch": 1.3081834576205098,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1829,
       "step": 80520
     },
     {
       "epoch": 1.3083459245178795,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.2288,
       "step": 80530
     },
     {
       "epoch": 1.308508391415249,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2216,
       "step": 80540
     },
     {
       "epoch": 1.3086708583126188,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.2099,
       "step": 80550
     },
     {
       "epoch": 1.3088333252099884,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3204,
+      "loss": 1.2287,
       "step": 80560
     },
     {
       "epoch": 1.308995792107358,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2085,
       "step": 80570
     },
     {
       "epoch": 1.3091582590047277,
-      "grad_norm": 1.6015625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2275,
       "step": 80580
     },
     {
       "epoch": 1.3093207259020976,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2105,
       "step": 80590
     },
     {
       "epoch": 1.3094831927994672,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2145,
       "step": 80600
     },
     {
       "epoch": 1.3096456596968369,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2234,
       "step": 80610
     },
     {
       "epoch": 1.3098081265942065,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.2389,
       "step": 80620
     },
     {
       "epoch": 1.3099705934915762,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3662,
+      "loss": 1.2534,
       "step": 80630
     },
     {
       "epoch": 1.3101330603889458,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.1787,
       "step": 80640
     },
     {
       "epoch": 1.3102955272863155,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.2346,
       "step": 80650
     },
     {
       "epoch": 1.3104579941836851,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1861,
       "step": 80660
     },
     {
       "epoch": 1.3106204610810548,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.188,
       "step": 80670
     },
     {
       "epoch": 1.3107829279784244,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2008,
       "step": 80680
     },
     {
       "epoch": 1.310945394875794,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3212,
+      "loss": 1.2439,
       "step": 80690
     },
     {
       "epoch": 1.3111078617731637,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.1979,
       "step": 80700
     },
     {
       "epoch": 1.3112703286705334,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.219,
       "step": 80710
     },
     {
       "epoch": 1.311432795567903,
-      "grad_norm": 2.625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.1819,
       "step": 80720
     },
     {
       "epoch": 1.3115952624652727,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2507,
       "step": 80730
     },
     {
       "epoch": 1.3117577293626423,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2078,
       "step": 80740
     },
     {
       "epoch": 1.311920196260012,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2155,
       "step": 80750
     },
     {
       "epoch": 1.3120826631573816,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3641,
+      "loss": 1.2432,
       "step": 80760
     },
     {
       "epoch": 1.3122451300547513,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.205,
       "step": 80770
     },
     {
       "epoch": 1.312407596952121,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.364,
+      "loss": 1.2052,
       "step": 80780
     },
     {
       "epoch": 1.3125700638494906,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3607,
+      "loss": 1.1787,
       "step": 80790
     },
     {
       "epoch": 1.3127325307468602,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.228,
       "step": 80800
     },
     {
       "epoch": 1.31289499764423,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2012,
       "step": 80810
     },
     {
       "epoch": 1.3130574645415996,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.2599,
       "step": 80820
     },
     {
       "epoch": 1.3132199314389692,
-      "grad_norm": 4.125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2749,
       "step": 80830
     },
     {
       "epoch": 1.3133823983363389,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2032,
       "step": 80840
     },
     {
       "epoch": 1.3135448652337085,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2176,
       "step": 80850
     },
     {
       "epoch": 1.3137073321310784,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2374,
       "step": 80860
     },
     {
       "epoch": 1.313869799028448,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2288,
       "step": 80870
     },
     {
       "epoch": 1.3140322659258177,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2367,
       "step": 80880
     },
     {
       "epoch": 1.3141947328231873,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3182,
+      "loss": 1.2032,
       "step": 80890
     },
     {
       "epoch": 1.314357199720557,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.1997,
       "step": 80900
     },
     {
       "epoch": 1.3145196666179266,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2192,
       "step": 80910
     },
     {
       "epoch": 1.3146821335152963,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.1787,
       "step": 80920
     },
     {
       "epoch": 1.314844600412666,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2112,
       "step": 80930
     },
     {
       "epoch": 1.3150070673100356,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2406,
       "step": 80940
     },
     {
       "epoch": 1.3151695342074052,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2165,
       "step": 80950
     },
     {
       "epoch": 1.3153320011047749,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.2117,
       "step": 80960
     },
     {
       "epoch": 1.3154944680021445,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.1826,
       "step": 80970
     },
     {
       "epoch": 1.3156569348995142,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2156,
       "step": 80980
     },
     {
       "epoch": 1.3158194017968838,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3742,
+      "loss": 1.1957,
       "step": 80990
     },
     {
       "epoch": 1.3159818686942535,
-      "grad_norm": 2.625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2127,
       "step": 81000
     },
     {
       "epoch": 1.3161443355916231,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2115,
       "step": 81010
     },
     {
       "epoch": 1.3163068024889928,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2432,
       "step": 81020
     },
     {
       "epoch": 1.3164692693863627,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2275,
       "step": 81030
     },
     {
       "epoch": 1.3166317362837323,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2591,
       "step": 81040
     },
     {
       "epoch": 1.316794203181102,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2639,
       "step": 81050
     },
     {
       "epoch": 1.3169566700784716,
-      "grad_norm": 3.71875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2087,
       "step": 81060
     },
     {
       "epoch": 1.3171191369758413,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.2228,
       "step": 81070
     },
     {
       "epoch": 1.317281603873211,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3649,
+      "loss": 1.206,
       "step": 81080
     },
     {
       "epoch": 1.3174440707705806,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.2367,
       "step": 81090
     },
     {
       "epoch": 1.3176065376679502,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2306,
       "step": 81100
     },
     {
       "epoch": 1.3177690045653199,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2188,
       "step": 81110
     },
     {
       "epoch": 1.3179314714626895,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.213,
       "step": 81120
     },
     {
       "epoch": 1.3180939383600592,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2381,
       "step": 81130
     },
     {
       "epoch": 1.3182564052574288,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2633,
       "step": 81140
     },
     {
       "epoch": 1.3184188721547985,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1857,
       "step": 81150
     },
     {
       "epoch": 1.3185813390521681,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2128,
       "step": 81160
     },
     {
       "epoch": 1.3187438059495378,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2182,
       "step": 81170
     },
     {
       "epoch": 1.3189062728469074,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.1843,
       "step": 81180
     },
     {
       "epoch": 1.319068739744277,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2039,
       "step": 81190
     },
     {
       "epoch": 1.3192312066416467,
-      "grad_norm": 2.109375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.2138,
       "step": 81200
     },
     {
       "epoch": 1.3193936735390164,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2191,
       "step": 81210
     },
     {
       "epoch": 1.319556140436386,
-      "grad_norm": 1.875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.1925,
       "step": 81220
     },
     {
       "epoch": 1.3197186073337557,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2199,
       "step": 81230
     },
     {
       "epoch": 1.3198810742311253,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.207,
       "step": 81240
     },
     {
       "epoch": 1.320043541128495,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3131,
+      "loss": 1.2179,
       "step": 81250
     },
     {
       "epoch": 1.3202060080258646,
-      "grad_norm": 2.625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2033,
       "step": 81260
     },
     {
       "epoch": 1.3203684749232343,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3201,
+      "loss": 1.2599,
       "step": 81270
     },
     {
       "epoch": 1.320530941820604,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2291,
       "step": 81280
     },
     {
       "epoch": 1.3206934087179736,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3189,
+      "loss": 1.1744,
       "step": 81290
     },
     {
       "epoch": 1.3208558756153435,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2194,
       "step": 81300
     },
     {
       "epoch": 1.321018342512713,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.1811,
       "step": 81310
     },
     {
       "epoch": 1.3211808094100828,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2277,
       "step": 81320
     },
     {
       "epoch": 1.3213432763074524,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2309,
       "step": 81330
     },
     {
       "epoch": 1.321505743204822,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3247,
+      "loss": 1.2127,
       "step": 81340
     },
     {
       "epoch": 1.3216682101021917,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.223,
       "step": 81350
     },
     {
       "epoch": 1.3218306769995614,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.1931,
       "step": 81360
     },
     {
       "epoch": 1.321993143896931,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2077,
       "step": 81370
     },
     {
       "epoch": 1.3221556107943007,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.1966,
       "step": 81380
     },
     {
       "epoch": 1.3223180776916703,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.23,
       "step": 81390
     },
     {
       "epoch": 1.32248054458904,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.2202,
       "step": 81400
     },
     {
       "epoch": 1.3226430114864096,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2521,
       "step": 81410
     },
     {
       "epoch": 1.3228054783837793,
-      "grad_norm": 2.5,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.1967,
       "step": 81420
     },
     {
       "epoch": 1.322967945281149,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.2259,
       "step": 81430
     },
     {
       "epoch": 1.3231304121785186,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.1842,
       "step": 81440
     },
     {
       "epoch": 1.3232928790758882,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.186,
       "step": 81450
     },
     {
       "epoch": 1.3234553459732579,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.231,
       "step": 81460
     },
     {
       "epoch": 1.3236178128706277,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2101,
       "step": 81470
     },
     {
       "epoch": 1.3237802797679974,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3747,
+      "loss": 1.2128,
       "step": 81480
     },
     {
       "epoch": 1.323942746665367,
-      "grad_norm": 2.625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2383,
       "step": 81490
     },
     {
       "epoch": 1.3241052135627367,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2085,
       "step": 81500
     },
     {
       "epoch": 1.3242676804601063,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2294,
       "step": 81510
     },
     {
       "epoch": 1.324430147357476,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.2301,
       "step": 81520
     },
     {
       "epoch": 1.3245926142548456,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2202,
       "step": 81530
     },
     {
       "epoch": 1.3247550811522153,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.1792,
       "step": 81540
     },
     {
       "epoch": 1.324917548049585,
-      "grad_norm": 2.5,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3566,
+      "loss": 1.1811,
       "step": 81550
     },
     {
       "epoch": 1.3250800149469546,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2367,
       "step": 81560
     },
     {
       "epoch": 1.3252424818443242,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2086,
       "step": 81570
     },
     {
       "epoch": 1.325404948741694,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2107,
       "step": 81580
     },
     {
       "epoch": 1.3255674156390636,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.1878,
       "step": 81590
     },
     {
       "epoch": 1.3257298825364332,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2178,
       "step": 81600
     },
     {
       "epoch": 1.3258923494338029,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.371,
+      "loss": 1.2124,
       "step": 81610
     },
     {
       "epoch": 1.3260548163311725,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.1976,
       "step": 81620
     },
     {
       "epoch": 1.3262172832285422,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2376,
       "step": 81630
     },
     {
       "epoch": 1.3263797501259118,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2047,
       "step": 81640
     },
     {
       "epoch": 1.3265422170232815,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2158,
       "step": 81650
     },
     {
       "epoch": 1.326704683920651,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2042,
       "step": 81660
     },
     {
       "epoch": 1.3268671508180208,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.1986,
       "step": 81670
     },
     {
       "epoch": 1.3270296177153904,
-      "grad_norm": 2.59375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2468,
       "step": 81680
     },
     {
       "epoch": 1.32719208461276,
-      "grad_norm": 6.8125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.1926,
       "step": 81690
     },
     {
       "epoch": 1.3273545515101297,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2247,
       "step": 81700
     },
     {
       "epoch": 1.3275170184074994,
-      "grad_norm": 2.625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2449,
       "step": 81710
     },
     {
       "epoch": 1.327679485304869,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2234,
       "step": 81720
     },
     {
       "epoch": 1.3278419522022387,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.197,
       "step": 81730
     },
     {
       "epoch": 1.3280044190996085,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.1884,
       "step": 81740
     },
     {
       "epoch": 1.3281668859969782,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2256,
       "step": 81750
     },
     {
       "epoch": 1.3283293528943478,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3108,
+      "loss": 1.2133,
       "step": 81760
     },
     {
       "epoch": 1.3284918197917175,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2107,
       "step": 81770
     },
     {
       "epoch": 1.3286542866890871,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2188,
       "step": 81780
     },
     {
       "epoch": 1.3288167535864568,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2156,
       "step": 81790
     },
     {
       "epoch": 1.3289792204838264,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.2169,
       "step": 81800
     },
     {
       "epoch": 1.329141687381196,
-      "grad_norm": 2.875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.2137,
       "step": 81810
     },
     {
       "epoch": 1.3293041542785657,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2008,
       "step": 81820
     },
     {
       "epoch": 1.3294666211759354,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2029,
       "step": 81830
     },
     {
       "epoch": 1.329629088073305,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2031,
       "step": 81840
     },
     {
       "epoch": 1.3297915549706747,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1849,
       "step": 81850
     },
     {
       "epoch": 1.3299540218680443,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2102,
       "step": 81860
     },
     {
       "epoch": 1.330116488765414,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2329,
       "step": 81870
     },
     {
       "epoch": 1.3302789556627836,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.22,
       "step": 81880
     },
     {
       "epoch": 1.3304414225601533,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2501,
       "step": 81890
     },
     {
       "epoch": 1.330603889457523,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2324,
       "step": 81900
     },
     {
       "epoch": 1.3307663563548928,
-      "grad_norm": 2.375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2085,
       "step": 81910
     },
     {
       "epoch": 1.3309288232522625,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.1954,
       "step": 81920
     },
     {
       "epoch": 1.3310912901496321,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2119,
       "step": 81930
     },
     {
       "epoch": 1.3312537570470018,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.1928,
       "step": 81940
     },
     {
       "epoch": 1.3314162239443714,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2216,
       "step": 81950
     },
     {
       "epoch": 1.331578690841741,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.2432,
       "step": 81960
     },
     {
       "epoch": 1.3317411577391107,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.219,
       "step": 81970
     },
     {
       "epoch": 1.3319036246364804,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2276,
       "step": 81980
     },
     {
       "epoch": 1.33206609153385,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2069,
       "step": 81990
     },
     {
       "epoch": 1.3322285584312197,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.213,
       "step": 82000
     },
     {
       "epoch": 1.3323910253285893,
-      "grad_norm": 3.125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.207,
       "step": 82010
     },
     {
       "epoch": 1.332553492225959,
-      "grad_norm": 4.1875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2328,
       "step": 82020
     },
     {
       "epoch": 1.3327159591233286,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1691,
       "step": 82030
     },
     {
       "epoch": 1.3328784260206983,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.238,
       "step": 82040
     },
     {
       "epoch": 1.333040892918068,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2359,
       "step": 82050
     },
     {
       "epoch": 1.3332033598154376,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2225,
       "step": 82060
     },
     {
       "epoch": 1.3333658267128072,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2417,
       "step": 82070
     },
     {
       "epoch": 1.3335282936101769,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2194,
       "step": 82080
     },
     {
       "epoch": 1.3336907605075465,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 1.193,
       "step": 82090
     },
     {
       "epoch": 1.3338532274049162,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.1904,
       "step": 82100
     },
     {
       "epoch": 1.3340156943022858,
-      "grad_norm": 2.125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2362,
       "step": 82110
     },
     {
       "epoch": 1.3341781611996555,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.1914,
       "step": 82120
     },
     {
       "epoch": 1.3343406280970251,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2106,
       "step": 82130
     },
     {
       "epoch": 1.3345030949943948,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2205,
       "step": 82140
     },
     {
       "epoch": 1.3346655618917644,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2173,
       "step": 82150
     },
     {
       "epoch": 1.334828028789134,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2193,
       "step": 82160
     },
     {
       "epoch": 1.3349904956865037,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2117,
       "step": 82170
     },
     {
       "epoch": 1.3351529625838736,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.1621,
       "step": 82180
     },
     {
       "epoch": 1.3353154294812433,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.246,
       "step": 82190
     },
     {
       "epoch": 1.335477896378613,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.2054,
       "step": 82200
     },
     {
       "epoch": 1.3356403632759826,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.1945,
       "step": 82210
     },
     {
       "epoch": 1.3358028301733522,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2238,
       "step": 82220
     },
     {
       "epoch": 1.3359652970707219,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2054,
       "step": 82230
     },
     {
       "epoch": 1.3361277639680915,
-      "grad_norm": 3.5,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.2497,
       "step": 82240
     },
     {
       "epoch": 1.3362902308654612,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3836,
+      "loss": 1.2334,
       "step": 82250
     },
     {
       "epoch": 1.3364526977628308,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3682,
+      "loss": 1.1831,
       "step": 82260
     },
     {
       "epoch": 1.3366151646602005,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2485,
       "step": 82270
     },
     {
       "epoch": 1.3367776315575701,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.1981,
       "step": 82280
     },
     {
       "epoch": 1.3369400984549398,
-      "grad_norm": 3.625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.1947,
       "step": 82290
     },
     {
       "epoch": 1.3371025653523094,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2186,
       "step": 82300
     },
     {
       "epoch": 1.337265032249679,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2363,
       "step": 82310
     },
     {
       "epoch": 1.3374274991470487,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2022,
       "step": 82320
     },
     {
       "epoch": 1.3375899660444184,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.224,
       "step": 82330
     },
     {
       "epoch": 1.337752432941788,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2181,
       "step": 82340
     },
     {
       "epoch": 1.337914899839158,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.229,
       "step": 82350
     },
     {
       "epoch": 1.3380773667365276,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3616,
+      "loss": 1.2176,
       "step": 82360
     },
     {
       "epoch": 1.3382398336338972,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2135,
       "step": 82370
     },
     {
       "epoch": 1.3384023005312669,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3195,
+      "loss": 1.2035,
       "step": 82380
     },
     {
       "epoch": 1.3385647674286365,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.1801,
       "step": 82390
     },
     {
       "epoch": 1.3387272343260062,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2119,
       "step": 82400
     },
     {
       "epoch": 1.3388897012233758,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2429,
       "step": 82410
     },
     {
       "epoch": 1.3390521681207455,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.205,
       "step": 82420
     },
     {
       "epoch": 1.339214635018115,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2328,
       "step": 82430
     },
     {
       "epoch": 1.3393771019154848,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.1922,
       "step": 82440
     },
     {
       "epoch": 1.3395395688128544,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2306,
       "step": 82450
     },
     {
       "epoch": 1.339702035710224,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.2305,
       "step": 82460
     },
     {
       "epoch": 1.3398645026075937,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.3183,
+      "loss": 1.2366,
       "step": 82470
     },
     {
       "epoch": 1.3400269695049634,
-      "grad_norm": 2.75,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.1604,
       "step": 82480
     },
     {
       "epoch": 1.340189436402333,
-      "grad_norm": 2.65625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.1756,
       "step": 82490
     },
     {
       "epoch": 1.3403519032997027,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2178,
       "step": 82500
     },
     {
       "epoch": 1.3405143701970723,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.205,
       "step": 82510
     },
     {
       "epoch": 1.340676837094442,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2341,
       "step": 82520
     },
     {
       "epoch": 1.3408393039918116,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2065,
       "step": 82530
     },
     {
       "epoch": 1.3410017708891813,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3029,
+      "loss": 1.2278,
       "step": 82540
     },
     {
       "epoch": 1.341164237786551,
-      "grad_norm": 3.125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.208,
       "step": 82550
     },
     {
       "epoch": 1.3413267046839206,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.2093,
       "step": 82560
     },
     {
       "epoch": 1.3414891715812902,
-      "grad_norm": 4.125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2225,
       "step": 82570
     },
     {
       "epoch": 1.3416516384786599,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2532,
       "step": 82580
     },
     {
       "epoch": 1.3418141053760295,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.1991,
       "step": 82590
     },
     {
       "epoch": 1.3419765722733992,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.1653,
       "step": 82600
     },
     {
       "epoch": 1.3421390391707688,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.306,
+      "loss": 1.181,
       "step": 82610
     },
     {
       "epoch": 1.3423015060681387,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2259,
       "step": 82620
     },
     {
       "epoch": 1.3424639729655083,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2651,
       "step": 82630
     },
     {
       "epoch": 1.342626439862878,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2421,
       "step": 82640
     },
     {
       "epoch": 1.3427889067602476,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.1957,
       "step": 82650
     },
     {
       "epoch": 1.3429513736576173,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.1443,
       "step": 82660
     },
     {
       "epoch": 1.343113840554987,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.198,
       "step": 82670
     },
     {
       "epoch": 1.3432763074523566,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2525,
       "step": 82680
     },
     {
       "epoch": 1.3434387743497263,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.1735,
       "step": 82690
     },
     {
       "epoch": 1.343601241247096,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.2477,
       "step": 82700
     },
     {
       "epoch": 1.3437637081444656,
-      "grad_norm": 3.375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2269,
       "step": 82710
     },
     {
       "epoch": 1.3439261750418352,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2006,
       "step": 82720
     },
     {
       "epoch": 1.3440886419392049,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2382,
       "step": 82730
     },
     {
       "epoch": 1.3442511088365745,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2113,
       "step": 82740
     },
     {
       "epoch": 1.3444135757339442,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2448,
       "step": 82750
     },
     {
       "epoch": 1.3445760426313138,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.1919,
       "step": 82760
     },
     {
       "epoch": 1.3447385095286835,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2283,
       "step": 82770
     },
     {
       "epoch": 1.3449009764260533,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 1.2134,
       "step": 82780
     },
     {
       "epoch": 1.345063443323423,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.239,
       "step": 82790
     },
     {
       "epoch": 1.3452259102207926,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3784,
+      "loss": 1.1885,
       "step": 82800
     },
     {
       "epoch": 1.3453883771181623,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2281,
       "step": 82810
     },
     {
       "epoch": 1.345550844015532,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2171,
       "step": 82820
     },
     {
       "epoch": 1.3457133109129016,
-      "grad_norm": 2.5,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2201,
       "step": 82830
     },
     {
       "epoch": 1.3458757778102712,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2491,
       "step": 82840
     },
     {
       "epoch": 1.3460382447076409,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2178,
       "step": 82850
     },
     {
       "epoch": 1.3462007116050105,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.1935,
       "step": 82860
     },
     {
       "epoch": 1.3463631785023802,
-      "grad_norm": 2.5,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2575,
       "step": 82870
     },
     {
       "epoch": 1.3465256453997498,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.1957,
       "step": 82880
     },
     {
       "epoch": 1.3466881122971195,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.1967,
       "step": 82890
     },
     {
       "epoch": 1.3468505791944891,
-      "grad_norm": 5.0625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2075,
       "step": 82900
     },
     {
       "epoch": 1.3470130460918588,
-      "grad_norm": 3.859375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.2261,
       "step": 82910
     },
     {
       "epoch": 1.3471755129892284,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.2091,
       "step": 82920
     },
     {
       "epoch": 1.347337979886598,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2374,
       "step": 82930
     },
     {
       "epoch": 1.3475004467839677,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2451,
       "step": 82940
     },
     {
       "epoch": 1.3476629136813374,
-      "grad_norm": 3.25,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3723,
+      "loss": 1.2331,
       "step": 82950
     },
     {
       "epoch": 1.347825380578707,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2264,
       "step": 82960
     },
     {
       "epoch": 1.3479878474760767,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.2087,
       "step": 82970
     },
     {
       "epoch": 1.3481503143734463,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3193,
+      "loss": 1.2224,
       "step": 82980
     },
     {
       "epoch": 1.348312781270816,
-      "grad_norm": 3.625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3744,
+      "loss": 1.2296,
       "step": 82990
     },
     {
       "epoch": 1.3484752481681856,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.1855,
       "step": 83000
     },
     {
       "epoch": 1.3486377150655553,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.2156,
       "step": 83010
     },
     {
       "epoch": 1.348800181962925,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2415,
       "step": 83020
     },
     {
       "epoch": 1.3489626488602946,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.228,
       "step": 83030
     },
     {
       "epoch": 1.3491251157576643,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.2127,
       "step": 83040
     },
     {
       "epoch": 1.349287582655034,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2058,
       "step": 83050
     },
     {
       "epoch": 1.3494500495524038,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2407,
       "step": 83060
     },
     {
       "epoch": 1.3496125164497734,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.1689,
       "step": 83070
     },
     {
       "epoch": 1.349774983347143,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2545,
       "step": 83080
     },
     {
       "epoch": 1.3499374502445127,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1936,
       "step": 83090
     },
     {
       "epoch": 1.3500999171418824,
-      "grad_norm": 2.875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2503,
       "step": 83100
     },
     {
       "epoch": 1.350262384039252,
-      "grad_norm": 3.375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3221,
+      "loss": 1.2243,
       "step": 83110
     },
     {
       "epoch": 1.3504248509366217,
-      "grad_norm": 8.25,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.1795,
       "step": 83120
     },
     {
       "epoch": 1.3505873178339913,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2348,
       "step": 83130
     },
     {
       "epoch": 1.350749784731361,
-      "grad_norm": 3.5,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.1895,
       "step": 83140
     },
     {
       "epoch": 1.3509122516287306,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.231,
       "step": 83150
     },
     {
       "epoch": 1.3510747185261003,
-      "grad_norm": 2.75,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2749,
       "step": 83160
     },
     {
       "epoch": 1.35123718542347,
-      "grad_norm": 3.640625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.1946,
       "step": 83170
     },
     {
       "epoch": 1.3513996523208396,
-      "grad_norm": 2.875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2399,
       "step": 83180
     },
     {
       "epoch": 1.3515621192182092,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2283,
       "step": 83190
     },
     {
       "epoch": 1.3517245861155789,
-      "grad_norm": 2.75,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2422,
       "step": 83200
     },
     {
       "epoch": 1.3518870530129485,
-      "grad_norm": 4.40625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2092,
       "step": 83210
     },
     {
       "epoch": 1.3520495199103184,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.1909,
       "step": 83220
     },
     {
       "epoch": 1.352211986807688,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2666,
       "step": 83230
     },
     {
       "epoch": 1.3523744537050577,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2266,
       "step": 83240
     },
     {
       "epoch": 1.3525369206024274,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2222,
       "step": 83250
     },
     {
       "epoch": 1.352699387499797,
-      "grad_norm": 4.71875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2282,
       "step": 83260
     },
     {
       "epoch": 1.3528618543971667,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.2189,
       "step": 83270
     },
     {
       "epoch": 1.3530243212945363,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3193,
+      "loss": 1.2204,
       "step": 83280
     },
     {
       "epoch": 1.353186788191906,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.1782,
       "step": 83290
     },
     {
       "epoch": 1.3533492550892756,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2166,
       "step": 83300
     },
     {
       "epoch": 1.3535117219866453,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2498,
       "step": 83310
     },
     {
       "epoch": 1.353674188884015,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2209,
       "step": 83320
     },
     {
       "epoch": 1.3538366557813846,
-      "grad_norm": 2.75,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.199,
       "step": 83330
     },
     {
       "epoch": 1.3539991226787542,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2044,
       "step": 83340
     },
     {
       "epoch": 1.3541615895761239,
-      "grad_norm": 2.75,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2074,
       "step": 83350
     },
     {
       "epoch": 1.3543240564734935,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2479,
       "step": 83360
     },
     {
       "epoch": 1.3544865233708632,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.1999,
       "step": 83370
     },
     {
       "epoch": 1.3546489902682328,
-      "grad_norm": 2.125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.2529,
       "step": 83380
     },
     {
       "epoch": 1.3548114571656025,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2405,
       "step": 83390
     },
     {
       "epoch": 1.3549739240629721,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2444,
       "step": 83400
     },
     {
       "epoch": 1.3551363909603418,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2011,
       "step": 83410
     },
     {
       "epoch": 1.3552988578577114,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.175,
       "step": 83420
     },
     {
       "epoch": 1.355461324755081,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2413,
       "step": 83430
     },
     {
       "epoch": 1.3556237916524507,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2175,
       "step": 83440
     },
     {
       "epoch": 1.3557862585498204,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2379,
       "step": 83450
     },
     {
       "epoch": 1.35594872544719,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2043,
       "step": 83460
     },
     {
       "epoch": 1.3561111923445597,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3609,
+      "loss": 1.2365,
       "step": 83470
     },
     {
       "epoch": 1.3562736592419293,
-      "grad_norm": 1.828125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3184,
+      "loss": 1.2142,
       "step": 83480
     },
     {
       "epoch": 1.356436126139299,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.2396,
       "step": 83490
     },
     {
       "epoch": 1.3565985930366689,
-      "grad_norm": 2.75,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.1522,
       "step": 83500
     },
     {
       "epoch": 1.3567610599340385,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2132,
       "step": 83510
     },
     {
       "epoch": 1.3569235268314082,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1985,
       "step": 83520
     },
     {
       "epoch": 1.3570859937287778,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2175,
       "step": 83530
     },
     {
       "epoch": 1.3572484606261475,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2078,
       "step": 83540
     },
     {
       "epoch": 1.357410927523517,
-      "grad_norm": 2.5,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2521,
       "step": 83550
     },
     {
       "epoch": 1.3575733944208868,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2515,
       "step": 83560
     },
     {
       "epoch": 1.3577358613182564,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2012,
       "step": 83570
     },
     {
       "epoch": 1.357898328215626,
-      "grad_norm": 2.25,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2388,
       "step": 83580
     },
     {
       "epoch": 1.3580607951129957,
-      "grad_norm": 2.0625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.1843,
       "step": 83590
     },
     {
       "epoch": 1.3582232620103654,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.235,
       "step": 83600
     },
     {
       "epoch": 1.358385728907735,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2161,
       "step": 83610
     },
     {
       "epoch": 1.3585481958051047,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2251,
       "step": 83620
     },
     {
       "epoch": 1.3587106627024743,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3238,
+      "loss": 1.2106,
       "step": 83630
     },
     {
       "epoch": 1.358873129599844,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3268,
+      "loss": 1.2214,
       "step": 83640
     },
     {
       "epoch": 1.3590355964972136,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.1756,
       "step": 83650
     },
     {
       "epoch": 1.3591980633945835,
-      "grad_norm": 4.15625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2223,
       "step": 83660
     },
     {
       "epoch": 1.3593605302919531,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3611,
+      "loss": 1.1991,
       "step": 83670
     },
     {
       "epoch": 1.3595229971893228,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2473,
       "step": 83680
     },
     {
       "epoch": 1.3596854640866924,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2234,
       "step": 83690
     },
     {
       "epoch": 1.359847930984062,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2,
       "step": 83700
     },
     {
       "epoch": 1.3600103978814317,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.1989,
       "step": 83710
     },
     {
       "epoch": 1.3601728647788014,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2406,
       "step": 83720
     },
     {
       "epoch": 1.360335331676171,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2407,
       "step": 83730
     },
     {
       "epoch": 1.3604977985735407,
-      "grad_norm": 3.125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.2311,
       "step": 83740
     },
     {
       "epoch": 1.3606602654709103,
-      "grad_norm": 2.875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.178,
       "step": 83750
     },
     {
       "epoch": 1.36082273236828,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2,
       "step": 83760
     },
     {
       "epoch": 1.3609851992656496,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2411,
       "step": 83770
     },
     {
       "epoch": 1.3611476661630193,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2064,
       "step": 83780
     },
     {
       "epoch": 1.361310133060389,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.1944,
       "step": 83790
     },
     {
       "epoch": 1.3614725999577586,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2049,
       "step": 83800
     },
     {
       "epoch": 1.3616350668551283,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.1891,
       "step": 83810
     },
     {
       "epoch": 1.361797533752498,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.2106,
       "step": 83820
     },
     {
       "epoch": 1.3619600006498676,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1754,
       "step": 83830
     },
     {
       "epoch": 1.3621224675472372,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3035,
+      "loss": 1.2441,
       "step": 83840
     },
     {
       "epoch": 1.3622849344446069,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2688,
       "step": 83850
     },
     {
       "epoch": 1.3624474013419765,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3607,
+      "loss": 1.2236,
       "step": 83860
     },
     {
       "epoch": 1.3626098682393462,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2444,
       "step": 83870
     },
     {
       "epoch": 1.3627723351367158,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3674,
+      "loss": 1.1956,
       "step": 83880
     },
     {
       "epoch": 1.3629348020340855,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2013,
       "step": 83890
     },
     {
       "epoch": 1.363097268931455,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.1887,
       "step": 83900
     },
     {
       "epoch": 1.3632597358288248,
-      "grad_norm": 4.65625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2894,
       "step": 83910
     },
     {
       "epoch": 1.3634222027261944,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2038,
       "step": 83920
     },
     {
       "epoch": 1.363584669623564,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.1837,
       "step": 83930
     },
     {
       "epoch": 1.363747136520934,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1998,
       "step": 83940
     },
     {
       "epoch": 1.3639096034183036,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2021,
       "step": 83950
     },
     {
       "epoch": 1.3640720703156732,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2527,
       "step": 83960
     },
     {
       "epoch": 1.3642345372130429,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2194,
       "step": 83970
     },
     {
       "epoch": 1.3643970041104125,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1858,
       "step": 83980
     },
     {
       "epoch": 1.3645594710077822,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3175,
+      "loss": 1.1982,
       "step": 83990
     },
     {
       "epoch": 1.3647219379051518,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.1997,
       "step": 84000
     },
     {
       "epoch": 1.3648844048025215,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2206,
       "step": 84010
     },
     {
       "epoch": 1.3650468716998911,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2143,
       "step": 84020
     },
     {
       "epoch": 1.3652093385972608,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.231,
       "step": 84030
     },
     {
       "epoch": 1.3653718054946304,
-      "grad_norm": 2.75,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2645,
       "step": 84040
     },
     {
       "epoch": 1.365534272392,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2077,
       "step": 84050
     },
     {
       "epoch": 1.3656967392893697,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2344,
       "step": 84060
     },
     {
       "epoch": 1.3658592061867394,
-      "grad_norm": 2.625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2071,
       "step": 84070
     },
     {
       "epoch": 1.366021673084109,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2153,
       "step": 84080
     },
     {
       "epoch": 1.3661841399814787,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2358,
       "step": 84090
     },
     {
       "epoch": 1.3663466068788486,
-      "grad_norm": 4.25,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.2133,
       "step": 84100
     },
     {
       "epoch": 1.3665090737762182,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.2137,
       "step": 84110
     },
     {
       "epoch": 1.3666715406735879,
-      "grad_norm": 2.875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.1791,
       "step": 84120
     },
     {
       "epoch": 1.3668340075709575,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2566,
       "step": 84130
     },
     {
       "epoch": 1.3669964744683272,
-      "grad_norm": 2.34375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.1801,
       "step": 84140
     },
     {
       "epoch": 1.3671589413656968,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2333,
       "step": 84150
     },
     {
       "epoch": 1.3673214082630665,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2823,
       "step": 84160
     },
     {
       "epoch": 1.3674838751604361,
-      "grad_norm": 2.875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.1631,
       "step": 84170
     },
     {
       "epoch": 1.3676463420578058,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2403,
       "step": 84180
     },
     {
       "epoch": 1.3678088089551754,
-      "grad_norm": 2.5,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2561,
       "step": 84190
     },
     {
       "epoch": 1.367971275852545,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2289,
       "step": 84200
     },
     {
       "epoch": 1.3681337427499147,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2179,
       "step": 84210
     },
     {
       "epoch": 1.3682962096472844,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.1696,
       "step": 84220
     },
     {
       "epoch": 1.368458676544654,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3125,
+      "loss": 1.193,
       "step": 84230
     },
     {
       "epoch": 1.3686211434420237,
-      "grad_norm": 2.75,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.1897,
       "step": 84240
     },
     {
       "epoch": 1.3687836103393933,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2197,
       "step": 84250
     },
     {
       "epoch": 1.368946077236763,
-      "grad_norm": 3.5,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2095,
       "step": 84260
     },
     {
       "epoch": 1.3691085441341326,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2308,
       "step": 84270
     },
     {
       "epoch": 1.3692710110315023,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.1688,
       "step": 84280
     },
     {
       "epoch": 1.369433477928872,
-      "grad_norm": 3.65625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.1862,
       "step": 84290
     },
     {
       "epoch": 1.3695959448262416,
-      "grad_norm": 2.5,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2149,
       "step": 84300
     },
     {
       "epoch": 1.3697584117236112,
-      "grad_norm": 3.0,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.1987,
       "step": 84310
     },
     {
       "epoch": 1.3699208786209809,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2144,
       "step": 84320
     },
     {
       "epoch": 1.3700833455183505,
-      "grad_norm": 3.578125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.218,
       "step": 84330
     },
     {
       "epoch": 1.3702458124157202,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3223,
+      "loss": 1.2542,
       "step": 84340
     },
     {
       "epoch": 1.3704082793130898,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2117,
       "step": 84350
     },
     {
       "epoch": 1.3705707462104595,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.25,
       "step": 84360
     },
     {
       "epoch": 1.3707332131078294,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2322,
       "step": 84370
     },
     {
       "epoch": 1.370895680005199,
-      "grad_norm": 2.25,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3205,
+      "loss": 1.204,
       "step": 84380
     },
     {
       "epoch": 1.3710581469025687,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.1925,
       "step": 84390
     },
     {
       "epoch": 1.3712206137999383,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3719,
+      "loss": 1.2251,
       "step": 84400
     },
     {
       "epoch": 1.371383080697308,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2105,
       "step": 84410
     },
     {
       "epoch": 1.3715455475946776,
-      "grad_norm": 3.03125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1795,
       "step": 84420
     },
     {
       "epoch": 1.3717080144920473,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3157,
+      "loss": 1.1684,
       "step": 84430
     },
     {
       "epoch": 1.371870481389417,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2098,
       "step": 84440
     },
     {
       "epoch": 1.3720329482867866,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1855,
       "step": 84450
     },
     {
       "epoch": 1.3721954151841562,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.2309,
       "step": 84460
     },
     {
       "epoch": 1.3723578820815259,
-      "grad_norm": 3.28125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2106,
       "step": 84470
     },
     {
       "epoch": 1.3725203489788955,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2489,
       "step": 84480
     },
     {
       "epoch": 1.3726828158762652,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2621,
       "step": 84490
     },
     {
       "epoch": 1.3728452827736348,
-      "grad_norm": 2.75,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.2491,
       "step": 84500
     },
     {
       "epoch": 1.3730077496710045,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2163,
       "step": 84510
     },
     {
       "epoch": 1.3731702165683741,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2113,
       "step": 84520
     },
     {
       "epoch": 1.3733326834657438,
-      "grad_norm": 2.75,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1857,
       "step": 84530
     },
     {
       "epoch": 1.3734951503631136,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2279,
       "step": 84540
     },
     {
       "epoch": 1.3736576172604833,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2501,
       "step": 84550
     },
     {
       "epoch": 1.373820084157853,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.1915,
       "step": 84560
     },
     {
       "epoch": 1.3739825510552226,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.2222,
       "step": 84570
     },
     {
       "epoch": 1.3741450179525923,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2227,
       "step": 84580
     },
     {
       "epoch": 1.374307484849962,
-      "grad_norm": 2.078125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2366,
       "step": 84590
     },
     {
       "epoch": 1.3744699517473316,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3698,
+      "loss": 1.2119,
       "step": 84600
     },
     {
       "epoch": 1.3746324186447012,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2336,
       "step": 84610
     },
     {
       "epoch": 1.3747948855420709,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3721,
+      "loss": 1.2546,
       "step": 84620
     },
     {
       "epoch": 1.3749573524394405,
-      "grad_norm": 2.625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.227,
       "step": 84630
     },
     {
       "epoch": 1.3751198193368102,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.1671,
       "step": 84640
     },
     {
       "epoch": 1.3752822862341798,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2147,
       "step": 84650
     },
     {
       "epoch": 1.3754447531315495,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2743,
       "step": 84660
     },
     {
       "epoch": 1.375607220028919,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2516,
       "step": 84670
     },
     {
       "epoch": 1.3757696869262888,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2233,
       "step": 84680
     },
     {
       "epoch": 1.3759321538236584,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2286,
       "step": 84690
     },
     {
       "epoch": 1.376094620721028,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2164,
       "step": 84700
     },
     {
       "epoch": 1.3762570876183977,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.2094,
       "step": 84710
     },
     {
       "epoch": 1.3764195545157674,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.1782,
       "step": 84720
     },
     {
       "epoch": 1.376582021413137,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.217,
       "step": 84730
     },
     {
       "epoch": 1.3767444883105067,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2043,
       "step": 84740
     },
     {
       "epoch": 1.3769069552078763,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.2245,
       "step": 84750
     },
     {
       "epoch": 1.377069422105246,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2055,
       "step": 84760
     },
     {
       "epoch": 1.3772318890026156,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.223,
       "step": 84770
     },
     {
       "epoch": 1.3773943558999853,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2314,
       "step": 84780
     },
     {
       "epoch": 1.377556822797355,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2233,
       "step": 84790
     },
     {
       "epoch": 1.3777192896947246,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2288,
       "step": 84800
     },
     {
       "epoch": 1.3778817565920944,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2558,
       "step": 84810
     },
     {
       "epoch": 1.378044223489464,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.205,
       "step": 84820
     },
     {
       "epoch": 1.3782066903868337,
-      "grad_norm": 3.546875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.1994,
       "step": 84830
     },
     {
       "epoch": 1.3783691572842034,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1995,
       "step": 84840
     },
     {
       "epoch": 1.378531624181573,
-      "grad_norm": 2.875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3176,
+      "loss": 1.2167,
       "step": 84850
     },
     {
       "epoch": 1.3786940910789427,
-      "grad_norm": 2.375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3271,
+      "loss": 1.2407,
       "step": 84860
     },
     {
       "epoch": 1.3788565579763123,
-      "grad_norm": 2.5,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.1958,
       "step": 84870
     },
     {
       "epoch": 1.379019024873682,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.1882,
       "step": 84880
     },
     {
       "epoch": 1.3791814917710516,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2182,
       "step": 84890
     },
     {
       "epoch": 1.3793439586684213,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2045,
       "step": 84900
     },
     {
       "epoch": 1.379506425565791,
-      "grad_norm": 2.625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2557,
       "step": 84910
     },
     {
       "epoch": 1.3796688924631606,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2452,
       "step": 84920
     },
     {
       "epoch": 1.3798313593605303,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2358,
       "step": 84930
     },
     {
       "epoch": 1.3799938262579,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2412,
       "step": 84940
     },
     {
       "epoch": 1.3801562931552696,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2418,
       "step": 84950
     },
     {
       "epoch": 1.3803187600526392,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2,
       "step": 84960
     },
     {
       "epoch": 1.3804812269500089,
-      "grad_norm": 3.375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2568,
       "step": 84970
     },
     {
       "epoch": 1.3806436938473787,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2568,
       "step": 84980
     },
     {
       "epoch": 1.3808061607447484,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2245,
       "step": 84990
     },
     {
       "epoch": 1.380968627642118,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2213,
       "step": 85000
     },
     {
       "epoch": 1.3811310945394877,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2067,
       "step": 85010
     },
     {
       "epoch": 1.3812935614368573,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2146,
       "step": 85020
     },
     {
       "epoch": 1.381456028334227,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2412,
       "step": 85030
     },
     {
       "epoch": 1.3816184952315966,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2283,
       "step": 85040
     },
     {
       "epoch": 1.3817809621289663,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2299,
       "step": 85050
     },
     {
       "epoch": 1.381943429026336,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3616,
+      "loss": 1.1621,
       "step": 85060
     },
     {
       "epoch": 1.3821058959237056,
-      "grad_norm": 1.6640625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.2433,
       "step": 85070
     },
     {
       "epoch": 1.3822683628210752,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2252,
       "step": 85080
     },
     {
       "epoch": 1.3824308297184449,
-      "grad_norm": 3.34375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.2205,
       "step": 85090
     },
     {
       "epoch": 1.3825932966158145,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3589,
+      "loss": 1.1977,
       "step": 85100
     },
     {
       "epoch": 1.3827557635131842,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.237,
       "step": 85110
     },
     {
       "epoch": 1.3829182304105538,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2303,
       "step": 85120
     },
     {
       "epoch": 1.3830806973079235,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2553,
       "step": 85130
     },
     {
       "epoch": 1.3832431642052931,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.2223,
       "step": 85140
     },
     {
       "epoch": 1.3834056311026628,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3107,
+      "loss": 1.1813,
       "step": 85150
     },
     {
       "epoch": 1.3835680980000324,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.2178,
       "step": 85160
     },
     {
       "epoch": 1.383730564897402,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2437,
       "step": 85170
     },
     {
       "epoch": 1.3838930317947717,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2361,
       "step": 85180
     },
     {
       "epoch": 1.3840554986921414,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2167,
       "step": 85190
     },
     {
       "epoch": 1.384217965589511,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.1636,
       "step": 85200
     },
     {
       "epoch": 1.3843804324868807,
-      "grad_norm": 4.375,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2351,
       "step": 85210
     },
     {
       "epoch": 1.3845428993842503,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2267,
       "step": 85220
     },
     {
       "epoch": 1.38470536628162,
-      "grad_norm": 2.875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2609,
       "step": 85230
     },
     {
       "epoch": 1.3848678331789896,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.2372,
       "step": 85240
     },
     {
       "epoch": 1.3850303000763595,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2099,
       "step": 85250
     },
     {
       "epoch": 1.3851927669737292,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.197,
       "step": 85260
     },
     {
       "epoch": 1.3853552338710988,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2356,
       "step": 85270
     },
     {
       "epoch": 1.3855177007684685,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.2145,
       "step": 85280
     },
     {
       "epoch": 1.3856801676658381,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.2085,
       "step": 85290
     },
     {
       "epoch": 1.3858426345632078,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2208,
       "step": 85300
     },
     {
       "epoch": 1.3860051014605774,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2272,
       "step": 85310
     },
     {
       "epoch": 1.386167568357947,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2021,
       "step": 85320
     },
     {
       "epoch": 1.3863300352553167,
-      "grad_norm": 2.25,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.2085,
       "step": 85330
     },
     {
       "epoch": 1.3864925021526864,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3645,
+      "loss": 1.224,
       "step": 85340
     },
     {
       "epoch": 1.386654969050056,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2293,
       "step": 85350
     },
     {
       "epoch": 1.3868174359474257,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2049,
       "step": 85360
     },
     {
       "epoch": 1.3869799028447953,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2129,
       "step": 85370
     },
     {
       "epoch": 1.387142369742165,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2124,
       "step": 85380
     },
     {
       "epoch": 1.3873048366395346,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2275,
       "step": 85390
     },
     {
       "epoch": 1.3874673035369043,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.224,
       "step": 85400
     },
     {
       "epoch": 1.387629770434274,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2094,
       "step": 85410
     },
     {
       "epoch": 1.3877922373316438,
-      "grad_norm": 4.34375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3683,
+      "loss": 1.2172,
       "step": 85420
     },
     {
       "epoch": 1.3879547042290135,
-      "grad_norm": 1.9921875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.1827,
       "step": 85430
     },
     {
       "epoch": 1.388117171126383,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2043,
       "step": 85440
     },
     {
       "epoch": 1.3882796380237528,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3078,
+      "loss": 1.2189,
       "step": 85450
     },
     {
       "epoch": 1.3884421049211224,
-      "grad_norm": 3.0,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.1735,
       "step": 85460
     },
     {
       "epoch": 1.388604571818492,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3634,
+      "loss": 1.2266,
       "step": 85470
     },
     {
       "epoch": 1.3887670387158617,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2337,
       "step": 85480
     },
     {
       "epoch": 1.3889295056132314,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3199,
+      "loss": 1.2592,
       "step": 85490
     },
     {
       "epoch": 1.389091972510601,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2233,
       "step": 85500
     },
     {
       "epoch": 1.3892544394079707,
-      "grad_norm": 2.625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1961,
       "step": 85510
     },
     {
       "epoch": 1.3894169063053403,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1817,
       "step": 85520
     },
     {
       "epoch": 1.38957937320271,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1796,
       "step": 85530
     },
     {
       "epoch": 1.3897418401000796,
-      "grad_norm": 2.125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2271,
       "step": 85540
     },
     {
       "epoch": 1.3899043069974493,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2259,
       "step": 85550
     },
     {
       "epoch": 1.390066773894819,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3123,
+      "loss": 1.2042,
       "step": 85560
     },
     {
       "epoch": 1.3902292407921886,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.319,
+      "loss": 1.2228,
       "step": 85570
     },
     {
       "epoch": 1.3903917076895582,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.2376,
       "step": 85580
     },
     {
       "epoch": 1.3905541745869279,
-      "grad_norm": 2.296875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3708,
+      "loss": 1.2276,
       "step": 85590
     },
     {
       "epoch": 1.3907166414842975,
-      "grad_norm": 1.671875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.2132,
       "step": 85600
     },
     {
       "epoch": 1.3908791083816672,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2289,
       "step": 85610
     },
     {
       "epoch": 1.3910415752790368,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2327,
       "step": 85620
     },
     {
       "epoch": 1.3912040421764065,
-      "grad_norm": 3.25,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.1778,
       "step": 85630
     },
     {
       "epoch": 1.3913665090737761,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2356,
       "step": 85640
     },
     {
       "epoch": 1.3915289759711458,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2529,
       "step": 85650
     },
     {
       "epoch": 1.3916914428685154,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2305,
       "step": 85660
     },
     {
       "epoch": 1.391853909765885,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2252,
       "step": 85670
     },
     {
       "epoch": 1.3920163766632547,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2162,
       "step": 85680
     },
     {
       "epoch": 1.3921788435606246,
-      "grad_norm": 3.5,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.2608,
       "step": 85690
     },
     {
       "epoch": 1.3923413104579943,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.1983,
       "step": 85700
     },
     {
       "epoch": 1.392503777355364,
-      "grad_norm": 3.125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2034,
       "step": 85710
     },
     {
       "epoch": 1.3926662442527336,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2041,
       "step": 85720
     },
     {
       "epoch": 1.3928287111501032,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2063,
       "step": 85730
     },
     {
       "epoch": 1.3929911780474729,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2075,
       "step": 85740
     },
     {
       "epoch": 1.3931536449448425,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.2466,
       "step": 85750
     },
     {
       "epoch": 1.3933161118422122,
-      "grad_norm": 3.125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3748,
+      "loss": 1.2541,
       "step": 85760
     },
     {
       "epoch": 1.3934785787395818,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.217,
       "step": 85770
     },
     {
       "epoch": 1.3936410456369515,
-      "grad_norm": 2.5,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3235,
+      "loss": 1.211,
       "step": 85780
     },
     {
       "epoch": 1.393803512534321,
-      "grad_norm": 2.75,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.1843,
       "step": 85790
     },
     {
       "epoch": 1.3939659794316908,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2156,
       "step": 85800
     },
     {
       "epoch": 1.3941284463290604,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3729,
+      "loss": 1.197,
       "step": 85810
     },
     {
       "epoch": 1.39429091322643,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2213,
       "step": 85820
     },
     {
       "epoch": 1.3944533801237997,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2226,
       "step": 85830
     },
     {
       "epoch": 1.3946158470211694,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2287,
       "step": 85840
     },
     {
       "epoch": 1.394778313918539,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.1988,
       "step": 85850
     },
     {
       "epoch": 1.3949407808159089,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2059,
       "step": 85860
     },
     {
       "epoch": 1.3951032477132785,
-      "grad_norm": 2.28125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.3091,
+      "loss": 1.2244,
       "step": 85870
     },
     {
       "epoch": 1.3952657146106482,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2151,
       "step": 85880
     },
     {
       "epoch": 1.3954281815080178,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.2169,
       "step": 85890
     },
     {
       "epoch": 1.3955906484053875,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.362,
+      "loss": 1.1699,
       "step": 85900
     },
     {
       "epoch": 1.3957531153027571,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.1969,
       "step": 85910
     },
     {
       "epoch": 1.3959155822001268,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.1998,
       "step": 85920
     },
     {
       "epoch": 1.3960780490974964,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.1731,
       "step": 85930
     },
     {
       "epoch": 1.396240515994866,
-      "grad_norm": 2.375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2017,
       "step": 85940
     },
     {
       "epoch": 1.3964029828922357,
-      "grad_norm": 1.9375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.1997,
       "step": 85950
     },
     {
       "epoch": 1.3965654497896054,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2587,
       "step": 85960
     },
     {
       "epoch": 1.396727916686975,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1932,
       "step": 85970
     },
     {
       "epoch": 1.3968903835843447,
-      "grad_norm": 4.03125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.212,
       "step": 85980
     },
     {
       "epoch": 1.3970528504817143,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2213,
       "step": 85990
     },
     {
       "epoch": 1.397215317379084,
-      "grad_norm": 2.625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2531,
       "step": 86000
     },
     {
       "epoch": 1.3973777842764536,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2412,
       "step": 86010
     },
     {
       "epoch": 1.3975402511738233,
-      "grad_norm": 3.234375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2307,
       "step": 86020
     },
     {
       "epoch": 1.397702718071193,
-      "grad_norm": 4.75,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3672,
+      "loss": 1.2317,
       "step": 86030
     },
     {
       "epoch": 1.3978651849685626,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2112,
       "step": 86040
     },
     {
       "epoch": 1.3980276518659323,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2287,
       "step": 86050
     },
     {
       "epoch": 1.398190118763302,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1971,
       "step": 86060
     },
     {
       "epoch": 1.3983525856606716,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.1959,
       "step": 86070
     },
     {
       "epoch": 1.3985150525580412,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.253,
       "step": 86080
     },
     {
       "epoch": 1.3986775194554109,
-      "grad_norm": 3.25,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.221,
       "step": 86090
     },
     {
       "epoch": 1.3988399863527805,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.32,
+      "loss": 1.2273,
       "step": 86100
     },
     {
       "epoch": 1.3990024532501502,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.2503,
       "step": 86110
     },
     {
       "epoch": 1.3991649201475198,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2441,
       "step": 86120
     },
     {
       "epoch": 1.3993273870448897,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2313,
       "step": 86130
     },
     {
       "epoch": 1.3994898539422593,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2101,
       "step": 86140
     },
     {
       "epoch": 1.399652320839629,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2132,
       "step": 86150
     },
     {
       "epoch": 1.3998147877369986,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2065,
       "step": 86160
     },
     {
       "epoch": 1.3999772546343683,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.1978,
       "step": 86170
     },
     {
       "epoch": 1.400139721531738,
-      "grad_norm": 2.75,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.272,
       "step": 86180
     },
     {
       "epoch": 1.4003021884291076,
-      "grad_norm": 2.25,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2432,
       "step": 86190
     },
     {
       "epoch": 1.4004646553264772,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2195,
       "step": 86200
     },
     {
       "epoch": 1.4006271222238469,
-      "grad_norm": 2.25,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2485,
       "step": 86210
     },
     {
       "epoch": 1.4007895891212165,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2582,
       "step": 86220
     },
     {
       "epoch": 1.4009520560185862,
-      "grad_norm": 3.375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2058,
       "step": 86230
     },
     {
       "epoch": 1.4011145229159558,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.1892,
       "step": 86240
     },
     {
       "epoch": 1.4012769898133255,
-      "grad_norm": 5.25,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2388,
       "step": 86250
     },
     {
       "epoch": 1.4014394567106951,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2227,
       "step": 86260
     },
     {
       "epoch": 1.4016019236080648,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2238,
       "step": 86270
     },
     {
       "epoch": 1.4017643905054344,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2153,
       "step": 86280
     },
     {
       "epoch": 1.4019268574028043,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2267,
       "step": 86290
     },
     {
       "epoch": 1.402089324300174,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.1783,
       "step": 86300
     },
     {
       "epoch": 1.4022517911975436,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2165,
       "step": 86310
     },
     {
       "epoch": 1.4024142580949133,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2837,
       "step": 86320
     },
     {
       "epoch": 1.402576724992283,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2371,
       "step": 86330
     },
     {
       "epoch": 1.4027391918896526,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3702,
+      "loss": 1.1896,
       "step": 86340
     },
     {
       "epoch": 1.4029016587870222,
-      "grad_norm": 2.25,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.1658,
       "step": 86350
     },
     {
       "epoch": 1.4030641256843919,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.2154,
       "step": 86360
     },
     {
       "epoch": 1.4032265925817615,
-      "grad_norm": 3.5,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2641,
       "step": 86370
     },
     {
       "epoch": 1.4033890594791312,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.175,
       "step": 86380
     },
     {
       "epoch": 1.4035515263765008,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.222,
       "step": 86390
     },
     {
       "epoch": 1.4037139932738705,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2536,
       "step": 86400
     },
     {
       "epoch": 1.4038764601712401,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.1999,
       "step": 86410
     },
     {
       "epoch": 1.4040389270686098,
-      "grad_norm": 3.421875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.194,
       "step": 86420
     },
     {
       "epoch": 1.4042013939659794,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3164,
+      "loss": 1.1751,
       "step": 86430
     },
     {
       "epoch": 1.404363860863349,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2441,
       "step": 86440
     },
     {
       "epoch": 1.4045263277607187,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.1922,
       "step": 86450
     },
     {
       "epoch": 1.4046887946580884,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1992,
       "step": 86460
     },
     {
       "epoch": 1.404851261555458,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2646,
       "step": 86470
     },
     {
       "epoch": 1.4050137284528277,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2097,
       "step": 86480
     },
     {
       "epoch": 1.4051761953501973,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2297,
       "step": 86490
     },
     {
       "epoch": 1.405338662247567,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.258,
       "step": 86500
     },
     {
       "epoch": 1.4055011291449366,
-      "grad_norm": 3.0,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2544,
       "step": 86510
     },
     {
       "epoch": 1.4056635960423063,
-      "grad_norm": 3.0,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2059,
       "step": 86520
     },
     {
       "epoch": 1.405826062939676,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2027,
       "step": 86530
     },
     {
       "epoch": 1.4059885298370456,
-      "grad_norm": 2.125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.1971,
       "step": 86540
     },
     {
       "epoch": 1.4061509967344152,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.1878,
       "step": 86550
     },
     {
       "epoch": 1.4063134636317849,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.1951,
       "step": 86560
     },
     {
       "epoch": 1.4064759305291548,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2464,
       "step": 86570
     },
     {
       "epoch": 1.4066383974265244,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2143,
       "step": 86580
     },
     {
       "epoch": 1.406800864323894,
-      "grad_norm": 2.34375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.221,
       "step": 86590
     },
     {
       "epoch": 1.4069633312212637,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2174,
       "step": 86600
     },
     {
       "epoch": 1.4071257981186334,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2325,
       "step": 86610
     },
     {
       "epoch": 1.407288265016003,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.224,
       "step": 86620
     },
     {
       "epoch": 1.4074507319133727,
-      "grad_norm": 3.765625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2059,
       "step": 86630
     },
     {
       "epoch": 1.4076131988107423,
-      "grad_norm": 5.375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1941,
       "step": 86640
     },
     {
       "epoch": 1.407775665708112,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2493,
       "step": 86650
     },
     {
       "epoch": 1.4079381326054816,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.1934,
       "step": 86660
     },
     {
       "epoch": 1.4081005995028513,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2094,
       "step": 86670
     },
     {
       "epoch": 1.408263066400221,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2658,
       "step": 86680
     },
     {
       "epoch": 1.4084255332975906,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.1878,
       "step": 86690
     },
     {
       "epoch": 1.4085880001949602,
-      "grad_norm": 2.421875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2182,
       "step": 86700
     },
     {
       "epoch": 1.4087504670923299,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.1934,
       "step": 86710
     },
     {
       "epoch": 1.4089129339896995,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.1597,
       "step": 86720
     },
     {
       "epoch": 1.4090754008870694,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2218,
       "step": 86730
     },
     {
       "epoch": 1.409237867784439,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2042,
       "step": 86740
     },
     {
       "epoch": 1.4094003346818087,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2395,
       "step": 86750
     },
     {
       "epoch": 1.4095628015791783,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2491,
       "step": 86760
     },
     {
       "epoch": 1.409725268476548,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2499,
       "step": 86770
     },
     {
       "epoch": 1.4098877353739176,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.2149,
       "step": 86780
     },
     {
       "epoch": 1.4100502022712873,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2176,
       "step": 86790
     },
     {
       "epoch": 1.410212669168657,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.1929,
       "step": 86800
     },
     {
       "epoch": 1.4103751360660266,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.1927,
       "step": 86810
     },
     {
       "epoch": 1.4105376029633963,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3129,
+      "loss": 1.2038,
       "step": 86820
     },
     {
       "epoch": 1.410700069860766,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2172,
       "step": 86830
     },
     {
       "epoch": 1.4108625367581356,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2344,
       "step": 86840
     },
     {
       "epoch": 1.4110250036555052,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2038,
       "step": 86850
     },
     {
       "epoch": 1.4111874705528749,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2381,
       "step": 86860
     },
     {
       "epoch": 1.4113499374502445,
-      "grad_norm": 3.25,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2576,
       "step": 86870
     },
     {
       "epoch": 1.4115124043476142,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2176,
       "step": 86880
     },
     {
       "epoch": 1.4116748712449838,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3587,
+      "loss": 1.271,
       "step": 86890
     },
     {
       "epoch": 1.4118373381423535,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2187,
       "step": 86900
     },
     {
       "epoch": 1.411999805039723,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2334,
       "step": 86910
     },
     {
       "epoch": 1.4121622719370928,
-      "grad_norm": 3.25,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2521,
       "step": 86920
     },
     {
       "epoch": 1.4123247388344624,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2587,
       "step": 86930
     },
     {
       "epoch": 1.412487205731832,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2253,
       "step": 86940
     },
     {
       "epoch": 1.4126496726292017,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2101,
       "step": 86950
     },
     {
       "epoch": 1.4128121395265714,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3178,
+      "loss": 1.2121,
       "step": 86960
     },
     {
       "epoch": 1.412974606423941,
-      "grad_norm": 3.0,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.1828,
       "step": 86970
     },
     {
       "epoch": 1.4131370733213107,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2109,
       "step": 86980
     },
     {
       "epoch": 1.4132995402186803,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2182,
       "step": 86990
     },
     {
       "epoch": 1.41346200711605,
-      "grad_norm": 1.90625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2394,
       "step": 87000
     },
     {
       "epoch": 1.4136244740134198,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.2539,
       "step": 87010
     },
     {
       "epoch": 1.4137869409107895,
-      "grad_norm": 3.65625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2078,
       "step": 87020
     },
     {
       "epoch": 1.4139494078081591,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2324,
       "step": 87030
     },
     {
       "epoch": 1.4141118747055288,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3697,
+      "loss": 1.1937,
       "step": 87040
     },
     {
       "epoch": 1.4142743416028984,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2311,
       "step": 87050
     },
     {
       "epoch": 1.414436808500268,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2122,
       "step": 87060
     },
     {
       "epoch": 1.4145992753976377,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2298,
       "step": 87070
     },
     {
       "epoch": 1.4147617422950074,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2293,
       "step": 87080
     },
     {
       "epoch": 1.414924209192377,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2638,
       "step": 87090
     },
     {
       "epoch": 1.4150866760897467,
-      "grad_norm": 3.25,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2012,
       "step": 87100
     },
     {
       "epoch": 1.4152491429871163,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2413,
       "step": 87110
     },
     {
       "epoch": 1.415411609884486,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2309,
       "step": 87120
     },
     {
       "epoch": 1.4155740767818557,
-      "grad_norm": 3.6875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2095,
       "step": 87130
     },
     {
       "epoch": 1.4157365436792253,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2079,
       "step": 87140
     },
     {
       "epoch": 1.415899010576595,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2128,
       "step": 87150
     },
     {
       "epoch": 1.4160614774739646,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.2478,
       "step": 87160
     },
     {
       "epoch": 1.4162239443713345,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.215,
       "step": 87170
     },
     {
       "epoch": 1.4163864112687041,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2455,
       "step": 87180
     },
     {
       "epoch": 1.4165488781660738,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2329,
       "step": 87190
     },
     {
       "epoch": 1.4167113450634434,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.2163,
       "step": 87200
     },
     {
       "epoch": 1.416873811960813,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2205,
       "step": 87210
     },
     {
       "epoch": 1.4170362788581827,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2532,
       "step": 87220
     },
     {
       "epoch": 1.4171987457555524,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.2368,
       "step": 87230
     },
     {
       "epoch": 1.417361212652922,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2179,
       "step": 87240
     },
     {
       "epoch": 1.4175236795502917,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1927,
       "step": 87250
     },
     {
       "epoch": 1.4176861464476613,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.311,
+      "loss": 1.2315,
       "step": 87260
     },
     {
       "epoch": 1.417848613345031,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2427,
       "step": 87270
     },
     {
       "epoch": 1.4180110802424006,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.1981,
       "step": 87280
     },
     {
       "epoch": 1.4181735471397703,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2079,
       "step": 87290
     },
     {
       "epoch": 1.41833601403714,
-      "grad_norm": 2.875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.213,
       "step": 87300
     },
     {
       "epoch": 1.4184984809345096,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.2073,
       "step": 87310
     },
     {
       "epoch": 1.4186609478318792,
-      "grad_norm": 3.0,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.22,
       "step": 87320
     },
     {
       "epoch": 1.4188234147292489,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.1894,
       "step": 87330
     },
     {
       "epoch": 1.4189858816266185,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.1938,
       "step": 87340
     },
     {
       "epoch": 1.4191483485239882,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2074,
       "step": 87350
     },
     {
       "epoch": 1.4193108154213578,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2333,
       "step": 87360
     },
     {
       "epoch": 1.4194732823187275,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2318,
       "step": 87370
     },
     {
       "epoch": 1.4196357492160971,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3646,
+      "loss": 1.2065,
       "step": 87380
     },
     {
       "epoch": 1.4197982161134668,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2347,
       "step": 87390
     },
     {
       "epoch": 1.4199606830108364,
-      "grad_norm": 2.75,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2215,
       "step": 87400
     },
     {
       "epoch": 1.420123149908206,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2002,
       "step": 87410
     },
     {
       "epoch": 1.4202856168055757,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2233,
       "step": 87420
     },
     {
       "epoch": 1.4204480837029454,
-      "grad_norm": 1.734375,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2091,
       "step": 87430
     },
     {
       "epoch": 1.420610550600315,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2394,
       "step": 87440
     },
     {
       "epoch": 1.420773017497685,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3574,
+      "loss": 1.2275,
       "step": 87450
     },
     {
       "epoch": 1.4209354843950546,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.1971,
       "step": 87460
     },
     {
       "epoch": 1.4210979512924242,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2019,
       "step": 87470
     },
     {
       "epoch": 1.4212604181897939,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2088,
       "step": 87480
     },
     {
       "epoch": 1.4214228850871635,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2282,
       "step": 87490
     },
     {
       "epoch": 1.4215853519845332,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3154,
+      "loss": 1.2577,
       "step": 87500
     },
     {
       "epoch": 1.4217478188819028,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.23,
       "step": 87510
     },
     {
       "epoch": 1.4219102857792725,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.2244,
       "step": 87520
     },
     {
       "epoch": 1.4220727526766421,
-      "grad_norm": 3.0,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.233,
       "step": 87530
     },
     {
       "epoch": 1.4222352195740118,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.1636,
       "step": 87540
     },
     {
       "epoch": 1.4223976864713814,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1972,
       "step": 87550
     },
     {
       "epoch": 1.422560153368751,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2124,
       "step": 87560
     },
     {
       "epoch": 1.4227226202661207,
-      "grad_norm": 4.65625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.2205,
       "step": 87570
     },
     {
       "epoch": 1.4228850871634904,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2074,
       "step": 87580
     },
     {
       "epoch": 1.42304755406086,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2138,
       "step": 87590
     },
     {
       "epoch": 1.4232100209582297,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1911,
       "step": 87600
     },
     {
       "epoch": 1.4233724878555996,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2196,
       "step": 87610
     },
     {
       "epoch": 1.4235349547529692,
-      "grad_norm": 2.5,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2466,
       "step": 87620
     },
     {
       "epoch": 1.4236974216503389,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.2311,
       "step": 87630
     },
     {
       "epoch": 1.4238598885477085,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2504,
       "step": 87640
     },
     {
       "epoch": 1.4240223554450782,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2136,
       "step": 87650
     },
     {
       "epoch": 1.4241848223424478,
-      "grad_norm": 2.5,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.205,
       "step": 87660
     },
     {
       "epoch": 1.4243472892398175,
-      "grad_norm": 1.984375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2237,
       "step": 87670
     },
     {
       "epoch": 1.424509756137187,
-      "grad_norm": 3.734375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2236,
       "step": 87680
     },
     {
       "epoch": 1.4246722230345568,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2474,
       "step": 87690
     },
     {
       "epoch": 1.4248346899319264,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.239,
       "step": 87700
     },
     {
       "epoch": 1.424997156829296,
-      "grad_norm": 3.25,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.1861,
       "step": 87710
     },
     {
       "epoch": 1.4251596237266657,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2185,
       "step": 87720
     },
     {
       "epoch": 1.4253220906240354,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3667,
+      "loss": 1.2142,
       "step": 87730
     },
     {
       "epoch": 1.425484557521405,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2097,
       "step": 87740
     },
     {
       "epoch": 1.4256470244187747,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.1946,
       "step": 87750
     },
     {
       "epoch": 1.4258094913161443,
-      "grad_norm": 2.375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2116,
       "step": 87760
     },
     {
       "epoch": 1.425971958213514,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.224,
       "step": 87770
     },
     {
       "epoch": 1.4261344251108836,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2017,
       "step": 87780
     },
     {
       "epoch": 1.4262968920082533,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2057,
       "step": 87790
     },
     {
       "epoch": 1.426459358905623,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.367,
+      "loss": 1.2444,
       "step": 87800
     },
     {
       "epoch": 1.4266218258029926,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2337,
       "step": 87810
     },
     {
       "epoch": 1.4267842927003622,
-      "grad_norm": 3.625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.233,
       "step": 87820
     },
     {
       "epoch": 1.4269467595977319,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.23,
       "step": 87830
     },
     {
       "epoch": 1.4271092264951015,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3613,
+      "loss": 1.2184,
       "step": 87840
     },
     {
       "epoch": 1.4272716933924712,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2074,
       "step": 87850
     },
     {
       "epoch": 1.4274341602898408,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2306,
       "step": 87860
     },
     {
       "epoch": 1.4275966271872105,
-      "grad_norm": 1.84375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.1766,
       "step": 87870
     },
     {
       "epoch": 1.4277590940845803,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2415,
       "step": 87880
     },
     {
       "epoch": 1.42792156098195,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1982,
       "step": 87890
     },
     {
       "epoch": 1.4280840278793197,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3626,
+      "loss": 1.2085,
       "step": 87900
     },
     {
       "epoch": 1.4282464947766893,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.2351,
       "step": 87910
     },
     {
       "epoch": 1.428408961674059,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.1983,
       "step": 87920
     },
     {
       "epoch": 1.4285714285714286,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.207,
       "step": 87930
     },
     {
       "epoch": 1.4287338954687983,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.1967,
       "step": 87940
     },
     {
       "epoch": 1.428896362366168,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3137,
+      "loss": 1.1861,
       "step": 87950
     },
     {
       "epoch": 1.4290588292635376,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2025,
       "step": 87960
     },
     {
       "epoch": 1.4292212961609072,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.2026,
       "step": 87970
     },
     {
       "epoch": 1.4293837630582769,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3146,
+      "loss": 1.1866,
       "step": 87980
     },
     {
       "epoch": 1.4295462299556465,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.1941,
       "step": 87990
     },
     {
       "epoch": 1.4297086968530162,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2243,
       "step": 88000
     },
     {
       "epoch": 1.4298711637503858,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2067,
       "step": 88010
     },
     {
       "epoch": 1.4300336306477555,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.1846,
       "step": 88020
     },
     {
       "epoch": 1.4301960975451251,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.3003,
       "step": 88030
     },
     {
       "epoch": 1.4303585644424948,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.1756,
       "step": 88040
     },
     {
       "epoch": 1.4305210313398646,
-      "grad_norm": 3.375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1945,
       "step": 88050
     },
     {
       "epoch": 1.4306834982372343,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2428,
       "step": 88060
     },
     {
       "epoch": 1.430845965134604,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2035,
       "step": 88070
     },
     {
       "epoch": 1.4310084320319736,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.21,
       "step": 88080
     },
     {
       "epoch": 1.4311708989293432,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2249,
       "step": 88090
     },
     {
       "epoch": 1.4313333658267129,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2143,
       "step": 88100
     },
     {
       "epoch": 1.4314958327240825,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2561,
       "step": 88110
     },
     {
       "epoch": 1.4316582996214522,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2499,
       "step": 88120
     },
     {
       "epoch": 1.4318207665188218,
-      "grad_norm": 3.0,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3601,
+      "loss": 1.2033,
       "step": 88130
     },
     {
       "epoch": 1.4319832334161915,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2412,
       "step": 88140
     },
     {
       "epoch": 1.4321457003135611,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2436,
       "step": 88150
     },
     {
       "epoch": 1.4323081672109308,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.1803,
       "step": 88160
     },
     {
       "epoch": 1.4324706341083004,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2055,
       "step": 88170
     },
     {
       "epoch": 1.43263310100567,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.1634,
       "step": 88180
     },
     {
       "epoch": 1.4327955679030397,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2431,
       "step": 88190
     },
     {
       "epoch": 1.4329580348004094,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2033,
       "step": 88200
     },
     {
       "epoch": 1.433120501697779,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2373,
       "step": 88210
     },
     {
       "epoch": 1.4332829685951487,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2048,
       "step": 88220
     },
     {
       "epoch": 1.4334454354925183,
-      "grad_norm": 2.5,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2077,
       "step": 88230
     },
     {
       "epoch": 1.433607902389888,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2487,
       "step": 88240
     },
     {
       "epoch": 1.4337703692872577,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.1528,
       "step": 88250
     },
     {
       "epoch": 1.4339328361846273,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2206,
       "step": 88260
     },
     {
       "epoch": 1.434095303081997,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3589,
+      "loss": 1.2145,
       "step": 88270
     },
     {
       "epoch": 1.4342577699793666,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2268,
       "step": 88280
     },
     {
       "epoch": 1.4344202368767363,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.1962,
       "step": 88290
     },
     {
       "epoch": 1.434582703774106,
-      "grad_norm": 4.25,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.2209,
       "step": 88300
     },
     {
       "epoch": 1.4347451706714756,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.2057,
       "step": 88310
     },
     {
       "epoch": 1.4349076375688454,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.2435,
       "step": 88320
     },
     {
       "epoch": 1.435070104466215,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2166,
       "step": 88330
     },
     {
       "epoch": 1.4352325713635847,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2277,
       "step": 88340
     },
     {
       "epoch": 1.4353950382609544,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2421,
       "step": 88350
     },
     {
       "epoch": 1.435557505158324,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3628,
+      "loss": 1.226,
       "step": 88360
     },
     {
       "epoch": 1.4357199720556937,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3626,
+      "loss": 1.2495,
       "step": 88370
     },
     {
       "epoch": 1.4358824389530633,
-      "grad_norm": 3.078125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2606,
       "step": 88380
     },
     {
       "epoch": 1.436044905850433,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.1923,
       "step": 88390
     },
     {
       "epoch": 1.4362073727478026,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.188,
       "step": 88400
     },
     {
       "epoch": 1.4363698396451723,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2158,
       "step": 88410
     },
     {
       "epoch": 1.436532306542542,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2328,
       "step": 88420
     },
     {
       "epoch": 1.4366947734399116,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.1828,
       "step": 88430
     },
     {
       "epoch": 1.4368572403372812,
-      "grad_norm": 3.125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3125,
+      "loss": 1.2332,
       "step": 88440
     },
     {
       "epoch": 1.437019707234651,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2718,
       "step": 88450
     },
     {
       "epoch": 1.4371821741320205,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.1747,
       "step": 88460
     },
     {
       "epoch": 1.4373446410293902,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2087,
       "step": 88470
     },
     {
       "epoch": 1.4375071079267598,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.2016,
       "step": 88480
     },
     {
       "epoch": 1.4376695748241297,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.2421,
       "step": 88490
     },
     {
       "epoch": 1.4378320417214994,
-      "grad_norm": 2.0,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2176,
       "step": 88500
     },
     {
       "epoch": 1.437994508618869,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2034,
       "step": 88510
     },
     {
       "epoch": 1.4381569755162387,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2283,
       "step": 88520
     },
     {
       "epoch": 1.4383194424136083,
-      "grad_norm": 3.71875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.2022,
       "step": 88530
     },
     {
       "epoch": 1.438481909310978,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.232,
       "step": 88540
     },
     {
       "epoch": 1.4386443762083476,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2108,
       "step": 88550
     },
     {
       "epoch": 1.4388068431057173,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.1758,
       "step": 88560
     },
     {
       "epoch": 1.438969310003087,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.369,
+      "loss": 1.255,
       "step": 88570
     },
     {
       "epoch": 1.4391317769004566,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2207,
       "step": 88580
     },
     {
       "epoch": 1.4392942437978262,
-      "grad_norm": 5.3125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.2019,
       "step": 88590
     },
     {
       "epoch": 1.4394567106951959,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2228,
       "step": 88600
     },
     {
       "epoch": 1.4396191775925655,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2008,
       "step": 88610
     },
     {
       "epoch": 1.4397816444899352,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.209,
       "step": 88620
     },
     {
       "epoch": 1.4399441113873048,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.1907,
       "step": 88630
     },
     {
       "epoch": 1.4401065782846745,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2541,
       "step": 88640
     },
     {
       "epoch": 1.4402690451820441,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2118,
       "step": 88650
     },
     {
       "epoch": 1.4404315120794138,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3679,
+      "loss": 1.2374,
       "step": 88660
     },
     {
       "epoch": 1.4405939789767834,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.211,
       "step": 88670
     },
     {
       "epoch": 1.440756445874153,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2364,
       "step": 88680
     },
     {
       "epoch": 1.4409189127715227,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2121,
       "step": 88690
     },
     {
       "epoch": 1.4410813796688924,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3614,
+      "loss": 1.2245,
       "step": 88700
     },
     {
       "epoch": 1.441243846566262,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2272,
       "step": 88710
     },
     {
       "epoch": 1.4414063134636317,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2053,
       "step": 88720
     },
     {
       "epoch": 1.4415687803610013,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.2133,
       "step": 88730
     },
     {
       "epoch": 1.441731247258371,
-      "grad_norm": 2.75,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.25,
       "step": 88740
     },
     {
       "epoch": 1.4418937141557406,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2187,
       "step": 88750
     },
     {
       "epoch": 1.4420561810531105,
-      "grad_norm": 3.453125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2241,
       "step": 88760
     },
     {
       "epoch": 1.4422186479504802,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2389,
       "step": 88770
     },
     {
       "epoch": 1.4423811148478498,
-      "grad_norm": 2.90625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2322,
       "step": 88780
     },
     {
       "epoch": 1.4425435817452195,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3627,
+      "loss": 1.2237,
       "step": 88790
     },
     {
       "epoch": 1.4427060486425891,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.1906,
       "step": 88800
     },
     {
       "epoch": 1.4428685155399588,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2463,
       "step": 88810
     },
     {
       "epoch": 1.4430309824373284,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2117,
       "step": 88820
     },
     {
       "epoch": 1.443193449334698,
-      "grad_norm": 2.109375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2195,
       "step": 88830
     },
     {
       "epoch": 1.4433559162320677,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1963,
       "step": 88840
     },
     {
       "epoch": 1.4435183831294374,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2029,
       "step": 88850
     },
     {
       "epoch": 1.443680850026807,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3238,
+      "loss": 1.2046,
       "step": 88860
     },
     {
       "epoch": 1.4438433169241767,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2382,
       "step": 88870
     },
     {
       "epoch": 1.4440057838215463,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2418,
       "step": 88880
     },
     {
       "epoch": 1.444168250718916,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2139,
       "step": 88890
     },
     {
       "epoch": 1.4443307176162856,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1967,
       "step": 88900
     },
     {
       "epoch": 1.4444931845136553,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2064,
       "step": 88910
     },
     {
       "epoch": 1.444655651411025,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3711,
+      "loss": 1.2186,
       "step": 88920
     },
     {
       "epoch": 1.4448181183083948,
-      "grad_norm": 4.1875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3702,
+      "loss": 1.2278,
       "step": 88930
     },
     {
       "epoch": 1.4449805852057644,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2021,
       "step": 88940
     },
     {
       "epoch": 1.445143052103134,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.1919,
       "step": 88950
     },
     {
       "epoch": 1.4453055190005037,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.2244,
       "step": 88960
     },
     {
       "epoch": 1.4454679858978734,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2391,
       "step": 88970
     },
     {
       "epoch": 1.445630452795243,
-      "grad_norm": 2.140625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2014,
       "step": 88980
     },
     {
       "epoch": 1.4457929196926127,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2256,
       "step": 88990
     },
     {
       "epoch": 1.4459553865899823,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.194,
       "step": 89000
     },
     {
       "epoch": 1.446117853487352,
-      "grad_norm": 3.5,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2464,
       "step": 89010
     },
     {
       "epoch": 1.4462803203847217,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2102,
       "step": 89020
     },
     {
       "epoch": 1.4464427872820913,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2167,
       "step": 89030
     },
     {
       "epoch": 1.446605254179461,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.1958,
       "step": 89040
     },
     {
       "epoch": 1.4467677210768306,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2205,
       "step": 89050
     },
     {
       "epoch": 1.4469301879742003,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2428,
       "step": 89060
     },
     {
       "epoch": 1.44709265487157,
-      "grad_norm": 3.375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3089,
+      "loss": 1.2576,
       "step": 89070
     },
     {
       "epoch": 1.4472551217689396,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2564,
       "step": 89080
     },
     {
       "epoch": 1.4474175886663092,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2442,
       "step": 89090
     },
     {
       "epoch": 1.4475800555636789,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2126,
       "step": 89100
     },
     {
       "epoch": 1.4477425224610485,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3565,
+      "loss": 1.1917,
       "step": 89110
     },
     {
       "epoch": 1.4479049893584182,
-      "grad_norm": 3.25,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.1945,
       "step": 89120
     },
     {
       "epoch": 1.4480674562557878,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3239,
+      "loss": 1.2539,
       "step": 89130
     },
     {
       "epoch": 1.4482299231531575,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2402,
       "step": 89140
     },
     {
       "epoch": 1.4483923900505271,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.1943,
       "step": 89150
     },
     {
       "epoch": 1.4485548569478968,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2022,
       "step": 89160
     },
     {
       "epoch": 1.4487173238452664,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2341,
       "step": 89170
     },
     {
       "epoch": 1.448879790742636,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2524,
       "step": 89180
     },
     {
       "epoch": 1.4490422576400057,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2218,
       "step": 89190
     },
     {
       "epoch": 1.4492047245373756,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2153,
       "step": 89200
     },
     {
       "epoch": 1.4493671914347452,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.2191,
       "step": 89210
     },
     {
       "epoch": 1.449529658332115,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2208,
       "step": 89220
     },
     {
       "epoch": 1.4496921252294845,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.1655,
       "step": 89230
     },
     {
       "epoch": 1.4498545921268542,
-      "grad_norm": 2.875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2216,
       "step": 89240
     },
     {
       "epoch": 1.4500170590242238,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.257,
       "step": 89250
     },
     {
       "epoch": 1.4501795259215935,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2138,
       "step": 89260
     },
     {
       "epoch": 1.4503419928189631,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.1346,
       "step": 89270
     },
     {
       "epoch": 1.4505044597163328,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3176,
+      "loss": 1.222,
       "step": 89280
     },
     {
       "epoch": 1.4506669266137024,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2249,
       "step": 89290
     },
     {
       "epoch": 1.450829393511072,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2127,
       "step": 89300
     },
     {
       "epoch": 1.4509918604084417,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2015,
       "step": 89310
     },
     {
       "epoch": 1.4511543273058114,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3185,
+      "loss": 1.1874,
       "step": 89320
     },
     {
       "epoch": 1.451316794203181,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3218,
+      "loss": 1.2087,
       "step": 89330
     },
     {
       "epoch": 1.4514792611005507,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.211,
       "step": 89340
     },
     {
       "epoch": 1.4516417279979204,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.265,
       "step": 89350
     },
     {
       "epoch": 1.4518041948952902,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3592,
+      "loss": 1.2115,
       "step": 89360
     },
     {
       "epoch": 1.4519666617926599,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2546,
       "step": 89370
     },
     {
       "epoch": 1.4521291286900295,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.2532,
       "step": 89380
     },
     {
       "epoch": 1.4522915955873992,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2207,
       "step": 89390
     },
     {
       "epoch": 1.4524540624847688,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2247,
       "step": 89400
     },
     {
       "epoch": 1.4526165293821385,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.1807,
       "step": 89410
     },
     {
       "epoch": 1.4527789962795081,
-      "grad_norm": 2.46875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2168,
       "step": 89420
     },
     {
       "epoch": 1.4529414631768778,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1672,
       "step": 89430
     },
     {
       "epoch": 1.4531039300742474,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2314,
       "step": 89440
     },
     {
       "epoch": 1.453266396971617,
-      "grad_norm": 2.75,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2283,
       "step": 89450
     },
     {
       "epoch": 1.4534288638689867,
-      "grad_norm": 6.09375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2248,
       "step": 89460
     },
     {
       "epoch": 1.4535913307663564,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2142,
       "step": 89470
     },
     {
       "epoch": 1.453753797663726,
-      "grad_norm": 2.625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2671,
       "step": 89480
     },
     {
       "epoch": 1.4539162645610957,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3155,
+      "loss": 1.2079,
       "step": 89490
     },
     {
       "epoch": 1.4540787314584653,
-      "grad_norm": 2.578125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.2359,
       "step": 89500
     },
     {
       "epoch": 1.454241198355835,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3566,
+      "loss": 1.1814,
       "step": 89510
     },
     {
       "epoch": 1.4544036652532046,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.2157,
       "step": 89520
     },
     {
       "epoch": 1.4545661321505743,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.204,
       "step": 89530
     },
     {
       "epoch": 1.454728599047944,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2647,
       "step": 89540
     },
     {
       "epoch": 1.4548910659453136,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1891,
       "step": 89550
     },
     {
       "epoch": 1.4550535328426832,
-      "grad_norm": 2.875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2084,
       "step": 89560
     },
     {
       "epoch": 1.455215999740053,
-      "grad_norm": 3.25,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2122,
       "step": 89570
     },
     {
       "epoch": 1.4553784666374225,
-      "grad_norm": 2.75,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.1893,
       "step": 89580
     },
     {
       "epoch": 1.4555409335347922,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2019,
       "step": 89590
     },
     {
       "epoch": 1.4557034004321618,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2461,
       "step": 89600
     },
     {
       "epoch": 1.4558658673295315,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2542,
       "step": 89610
     },
     {
       "epoch": 1.4560283342269011,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2154,
       "step": 89620
     },
     {
       "epoch": 1.4561908011242708,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2109,
       "step": 89630
     },
     {
       "epoch": 1.4563532680216407,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2351,
       "step": 89640
     },
     {
       "epoch": 1.4565157349190103,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2052,
       "step": 89650
     },
     {
       "epoch": 1.45667820181638,
-      "grad_norm": 4.0,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2039,
       "step": 89660
     },
     {
       "epoch": 1.4568406687137496,
-      "grad_norm": 3.125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.1708,
       "step": 89670
     },
     {
       "epoch": 1.4570031356111193,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.2087,
       "step": 89680
     },
     {
       "epoch": 1.457165602508489,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1841,
       "step": 89690
     },
     {
       "epoch": 1.4573280694058586,
-      "grad_norm": 2.734375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3184,
+      "loss": 1.2361,
       "step": 89700
     },
     {
       "epoch": 1.4574905363032282,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2378,
       "step": 89710
     },
     {
       "epoch": 1.4576530032005979,
-      "grad_norm": 2.75,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3268,
+      "loss": 1.2353,
       "step": 89720
     },
     {
       "epoch": 1.4578154700979675,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.2234,
       "step": 89730
     },
     {
       "epoch": 1.4579779369953372,
-      "grad_norm": 3.65625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.202,
       "step": 89740
     },
     {
       "epoch": 1.4581404038927068,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.224,
       "step": 89750
     },
     {
       "epoch": 1.4583028707900765,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2223,
       "step": 89760
     },
     {
       "epoch": 1.4584653376874461,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2416,
       "step": 89770
     },
     {
       "epoch": 1.4586278045848158,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.1979,
       "step": 89780
     },
     {
       "epoch": 1.4587902714821854,
-      "grad_norm": 3.125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.1996,
       "step": 89790
     },
     {
       "epoch": 1.4589527383795553,
-      "grad_norm": 1.890625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2338,
       "step": 89800
     },
     {
       "epoch": 1.459115205276925,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2106,
       "step": 89810
     },
     {
       "epoch": 1.4592776721742946,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.239,
       "step": 89820
     },
     {
       "epoch": 1.4594401390716643,
-      "grad_norm": 3.515625,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.1903,
       "step": 89830
     },
     {
       "epoch": 1.459602605969034,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2168,
       "step": 89840
     },
     {
       "epoch": 1.4597650728664036,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2583,
       "step": 89850
     },
     {
       "epoch": 1.4599275397637732,
-      "grad_norm": 3.984375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2312,
       "step": 89860
     },
     {
       "epoch": 1.4600900066611429,
-      "grad_norm": 2.0,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.232,
       "step": 89870
     },
     {
       "epoch": 1.4602524735585125,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2137,
       "step": 89880
     },
     {
       "epoch": 1.4604149404558822,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.1903,
       "step": 89890
     },
     {
       "epoch": 1.4605774073532518,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2175,
       "step": 89900
     },
     {
       "epoch": 1.4607398742506215,
-      "grad_norm": 1.9140625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.1999,
       "step": 89910
     },
     {
       "epoch": 1.4609023411479911,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2203,
       "step": 89920
     },
     {
       "epoch": 1.4610648080453608,
-      "grad_norm": 2.53125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2173,
       "step": 89930
     },
     {
       "epoch": 1.4612272749427304,
-      "grad_norm": 3.5,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2262,
       "step": 89940
     },
     {
       "epoch": 1.4613897418401,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3647,
+      "loss": 1.2574,
       "step": 89950
     },
     {
       "epoch": 1.4615522087374697,
-      "grad_norm": 1.75,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1898,
       "step": 89960
     },
     {
       "epoch": 1.4617146756348394,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2432,
       "step": 89970
     },
     {
       "epoch": 1.461877142532209,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2415,
       "step": 89980
     },
     {
       "epoch": 1.4620396094295787,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2211,
       "step": 89990
     },
     {
       "epoch": 1.4622020763269483,
-      "grad_norm": 2.734375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2198,
       "step": 90000
     },
     {
       "epoch": 1.462364543224318,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.213,
       "step": 90010
     },
     {
       "epoch": 1.4625270101216876,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2151,
       "step": 90020
     },
     {
       "epoch": 1.4626894770190573,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2397,
       "step": 90030
     },
     {
       "epoch": 1.462851943916427,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.1869,
       "step": 90040
     },
     {
       "epoch": 1.4630144108137966,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.32,
+      "loss": 1.2192,
       "step": 90050
     },
     {
       "epoch": 1.4631768777111662,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2472,
       "step": 90060
     },
     {
       "epoch": 1.4633393446085359,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.189,
       "step": 90070
     },
     {
       "epoch": 1.4635018115059057,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2133,
       "step": 90080
     },
     {
       "epoch": 1.4636642784032754,
-      "grad_norm": 4.28125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2131,
       "step": 90090
     },
     {
       "epoch": 1.463826745300645,
-      "grad_norm": 3.0,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3127,
+      "loss": 1.1937,
       "step": 90100
     },
     {
       "epoch": 1.4639892121980147,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2048,
       "step": 90110
     },
     {
       "epoch": 1.4641516790953844,
-      "grad_norm": 2.25,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2309,
       "step": 90120
     },
     {
       "epoch": 1.464314145992754,
-      "grad_norm": 2.5,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2109,
       "step": 90130
     },
     {
       "epoch": 1.4644766128901237,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.2027,
       "step": 90140
     },
     {
       "epoch": 1.4646390797874933,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2117,
       "step": 90150
     },
     {
       "epoch": 1.464801546684863,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2328,
       "step": 90160
     },
     {
       "epoch": 1.4649640135822326,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.1887,
       "step": 90170
     },
     {
       "epoch": 1.4651264804796023,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.2211,
       "step": 90180
     },
     {
       "epoch": 1.465288947376972,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.1855,
       "step": 90190
     },
     {
       "epoch": 1.4654514142743416,
-      "grad_norm": 3.65625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1707,
       "step": 90200
     },
     {
       "epoch": 1.4656138811717112,
-      "grad_norm": 3.0,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.2164,
       "step": 90210
     },
     {
       "epoch": 1.4657763480690809,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.1792,
       "step": 90220
     },
     {
       "epoch": 1.4659388149664505,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.233,
       "step": 90230
     },
     {
       "epoch": 1.4661012818638204,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.1357,
       "step": 90240
     },
     {
       "epoch": 1.46626374876119,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2171,
       "step": 90250
     },
     {
       "epoch": 1.4664262156585597,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2043,
       "step": 90260
     },
     {
       "epoch": 1.4665886825559293,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2306,
       "step": 90270
     },
     {
       "epoch": 1.466751149453299,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2333,
       "step": 90280
     },
     {
       "epoch": 1.4669136163506686,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2071,
       "step": 90290
     },
     {
       "epoch": 1.4670760832480383,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2102,
       "step": 90300
     },
     {
       "epoch": 1.467238550145408,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2254,
       "step": 90310
     },
     {
       "epoch": 1.4674010170427776,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.2437,
       "step": 90320
     },
     {
       "epoch": 1.4675634839401472,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2342,
       "step": 90330
     },
     {
       "epoch": 1.467725950837517,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2582,
       "step": 90340
     },
     {
       "epoch": 1.4678884177348865,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.1851,
       "step": 90350
     },
     {
       "epoch": 1.4680508846322562,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.195,
       "step": 90360
     },
     {
       "epoch": 1.4682133515296258,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2061,
       "step": 90370
     },
     {
       "epoch": 1.4683758184269955,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2124,
       "step": 90380
     },
     {
       "epoch": 1.4685382853243651,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2188,
       "step": 90390
     },
     {
       "epoch": 1.4687007522217348,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.201,
       "step": 90400
     },
     {
       "epoch": 1.4688632191191044,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2185,
       "step": 90410
     },
     {
       "epoch": 1.469025686016474,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2603,
       "step": 90420
     },
     {
       "epoch": 1.4691881529138437,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2092,
       "step": 90430
     },
     {
       "epoch": 1.4693506198112134,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2545,
       "step": 90440
     },
     {
       "epoch": 1.469513086708583,
-      "grad_norm": 2.5,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2099,
       "step": 90450
     },
     {
       "epoch": 1.4696755536059527,
-      "grad_norm": 3.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2316,
       "step": 90460
     },
     {
       "epoch": 1.4698380205033224,
-      "grad_norm": 2.625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1924,
       "step": 90470
     },
     {
       "epoch": 1.470000487400692,
-      "grad_norm": 2.625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.2236,
       "step": 90480
     },
     {
       "epoch": 1.4701629542980617,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2187,
       "step": 90490
     },
     {
       "epoch": 1.4703254211954313,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2119,
       "step": 90500
     },
     {
       "epoch": 1.470487888092801,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.201,
       "step": 90510
     },
     {
       "epoch": 1.4706503549901708,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.196,
       "step": 90520
     },
     {
       "epoch": 1.4708128218875405,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2185,
       "step": 90530
     },
     {
       "epoch": 1.4709752887849101,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2398,
       "step": 90540
     },
     {
       "epoch": 1.4711377556822798,
-      "grad_norm": 2.125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2722,
       "step": 90550
     },
     {
       "epoch": 1.4713002225796494,
-      "grad_norm": 2.875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2351,
       "step": 90560
     },
     {
       "epoch": 1.471462689477019,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.2191,
       "step": 90570
     },
     {
       "epoch": 1.4716251563743887,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2336,
       "step": 90580
     },
     {
       "epoch": 1.4717876232717584,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.1899,
       "step": 90590
     },
     {
       "epoch": 1.471950090169128,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.2026,
       "step": 90600
     },
     {
       "epoch": 1.4721125570664977,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2857,
       "step": 90610
     },
     {
       "epoch": 1.4722750239638673,
-      "grad_norm": 5.875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3224,
+      "loss": 1.1889,
       "step": 90620
     },
     {
       "epoch": 1.472437490861237,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.2474,
       "step": 90630
     },
     {
       "epoch": 1.4725999577586066,
-      "grad_norm": 4.40625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2117,
       "step": 90640
     },
     {
       "epoch": 1.4727624246559763,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.307,
+      "loss": 1.2083,
       "step": 90650
     },
     {
       "epoch": 1.472924891553346,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2701,
       "step": 90660
     },
     {
       "epoch": 1.4730873584507156,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3235,
+      "loss": 1.2277,
       "step": 90670
     },
     {
       "epoch": 1.4732498253480855,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1897,
       "step": 90680
     },
     {
       "epoch": 1.4734122922454551,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.1925,
       "step": 90690
     },
     {
       "epoch": 1.4735747591428248,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1964,
       "step": 90700
     },
     {
       "epoch": 1.4737372260401944,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2323,
       "step": 90710
     },
     {
       "epoch": 1.473899692937564,
-      "grad_norm": 2.390625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2528,
       "step": 90720
     },
     {
       "epoch": 1.4740621598349337,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3765,
+      "loss": 1.2518,
       "step": 90730
     },
     {
       "epoch": 1.4742246267323034,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3693,
+      "loss": 1.2426,
       "step": 90740
     },
     {
       "epoch": 1.474387093629673,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3189,
+      "loss": 1.2511,
       "step": 90750
     },
     {
       "epoch": 1.4745495605270427,
-      "grad_norm": 2.25,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.211,
       "step": 90760
     },
     {
       "epoch": 1.4747120274244123,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3152,
+      "loss": 1.2241,
       "step": 90770
     },
     {
       "epoch": 1.474874494321782,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.1988,
       "step": 90780
     },
     {
       "epoch": 1.4750369612191516,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.1763,
       "step": 90790
     },
     {
       "epoch": 1.4751994281165213,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2173,
       "step": 90800
     },
     {
       "epoch": 1.475361895013891,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.1953,
       "step": 90810
     },
     {
       "epoch": 1.4755243619112606,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2461,
       "step": 90820
     },
     {
       "epoch": 1.4756868288086302,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.2369,
       "step": 90830
     },
     {
       "epoch": 1.4758492957059999,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2136,
       "step": 90840
     },
     {
       "epoch": 1.4760117626033695,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.2358,
       "step": 90850
     },
     {
       "epoch": 1.4761742295007392,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2191,
       "step": 90860
     },
     {
       "epoch": 1.4763366963981088,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2037,
       "step": 90870
     },
     {
       "epoch": 1.4764991632954785,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2425,
       "step": 90880
     },
     {
       "epoch": 1.4766616301928481,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2147,
       "step": 90890
     },
     {
       "epoch": 1.4768240970902178,
-      "grad_norm": 3.0,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.1974,
       "step": 90900
     },
     {
       "epoch": 1.4769865639875874,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.2233,
       "step": 90910
     },
     {
       "epoch": 1.477149030884957,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.24,
       "step": 90920
     },
     {
       "epoch": 1.4773114977823267,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2398,
       "step": 90930
     },
     {
       "epoch": 1.4774739646796964,
-      "grad_norm": 2.25,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1691,
       "step": 90940
     },
     {
       "epoch": 1.477636431577066,
-      "grad_norm": 2.921875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3609,
+      "loss": 1.1948,
       "step": 90950
     },
     {
       "epoch": 1.477798898474436,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2198,
       "step": 90960
     },
     {
       "epoch": 1.4779613653718056,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.2361,
       "step": 90970
     },
     {
       "epoch": 1.4781238322691752,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2359,
       "step": 90980
     },
     {
       "epoch": 1.4782862991665449,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2074,
       "step": 90990
     },
     {
       "epoch": 1.4784487660639145,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2154,
       "step": 91000
     },
     {
       "epoch": 1.4786112329612842,
-      "grad_norm": 2.65625,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2266,
       "step": 91010
     },
     {
       "epoch": 1.4787736998586538,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2197,
       "step": 91020
     },
     {
       "epoch": 1.4789361667560235,
-      "grad_norm": 2.875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3281,
+      "loss": 1.1703,
       "step": 91030
     },
     {
       "epoch": 1.4790986336533931,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.1309,
       "step": 91040
     },
     {
       "epoch": 1.4792611005507628,
-      "grad_norm": 3.265625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.1946,
       "step": 91050
     },
     {
       "epoch": 1.4794235674481324,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2053,
       "step": 91060
     },
     {
       "epoch": 1.479586034345502,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2129,
       "step": 91070
     },
     {
       "epoch": 1.4797485012428717,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.1954,
       "step": 91080
     },
     {
       "epoch": 1.4799109681402414,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3727,
+      "loss": 1.2226,
       "step": 91090
     },
     {
       "epoch": 1.480073435037611,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.202,
       "step": 91100
     },
     {
       "epoch": 1.4802359019349807,
-      "grad_norm": 4.09375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.184,
       "step": 91110
     },
     {
       "epoch": 1.4803983688323505,
-      "grad_norm": 3.578125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2186,
       "step": 91120
     },
     {
       "epoch": 1.4805608357297202,
-      "grad_norm": 2.15625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2255,
       "step": 91130
     },
     {
       "epoch": 1.4807233026270898,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.1926,
       "step": 91140
     },
     {
       "epoch": 1.4808857695244595,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.2093,
       "step": 91150
     },
     {
       "epoch": 1.4810482364218291,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2163,
       "step": 91160
     },
     {
       "epoch": 1.4812107033191988,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2594,
       "step": 91170
     },
     {
       "epoch": 1.4813731702165684,
-      "grad_norm": 3.84375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.1982,
       "step": 91180
     },
     {
       "epoch": 1.481535637113938,
-      "grad_norm": 2.75,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2156,
       "step": 91190
     },
     {
       "epoch": 1.4816981040113077,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1977,
       "step": 91200
     },
     {
       "epoch": 1.4818605709086774,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3663,
+      "loss": 1.1812,
       "step": 91210
     },
     {
       "epoch": 1.482023037806047,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2547,
       "step": 91220
     },
     {
       "epoch": 1.4821855047034167,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2214,
       "step": 91230
     },
     {
       "epoch": 1.4823479716007864,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2461,
       "step": 91240
     },
     {
       "epoch": 1.482510438498156,
-      "grad_norm": 4.40625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.1909,
       "step": 91250
     },
     {
       "epoch": 1.4826729053955257,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2642,
       "step": 91260
     },
     {
       "epoch": 1.4828353722928953,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.1799,
       "step": 91270
     },
     {
       "epoch": 1.482997839190265,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.236,
       "step": 91280
     },
     {
       "epoch": 1.4831603060876346,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3574,
+      "loss": 1.199,
       "step": 91290
     },
     {
       "epoch": 1.4833227729850043,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2789,
       "step": 91300
     },
     {
       "epoch": 1.483485239882374,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2172,
       "step": 91310
     },
     {
       "epoch": 1.4836477067797436,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2252,
       "step": 91320
     },
     {
       "epoch": 1.4838101736771132,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2103,
       "step": 91330
     },
     {
       "epoch": 1.4839726405744829,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2651,
       "step": 91340
     },
     {
       "epoch": 1.4841351074718525,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2284,
       "step": 91350
     },
     {
       "epoch": 1.4842975743692222,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2303,
       "step": 91360
     },
     {
       "epoch": 1.4844600412665918,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2095,
       "step": 91370
     },
     {
       "epoch": 1.4846225081639615,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.213,
       "step": 91380
     },
     {
       "epoch": 1.4847849750613313,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2085,
       "step": 91390
     },
     {
       "epoch": 1.484947441958701,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2132,
       "step": 91400
     },
     {
       "epoch": 1.4851099088560706,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2448,
       "step": 91410
     },
     {
       "epoch": 1.4852723757534403,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2337,
       "step": 91420
     },
     {
       "epoch": 1.48543484265081,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2079,
       "step": 91430
     },
     {
       "epoch": 1.4855973095481796,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2229,
       "step": 91440
     },
     {
       "epoch": 1.4857597764455492,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2358,
       "step": 91450
     },
     {
       "epoch": 1.485922243342919,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2502,
       "step": 91460
     },
     {
       "epoch": 1.4860847102402885,
-      "grad_norm": 4.21875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.2799,
       "step": 91470
     },
     {
       "epoch": 1.4862471771376582,
-      "grad_norm": 2.25,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2114,
       "step": 91480
     },
     {
       "epoch": 1.4864096440350278,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2318,
       "step": 91490
     },
     {
       "epoch": 1.4865721109323975,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2263,
       "step": 91500
     },
     {
       "epoch": 1.4867345778297671,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.369,
+      "loss": 1.1982,
       "step": 91510
     },
     {
       "epoch": 1.4868970447271368,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2217,
       "step": 91520
     },
     {
       "epoch": 1.4870595116245064,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3579,
+      "loss": 1.1921,
       "step": 91530
     },
     {
       "epoch": 1.487221978521876,
-      "grad_norm": 3.296875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2268,
       "step": 91540
     },
     {
       "epoch": 1.4873844454192457,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2467,
       "step": 91550
     },
     {
       "epoch": 1.4875469123166156,
-      "grad_norm": 3.359375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2148,
       "step": 91560
     },
     {
       "epoch": 1.4877093792139853,
-      "grad_norm": 5.09375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.1952,
       "step": 91570
     },
     {
       "epoch": 1.487871846111355,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2148,
       "step": 91580
     },
     {
       "epoch": 1.4880343130087246,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2295,
       "step": 91590
     },
     {
       "epoch": 1.4881967799060942,
-      "grad_norm": 2.046875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2141,
       "step": 91600
     },
     {
       "epoch": 1.4883592468034639,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2398,
       "step": 91610
     },
     {
       "epoch": 1.4885217137008335,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2259,
       "step": 91620
     },
     {
       "epoch": 1.4886841805982032,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2153,
       "step": 91630
     },
     {
       "epoch": 1.4888466474955728,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2319,
       "step": 91640
     },
     {
       "epoch": 1.4890091143929425,
-      "grad_norm": 3.0,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.238,
       "step": 91650
     },
     {
       "epoch": 1.4891715812903121,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2102,
       "step": 91660
     },
     {
       "epoch": 1.4893340481876818,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3663,
+      "loss": 1.1853,
       "step": 91670
     },
     {
       "epoch": 1.4894965150850514,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.259,
       "step": 91680
     },
     {
       "epoch": 1.489658981982421,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.1909,
       "step": 91690
     },
     {
       "epoch": 1.4898214488797907,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.2121,
       "step": 91700
     },
     {
       "epoch": 1.4899839157771604,
-      "grad_norm": 4.28125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.2344,
       "step": 91710
     },
     {
       "epoch": 1.49014638267453,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.2195,
       "step": 91720
     },
     {
       "epoch": 1.4903088495718997,
-      "grad_norm": 2.5,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.1986,
       "step": 91730
     },
     {
       "epoch": 1.4904713164692693,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2296,
       "step": 91740
     },
     {
       "epoch": 1.490633783366639,
-      "grad_norm": 2.625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.1683,
       "step": 91750
     },
     {
       "epoch": 1.4907962502640086,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3632,
+      "loss": 1.203,
       "step": 91760
     },
     {
       "epoch": 1.4909587171613783,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.1752,
       "step": 91770
     },
     {
       "epoch": 1.491121184058748,
-      "grad_norm": 4.03125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2503,
       "step": 91780
     },
     {
       "epoch": 1.4912836509561176,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2193,
       "step": 91790
     },
     {
       "epoch": 1.4914461178534872,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2434,
       "step": 91800
     },
     {
       "epoch": 1.491608584750857,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2559,
       "step": 91810
     },
     {
       "epoch": 1.4917710516482265,
-      "grad_norm": 4.34375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.1814,
       "step": 91820
     },
     {
       "epoch": 1.4919335185455964,
-      "grad_norm": 3.25,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2225,
       "step": 91830
     },
     {
       "epoch": 1.492095985442966,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2217,
       "step": 91840
     },
     {
       "epoch": 1.4922584523403357,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.2216,
       "step": 91850
     },
     {
       "epoch": 1.4924209192377054,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2358,
       "step": 91860
     },
     {
       "epoch": 1.492583386135075,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.1898,
       "step": 91870
     },
     {
       "epoch": 1.4927458530324447,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.173,
       "step": 91880
     },
     {
       "epoch": 1.4929083199298143,
-      "grad_norm": 3.78125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.1998,
       "step": 91890
     },
     {
       "epoch": 1.493070786827184,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.2091,
       "step": 91900
     },
     {
       "epoch": 1.4932332537245536,
-      "grad_norm": 1.796875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.2128,
       "step": 91910
     },
     {
       "epoch": 1.4933957206219233,
-      "grad_norm": 3.375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2098,
       "step": 91920
     },
     {
       "epoch": 1.493558187519293,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.1855,
       "step": 91930
     },
     {
       "epoch": 1.4937206544166626,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3028,
+      "loss": 1.1825,
       "step": 91940
     },
     {
       "epoch": 1.4938831213140322,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2077,
       "step": 91950
     },
     {
       "epoch": 1.4940455882114019,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2026,
       "step": 91960
     },
     {
       "epoch": 1.4942080551087715,
-      "grad_norm": 2.3125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2611,
       "step": 91970
     },
     {
       "epoch": 1.4943705220061412,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2532,
       "step": 91980
     },
     {
       "epoch": 1.4945329889035108,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3685,
+      "loss": 1.2184,
       "step": 91990
     },
     {
       "epoch": 1.4946954558008807,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3088,
+      "loss": 1.1648,
       "step": 92000
     },
     {
       "epoch": 1.4948579226982504,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2009,
       "step": 92010
     },
     {
       "epoch": 1.49502038959562,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2565,
       "step": 92020
     },
     {
       "epoch": 1.4951828564929897,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.2366,
       "step": 92030
     },
     {
       "epoch": 1.4953453233903593,
-      "grad_norm": 2.75,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.2184,
       "step": 92040
     },
     {
       "epoch": 1.495507790287729,
-      "grad_norm": 3.125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2279,
       "step": 92050
     },
     {
       "epoch": 1.4956702571850986,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.236,
       "step": 92060
     },
     {
       "epoch": 1.4958327240824683,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.364,
+      "loss": 1.2052,
       "step": 92070
     },
     {
       "epoch": 1.495995190979838,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.264,
       "step": 92080
     },
     {
       "epoch": 1.4961576578772076,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2226,
       "step": 92090
     },
     {
       "epoch": 1.4963201247745772,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2575,
       "step": 92100
     },
     {
       "epoch": 1.4964825916719469,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.204,
       "step": 92110
     },
     {
       "epoch": 1.4966450585693165,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.199,
       "step": 92120
     },
     {
       "epoch": 1.4968075254666862,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2132,
       "step": 92130
     },
     {
       "epoch": 1.4969699923640558,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3613,
+      "loss": 1.1712,
       "step": 92140
     },
     {
       "epoch": 1.4971324592614255,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2595,
       "step": 92150
     },
     {
       "epoch": 1.4972949261587951,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2233,
       "step": 92160
     },
     {
       "epoch": 1.4974573930561648,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.32,
+      "loss": 1.2122,
       "step": 92170
     },
     {
       "epoch": 1.4976198599535344,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3218,
+      "loss": 1.219,
       "step": 92180
     },
     {
       "epoch": 1.497782326850904,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2211,
       "step": 92190
     },
     {
       "epoch": 1.4979447937482737,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.1842,
       "step": 92200
     },
     {
       "epoch": 1.4981072606456434,
-      "grad_norm": 4.125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2388,
       "step": 92210
     },
     {
       "epoch": 1.498269727543013,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.2199,
       "step": 92220
     },
     {
       "epoch": 1.4984321944403827,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2076,
       "step": 92230
     },
     {
       "epoch": 1.4985946613377523,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2016,
       "step": 92240
     },
     {
       "epoch": 1.498757128235122,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2423,
       "step": 92250
     },
     {
       "epoch": 1.4989195951324916,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3639,
+      "loss": 1.2474,
       "step": 92260
     },
     {
       "epoch": 1.4990820620298615,
-      "grad_norm": 2.1875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2478,
       "step": 92270
     },
     {
       "epoch": 1.4992445289272311,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2618,
       "step": 92280
     },
     {
       "epoch": 1.4994069958246008,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.201,
       "step": 92290
     },
     {
       "epoch": 1.4995694627219704,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2397,
       "step": 92300
     },
     {
       "epoch": 1.49973192961934,
-      "grad_norm": 3.78125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2473,
       "step": 92310
     },
     {
       "epoch": 1.4998943965167097,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.1917,
       "step": 92320
     },
     {
       "epoch": 1.5000568634140794,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2331,
       "step": 92330
     },
     {
       "epoch": 1.500219330311449,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.229,
       "step": 92340
     },
     {
       "epoch": 1.5003817972088187,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2051,
       "step": 92350
     },
     {
       "epoch": 1.5005442641061884,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2258,
       "step": 92360
     },
     {
       "epoch": 1.500706731003558,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.2211,
       "step": 92370
     },
     {
       "epoch": 1.5008691979009277,
-      "grad_norm": 2.953125,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2271,
       "step": 92380
     },
     {
       "epoch": 1.5010316647982973,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2073,
       "step": 92390
     },
     {
       "epoch": 1.501194131695667,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2267,
       "step": 92400
     },
     {
       "epoch": 1.5013565985930368,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.1972,
       "step": 92410
     },
     {
       "epoch": 1.5015190654904065,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2317,
       "step": 92420
     },
     {
       "epoch": 1.5016815323877761,
-      "grad_norm": 4.3125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2723,
       "step": 92430
     },
     {
       "epoch": 1.5018439992851458,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.1983,
       "step": 92440
     },
     {
       "epoch": 1.5020064661825154,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.315,
+      "loss": 1.2086,
       "step": 92450
     },
     {
       "epoch": 1.502168933079885,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.198,
       "step": 92460
     },
     {
       "epoch": 1.5023313999772547,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3799,
+      "loss": 1.2754,
       "step": 92470
     },
     {
       "epoch": 1.5024938668746244,
-      "grad_norm": 2.5,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2626,
       "step": 92480
     },
     {
       "epoch": 1.502656333771994,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2337,
       "step": 92490
     },
     {
       "epoch": 1.5028188006693637,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2036,
       "step": 92500
     },
     {
       "epoch": 1.5029812675667333,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.231,
       "step": 92510
     },
     {
       "epoch": 1.503143734464103,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2299,
       "step": 92520
     },
     {
       "epoch": 1.5033062013614726,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.262,
       "step": 92530
     },
     {
       "epoch": 1.5034686682588423,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.295,
       "step": 92540
     },
     {
       "epoch": 1.503631135156212,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.1908,
       "step": 92550
     },
     {
       "epoch": 1.5037936020535816,
-      "grad_norm": 3.125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.1799,
       "step": 92560
     },
     {
       "epoch": 1.5039560689509512,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2136,
       "step": 92570
     },
     {
       "epoch": 1.504118535848321,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2056,
       "step": 92580
     },
     {
       "epoch": 1.5042810027456905,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2258,
       "step": 92590
     },
     {
       "epoch": 1.5044434696430602,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2093,
       "step": 92600
     },
     {
       "epoch": 1.5046059365404298,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3121,
+      "loss": 1.2495,
       "step": 92610
     },
     {
       "epoch": 1.5047684034377995,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.2367,
       "step": 92620
     },
     {
       "epoch": 1.5049308703351691,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2449,
       "step": 92630
     },
     {
       "epoch": 1.5050933372325388,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2424,
       "step": 92640
     },
     {
       "epoch": 1.5052558041299084,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.294,
       "step": 92650
     },
     {
       "epoch": 1.505418271027278,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2701,
       "step": 92660
     },
     {
       "epoch": 1.5055807379246477,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.21,
       "step": 92670
     },
     {
       "epoch": 1.5057432048220174,
-      "grad_norm": 3.859375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.2037,
       "step": 92680
     },
     {
       "epoch": 1.505905671719387,
-      "grad_norm": 2.9375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2321,
       "step": 92690
     },
     {
       "epoch": 1.5060681386167567,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.1862,
       "step": 92700
     },
     {
       "epoch": 1.5062306055141264,
-      "grad_norm": 3.25,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.2317,
       "step": 92710
     },
     {
       "epoch": 1.506393072411496,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2301,
       "step": 92720
     },
     {
       "epoch": 1.5065555393088657,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2023,
       "step": 92730
     },
     {
       "epoch": 1.5067180062062355,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2786,
       "step": 92740
     },
     {
       "epoch": 1.5068804731036052,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3632,
+      "loss": 1.2028,
       "step": 92750
     },
     {
       "epoch": 1.5070429400009748,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.213,
       "step": 92760
     },
     {
       "epoch": 1.5072054068983445,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3068,
+      "loss": 1.2286,
       "step": 92770
     },
     {
       "epoch": 1.5073678737957141,
-      "grad_norm": 2.5,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.1909,
       "step": 92780
     },
     {
       "epoch": 1.5075303406930838,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.1813,
       "step": 92790
     },
     {
       "epoch": 1.5076928075904534,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.1656,
       "step": 92800
     },
     {
       "epoch": 1.507855274487823,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2165,
       "step": 92810
     },
     {
       "epoch": 1.5080177413851927,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3695,
+      "loss": 1.2086,
       "step": 92820
     },
     {
       "epoch": 1.5081802082825624,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2222,
       "step": 92830
     },
     {
       "epoch": 1.508342675179932,
-      "grad_norm": 3.5,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2142,
       "step": 92840
     },
     {
       "epoch": 1.508505142077302,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2329,
       "step": 92850
     },
     {
       "epoch": 1.5086676089746716,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2237,
       "step": 92860
     },
     {
       "epoch": 1.5088300758720412,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2028,
       "step": 92870
     },
     {
       "epoch": 1.5089925427694109,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.1722,
       "step": 92880
     },
     {
       "epoch": 1.5091550096667805,
-      "grad_norm": 3.625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.213,
       "step": 92890
     },
     {
       "epoch": 1.5093174765641502,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2689,
       "step": 92900
     },
     {
       "epoch": 1.5094799434615198,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.259,
       "step": 92910
     },
     {
       "epoch": 1.5096424103588895,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3628,
+      "loss": 1.1886,
       "step": 92920
     },
     {
       "epoch": 1.5098048772562591,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2241,
       "step": 92930
     },
     {
       "epoch": 1.5099673441536288,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.234,
       "step": 92940
     },
     {
       "epoch": 1.5101298110509984,
-      "grad_norm": 2.625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2301,
       "step": 92950
     },
     {
       "epoch": 1.510292277948368,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2097,
       "step": 92960
     },
     {
       "epoch": 1.5104547448457377,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2226,
       "step": 92970
     },
     {
       "epoch": 1.5106172117431074,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1911,
       "step": 92980
     },
     {
       "epoch": 1.510779678640477,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.1924,
       "step": 92990
     },
     {
       "epoch": 1.5109421455378467,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2499,
       "step": 93000
     },
     {
       "epoch": 1.5111046124352163,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.207,
       "step": 93010
     },
     {
       "epoch": 1.511267079332586,
-      "grad_norm": 2.875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.1678,
       "step": 93020
     },
     {
       "epoch": 1.5114295462299556,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.1876,
       "step": 93030
     },
     {
       "epoch": 1.5115920131273253,
-      "grad_norm": 2.25,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2231,
       "step": 93040
     },
     {
       "epoch": 1.511754480024695,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2307,
       "step": 93050
     },
     {
       "epoch": 1.5119169469220646,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2142,
       "step": 93060
     },
     {
       "epoch": 1.5120794138194342,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2014,
       "step": 93070
     },
     {
       "epoch": 1.5122418807168039,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.1702,
       "step": 93080
     },
     {
       "epoch": 1.5124043476141735,
-      "grad_norm": 2.5,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.1769,
       "step": 93090
     },
     {
       "epoch": 1.5125668145115432,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2571,
       "step": 93100
     },
     {
       "epoch": 1.5127292814089128,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2102,
       "step": 93110
     },
     {
       "epoch": 1.5128917483062825,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2382,
       "step": 93120
     },
     {
       "epoch": 1.5130542152036521,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2231,
       "step": 93130
     },
     {
       "epoch": 1.5132166821010218,
-      "grad_norm": 3.625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2272,
       "step": 93140
     },
     {
       "epoch": 1.5133791489983914,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.1539,
       "step": 93150
     },
     {
       "epoch": 1.513541615895761,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2416,
       "step": 93160
     },
     {
       "epoch": 1.5137040827931307,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.186,
       "step": 93170
     },
     {
       "epoch": 1.5138665496905006,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2181,
       "step": 93180
     },
     {
       "epoch": 1.5140290165878703,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.1814,
       "step": 93190
     },
     {
       "epoch": 1.51419148348524,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3574,
+      "loss": 1.2509,
       "step": 93200
     },
     {
       "epoch": 1.5143539503826096,
-      "grad_norm": 3.25,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.1979,
       "step": 93210
     },
     {
       "epoch": 1.5145164172799792,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2075,
       "step": 93220
     },
     {
       "epoch": 1.5146788841773489,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1978,
       "step": 93230
     },
     {
       "epoch": 1.5148413510747185,
-      "grad_norm": 3.65625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2163,
       "step": 93240
     },
     {
       "epoch": 1.5150038179720882,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3622,
+      "loss": 1.1702,
       "step": 93250
     },
     {
       "epoch": 1.5151662848694578,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2588,
       "step": 93260
     },
     {
       "epoch": 1.5153287517668275,
-      "grad_norm": 3.5,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2101,
       "step": 93270
     },
     {
       "epoch": 1.5154912186641971,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.2233,
       "step": 93280
     },
     {
       "epoch": 1.515653685561567,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3613,
+      "loss": 1.2363,
       "step": 93290
     },
     {
       "epoch": 1.5158161524589366,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3668,
+      "loss": 1.198,
       "step": 93300
     },
     {
       "epoch": 1.5159786193563063,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.1885,
       "step": 93310
     },
     {
       "epoch": 1.516141086253676,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2509,
       "step": 93320
     },
     {
       "epoch": 1.5163035531510456,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2416,
       "step": 93330
     },
     {
       "epoch": 1.5164660200484152,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3589,
+      "loss": 1.1983,
       "step": 93340
     },
     {
       "epoch": 1.516628486945785,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.1939,
       "step": 93350
     },
     {
       "epoch": 1.5167909538431545,
-      "grad_norm": 2.375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.2022,
       "step": 93360
     },
     {
       "epoch": 1.5169534207405242,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3574,
+      "loss": 1.2323,
       "step": 93370
     },
     {
       "epoch": 1.5171158876378938,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.205,
       "step": 93380
     },
     {
       "epoch": 1.5172783545352635,
-      "grad_norm": 2.625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.2109,
       "step": 93390
     },
     {
       "epoch": 1.5174408214326331,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.366,
+      "loss": 1.2042,
       "step": 93400
     },
     {
       "epoch": 1.5176032883300028,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.1966,
       "step": 93410
     },
     {
       "epoch": 1.5177657552273724,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.1849,
       "step": 93420
     },
     {
       "epoch": 1.517928222124742,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.1996,
       "step": 93430
     },
     {
       "epoch": 1.5180906890221118,
-      "grad_norm": 2.8125,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.208,
       "step": 93440
     },
     {
       "epoch": 1.5182531559194814,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.2257,
       "step": 93450
     },
     {
       "epoch": 1.518415622816851,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.1896,
       "step": 93460
     },
     {
       "epoch": 1.5185780897142207,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.227,
       "step": 93470
     },
     {
       "epoch": 1.5187405566115904,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.1871,
       "step": 93480
     },
     {
       "epoch": 1.51890302350896,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1837,
       "step": 93490
     },
     {
       "epoch": 1.5190654904063297,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2198,
       "step": 93500
     },
     {
       "epoch": 1.5192279573036993,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2352,
       "step": 93510
     },
     {
       "epoch": 1.519390424201069,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2096,
       "step": 93520
     },
     {
       "epoch": 1.5195528910984386,
-      "grad_norm": 2.75,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1951,
       "step": 93530
     },
     {
       "epoch": 1.5197153579958083,
-      "grad_norm": 2.625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2479,
       "step": 93540
     },
     {
       "epoch": 1.519877824893178,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2416,
       "step": 93550
     },
     {
       "epoch": 1.5200402917905476,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2235,
       "step": 93560
     },
     {
       "epoch": 1.5202027586879172,
-      "grad_norm": 2.90625,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.321,
+      "loss": 1.2069,
       "step": 93570
     },
     {
       "epoch": 1.5203652255852869,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.2415,
       "step": 93580
     },
     {
       "epoch": 1.5205276924826565,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.1875,
       "step": 93590
     },
     {
       "epoch": 1.5206901593800262,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2491,
       "step": 93600
     },
     {
       "epoch": 1.520852626277396,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.1749,
       "step": 93610
     },
     {
       "epoch": 1.5210150931747657,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2149,
       "step": 93620
     },
     {
       "epoch": 1.5211775600721353,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.205,
       "step": 93630
     },
     {
       "epoch": 1.521340026969505,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2463,
       "step": 93640
     },
     {
       "epoch": 1.5215024938668746,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2086,
       "step": 93650
     },
     {
       "epoch": 1.5216649607642443,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2634,
       "step": 93660
     },
     {
       "epoch": 1.521827427661614,
-      "grad_norm": 4.375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2093,
       "step": 93670
     },
     {
       "epoch": 1.5219898945589836,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.222,
       "step": 93680
     },
     {
       "epoch": 1.5221523614563532,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.2011,
       "step": 93690
     },
     {
       "epoch": 1.522314828353723,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2002,
       "step": 93700
     },
     {
       "epoch": 1.5224772952510925,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.2233,
       "step": 93710
     },
     {
       "epoch": 1.5226397621484624,
-      "grad_norm": 2.125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.1783,
       "step": 93720
     },
     {
       "epoch": 1.522802229045832,
-      "grad_norm": 2.25,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.1778,
       "step": 93730
     },
     {
       "epoch": 1.5229646959432017,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1991,
       "step": 93740
     },
     {
       "epoch": 1.5231271628405714,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.1919,
       "step": 93750
     },
     {
       "epoch": 1.523289629737941,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.1946,
       "step": 93760
     },
     {
       "epoch": 1.5234520966353107,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2416,
       "step": 93770
     },
     {
       "epoch": 1.5236145635326803,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2302,
       "step": 93780
     },
     {
       "epoch": 1.52377703043005,
-      "grad_norm": 2.25,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.1913,
       "step": 93790
     },
     {
       "epoch": 1.5239394973274196,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2287,
       "step": 93800
     },
     {
       "epoch": 1.5241019642247893,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3106,
+      "loss": 1.25,
       "step": 93810
     },
     {
       "epoch": 1.524264431122159,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3252,
+      "loss": 1.2033,
       "step": 93820
     },
     {
       "epoch": 1.5244268980195286,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2164,
       "step": 93830
     },
     {
       "epoch": 1.5245893649168982,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2451,
       "step": 93840
     },
     {
       "epoch": 1.5247518318142679,
-      "grad_norm": 2.625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3794,
+      "loss": 1.2116,
       "step": 93850
     },
     {
       "epoch": 1.5249142987116375,
-      "grad_norm": 4.59375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.3647,
+      "loss": 1.1865,
       "step": 93860
     },
     {
       "epoch": 1.5250767656090072,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1979,
       "step": 93870
     },
     {
       "epoch": 1.5252392325063768,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3178,
+      "loss": 1.182,
       "step": 93880
     },
     {
       "epoch": 1.5254016994037465,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.2213,
       "step": 93890
     },
     {
       "epoch": 1.5255641663011161,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2123,
       "step": 93900
     },
     {
       "epoch": 1.5257266331984858,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.2266,
       "step": 93910
     },
     {
       "epoch": 1.5258891000958554,
-      "grad_norm": 2.40625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2067,
       "step": 93920
     },
     {
       "epoch": 1.526051566993225,
-      "grad_norm": 1.8828125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.1864,
       "step": 93930
     },
     {
       "epoch": 1.5262140338905947,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.2014,
       "step": 93940
     },
     {
       "epoch": 1.5263765007879644,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.1917,
       "step": 93950
     },
     {
       "epoch": 1.526538967685334,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2736,
       "step": 93960
     },
     {
       "epoch": 1.5267014345827037,
-      "grad_norm": 2.28125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2145,
       "step": 93970
     },
     {
       "epoch": 1.5268639014800733,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.2069,
       "step": 93980
     },
     {
       "epoch": 1.527026368377443,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2362,
       "step": 93990
     },
     {
       "epoch": 1.5271888352748126,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2302,
       "step": 94000
     },
     {
       "epoch": 1.5273513021721823,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2032,
       "step": 94010
     },
     {
       "epoch": 1.527513769069552,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.1995,
       "step": 94020
     },
     {
       "epoch": 1.5276762359669216,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2448,
       "step": 94030
     },
     {
       "epoch": 1.5278387028642912,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.1916,
       "step": 94040
     },
     {
       "epoch": 1.5280011697616611,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2155,
       "step": 94050
     },
     {
       "epoch": 1.5281636366590308,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2138,
       "step": 94060
     },
     {
       "epoch": 1.5283261035564004,
-      "grad_norm": 2.5,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2135,
       "step": 94070
     },
     {
       "epoch": 1.52848857045377,
-      "grad_norm": 3.5,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2291,
       "step": 94080
     },
     {
       "epoch": 1.5286510373511397,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.206,
       "step": 94090
     },
     {
       "epoch": 1.5288135042485094,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.2674,
       "step": 94100
     },
     {
       "epoch": 1.528975971145879,
-      "grad_norm": 3.25,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.21,
       "step": 94110
     },
     {
       "epoch": 1.5291384380432487,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2208,
       "step": 94120
     },
     {
       "epoch": 1.5293009049406183,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2011,
       "step": 94130
     },
     {
       "epoch": 1.529463371837988,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.2137,
       "step": 94140
     },
     {
       "epoch": 1.5296258387353576,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2166,
       "step": 94150
     },
     {
       "epoch": 1.5297883056327275,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.1805,
       "step": 94160
     },
     {
       "epoch": 1.5299507725300971,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.2146,
       "step": 94170
     },
     {
       "epoch": 1.5301132394274668,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2246,
       "step": 94180
     },
     {
       "epoch": 1.5302757063248364,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2174,
       "step": 94190
     },
     {
       "epoch": 1.530438173222206,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2495,
       "step": 94200
     },
     {
       "epoch": 1.5306006401195758,
-      "grad_norm": 1.6796875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2362,
       "step": 94210
     },
     {
       "epoch": 1.5307631070169454,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.2035,
       "step": 94220
     },
     {
       "epoch": 1.530925573914315,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1792,
       "step": 94230
     },
     {
       "epoch": 1.5310880408116847,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.315,
+      "loss": 1.2291,
       "step": 94240
     },
     {
       "epoch": 1.5312505077090544,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2196,
       "step": 94250
     },
     {
       "epoch": 1.531412974606424,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2295,
       "step": 94260
     },
     {
       "epoch": 1.5315754415037937,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3184,
+      "loss": 1.2304,
       "step": 94270
     },
     {
       "epoch": 1.5317379084011633,
-      "grad_norm": 3.25,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2184,
       "step": 94280
     },
     {
       "epoch": 1.531900375298533,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2125,
       "step": 94290
     },
     {
       "epoch": 1.5320628421959026,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1957,
       "step": 94300
     },
     {
       "epoch": 1.5322253090932723,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2188,
       "step": 94310
     },
     {
       "epoch": 1.532387775990642,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2369,
       "step": 94320
     },
     {
       "epoch": 1.5325502428880116,
-      "grad_norm": 2.25,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2402,
       "step": 94330
     },
     {
       "epoch": 1.5327127097853812,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.1894,
       "step": 94340
     },
     {
       "epoch": 1.5328751766827509,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2272,
       "step": 94350
     },
     {
       "epoch": 1.5330376435801205,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3141,
+      "loss": 1.2268,
       "step": 94360
     },
     {
       "epoch": 1.5332001104774902,
-      "grad_norm": 2.125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3208,
+      "loss": 1.2283,
       "step": 94370
     },
     {
       "epoch": 1.5333625773748598,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.179,
       "step": 94380
     },
     {
       "epoch": 1.5335250442722295,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.181,
       "step": 94390
     },
     {
       "epoch": 1.5336875111695991,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3718,
+      "loss": 1.1988,
       "step": 94400
     },
     {
       "epoch": 1.5338499780669688,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2057,
       "step": 94410
     },
     {
       "epoch": 1.5340124449643384,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.239,
       "step": 94420
     },
     {
       "epoch": 1.534174911861708,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.187,
       "step": 94430
     },
     {
       "epoch": 1.5343373787590777,
-      "grad_norm": 2.875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.1713,
       "step": 94440
     },
     {
       "epoch": 1.5344998456564474,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2421,
       "step": 94450
     },
     {
       "epoch": 1.534662312553817,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.1965,
       "step": 94460
     },
     {
       "epoch": 1.5348247794511867,
-      "grad_norm": 3.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2176,
       "step": 94470
     },
     {
       "epoch": 1.5349872463485563,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3177,
+      "loss": 1.2593,
       "step": 94480
     },
     {
       "epoch": 1.5351497132459262,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2009,
       "step": 94490
     },
     {
       "epoch": 1.5353121801432958,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.1853,
       "step": 94500
     },
     {
       "epoch": 1.5354746470406655,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.1975,
       "step": 94510
     },
     {
       "epoch": 1.5356371139380351,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.2124,
       "step": 94520
     },
     {
       "epoch": 1.5357995808354048,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3642,
+      "loss": 1.1835,
       "step": 94530
     },
     {
       "epoch": 1.5359620477327744,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1987,
       "step": 94540
     },
     {
       "epoch": 1.536124514630144,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3205,
+      "loss": 1.2287,
       "step": 94550
     },
     {
       "epoch": 1.5362869815275138,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.198,
       "step": 94560
     },
     {
       "epoch": 1.5364494484248834,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.182,
       "step": 94570
     },
     {
       "epoch": 1.536611915322253,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.2345,
       "step": 94580
     },
     {
       "epoch": 1.5367743822196227,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2516,
       "step": 94590
     },
     {
       "epoch": 1.5369368491169926,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2092,
       "step": 94600
     },
     {
       "epoch": 1.5370993160143622,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3143,
+      "loss": 1.2164,
       "step": 94610
     },
     {
       "epoch": 1.5372617829117319,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3201,
+      "loss": 1.2474,
       "step": 94620
     },
     {
       "epoch": 1.5374242498091015,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2126,
       "step": 94630
     },
     {
       "epoch": 1.5375867167064712,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3786,
+      "loss": 1.1908,
       "step": 94640
     },
     {
       "epoch": 1.5377491836038408,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.1992,
       "step": 94650
     },
     {
       "epoch": 1.5379116505012105,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3146,
+      "loss": 1.1956,
       "step": 94660
     },
     {
       "epoch": 1.5380741173985801,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3153,
+      "loss": 1.2065,
       "step": 94670
     },
     {
       "epoch": 1.5382365842959498,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2192,
       "step": 94680
     },
     {
       "epoch": 1.5383990511933194,
-      "grad_norm": 3.390625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.2342,
       "step": 94690
     },
     {
       "epoch": 1.538561518090689,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2427,
       "step": 94700
     },
     {
       "epoch": 1.5387239849880587,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2097,
       "step": 94710
     },
     {
       "epoch": 1.5388864518854284,
-      "grad_norm": 2.75,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2149,
       "step": 94720
     },
     {
       "epoch": 1.539048918782798,
-      "grad_norm": 3.5,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2123,
       "step": 94730
     },
     {
       "epoch": 1.5392113856801677,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2021,
       "step": 94740
     },
     {
       "epoch": 1.5393738525775373,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.1661,
       "step": 94750
     },
     {
       "epoch": 1.539536319474907,
-      "grad_norm": 3.25,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2497,
       "step": 94760
     },
     {
       "epoch": 1.5396987863722766,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.1687,
       "step": 94770
     },
     {
       "epoch": 1.5398612532696463,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2507,
       "step": 94780
     },
     {
       "epoch": 1.540023720167016,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.1661,
       "step": 94790
     },
     {
       "epoch": 1.5401861870643856,
-      "grad_norm": 4.25,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2083,
       "step": 94800
     },
     {
       "epoch": 1.5403486539617552,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.252,
       "step": 94810
     },
     {
       "epoch": 1.540511120859125,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3137,
+      "loss": 1.2017,
       "step": 94820
     },
     {
       "epoch": 1.5406735877564945,
-      "grad_norm": 2.625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.215,
       "step": 94830
     },
     {
       "epoch": 1.5408360546538642,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2252,
       "step": 94840
     },
     {
       "epoch": 1.5409985215512338,
-      "grad_norm": 2.75,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.2114,
       "step": 94850
     },
     {
       "epoch": 1.5411609884486035,
-      "grad_norm": 2.75,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2543,
       "step": 94860
     },
     {
       "epoch": 1.5413234553459731,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3727,
+      "loss": 1.2061,
       "step": 94870
     },
     {
       "epoch": 1.5414859222433428,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2163,
       "step": 94880
     },
     {
       "epoch": 1.5416483891407124,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.2001,
       "step": 94890
     },
     {
       "epoch": 1.541810856038082,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2471,
       "step": 94900
     },
     {
       "epoch": 1.5419733229354518,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2365,
       "step": 94910
     },
     {
       "epoch": 1.5421357898328214,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1977,
       "step": 94920
     },
     {
       "epoch": 1.5422982567301913,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1949,
       "step": 94930
     },
     {
       "epoch": 1.542460723627561,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1989,
       "step": 94940
     },
     {
       "epoch": 1.5426231905249306,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2468,
       "step": 94950
     },
     {
       "epoch": 1.5427856574223002,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.2396,
       "step": 94960
     },
     {
       "epoch": 1.5429481243196699,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2135,
       "step": 94970
     },
     {
       "epoch": 1.5431105912170395,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2197,
       "step": 94980
     },
     {
       "epoch": 1.5432730581144092,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2364,
       "step": 94990
     },
     {
       "epoch": 1.5434355250117788,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2314,
       "step": 95000
     },
     {
       "epoch": 1.5435979919091485,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2679,
       "step": 95010
     },
     {
       "epoch": 1.5437604588065181,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2602,
       "step": 95020
     },
     {
       "epoch": 1.5439229257038878,
-      "grad_norm": 2.5,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.1715,
       "step": 95030
     },
     {
       "epoch": 1.5440853926012577,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2203,
       "step": 95040
     },
     {
       "epoch": 1.5442478594986273,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2364,
       "step": 95050
     },
     {
       "epoch": 1.544410326395997,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1941,
       "step": 95060
     },
     {
       "epoch": 1.5445727932933666,
-      "grad_norm": 2.0,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2067,
       "step": 95070
     },
     {
       "epoch": 1.5447352601907363,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3178,
+      "loss": 1.2459,
       "step": 95080
     },
     {
       "epoch": 1.544897727088106,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1942,
       "step": 95090
     },
     {
       "epoch": 1.5450601939854756,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2184,
       "step": 95100
     },
     {
       "epoch": 1.5452226608828452,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.1921,
       "step": 95110
     },
     {
       "epoch": 1.5453851277802149,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.1856,
       "step": 95120
     },
     {
       "epoch": 1.5455475946775845,
-      "grad_norm": 4.21875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2304,
       "step": 95130
     },
     {
       "epoch": 1.5457100615749542,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.2398,
       "step": 95140
     },
     {
       "epoch": 1.5458725284723238,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2078,
       "step": 95150
     },
     {
       "epoch": 1.5460349953696935,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2107,
       "step": 95160
     },
     {
       "epoch": 1.5461974622670631,
-      "grad_norm": 2.875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3686,
+      "loss": 1.2531,
       "step": 95170
     },
     {
       "epoch": 1.5463599291644328,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2456,
       "step": 95180
     },
     {
       "epoch": 1.5465223960618024,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.253,
       "step": 95190
     },
     {
       "epoch": 1.546684862959172,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2496,
       "step": 95200
     },
     {
       "epoch": 1.5468473298565417,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2121,
       "step": 95210
     },
     {
       "epoch": 1.5470097967539114,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3645,
+      "loss": 1.1909,
       "step": 95220
     },
     {
       "epoch": 1.547172263651281,
-      "grad_norm": 2.875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.1738,
       "step": 95230
     },
     {
       "epoch": 1.5473347305486507,
-      "grad_norm": 3.21875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.2113,
       "step": 95240
     },
     {
       "epoch": 1.5474971974460203,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2569,
       "step": 95250
     },
     {
       "epoch": 1.54765966434339,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.2428,
       "step": 95260
     },
     {
       "epoch": 1.5478221312407596,
-      "grad_norm": 2.75,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.1826,
       "step": 95270
     },
     {
       "epoch": 1.5479845981381293,
-      "grad_norm": 4.40625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.1856,
       "step": 95280
     },
     {
       "epoch": 1.548147065035499,
-      "grad_norm": 3.375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2185,
       "step": 95290
     },
     {
       "epoch": 1.5483095319328686,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.2122,
       "step": 95300
     },
     {
       "epoch": 1.5484719988302382,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2497,
       "step": 95310
     },
     {
       "epoch": 1.5486344657276079,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.1968,
       "step": 95320
     },
     {
       "epoch": 1.5487969326249775,
-      "grad_norm": 4.03125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2089,
       "step": 95330
     },
     {
       "epoch": 1.5489593995223472,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.376,
+      "loss": 1.1976,
       "step": 95340
     },
     {
       "epoch": 1.5491218664197168,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2241,
       "step": 95350
     },
     {
       "epoch": 1.5492843333170865,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.1885,
       "step": 95360
     },
     {
       "epoch": 1.5494468002144564,
-      "grad_norm": 2.40625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2155,
       "step": 95370
     },
     {
       "epoch": 1.549609267111826,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2451,
       "step": 95380
     },
     {
       "epoch": 1.5497717340091957,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2209,
       "step": 95390
     },
     {
       "epoch": 1.5499342009065653,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2222,
       "step": 95400
     },
     {
       "epoch": 1.550096667803935,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2132,
       "step": 95410
     },
     {
       "epoch": 1.5502591347013046,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2143,
       "step": 95420
     },
     {
       "epoch": 1.5504216015986743,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3145,
+      "loss": 1.2176,
       "step": 95430
     },
     {
       "epoch": 1.550584068496044,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1966,
       "step": 95440
     },
     {
       "epoch": 1.5507465353934136,
-      "grad_norm": 3.84375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2399,
       "step": 95450
     },
     {
       "epoch": 1.5509090022907832,
-      "grad_norm": 1.953125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.2066,
       "step": 95460
     },
     {
       "epoch": 1.5510714691881529,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2008,
       "step": 95470
     },
     {
       "epoch": 1.5512339360855227,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2045,
       "step": 95480
     },
     {
       "epoch": 1.5513964029828924,
-      "grad_norm": 3.4375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.1942,
       "step": 95490
     },
     {
       "epoch": 1.551558869880262,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2407,
       "step": 95500
     },
     {
       "epoch": 1.5517213367776317,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2531,
       "step": 95510
     },
     {
       "epoch": 1.5518838036750013,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.1898,
       "step": 95520
     },
     {
       "epoch": 1.552046270572371,
-      "grad_norm": 3.125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.1856,
       "step": 95530
     },
     {
       "epoch": 1.5522087374697406,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.2232,
       "step": 95540
     },
     {
       "epoch": 1.5523712043671103,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.1927,
       "step": 95550
     },
     {
       "epoch": 1.55253367126448,
-      "grad_norm": 2.234375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2248,
       "step": 95560
     },
     {
       "epoch": 1.5526961381618496,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.235,
       "step": 95570
     },
     {
       "epoch": 1.5528586050592192,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2497,
       "step": 95580
     },
     {
       "epoch": 1.553021071956589,
-      "grad_norm": 3.765625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.2111,
       "step": 95590
     },
     {
       "epoch": 1.5531835388539585,
-      "grad_norm": 4.15625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2477,
       "step": 95600
     },
     {
       "epoch": 1.5533460057513282,
-      "grad_norm": 3.375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2452,
       "step": 95610
     },
     {
       "epoch": 1.5535084726486978,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2223,
       "step": 95620
     },
     {
       "epoch": 1.5536709395460675,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.2272,
       "step": 95630
     },
     {
       "epoch": 1.5538334064434371,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2094,
       "step": 95640
     },
     {
       "epoch": 1.5539958733408068,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.1566,
       "step": 95650
     },
     {
       "epoch": 1.5541583402381765,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2687,
       "step": 95660
     },
     {
       "epoch": 1.554320807135546,
-      "grad_norm": 3.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3145,
+      "loss": 1.1829,
       "step": 95670
     },
     {
       "epoch": 1.5544832740329158,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2068,
       "step": 95680
     },
     {
       "epoch": 1.5546457409302854,
-      "grad_norm": 3.015625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.2064,
       "step": 95690
     },
     {
       "epoch": 1.554808207827655,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2504,
       "step": 95700
     },
     {
       "epoch": 1.5549706747250247,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.2016,
       "step": 95710
     },
     {
       "epoch": 1.5551331416223944,
-      "grad_norm": 3.5,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2631,
       "step": 95720
     },
     {
       "epoch": 1.555295608519764,
-      "grad_norm": 3.125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3661,
+      "loss": 1.2284,
       "step": 95730
     },
     {
       "epoch": 1.5554580754171337,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2219,
       "step": 95740
     },
     {
       "epoch": 1.5556205423145033,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.1964,
       "step": 95750
     },
     {
       "epoch": 1.555783009211873,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2136,
       "step": 95760
     },
     {
       "epoch": 1.5559454761092426,
-      "grad_norm": 1.8828125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2299,
       "step": 95770
     },
     {
       "epoch": 1.5561079430066123,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.2543,
       "step": 95780
     },
     {
       "epoch": 1.556270409903982,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2315,
       "step": 95790
     },
     {
       "epoch": 1.5564328768013516,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2117,
       "step": 95800
     },
     {
       "epoch": 1.5565953436987214,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2049,
       "step": 95810
     },
     {
       "epoch": 1.556757810596091,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.209,
       "step": 95820
     },
     {
       "epoch": 1.5569202774934607,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2401,
       "step": 95830
     },
     {
       "epoch": 1.5570827443908304,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2681,
       "step": 95840
     },
     {
       "epoch": 1.5572452112882,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2223,
       "step": 95850
     },
     {
       "epoch": 1.5574076781855697,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2163,
       "step": 95860
     },
     {
       "epoch": 1.5575701450829393,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2199,
       "step": 95870
     },
     {
       "epoch": 1.557732611980309,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3247,
+      "loss": 1.1967,
       "step": 95880
     },
     {
       "epoch": 1.5578950788776786,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3217,
+      "loss": 1.1828,
       "step": 95890
     },
     {
       "epoch": 1.5580575457750483,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2302,
       "step": 95900
     },
     {
       "epoch": 1.558220012672418,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.1839,
       "step": 95910
     },
     {
       "epoch": 1.5583824795697878,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2355,
       "step": 95920
     },
     {
       "epoch": 1.5585449464671575,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2371,
       "step": 95930
     },
     {
       "epoch": 1.5587074133645271,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3616,
+      "loss": 1.2233,
       "step": 95940
     },
     {
       "epoch": 1.5588698802618968,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.1903,
       "step": 95950
     },
     {
       "epoch": 1.5590323471592664,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3614,
+      "loss": 1.2517,
       "step": 95960
     },
     {
       "epoch": 1.559194814056636,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1982,
       "step": 95970
     },
     {
       "epoch": 1.5593572809540057,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2308,
       "step": 95980
     },
     {
       "epoch": 1.5595197478513754,
-      "grad_norm": 1.90625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2197,
       "step": 95990
     },
     {
       "epoch": 1.559682214748745,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2495,
       "step": 96000
     },
     {
       "epoch": 1.5598446816461147,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2128,
       "step": 96010
     },
     {
       "epoch": 1.5600071485434843,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2195,
       "step": 96020
     },
     {
       "epoch": 1.560169615440854,
-      "grad_norm": 2.703125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2064,
       "step": 96030
     },
     {
       "epoch": 1.5603320823382236,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3093,
+      "loss": 1.2327,
       "step": 96040
     },
     {
       "epoch": 1.5604945492355933,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2329,
       "step": 96050
     },
     {
       "epoch": 1.560657016132963,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2341,
       "step": 96060
     },
     {
       "epoch": 1.5608194830303326,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.256,
       "step": 96070
     },
     {
       "epoch": 1.5609819499277022,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2115,
       "step": 96080
     },
     {
       "epoch": 1.5611444168250719,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2064,
       "step": 96090
     },
     {
       "epoch": 1.5613068837224415,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.191,
       "step": 96100
     },
     {
       "epoch": 1.5614693506198112,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.2013,
       "step": 96110
     },
     {
       "epoch": 1.5616318175171808,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2076,
       "step": 96120
     },
     {
       "epoch": 1.5617942844145505,
-      "grad_norm": 3.125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3687,
+      "loss": 1.2205,
       "step": 96130
     },
     {
       "epoch": 1.5619567513119201,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.2161,
       "step": 96140
     },
     {
       "epoch": 1.5621192182092898,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2036,
       "step": 96150
     },
     {
       "epoch": 1.5622816851066594,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2083,
       "step": 96160
     },
     {
       "epoch": 1.562444152004029,
-      "grad_norm": 1.9453125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2347,
       "step": 96170
     },
     {
       "epoch": 1.5626066189013987,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2579,
       "step": 96180
     },
     {
       "epoch": 1.5627690857987684,
-      "grad_norm": 3.453125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.366,
+      "loss": 1.2135,
       "step": 96190
     },
     {
       "epoch": 1.562931552696138,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.2139,
       "step": 96200
     },
     {
       "epoch": 1.5630940195935077,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2468,
       "step": 96210
     },
     {
       "epoch": 1.5632564864908773,
-      "grad_norm": 3.25,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2396,
       "step": 96220
     },
     {
       "epoch": 1.563418953388247,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2055,
       "step": 96230
     },
     {
       "epoch": 1.5635814202856166,
-      "grad_norm": 2.8125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.1956,
       "step": 96240
     },
     {
       "epoch": 1.5637438871829865,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.218,
       "step": 96250
     },
     {
       "epoch": 1.5639063540803562,
-      "grad_norm": 2.125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2394,
       "step": 96260
     },
     {
       "epoch": 1.5640688209777258,
-      "grad_norm": 3.921875,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2214,
       "step": 96270
     },
     {
       "epoch": 1.5642312878750955,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3726,
+      "loss": 1.2646,
       "step": 96280
     },
     {
       "epoch": 1.5643937547724651,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.22,
       "step": 96290
     },
     {
       "epoch": 1.5645562216698348,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3579,
+      "loss": 1.1956,
       "step": 96300
     },
     {
       "epoch": 1.5647186885672044,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2058,
       "step": 96310
     },
     {
       "epoch": 1.564881155464574,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.1908,
       "step": 96320
     },
     {
       "epoch": 1.5650436223619437,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 1.2419,
       "step": 96330
     },
     {
       "epoch": 1.5652060892593134,
-      "grad_norm": 3.0,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.2216,
       "step": 96340
     },
     {
       "epoch": 1.565368556156683,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.1998,
       "step": 96350
     },
     {
       "epoch": 1.565531023054053,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.1965,
       "step": 96360
     },
     {
       "epoch": 1.5656934899514225,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.2554,
       "step": 96370
     },
     {
       "epoch": 1.5658559568487922,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2254,
       "step": 96380
     },
     {
       "epoch": 1.5660184237461618,
-      "grad_norm": 3.125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.1941,
       "step": 96390
     },
     {
       "epoch": 1.5661808906435315,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3626,
+      "loss": 1.1775,
       "step": 96400
     },
     {
       "epoch": 1.5663433575409011,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.2103,
       "step": 96410
     },
     {
       "epoch": 1.5665058244382708,
-      "grad_norm": 3.125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2077,
       "step": 96420
     },
     {
       "epoch": 1.5666682913356405,
-      "grad_norm": 2.375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3119,
+      "loss": 1.2319,
       "step": 96430
     },
     {
       "epoch": 1.56683075823301,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2752,
       "step": 96440
     },
     {
       "epoch": 1.5669932251303798,
-      "grad_norm": 4.0,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.246,
       "step": 96450
     },
     {
       "epoch": 1.5671556920277494,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2344,
       "step": 96460
     },
     {
       "epoch": 1.567318158925119,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.2229,
       "step": 96470
     },
     {
       "epoch": 1.5674806258224887,
-      "grad_norm": 2.25,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2205,
       "step": 96480
     },
     {
       "epoch": 1.5676430927198584,
-      "grad_norm": 2.0,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.199,
       "step": 96490
     },
     {
       "epoch": 1.567805559617228,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2554,
       "step": 96500
     },
     {
       "epoch": 1.5679680265145977,
-      "grad_norm": 2.625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.1753,
       "step": 96510
     },
     {
       "epoch": 1.5681304934119673,
-      "grad_norm": 1.984375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2818,
       "step": 96520
     },
     {
       "epoch": 1.568292960309337,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2204,
       "step": 96530
     },
     {
       "epoch": 1.5684554272067066,
-      "grad_norm": 2.875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3636,
+      "loss": 1.2388,
       "step": 96540
     },
     {
       "epoch": 1.5686178941040763,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.2451,
       "step": 96550
     },
     {
       "epoch": 1.568780361001446,
-      "grad_norm": 4.125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2481,
       "step": 96560
     },
     {
       "epoch": 1.5689428278988156,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3141,
+      "loss": 1.2139,
       "step": 96570
     },
     {
       "epoch": 1.5691052947961852,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2372,
       "step": 96580
     },
     {
       "epoch": 1.5692677616935549,
-      "grad_norm": 2.25,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2299,
       "step": 96590
     },
     {
       "epoch": 1.5694302285909245,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2739,
       "step": 96600
     },
     {
       "epoch": 1.5695926954882942,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2149,
       "step": 96610
     },
     {
       "epoch": 1.5697551623856638,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2471,
       "step": 96620
     },
     {
       "epoch": 1.5699176292830335,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2111,
       "step": 96630
     },
     {
       "epoch": 1.5700800961804031,
-      "grad_norm": 2.625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2322,
       "step": 96640
     },
     {
       "epoch": 1.5702425630777728,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2074,
       "step": 96650
     },
     {
       "epoch": 1.5704050299751424,
-      "grad_norm": 2.625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2419,
       "step": 96660
     },
     {
       "epoch": 1.570567496872512,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2265,
       "step": 96670
     },
     {
       "epoch": 1.5707299637698817,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2342,
       "step": 96680
     },
     {
       "epoch": 1.5708924306672516,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.216,
       "step": 96690
     },
     {
       "epoch": 1.5710548975646212,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2053,
       "step": 96700
     },
     {
       "epoch": 1.571217364461991,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2637,
       "step": 96710
     },
     {
       "epoch": 1.5713798313593605,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2278,
       "step": 96720
     },
     {
       "epoch": 1.5715422982567302,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.2326,
       "step": 96730
     },
     {
       "epoch": 1.5717047651540998,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2245,
       "step": 96740
     },
     {
       "epoch": 1.5718672320514695,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2096,
       "step": 96750
     },
     {
       "epoch": 1.5720296989488391,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.1906,
       "step": 96760
     },
     {
       "epoch": 1.5721921658462088,
-      "grad_norm": 3.25,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2621,
       "step": 96770
     },
     {
       "epoch": 1.5723546327435785,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.1793,
       "step": 96780
     },
     {
       "epoch": 1.572517099640948,
-      "grad_norm": 3.25,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2553,
       "step": 96790
     },
     {
       "epoch": 1.572679566538318,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2398,
       "step": 96800
     },
     {
       "epoch": 1.5728420334356876,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3715,
+      "loss": 1.1865,
       "step": 96810
     },
     {
       "epoch": 1.5730045003330573,
-      "grad_norm": 2.5,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2045,
       "step": 96820
     },
     {
       "epoch": 1.573166967230427,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2494,
       "step": 96830
     },
     {
       "epoch": 1.5733294341277966,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.1817,
       "step": 96840
     },
     {
       "epoch": 1.5734919010251662,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2323,
       "step": 96850
     },
     {
       "epoch": 1.5736543679225359,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2476,
       "step": 96860
     },
     {
       "epoch": 1.5738168348199055,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3599,
+      "loss": 1.2139,
       "step": 96870
     },
     {
       "epoch": 1.5739793017172752,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2074,
       "step": 96880
     },
     {
       "epoch": 1.5741417686146448,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2333,
       "step": 96890
     },
     {
       "epoch": 1.5743042355120145,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2225,
       "step": 96900
     },
     {
       "epoch": 1.5744667024093841,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2065,
       "step": 96910
     },
     {
       "epoch": 1.5746291693067538,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2319,
       "step": 96920
     },
     {
       "epoch": 1.5747916362041234,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2593,
       "step": 96930
     },
     {
       "epoch": 1.574954103101493,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.2237,
       "step": 96940
     },
     {
       "epoch": 1.5751165699988627,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2092,
       "step": 96950
     },
     {
       "epoch": 1.5752790368962324,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2501,
       "step": 96960
     },
     {
       "epoch": 1.575441503793602,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2045,
       "step": 96970
     },
     {
       "epoch": 1.5756039706909717,
-      "grad_norm": 2.75,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.2296,
       "step": 96980
     },
     {
       "epoch": 1.5757664375883413,
-      "grad_norm": 4.40625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.1942,
       "step": 96990
     },
     {
       "epoch": 1.575928904485711,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2206,
       "step": 97000
     },
     {
       "epoch": 1.5760913713830806,
-      "grad_norm": 2.25,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2808,
       "step": 97010
     },
     {
       "epoch": 1.5762538382804503,
-      "grad_norm": 4.15625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.2224,
       "step": 97020
     },
     {
       "epoch": 1.57641630517782,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3249,
+      "loss": 1.2113,
       "step": 97030
     },
     {
       "epoch": 1.5765787720751896,
-      "grad_norm": 4.1875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2689,
       "step": 97040
     },
     {
       "epoch": 1.5767412389725592,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.1926,
       "step": 97050
     },
     {
       "epoch": 1.576903705869929,
-      "grad_norm": 2.5,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.1847,
       "step": 97060
     },
     {
       "epoch": 1.5770661727672985,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2427,
       "step": 97070
     },
     {
       "epoch": 1.5772286396646682,
-      "grad_norm": 2.125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.2052,
       "step": 97080
     },
     {
       "epoch": 1.5773911065620378,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2733,
       "step": 97090
     },
     {
       "epoch": 1.5775535734594075,
-      "grad_norm": 4.46875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.197,
       "step": 97100
     },
     {
       "epoch": 1.5777160403567771,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.1841,
       "step": 97110
     },
     {
       "epoch": 1.577878507254147,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3051,
+      "loss": 1.2299,
       "step": 97120
     },
     {
       "epoch": 1.5780409741515167,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2001,
       "step": 97130
     },
     {
       "epoch": 1.5782034410488863,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2029,
       "step": 97140
     },
     {
       "epoch": 1.578365907946256,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2117,
       "step": 97150
     },
     {
       "epoch": 1.5785283748436256,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2466,
       "step": 97160
     },
     {
       "epoch": 1.5786908417409953,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1875,
       "step": 97170
     },
     {
       "epoch": 1.578853308638365,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.218,
       "step": 97180
     },
     {
       "epoch": 1.5790157755357346,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2029,
       "step": 97190
     },
     {
       "epoch": 1.5791782424331042,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1898,
       "step": 97200
     },
     {
       "epoch": 1.5793407093304739,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.1862,
       "step": 97210
     },
     {
       "epoch": 1.5795031762278435,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2226,
       "step": 97220
     },
     {
       "epoch": 1.5796656431252134,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.2229,
       "step": 97230
     },
     {
       "epoch": 1.579828110022583,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.194,
       "step": 97240
     },
     {
       "epoch": 1.5799905769199527,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2277,
       "step": 97250
     },
     {
       "epoch": 1.5801530438173224,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3121,
+      "loss": 1.2336,
       "step": 97260
     },
     {
       "epoch": 1.580315510714692,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.249,
       "step": 97270
     },
     {
       "epoch": 1.5804779776120617,
-      "grad_norm": 3.125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.268,
       "step": 97280
     },
     {
       "epoch": 1.5806404445094313,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2306,
       "step": 97290
     },
     {
       "epoch": 1.580802911406801,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2104,
       "step": 97300
     },
     {
       "epoch": 1.5809653783041706,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2208,
       "step": 97310
     },
     {
       "epoch": 1.5811278452015403,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.1961,
       "step": 97320
     },
     {
       "epoch": 1.58129031209891,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2276,
       "step": 97330
     },
     {
       "epoch": 1.5814527789962796,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3206,
+      "loss": 1.2364,
       "step": 97340
     },
     {
       "epoch": 1.5816152458936492,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2159,
       "step": 97350
     },
     {
       "epoch": 1.5817777127910189,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2223,
       "step": 97360
     },
     {
       "epoch": 1.5819401796883885,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2201,
       "step": 97370
     },
     {
       "epoch": 1.5821026465857582,
-      "grad_norm": 2.75,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3598,
+      "loss": 1.2375,
       "step": 97380
     },
     {
       "epoch": 1.5822651134831278,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2339,
       "step": 97390
     },
     {
       "epoch": 1.5824275803804975,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.2133,
       "step": 97400
     },
     {
       "epoch": 1.5825900472778671,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2175,
       "step": 97410
     },
     {
       "epoch": 1.5827525141752368,
-      "grad_norm": 2.5,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.2091,
       "step": 97420
     },
     {
       "epoch": 1.5829149810726064,
-      "grad_norm": 2.328125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2518,
       "step": 97430
     },
     {
       "epoch": 1.583077447969976,
-      "grad_norm": 2.875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2072,
       "step": 97440
     },
     {
       "epoch": 1.5832399148673457,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2067,
       "step": 97450
     },
     {
       "epoch": 1.5834023817647154,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.215,
       "step": 97460
     },
     {
       "epoch": 1.583564848662085,
-      "grad_norm": 3.796875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.2436,
       "step": 97470
     },
     {
       "epoch": 1.5837273155594547,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.1693,
       "step": 97480
     },
     {
       "epoch": 1.5838897824568243,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2432,
       "step": 97490
     },
     {
       "epoch": 1.584052249354194,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2021,
       "step": 97500
     },
     {
       "epoch": 1.5842147162515636,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2172,
       "step": 97510
     },
     {
       "epoch": 1.5843771831489333,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.1977,
       "step": 97520
     },
     {
       "epoch": 1.584539650046303,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2392,
       "step": 97530
     },
     {
       "epoch": 1.5847021169436726,
-      "grad_norm": 4.375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2176,
       "step": 97540
     },
     {
       "epoch": 1.5848645838410422,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2245,
       "step": 97550
     },
     {
       "epoch": 1.585027050738412,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2336,
       "step": 97560
     },
     {
       "epoch": 1.5851895176357818,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2335,
       "step": 97570
     },
     {
       "epoch": 1.5853519845331514,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2384,
       "step": 97580
     },
     {
       "epoch": 1.585514451430521,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.1819,
       "step": 97590
     },
     {
       "epoch": 1.5856769183278907,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2299,
       "step": 97600
     },
     {
       "epoch": 1.5858393852252604,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2243,
       "step": 97610
     },
     {
       "epoch": 1.58600185212263,
-      "grad_norm": 2.875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.236,
       "step": 97620
     },
     {
       "epoch": 1.5861643190199997,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2452,
       "step": 97630
     },
     {
       "epoch": 1.5863267859173693,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.1761,
       "step": 97640
     },
     {
       "epoch": 1.586489252814739,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.24,
       "step": 97650
     },
     {
       "epoch": 1.5866517197121086,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2271,
       "step": 97660
     },
     {
       "epoch": 1.5868141866094785,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2493,
       "step": 97670
     },
     {
       "epoch": 1.5869766535068481,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2212,
       "step": 97680
     },
     {
       "epoch": 1.5871391204042178,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2182,
       "step": 97690
     },
     {
       "epoch": 1.5873015873015874,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2082,
       "step": 97700
     },
     {
       "epoch": 1.587464054198957,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.2239,
       "step": 97710
     },
     {
       "epoch": 1.5876265210963267,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2493,
       "step": 97720
     },
     {
       "epoch": 1.5877889879936964,
-      "grad_norm": 6.75,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2408,
       "step": 97730
     },
     {
       "epoch": 1.587951454891066,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2261,
       "step": 97740
     },
     {
       "epoch": 1.5881139217884357,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2273,
       "step": 97750
     },
     {
       "epoch": 1.5882763886858053,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.1774,
       "step": 97760
     },
     {
       "epoch": 1.588438855583175,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2452,
       "step": 97770
     },
     {
       "epoch": 1.5886013224805446,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.1898,
       "step": 97780
     },
     {
       "epoch": 1.5887637893779143,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3106,
+      "loss": 1.2218,
       "step": 97790
     },
     {
       "epoch": 1.588926256275284,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2084,
       "step": 97800
     },
     {
       "epoch": 1.5890887231726536,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2236,
       "step": 97810
     },
     {
       "epoch": 1.5892511900700232,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.207,
       "step": 97820
     },
     {
       "epoch": 1.589413656967393,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3669,
+      "loss": 1.2564,
       "step": 97830
     },
     {
       "epoch": 1.5895761238647625,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2143,
       "step": 97840
     },
     {
       "epoch": 1.5897385907621322,
-      "grad_norm": 3.75,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.1956,
       "step": 97850
     },
     {
       "epoch": 1.5899010576595018,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3676,
+      "loss": 1.223,
       "step": 97860
     },
     {
       "epoch": 1.5900635245568715,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3601,
+      "loss": 1.2207,
       "step": 97870
     },
     {
       "epoch": 1.5902259914542412,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1759,
       "step": 97880
     },
     {
       "epoch": 1.5903884583516108,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2365,
       "step": 97890
     },
     {
       "epoch": 1.5905509252489805,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.1914,
       "step": 97900
     },
     {
       "epoch": 1.59071339214635,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2012,
       "step": 97910
     },
     {
       "epoch": 1.5908758590437198,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2139,
       "step": 97920
     },
     {
       "epoch": 1.5910383259410894,
-      "grad_norm": 2.765625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2316,
       "step": 97930
     },
     {
       "epoch": 1.591200792838459,
-      "grad_norm": 3.125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2418,
       "step": 97940
     },
     {
       "epoch": 1.5913632597358287,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2543,
       "step": 97950
     },
     {
       "epoch": 1.5915257266331984,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.1935,
       "step": 97960
     },
     {
       "epoch": 1.591688193530568,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.234,
       "step": 97970
     },
     {
       "epoch": 1.5918506604279377,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2116,
       "step": 97980
     },
     {
       "epoch": 1.5920131273253073,
-      "grad_norm": 3.4375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2471,
       "step": 97990
     },
     {
       "epoch": 1.5921755942226772,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.2231,
       "step": 98000
     },
     {
       "epoch": 1.5923380611200468,
-      "grad_norm": 4.625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3594,
+      "loss": 1.2252,
       "step": 98010
     },
     {
       "epoch": 1.5925005280174165,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2171,
       "step": 98020
     },
     {
       "epoch": 1.5926629949147861,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.1965,
       "step": 98030
     },
     {
       "epoch": 1.5928254618121558,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2092,
       "step": 98040
     },
     {
       "epoch": 1.5929879287095254,
-      "grad_norm": 2.875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2144,
       "step": 98050
     },
     {
       "epoch": 1.593150395606895,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2152,
       "step": 98060
     },
     {
       "epoch": 1.5933128625042647,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.1924,
       "step": 98070
     },
     {
       "epoch": 1.5934753294016344,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2097,
       "step": 98080
     },
     {
       "epoch": 1.593637796299004,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.206,
       "step": 98090
     },
     {
       "epoch": 1.5938002631963737,
-      "grad_norm": 3.0,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.1986,
       "step": 98100
     },
     {
       "epoch": 1.5939627300937436,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.202,
       "step": 98110
     },
     {
       "epoch": 1.5941251969911132,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.2286,
       "step": 98120
     },
     {
       "epoch": 1.5942876638884829,
-      "grad_norm": 2.0,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1981,
       "step": 98130
     },
     {
       "epoch": 1.5944501307858525,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.1665,
       "step": 98140
     },
     {
       "epoch": 1.5946125976832222,
-      "grad_norm": 3.0,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2367,
       "step": 98150
     },
     {
       "epoch": 1.5947750645805918,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2282,
       "step": 98160
     },
     {
       "epoch": 1.5949375314779615,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2224,
       "step": 98170
     },
     {
       "epoch": 1.5950999983753311,
-      "grad_norm": 2.25,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.1831,
       "step": 98180
     },
     {
       "epoch": 1.5952624652727008,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2723,
       "step": 98190
     },
     {
       "epoch": 1.5954249321700704,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.2375,
       "step": 98200
     },
     {
       "epoch": 1.59558739906744,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3635,
+      "loss": 1.1683,
       "step": 98210
     },
     {
       "epoch": 1.5957498659648097,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2375,
       "step": 98220
     },
     {
       "epoch": 1.5959123328621794,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.234,
       "step": 98230
     },
     {
       "epoch": 1.596074799759549,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.2749,
       "step": 98240
     },
     {
       "epoch": 1.5962372666569187,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.1723,
       "step": 98250
     },
     {
       "epoch": 1.5963997335542883,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2505,
       "step": 98260
     },
     {
       "epoch": 1.596562200451658,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2412,
       "step": 98270
     },
     {
       "epoch": 1.5967246673490276,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.2584,
       "step": 98280
     },
     {
       "epoch": 1.5968871342463973,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.247,
       "step": 98290
     },
     {
       "epoch": 1.597049601143767,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3091,
+      "loss": 1.2147,
       "step": 98300
     },
     {
       "epoch": 1.5972120680411366,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.243,
       "step": 98310
     },
     {
       "epoch": 1.5973745349385062,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.1833,
       "step": 98320
     },
     {
       "epoch": 1.5975370018358759,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.1926,
       "step": 98330
     },
     {
       "epoch": 1.5976994687332455,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2378,
       "step": 98340
     },
     {
       "epoch": 1.5978619356306152,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2441,
       "step": 98350
     },
     {
       "epoch": 1.5980244025279848,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.241,
       "step": 98360
     },
     {
       "epoch": 1.5981868694253545,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2199,
       "step": 98370
     },
     {
       "epoch": 1.5983493363227241,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.2272,
       "step": 98380
     },
     {
       "epoch": 1.5985118032200938,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.2038,
       "step": 98390
     },
     {
       "epoch": 1.5986742701174634,
-      "grad_norm": 2.0,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.1897,
       "step": 98400
     },
     {
       "epoch": 1.598836737014833,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2454,
       "step": 98410
     },
     {
       "epoch": 1.5989992039122027,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.2079,
       "step": 98420
     },
     {
       "epoch": 1.5991616708095724,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2121,
       "step": 98430
     },
     {
       "epoch": 1.5993241377069423,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2064,
       "step": 98440
     },
     {
       "epoch": 1.599486604604312,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2157,
       "step": 98450
     },
     {
       "epoch": 1.5996490715016816,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.2291,
       "step": 98460
     },
     {
       "epoch": 1.5998115383990512,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2298,
       "step": 98470
     },
     {
       "epoch": 1.5999740052964209,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3675,
+      "loss": 1.1992,
       "step": 98480
     },
     {
       "epoch": 1.6001364721937905,
-      "grad_norm": 3.25,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1275,
       "step": 98490
     },
     {
       "epoch": 1.6002989390911602,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3558,
+      "loss": 1.2078,
       "step": 98500
     },
     {
       "epoch": 1.6004614059885298,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2562,
       "step": 98510
     },
     {
       "epoch": 1.6006238728858995,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2068,
       "step": 98520
     },
     {
       "epoch": 1.6007863397832691,
-      "grad_norm": 3.5,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2022,
       "step": 98530
     },
     {
       "epoch": 1.6009488066806388,
-      "grad_norm": 3.984375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2196,
       "step": 98540
     },
     {
       "epoch": 1.6011112735780086,
-      "grad_norm": 2.5,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.1954,
       "step": 98550
     },
     {
       "epoch": 1.6012737404753783,
-      "grad_norm": 2.375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.1936,
       "step": 98560
     },
     {
       "epoch": 1.601436207372748,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2367,
       "step": 98570
     },
     {
       "epoch": 1.6015986742701176,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.227,
       "step": 98580
     },
     {
       "epoch": 1.6017611411674872,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3221,
+      "loss": 1.2337,
       "step": 98590
     },
     {
       "epoch": 1.601923608064857,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2245,
       "step": 98600
     },
     {
       "epoch": 1.6020860749622265,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2345,
       "step": 98610
     },
     {
       "epoch": 1.6022485418595962,
-      "grad_norm": 3.25,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.1829,
       "step": 98620
     },
     {
       "epoch": 1.6024110087569658,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.1975,
       "step": 98630
     },
     {
       "epoch": 1.6025734756543355,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2193,
       "step": 98640
     },
     {
       "epoch": 1.6027359425517052,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2187,
       "step": 98650
     },
     {
       "epoch": 1.6028984094490748,
-      "grad_norm": 4.375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3592,
+      "loss": 1.2415,
       "step": 98660
     },
     {
       "epoch": 1.6030608763464445,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2488,
       "step": 98670
     },
     {
       "epoch": 1.603223343243814,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2149,
       "step": 98680
     },
     {
       "epoch": 1.6033858101411838,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2179,
       "step": 98690
     },
     {
       "epoch": 1.6035482770385534,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2066,
       "step": 98700
     },
     {
       "epoch": 1.603710743935923,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.1733,
       "step": 98710
     },
     {
       "epoch": 1.6038732108332927,
-      "grad_norm": 2.875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2104,
       "step": 98720
     },
     {
       "epoch": 1.6040356777306624,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3093,
+      "loss": 1.2239,
       "step": 98730
     },
     {
       "epoch": 1.604198144628032,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.2461,
       "step": 98740
     },
     {
       "epoch": 1.6043606115254017,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3254,
+      "loss": 1.2329,
       "step": 98750
     },
     {
       "epoch": 1.6045230784227713,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3712,
+      "loss": 1.2303,
       "step": 98760
     },
     {
       "epoch": 1.604685545320141,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2123,
       "step": 98770
     },
     {
       "epoch": 1.6048480122175106,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.1842,
       "step": 98780
     },
     {
       "epoch": 1.6050104791148803,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.1834,
       "step": 98790
     },
     {
       "epoch": 1.60517294601225,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.239,
       "step": 98800
     },
     {
       "epoch": 1.6053354129096196,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2587,
       "step": 98810
     },
     {
       "epoch": 1.6054978798069892,
-      "grad_norm": 1.9609375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2387,
       "step": 98820
     },
     {
       "epoch": 1.6056603467043589,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3693,
+      "loss": 1.2342,
       "step": 98830
     },
     {
       "epoch": 1.6058228136017285,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.2031,
       "step": 98840
     },
     {
       "epoch": 1.6059852804990982,
-      "grad_norm": 4.25,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.248,
       "step": 98850
     },
     {
       "epoch": 1.6061477473964678,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2611,
       "step": 98860
     },
     {
       "epoch": 1.6063102142938375,
-      "grad_norm": 2.5,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2437,
       "step": 98870
     },
     {
       "epoch": 1.6064726811912073,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.2297,
       "step": 98880
     },
     {
       "epoch": 1.606635148088577,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2299,
       "step": 98890
     },
     {
       "epoch": 1.6067976149859466,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2268,
       "step": 98900
     },
     {
       "epoch": 1.6069600818833163,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.232,
       "step": 98910
     },
     {
       "epoch": 1.607122548780686,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.187,
       "step": 98920
     },
     {
       "epoch": 1.6072850156780556,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3155,
+      "loss": 1.254,
       "step": 98930
     },
     {
       "epoch": 1.6074474825754252,
-      "grad_norm": 3.125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2353,
       "step": 98940
     },
     {
       "epoch": 1.607609949472795,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2358,
       "step": 98950
     },
     {
       "epoch": 1.6077724163701645,
-      "grad_norm": 2.875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3564,
+      "loss": 1.2039,
       "step": 98960
     },
     {
       "epoch": 1.6079348832675342,
-      "grad_norm": 2.75,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2287,
       "step": 98970
     },
     {
       "epoch": 1.6080973501649038,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2878,
       "step": 98980
     },
     {
       "epoch": 1.6082598170622737,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2321,
       "step": 98990
     },
     {
       "epoch": 1.6084222839596434,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2308,
       "step": 99000
     },
     {
       "epoch": 1.608584750857013,
-      "grad_norm": 2.5,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2342,
       "step": 99010
     },
     {
       "epoch": 1.6087472177543827,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2065,
       "step": 99020
     },
     {
       "epoch": 1.6089096846517523,
-      "grad_norm": 1.4921875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2371,
       "step": 99030
     },
     {
       "epoch": 1.609072151549122,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2007,
       "step": 99040
     },
     {
       "epoch": 1.6092346184464916,
-      "grad_norm": 3.90625,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.1799,
       "step": 99050
     },
     {
       "epoch": 1.6093970853438613,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.1971,
       "step": 99060
     },
     {
       "epoch": 1.609559552241231,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.1918,
       "step": 99070
     },
     {
       "epoch": 1.6097220191386006,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.234,
       "step": 99080
     },
     {
       "epoch": 1.6098844860359702,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.1802,
       "step": 99090
     },
     {
       "epoch": 1.6100469529333399,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2247,
       "step": 99100
     },
     {
       "epoch": 1.6102094198307095,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.2164,
       "step": 99110
     },
     {
       "epoch": 1.6103718867280792,
-      "grad_norm": 1.8671875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.1939,
       "step": 99120
     },
     {
       "epoch": 1.6105343536254488,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3589,
+      "loss": 1.2161,
       "step": 99130
     },
     {
       "epoch": 1.6106968205228185,
-      "grad_norm": 2.75,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1668,
       "step": 99140
     },
     {
       "epoch": 1.6108592874201881,
-      "grad_norm": 2.203125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2277,
       "step": 99150
     },
     {
       "epoch": 1.6110217543175578,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2186,
       "step": 99160
     },
     {
       "epoch": 1.6111842212149274,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.1761,
       "step": 99170
     },
     {
       "epoch": 1.611346688112297,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2017,
       "step": 99180
     },
     {
       "epoch": 1.6115091550096667,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3185,
+      "loss": 1.1938,
       "step": 99190
     },
     {
       "epoch": 1.6116716219070364,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2136,
       "step": 99200
     },
     {
       "epoch": 1.611834088804406,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2331,
       "step": 99210
     },
     {
       "epoch": 1.6119965557017757,
-      "grad_norm": 2.875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2249,
       "step": 99220
     },
     {
       "epoch": 1.6121590225991453,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2313,
       "step": 99230
     },
     {
       "epoch": 1.612321489496515,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2133,
       "step": 99240
     },
     {
       "epoch": 1.6124839563938846,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2052,
       "step": 99250
     },
     {
       "epoch": 1.6126464232912543,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2449,
       "step": 99260
     },
     {
       "epoch": 1.612808890188624,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2272,
       "step": 99270
     },
     {
       "epoch": 1.6129713570859936,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2204,
       "step": 99280
     },
     {
       "epoch": 1.6131338239833632,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2252,
       "step": 99290
     },
     {
       "epoch": 1.613296290880733,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.1719,
       "step": 99300
     },
     {
       "epoch": 1.6134587577781025,
-      "grad_norm": 3.25,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3601,
+      "loss": 1.1738,
       "step": 99310
     },
     {
       "epoch": 1.6136212246754724,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2104,
       "step": 99320
     },
     {
       "epoch": 1.613783691572842,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2378,
       "step": 99330
     },
     {
       "epoch": 1.6139461584702117,
-      "grad_norm": 4.0,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2054,
       "step": 99340
     },
     {
       "epoch": 1.6141086253675814,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2024,
       "step": 99350
     },
     {
       "epoch": 1.614271092264951,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.236,
       "step": 99360
     },
     {
       "epoch": 1.6144335591623207,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2207,
       "step": 99370
     },
     {
       "epoch": 1.6145960260596903,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.254,
       "step": 99380
     },
     {
       "epoch": 1.61475849295706,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2558,
       "step": 99390
     },
     {
       "epoch": 1.6149209598544296,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.1913,
       "step": 99400
     },
     {
       "epoch": 1.6150834267517993,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2027,
       "step": 99410
     },
     {
       "epoch": 1.615245893649169,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2575,
       "step": 99420
     },
     {
       "epoch": 1.6154083605465388,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2017,
       "step": 99430
     },
     {
       "epoch": 1.6155708274439085,
-      "grad_norm": 3.0,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3574,
+      "loss": 1.2074,
       "step": 99440
     },
     {
       "epoch": 1.615733294341278,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2241,
       "step": 99450
     },
     {
       "epoch": 1.6158957612386478,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.217,
       "step": 99460
     },
     {
       "epoch": 1.6160582281360174,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.2321,
       "step": 99470
     },
     {
       "epoch": 1.616220695033387,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2043,
       "step": 99480
     },
     {
       "epoch": 1.6163831619307567,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.364,
+      "loss": 1.2523,
       "step": 99490
     },
     {
       "epoch": 1.6165456288281264,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2127,
       "step": 99500
     },
     {
       "epoch": 1.616708095725496,
-      "grad_norm": 3.5,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2329,
       "step": 99510
     },
     {
       "epoch": 1.6168705626228657,
-      "grad_norm": 4.3125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.1896,
       "step": 99520
     },
     {
       "epoch": 1.6170330295202353,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.2226,
       "step": 99530
     },
     {
       "epoch": 1.617195496417605,
-      "grad_norm": 4.0,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2501,
       "step": 99540
     },
     {
       "epoch": 1.6173579633149746,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.206,
       "step": 99550
     },
     {
       "epoch": 1.6175204302123443,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2074,
       "step": 99560
     },
     {
       "epoch": 1.617682897109714,
-      "grad_norm": 2.625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1938,
       "step": 99570
     },
     {
       "epoch": 1.6178453640070836,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.2045,
       "step": 99580
     },
     {
       "epoch": 1.6180078309044532,
-      "grad_norm": 2.875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2153,
       "step": 99590
     },
     {
       "epoch": 1.6181702978018229,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2097,
       "step": 99600
     },
     {
       "epoch": 1.6183327646991925,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2065,
       "step": 99610
     },
     {
       "epoch": 1.6184952315965622,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2153,
       "step": 99620
     },
     {
       "epoch": 1.6186576984939318,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1926,
       "step": 99630
     },
     {
       "epoch": 1.6188201653913015,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2199,
       "step": 99640
     },
     {
       "epoch": 1.6189826322886711,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2002,
       "step": 99650
     },
     {
       "epoch": 1.6191450991860408,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.1941,
       "step": 99660
     },
     {
       "epoch": 1.6193075660834104,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.2236,
       "step": 99670
     },
     {
       "epoch": 1.61947003298078,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2114,
       "step": 99680
     },
     {
       "epoch": 1.6196324998781497,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2311,
       "step": 99690
     },
     {
       "epoch": 1.6197949667755194,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2432,
       "step": 99700
     },
     {
       "epoch": 1.619957433672889,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3129,
+      "loss": 1.2189,
       "step": 99710
     },
     {
       "epoch": 1.6201199005702587,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.1896,
       "step": 99720
     },
     {
       "epoch": 1.6202823674676283,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2233,
       "step": 99730
     },
     {
       "epoch": 1.620444834364998,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1911,
       "step": 99740
     },
     {
       "epoch": 1.6206073012623676,
-      "grad_norm": 2.46875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.232,
       "step": 99750
     },
     {
       "epoch": 1.6207697681597375,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.229,
       "step": 99760
     },
     {
       "epoch": 1.6209322350571072,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3195,
+      "loss": 1.191,
       "step": 99770
     },
     {
       "epoch": 1.6210947019544768,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2071,
       "step": 99780
     },
     {
       "epoch": 1.6212571688518465,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2069,
       "step": 99790
     },
     {
       "epoch": 1.621419635749216,
-      "grad_norm": 2.625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.2101,
       "step": 99800
     },
     {
       "epoch": 1.6215821026465858,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2398,
       "step": 99810
     },
     {
       "epoch": 1.6217445695439554,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.216,
       "step": 99820
     },
     {
       "epoch": 1.621907036441325,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.1952,
       "step": 99830
     },
     {
       "epoch": 1.6220695033386947,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2647,
       "step": 99840
     },
     {
       "epoch": 1.6222319702360644,
-      "grad_norm": 3.125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2454,
       "step": 99850
     },
     {
       "epoch": 1.622394437133434,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2115,
       "step": 99860
     },
     {
       "epoch": 1.6225569040308039,
-      "grad_norm": 3.703125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2244,
       "step": 99870
     },
     {
       "epoch": 1.6227193709281735,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2108,
       "step": 99880
     },
     {
       "epoch": 1.6228818378255432,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.2321,
       "step": 99890
     },
     {
       "epoch": 1.6230443047229128,
-      "grad_norm": 3.953125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.1811,
       "step": 99900
     },
     {
       "epoch": 1.6232067716202825,
-      "grad_norm": 3.875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2237,
       "step": 99910
     },
     {
       "epoch": 1.6233692385176521,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2503,
       "step": 99920
     },
     {
       "epoch": 1.6235317054150218,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.2288,
       "step": 99930
     },
     {
       "epoch": 1.6236941723123914,
-      "grad_norm": 3.6875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3582,
+      "loss": 1.2352,
       "step": 99940
     },
     {
       "epoch": 1.623856639209761,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.219,
       "step": 99950
     },
     {
       "epoch": 1.6240191061071307,
-      "grad_norm": 1.7578125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.227,
       "step": 99960
     },
     {
       "epoch": 1.6241815730045004,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3658,
+      "loss": 1.2317,
       "step": 99970
     },
     {
       "epoch": 1.62434403990187,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1796,
       "step": 99980
     },
     {
       "epoch": 1.6245065067992397,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.1989,
       "step": 99990
     },
     {
       "epoch": 1.6246689736966093,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3704,
+      "loss": 1.1921,
       "step": 100000
     },
     {
       "epoch": 1.624831440593979,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.186,
       "step": 100010
     },
     {
       "epoch": 1.6249939074913486,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2049,
       "step": 100020
     },
     {
       "epoch": 1.6251563743887183,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2323,
       "step": 100030
     },
     {
       "epoch": 1.625318841286088,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2023,
       "step": 100040
     },
     {
       "epoch": 1.6254813081834576,
-      "grad_norm": 4.1875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2001,
       "step": 100050
     },
     {
       "epoch": 1.6256437750808272,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2238,
       "step": 100060
     },
     {
       "epoch": 1.625806241978197,
-      "grad_norm": 2.625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3144,
+      "loss": 1.2108,
       "step": 100070
     },
     {
       "epoch": 1.6259687088755665,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.2275,
       "step": 100080
     },
     {
       "epoch": 1.6261311757729362,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2125,
       "step": 100090
     },
     {
       "epoch": 1.6262936426703059,
-      "grad_norm": 3.0,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.2178,
       "step": 100100
     },
     {
       "epoch": 1.6264561095676755,
-      "grad_norm": 3.390625,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2329,
       "step": 100110
     },
     {
       "epoch": 1.6266185764650452,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.224,
       "step": 100120
     },
     {
       "epoch": 1.6267810433624148,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3202,
+      "loss": 1.2187,
       "step": 100130
     },
     {
       "epoch": 1.6269435102597845,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2441,
       "step": 100140
     },
     {
       "epoch": 1.627105977157154,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3589,
+      "loss": 1.2256,
       "step": 100150
     },
     {
       "epoch": 1.6272684440545238,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.1536,
       "step": 100160
     },
     {
       "epoch": 1.6274309109518934,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.261,
       "step": 100170
     },
     {
       "epoch": 1.627593377849263,
-      "grad_norm": 3.875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2349,
       "step": 100180
     },
     {
       "epoch": 1.6277558447466327,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2652,
       "step": 100190
     },
     {
       "epoch": 1.6279183116440026,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2417,
       "step": 100200
     },
     {
       "epoch": 1.6280807785413722,
-      "grad_norm": 2.40625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2385,
       "step": 100210
     },
     {
       "epoch": 1.6282432454387419,
-      "grad_norm": 3.765625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.2184,
       "step": 100220
     },
     {
       "epoch": 1.6284057123361115,
-      "grad_norm": 2.875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2399,
       "step": 100230
     },
     {
       "epoch": 1.6285681792334812,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2159,
       "step": 100240
     },
     {
       "epoch": 1.6287306461308508,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2689,
       "step": 100250
     },
     {
       "epoch": 1.6288931130282205,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1732,
       "step": 100260
     },
     {
       "epoch": 1.6290555799255901,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2093,
       "step": 100270
     },
     {
       "epoch": 1.6292180468229598,
-      "grad_norm": 2.71875,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.1974,
       "step": 100280
     },
     {
       "epoch": 1.6293805137203294,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2356,
       "step": 100290
     },
     {
       "epoch": 1.629542980617699,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.1792,
       "step": 100300
     },
     {
       "epoch": 1.629705447515069,
-      "grad_norm": 4.71875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.1911,
       "step": 100310
     },
     {
       "epoch": 1.6298679144124386,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2509,
       "step": 100320
     },
     {
       "epoch": 1.6300303813098083,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2481,
       "step": 100330
     },
     {
       "epoch": 1.630192848207178,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2002,
       "step": 100340
     },
     {
       "epoch": 1.6303553151045476,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.176,
       "step": 100350
     },
     {
       "epoch": 1.6305177820019172,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2302,
       "step": 100360
     },
     {
       "epoch": 1.6306802488992869,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.1964,
       "step": 100370
     },
     {
       "epoch": 1.6308427157966565,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3634,
+      "loss": 1.221,
       "step": 100380
     },
     {
       "epoch": 1.6310051826940262,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.2417,
       "step": 100390
     },
     {
       "epoch": 1.6311676495913958,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.1958,
       "step": 100400
     },
     {
       "epoch": 1.6313301164887655,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2077,
       "step": 100410
     },
     {
       "epoch": 1.6314925833861351,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.2021,
       "step": 100420
     },
     {
       "epoch": 1.6316550502835048,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3609,
+      "loss": 1.2082,
       "step": 100430
     },
     {
       "epoch": 1.6318175171808744,
-      "grad_norm": 2.5,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.233,
       "step": 100440
     },
     {
       "epoch": 1.631979984078244,
-      "grad_norm": 2.046875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.2357,
       "step": 100450
     },
     {
       "epoch": 1.6321424509756137,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2024,
       "step": 100460
     },
     {
       "epoch": 1.6323049178729834,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2491,
       "step": 100470
     },
     {
       "epoch": 1.632467384770353,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2012,
       "step": 100480
     },
     {
       "epoch": 1.6326298516677227,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.2112,
       "step": 100490
     },
     {
       "epoch": 1.6327923185650923,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.231,
       "step": 100500
     },
     {
       "epoch": 1.632954785462462,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2525,
       "step": 100510
     },
     {
       "epoch": 1.6331172523598316,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2251,
       "step": 100520
     },
     {
       "epoch": 1.6332797192572013,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 1.2153,
       "step": 100530
     },
     {
       "epoch": 1.633442186154571,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2097,
       "step": 100540
     },
     {
       "epoch": 1.6336046530519406,
-      "grad_norm": 3.234375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3247,
+      "loss": 1.2331,
       "step": 100550
     },
     {
       "epoch": 1.6337671199493102,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.1929,
       "step": 100560
     },
     {
       "epoch": 1.6339295868466799,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3638,
+      "loss": 1.2021,
       "step": 100570
     },
     {
       "epoch": 1.6340920537440495,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2265,
       "step": 100580
     },
     {
       "epoch": 1.6342545206414192,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2019,
       "step": 100590
     },
     {
       "epoch": 1.6344169875387888,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2037,
       "step": 100600
     },
     {
       "epoch": 1.6345794544361585,
-      "grad_norm": 3.078125,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2208,
       "step": 100610
     },
     {
       "epoch": 1.6347419213335281,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.2578,
       "step": 100620
     },
     {
       "epoch": 1.634904388230898,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.1705,
       "step": 100630
     },
     {
       "epoch": 1.6350668551282677,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2222,
       "step": 100640
     },
     {
       "epoch": 1.6352293220256373,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.1801,
       "step": 100650
     },
     {
       "epoch": 1.635391788923007,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2618,
       "step": 100660
     },
     {
       "epoch": 1.6355542558203766,
-      "grad_norm": 3.375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2216,
       "step": 100670
     },
     {
       "epoch": 1.6357167227177463,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2314,
       "step": 100680
     },
     {
       "epoch": 1.635879189615116,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.2276,
       "step": 100690
     },
     {
       "epoch": 1.6360416565124856,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2263,
       "step": 100700
     },
     {
       "epoch": 1.6362041234098552,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2034,
       "step": 100710
     },
     {
       "epoch": 1.6363665903072249,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2093,
       "step": 100720
     },
     {
       "epoch": 1.6365290572045945,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2272,
       "step": 100730
     },
     {
       "epoch": 1.6366915241019644,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2157,
       "step": 100740
     },
     {
       "epoch": 1.636853990999334,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.237,
       "step": 100750
     },
     {
       "epoch": 1.6370164578967037,
-      "grad_norm": 3.125,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2272,
       "step": 100760
     },
     {
       "epoch": 1.6371789247940733,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3577,
+      "loss": 1.1932,
       "step": 100770
     },
     {
       "epoch": 1.637341391691443,
-      "grad_norm": 3.375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.251,
       "step": 100780
     },
     {
       "epoch": 1.6375038585888126,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3641,
+      "loss": 1.205,
       "step": 100790
     },
     {
       "epoch": 1.6376663254861823,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2128,
       "step": 100800
     },
     {
       "epoch": 1.637828792383552,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2213,
       "step": 100810
     },
     {
       "epoch": 1.6379912592809216,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2385,
       "step": 100820
     },
     {
       "epoch": 1.6381537261782912,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.1999,
       "step": 100830
     },
     {
       "epoch": 1.638316193075661,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.19,
       "step": 100840
     },
     {
       "epoch": 1.6384786599730305,
-      "grad_norm": 2.859375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2404,
       "step": 100850
     },
     {
       "epoch": 1.6386411268704002,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.2735,
       "step": 100860
     },
     {
       "epoch": 1.6388035937677699,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2517,
       "step": 100870
     },
     {
       "epoch": 1.6389660606651395,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.1915,
       "step": 100880
     },
     {
       "epoch": 1.6391285275625092,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2474,
       "step": 100890
     },
     {
       "epoch": 1.6392909944598788,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2165,
       "step": 100900
     },
     {
       "epoch": 1.6394534613572485,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.1782,
       "step": 100910
     },
     {
       "epoch": 1.639615928254618,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.311,
+      "loss": 1.2315,
       "step": 100920
     },
     {
       "epoch": 1.6397783951519878,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2259,
       "step": 100930
     },
     {
       "epoch": 1.6399408620493574,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.1887,
       "step": 100940
     },
     {
       "epoch": 1.640103328946727,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2096,
       "step": 100950
     },
     {
       "epoch": 1.6402657958440967,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3198,
+      "loss": 1.249,
       "step": 100960
     },
     {
       "epoch": 1.6404282627414664,
-      "grad_norm": 3.9375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2202,
       "step": 100970
     },
     {
       "epoch": 1.640590729638836,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.2375,
       "step": 100980
     },
     {
       "epoch": 1.6407531965362057,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.202,
       "step": 100990
     },
     {
       "epoch": 1.6409156634335753,
-      "grad_norm": 2.625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1937,
       "step": 101000
     },
     {
       "epoch": 1.641078130330945,
-      "grad_norm": 3.8125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2512,
       "step": 101010
     },
     {
       "epoch": 1.6412405972283146,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.2189,
       "step": 101020
     },
     {
       "epoch": 1.6414030641256843,
-      "grad_norm": 1.9296875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2075,
       "step": 101030
     },
     {
       "epoch": 1.641565531023054,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2222,
       "step": 101040
     },
     {
       "epoch": 1.6417279979204236,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1966,
       "step": 101050
     },
     {
       "epoch": 1.6418904648177932,
-      "grad_norm": 2.75,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.1793,
       "step": 101060
     },
     {
       "epoch": 1.642052931715163,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2042,
       "step": 101070
     },
     {
       "epoch": 1.6422153986125327,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2325,
       "step": 101080
     },
     {
       "epoch": 1.6423778655099024,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2087,
       "step": 101090
     },
     {
       "epoch": 1.642540332407272,
-      "grad_norm": 2.9375,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.3723,
+      "loss": 1.2266,
       "step": 101100
     },
     {
       "epoch": 1.6427027993046417,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2283,
       "step": 101110
     },
     {
       "epoch": 1.6428652662020113,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2306,
       "step": 101120
     },
     {
       "epoch": 1.643027733099381,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.203,
       "step": 101130
     },
     {
       "epoch": 1.6431901999967506,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.1944,
       "step": 101140
     },
     {
       "epoch": 1.6433526668941203,
-      "grad_norm": 3.921875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1863,
       "step": 101150
     },
     {
       "epoch": 1.64351513379149,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2256,
       "step": 101160
     },
     {
       "epoch": 1.6436776006888596,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2249,
       "step": 101170
     },
     {
       "epoch": 1.6438400675862295,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.2095,
       "step": 101180
     },
     {
       "epoch": 1.6440025344835991,
-      "grad_norm": 2.625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.1991,
       "step": 101190
     },
     {
       "epoch": 1.6441650013809688,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2499,
       "step": 101200
     },
     {
       "epoch": 1.6443274682783384,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2164,
       "step": 101210
     },
     {
       "epoch": 1.644489935175708,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2198,
       "step": 101220
     },
     {
       "epoch": 1.6446524020730777,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2531,
       "step": 101230
     },
     {
       "epoch": 1.6448148689704474,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.2481,
       "step": 101240
     },
     {
       "epoch": 1.644977335867817,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2312,
       "step": 101250
     },
     {
       "epoch": 1.6451398027651867,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2026,
       "step": 101260
     },
     {
       "epoch": 1.6453022696625563,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.2598,
       "step": 101270
     },
     {
       "epoch": 1.645464736559926,
-      "grad_norm": 2.5,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2205,
       "step": 101280
     },
     {
       "epoch": 1.6456272034572956,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2366,
       "step": 101290
     },
     {
       "epoch": 1.6457896703546653,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3731,
+      "loss": 1.2171,
       "step": 101300
     },
     {
       "epoch": 1.645952137252035,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1943,
       "step": 101310
     },
     {
       "epoch": 1.6461146041494046,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.261,
       "step": 101320
     },
     {
       "epoch": 1.6462770710467742,
-      "grad_norm": 4.40625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.2154,
       "step": 101330
     },
     {
       "epoch": 1.6464395379441439,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.213,
       "step": 101340
     },
     {
       "epoch": 1.6466020048415135,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.251,
       "step": 101350
     },
     {
       "epoch": 1.6467644717388832,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.212,
       "step": 101360
     },
     {
       "epoch": 1.6469269386362528,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2163,
       "step": 101370
     },
     {
       "epoch": 1.6470894055336225,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2024,
       "step": 101380
     },
     {
       "epoch": 1.6472518724309921,
-      "grad_norm": 3.65625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2089,
       "step": 101390
     },
     {
       "epoch": 1.6474143393283618,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3217,
+      "loss": 1.2004,
       "step": 101400
     },
     {
       "epoch": 1.6475768062257314,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2245,
       "step": 101410
     },
     {
       "epoch": 1.647739273123101,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2231,
       "step": 101420
     },
     {
       "epoch": 1.6479017400204707,
-      "grad_norm": 2.5,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.265,
       "step": 101430
     },
     {
       "epoch": 1.6480642069178404,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2231,
       "step": 101440
     },
     {
       "epoch": 1.64822667381521,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3208,
+      "loss": 1.2146,
       "step": 101450
     },
     {
       "epoch": 1.6483891407125797,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2511,
       "step": 101460
     },
     {
       "epoch": 1.6485516076099493,
-      "grad_norm": 3.0,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2197,
       "step": 101470
     },
     {
       "epoch": 1.648714074507319,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3626,
+      "loss": 1.2238,
       "step": 101480
     },
     {
       "epoch": 1.6488765414046886,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.2093,
       "step": 101490
     },
     {
       "epoch": 1.6490390083020583,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.1963,
       "step": 101500
     },
     {
       "epoch": 1.6492014751994282,
-      "grad_norm": 3.125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2443,
       "step": 101510
     },
     {
       "epoch": 1.6493639420967978,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2243,
       "step": 101520
     },
     {
       "epoch": 1.6495264089941675,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3102,
+      "loss": 1.227,
       "step": 101530
     },
     {
       "epoch": 1.6496888758915371,
-      "grad_norm": 4.0,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2262,
       "step": 101540
     },
     {
       "epoch": 1.6498513427889068,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2205,
       "step": 101550
     },
     {
       "epoch": 1.6500138096862764,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2435,
       "step": 101560
     },
     {
       "epoch": 1.650176276583646,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2253,
       "step": 101570
     },
     {
       "epoch": 1.6503387434810157,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2088,
       "step": 101580
     },
     {
       "epoch": 1.6505012103783854,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2089,
       "step": 101590
     },
     {
       "epoch": 1.650663677275755,
-      "grad_norm": 3.0,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.1871,
       "step": 101600
     },
     {
       "epoch": 1.6508261441731247,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2195,
       "step": 101610
     },
     {
       "epoch": 1.6509886110704945,
-      "grad_norm": 3.78125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.1994,
       "step": 101620
     },
     {
       "epoch": 1.6511510779678642,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.362,
+      "loss": 1.2378,
       "step": 101630
     },
     {
       "epoch": 1.6513135448652339,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.191,
       "step": 101640
     },
     {
       "epoch": 1.6514760117626035,
-      "grad_norm": 3.828125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2226,
       "step": 101650
     },
     {
       "epoch": 1.6516384786599732,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2007,
       "step": 101660
     },
     {
       "epoch": 1.6518009455573428,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.1875,
       "step": 101670
     },
     {
       "epoch": 1.6519634124547125,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3221,
+      "loss": 1.1519,
       "step": 101680
     },
     {
       "epoch": 1.652125879352082,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2199,
       "step": 101690
     },
     {
       "epoch": 1.6522883462494518,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3639,
+      "loss": 1.2458,
       "step": 101700
     },
     {
       "epoch": 1.6524508131468214,
-      "grad_norm": 3.625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2101,
       "step": 101710
     },
     {
       "epoch": 1.652613280044191,
-      "grad_norm": 2.546875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.216,
       "step": 101720
     },
     {
       "epoch": 1.6527757469415607,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2355,
       "step": 101730
     },
     {
       "epoch": 1.6529382138389304,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.1968,
       "step": 101740
     },
     {
       "epoch": 1.6531006807363,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2344,
       "step": 101750
     },
     {
       "epoch": 1.6532631476336697,
-      "grad_norm": 2.625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2608,
       "step": 101760
     },
     {
       "epoch": 1.6534256145310393,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.1925,
       "step": 101770
     },
     {
       "epoch": 1.653588081428409,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2544,
       "step": 101780
     },
     {
       "epoch": 1.6537505483257786,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.1875,
       "step": 101790
     },
     {
       "epoch": 1.6539130152231483,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2442,
       "step": 101800
     },
     {
       "epoch": 1.654075482120518,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2297,
       "step": 101810
     },
     {
       "epoch": 1.6542379490178876,
-      "grad_norm": 2.875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.234,
       "step": 101820
     },
     {
       "epoch": 1.6544004159152572,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2324,
       "step": 101830
     },
     {
       "epoch": 1.6545628828126269,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.243,
       "step": 101840
     },
     {
       "epoch": 1.6547253497099965,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.1924,
       "step": 101850
     },
     {
       "epoch": 1.6548878166073662,
-      "grad_norm": 3.5,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.2657,
       "step": 101860
     },
     {
       "epoch": 1.6550502835047358,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2646,
       "step": 101870
     },
     {
       "epoch": 1.6552127504021055,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.1748,
       "step": 101880
     },
     {
       "epoch": 1.6553752172994751,
-      "grad_norm": 3.890625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.2437,
       "step": 101890
     },
     {
       "epoch": 1.6555376841968448,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2333,
       "step": 101900
     },
     {
       "epoch": 1.6557001510942144,
-      "grad_norm": 4.125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.1972,
       "step": 101910
     },
     {
       "epoch": 1.655862617991584,
-      "grad_norm": 4.46875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3725,
+      "loss": 1.2243,
       "step": 101920
     },
     {
       "epoch": 1.6560250848889537,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.26,
       "step": 101930
     },
     {
       "epoch": 1.6561875517863234,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2304,
       "step": 101940
     },
     {
       "epoch": 1.6563500186836932,
-      "grad_norm": 1.828125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.2125,
       "step": 101950
     },
     {
       "epoch": 1.656512485581063,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.1991,
       "step": 101960
     },
     {
       "epoch": 1.6566749524784325,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2152,
       "step": 101970
     },
     {
       "epoch": 1.6568374193758022,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.202,
       "step": 101980
     },
     {
       "epoch": 1.6569998862731719,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.251,
       "step": 101990
     },
     {
       "epoch": 1.6571623531705415,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2293,
       "step": 102000
     },
     {
       "epoch": 1.6573248200679112,
-      "grad_norm": 2.625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.1831,
       "step": 102010
     },
     {
       "epoch": 1.6574872869652808,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2148,
       "step": 102020
     },
     {
       "epoch": 1.6576497538626505,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2285,
       "step": 102030
     },
     {
       "epoch": 1.65781222076002,
-      "grad_norm": 2.625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.2249,
       "step": 102040
     },
     {
       "epoch": 1.6579746876573898,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2277,
       "step": 102050
     },
     {
       "epoch": 1.6581371545547596,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2206,
       "step": 102060
     },
     {
       "epoch": 1.6582996214521293,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.1722,
       "step": 102070
     },
     {
       "epoch": 1.658462088349499,
-      "grad_norm": 4.84375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2356,
       "step": 102080
     },
     {
       "epoch": 1.6586245552468686,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2494,
       "step": 102090
     },
     {
       "epoch": 1.6587870221442382,
-      "grad_norm": 2.5,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2391,
       "step": 102100
     },
     {
       "epoch": 1.6589494890416079,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.2355,
       "step": 102110
     },
     {
       "epoch": 1.6591119559389775,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2312,
       "step": 102120
     },
     {
       "epoch": 1.6592744228363472,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1876,
       "step": 102130
     },
     {
       "epoch": 1.6594368897337168,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3701,
+      "loss": 1.2593,
       "step": 102140
     },
     {
       "epoch": 1.6595993566310865,
-      "grad_norm": 2.140625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2059,
       "step": 102150
     },
     {
       "epoch": 1.6597618235284561,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.216,
       "step": 102160
     },
     {
       "epoch": 1.6599242904258258,
-      "grad_norm": 3.78125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.2562,
       "step": 102170
     },
     {
       "epoch": 1.6600867573231954,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2179,
       "step": 102180
     },
     {
       "epoch": 1.660249224220565,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3235,
+      "loss": 1.2556,
       "step": 102190
     },
     {
       "epoch": 1.6604116911179347,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.2482,
       "step": 102200
     },
     {
       "epoch": 1.6605741580153044,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2168,
       "step": 102210
     },
     {
       "epoch": 1.660736624912674,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.1475,
       "step": 102220
     },
     {
       "epoch": 1.6608990918100437,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2397,
       "step": 102230
     },
     {
       "epoch": 1.6610615587074133,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.1994,
       "step": 102240
     },
     {
       "epoch": 1.661224025604783,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3671,
+      "loss": 1.2449,
       "step": 102250
     },
     {
       "epoch": 1.6613864925021526,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2083,
       "step": 102260
     },
     {
       "epoch": 1.6615489593995223,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.141,
       "step": 102270
     },
     {
       "epoch": 1.661711426296892,
-      "grad_norm": 2.90625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2292,
       "step": 102280
     },
     {
       "epoch": 1.6618738931942616,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2557,
       "step": 102290
     },
     {
       "epoch": 1.6620363600916312,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2199,
       "step": 102300
     },
     {
       "epoch": 1.662198826989001,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.1936,
       "step": 102310
     },
     {
       "epoch": 1.6623612938863706,
-      "grad_norm": 2.5,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2178,
       "step": 102320
     },
     {
       "epoch": 1.6625237607837402,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 1.2318,
       "step": 102330
     },
     {
       "epoch": 1.6626862276811099,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.2491,
       "step": 102340
     },
     {
       "epoch": 1.6628486945784795,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.324,
+      "loss": 1.257,
       "step": 102350
     },
     {
       "epoch": 1.6630111614758492,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2453,
       "step": 102360
     },
     {
       "epoch": 1.6631736283732188,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2275,
       "step": 102370
     },
     {
       "epoch": 1.6633360952705885,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2246,
       "step": 102380
     },
     {
       "epoch": 1.6634985621679583,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2349,
       "step": 102390
     },
     {
       "epoch": 1.663661029065328,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.1945,
       "step": 102400
     },
     {
       "epoch": 1.6638234959626976,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.1992,
       "step": 102410
     },
     {
       "epoch": 1.6639859628600673,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2383,
       "step": 102420
     },
     {
       "epoch": 1.664148429757437,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2323,
       "step": 102430
     },
     {
       "epoch": 1.6643108966548066,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2144,
       "step": 102440
     },
     {
       "epoch": 1.6644733635521762,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2156,
       "step": 102450
     },
     {
       "epoch": 1.6646358304495459,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.227,
       "step": 102460
     },
     {
       "epoch": 1.6647982973469155,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3566,
+      "loss": 1.2399,
       "step": 102470
     },
     {
       "epoch": 1.6649607642442852,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1798,
       "step": 102480
     },
     {
       "epoch": 1.6651232311416548,
-      "grad_norm": 3.890625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.1836,
       "step": 102490
     },
     {
       "epoch": 1.6652856980390247,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2264,
       "step": 102500
     },
     {
       "epoch": 1.6654481649363944,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2368,
       "step": 102510
     },
     {
       "epoch": 1.665610631833764,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.1737,
       "step": 102520
     },
     {
       "epoch": 1.6657730987311337,
-      "grad_norm": 3.25,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2526,
       "step": 102530
     },
     {
       "epoch": 1.6659355656285033,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2182,
       "step": 102540
     },
     {
       "epoch": 1.666098032525873,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2237,
       "step": 102550
     },
     {
       "epoch": 1.6662604994232426,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2077,
       "step": 102560
     },
     {
       "epoch": 1.6664229663206123,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.2057,
       "step": 102570
     },
     {
       "epoch": 1.666585433217982,
-      "grad_norm": 3.625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.2369,
       "step": 102580
     },
     {
       "epoch": 1.6667479001153516,
-      "grad_norm": 2.25,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.3237,
+      "loss": 1.199,
       "step": 102590
     },
     {
       "epoch": 1.6669103670127212,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.1483,
       "step": 102600
     },
     {
       "epoch": 1.6670728339100909,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.1947,
       "step": 102610
     },
     {
       "epoch": 1.6672353008074605,
-      "grad_norm": 1.75,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.1893,
       "step": 102620
     },
     {
       "epoch": 1.6673977677048302,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2282,
       "step": 102630
     },
     {
       "epoch": 1.6675602346021998,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.2505,
       "step": 102640
     },
     {
       "epoch": 1.6677227014995695,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3237,
+      "loss": 1.2061,
       "step": 102650
     },
     {
       "epoch": 1.6678851683969391,
-      "grad_norm": 3.9375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2188,
       "step": 102660
     },
     {
       "epoch": 1.6680476352943088,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3654,
+      "loss": 1.2011,
       "step": 102670
     },
     {
       "epoch": 1.6682101021916784,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2196,
       "step": 102680
     },
     {
       "epoch": 1.668372569089048,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.1539,
       "step": 102690
     },
     {
       "epoch": 1.6685350359864177,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2168,
       "step": 102700
     },
     {
       "epoch": 1.6686975028837874,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2105,
       "step": 102710
     },
     {
       "epoch": 1.668859969781157,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3624,
+      "loss": 1.2433,
       "step": 102720
     },
     {
       "epoch": 1.6690224366785267,
-      "grad_norm": 2.375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.1956,
       "step": 102730
     },
     {
       "epoch": 1.6691849035758963,
-      "grad_norm": 2.875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.1873,
       "step": 102740
     },
     {
       "epoch": 1.669347370473266,
-      "grad_norm": 1.9609375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3214,
+      "loss": 1.2204,
       "step": 102750
     },
     {
       "epoch": 1.6695098373706356,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2092,
       "step": 102760
     },
     {
       "epoch": 1.6696723042680053,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2147,
       "step": 102770
     },
     {
       "epoch": 1.669834771165375,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.269,
       "step": 102780
     },
     {
       "epoch": 1.6699972380627446,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2557,
       "step": 102790
     },
     {
       "epoch": 1.6701597049601142,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.1958,
       "step": 102800
     },
     {
       "epoch": 1.6703221718574839,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.2149,
       "step": 102810
     },
     {
       "epoch": 1.6704846387548535,
-      "grad_norm": 2.75,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2456,
       "step": 102820
     },
     {
       "epoch": 1.6706471056522234,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.2315,
       "step": 102830
     },
     {
       "epoch": 1.670809572549593,
-      "grad_norm": 1.71875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.199,
       "step": 102840
     },
     {
       "epoch": 1.6709720394469627,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.324,
+      "loss": 1.2311,
       "step": 102850
     },
     {
       "epoch": 1.6711345063443324,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.196,
       "step": 102860
     },
     {
       "epoch": 1.671296973241702,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.1911,
       "step": 102870
     },
     {
       "epoch": 1.6714594401390717,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2522,
       "step": 102880
     },
     {
       "epoch": 1.6716219070364413,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2276,
       "step": 102890
     },
     {
       "epoch": 1.671784373933811,
-      "grad_norm": 2.0,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3167,
+      "loss": 1.2432,
       "step": 102900
     },
     {
       "epoch": 1.6719468408311806,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3129,
+      "loss": 1.2612,
       "step": 102910
     },
     {
       "epoch": 1.6721093077285503,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2108,
       "step": 102920
     },
     {
       "epoch": 1.67227177462592,
-      "grad_norm": 4.28125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.207,
       "step": 102930
     },
     {
       "epoch": 1.6724342415232898,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.1933,
       "step": 102940
     },
     {
       "epoch": 1.6725967084206594,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2022,
       "step": 102950
     },
     {
       "epoch": 1.672759175318029,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2337,
       "step": 102960
     },
     {
       "epoch": 1.6729216422153987,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.1591,
       "step": 102970
     },
     {
       "epoch": 1.6730841091127684,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.225,
       "step": 102980
     },
     {
       "epoch": 1.673246576010138,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3582,
+      "loss": 1.2677,
       "step": 102990
     },
     {
       "epoch": 1.6734090429075077,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2557,
       "step": 103000
     },
     {
       "epoch": 1.6735715098048773,
-      "grad_norm": 2.09375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.2433,
       "step": 103010
     },
     {
       "epoch": 1.673733976702247,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.1873,
       "step": 103020
     },
     {
       "epoch": 1.6738964435996166,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.1759,
       "step": 103030
     },
     {
       "epoch": 1.6740589104969863,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2058,
       "step": 103040
     },
     {
       "epoch": 1.674221377394356,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2113,
       "step": 103050
     },
     {
       "epoch": 1.6743838442917256,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.234,
       "step": 103060
     },
     {
       "epoch": 1.6745463111890952,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3622,
+      "loss": 1.2228,
       "step": 103070
     },
     {
       "epoch": 1.674708778086465,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2555,
       "step": 103080
     },
     {
       "epoch": 1.6748712449838346,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.2453,
       "step": 103090
     },
     {
       "epoch": 1.6750337118812042,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3208,
+      "loss": 1.2166,
       "step": 103100
     },
     {
       "epoch": 1.6751961787785739,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2127,
       "step": 103110
     },
     {
       "epoch": 1.6753586456759435,
-      "grad_norm": 4.40625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2054,
       "step": 103120
     },
     {
       "epoch": 1.6755211125733132,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2232,
       "step": 103130
     },
     {
       "epoch": 1.6756835794706828,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.2488,
       "step": 103140
     },
     {
       "epoch": 1.6758460463680525,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2305,
       "step": 103150
     },
     {
       "epoch": 1.676008513265422,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.1942,
       "step": 103160
     },
     {
       "epoch": 1.6761709801627918,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1854,
       "step": 103170
     },
     {
       "epoch": 1.6763334470601614,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.2283,
       "step": 103180
     },
     {
       "epoch": 1.676495913957531,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2374,
       "step": 103190
     },
     {
       "epoch": 1.6766583808549007,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.276,
       "step": 103200
     },
     {
       "epoch": 1.6768208477522704,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2342,
       "step": 103210
     },
     {
       "epoch": 1.67698331464964,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3258,
+      "loss": 1.1818,
       "step": 103220
     },
     {
       "epoch": 1.6771457815470097,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2219,
       "step": 103230
     },
     {
       "epoch": 1.6773082484443793,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.225,
       "step": 103240
     },
     {
       "epoch": 1.677470715341749,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2364,
       "step": 103250
     },
     {
       "epoch": 1.6776331822391186,
-      "grad_norm": 3.328125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2206,
       "step": 103260
     },
     {
       "epoch": 1.6777956491364885,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2243,
       "step": 103270
     },
     {
       "epoch": 1.6779581160338581,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2241,
       "step": 103280
     },
     {
       "epoch": 1.6781205829312278,
-      "grad_norm": 3.515625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.1958,
       "step": 103290
     },
     {
       "epoch": 1.6782830498285974,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.217,
       "step": 103300
     },
     {
       "epoch": 1.678445516725967,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2246,
       "step": 103310
     },
     {
       "epoch": 1.6786079836233367,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2222,
       "step": 103320
     },
     {
       "epoch": 1.6787704505207064,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2461,
       "step": 103330
     },
     {
       "epoch": 1.678932917418076,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2192,
       "step": 103340
     },
     {
       "epoch": 1.6790953843154457,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2411,
       "step": 103350
     },
     {
       "epoch": 1.6792578512128153,
-      "grad_norm": 3.375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2104,
       "step": 103360
     },
     {
       "epoch": 1.679420318110185,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.1619,
       "step": 103370
     },
     {
       "epoch": 1.6795827850075549,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2394,
       "step": 103380
     },
     {
       "epoch": 1.6797452519049245,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.209,
       "step": 103390
     },
     {
       "epoch": 1.6799077188022942,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3221,
+      "loss": 1.2533,
       "step": 103400
     },
     {
       "epoch": 1.6800701856996638,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2513,
       "step": 103410
     },
     {
       "epoch": 1.6802326525970335,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2154,
       "step": 103420
     },
     {
       "epoch": 1.6803951194944031,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3141,
+      "loss": 1.2173,
       "step": 103430
     },
     {
       "epoch": 1.6805575863917728,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.2628,
       "step": 103440
     },
     {
       "epoch": 1.6807200532891424,
-      "grad_norm": 3.25,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.2601,
       "step": 103450
     },
     {
       "epoch": 1.680882520186512,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.1931,
       "step": 103460
     },
     {
       "epoch": 1.6810449870838817,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.267,
       "step": 103470
     },
     {
       "epoch": 1.6812074539812514,
-      "grad_norm": 3.34375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3699,
+      "loss": 1.1793,
       "step": 103480
     },
     {
       "epoch": 1.681369920878621,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2476,
       "step": 103490
     },
     {
       "epoch": 1.6815323877759907,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.2294,
       "step": 103500
     },
     {
       "epoch": 1.6816948546733603,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2042,
       "step": 103510
     },
     {
       "epoch": 1.68185732157073,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2383,
       "step": 103520
     },
     {
       "epoch": 1.6820197884680996,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2025,
       "step": 103530
     },
     {
       "epoch": 1.6821822553654693,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2341,
       "step": 103540
     },
     {
       "epoch": 1.682344722262839,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2526,
       "step": 103550
     },
     {
       "epoch": 1.6825071891602086,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3254,
+      "loss": 1.1738,
       "step": 103560
     },
     {
       "epoch": 1.6826696560575782,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2339,
       "step": 103570
     },
     {
       "epoch": 1.6828321229549479,
-      "grad_norm": 2.25,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.2027,
       "step": 103580
     },
     {
       "epoch": 1.6829945898523175,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2004,
       "step": 103590
     },
     {
       "epoch": 1.6831570567496872,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2052,
       "step": 103600
     },
     {
       "epoch": 1.6833195236470568,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.1934,
       "step": 103610
     },
     {
       "epoch": 1.6834819905444265,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2078,
       "step": 103620
     },
     {
       "epoch": 1.6836444574417961,
-      "grad_norm": 2.546875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2467,
       "step": 103630
     },
     {
       "epoch": 1.6838069243391658,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2597,
       "step": 103640
     },
     {
       "epoch": 1.6839693912365354,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.1864,
       "step": 103650
     },
     {
       "epoch": 1.684131858133905,
-      "grad_norm": 2.875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3092,
+      "loss": 1.227,
       "step": 103660
     },
     {
       "epoch": 1.6842943250312747,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2404,
       "step": 103670
     },
     {
       "epoch": 1.6844567919286444,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2108,
       "step": 103680
     },
     {
       "epoch": 1.684619258826014,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2207,
       "step": 103690
     },
     {
       "epoch": 1.6847817257233837,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2351,
       "step": 103700
     },
     {
       "epoch": 1.6849441926207536,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2238,
       "step": 103710
     },
     {
       "epoch": 1.6851066595181232,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2158,
       "step": 103720
     },
     {
       "epoch": 1.6852691264154929,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2863,
       "step": 103730
     },
     {
       "epoch": 1.6854315933128625,
-      "grad_norm": 4.09375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2203,
       "step": 103740
     },
     {
       "epoch": 1.6855940602102322,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3264,
+      "loss": 1.2023,
       "step": 103750
     },
     {
       "epoch": 1.6857565271076018,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.196,
       "step": 103760
     },
     {
       "epoch": 1.6859189940049715,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3281,
+      "loss": 1.2347,
       "step": 103770
     },
     {
       "epoch": 1.6860814609023411,
-      "grad_norm": 2.625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3238,
+      "loss": 1.1768,
       "step": 103780
     },
     {
       "epoch": 1.6862439277997108,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2127,
       "step": 103790
     },
     {
       "epoch": 1.6864063946970804,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2256,
       "step": 103800
     },
     {
       "epoch": 1.6865688615944503,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2052,
       "step": 103810
     },
     {
       "epoch": 1.68673132849182,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2117,
       "step": 103820
     },
     {
       "epoch": 1.6868937953891896,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2312,
       "step": 103830
     },
     {
       "epoch": 1.6870562622865592,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2294,
       "step": 103840
     },
     {
       "epoch": 1.687218729183929,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1996,
       "step": 103850
     },
     {
       "epoch": 1.6873811960812986,
-      "grad_norm": 2.375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2354,
       "step": 103860
     },
     {
       "epoch": 1.6875436629786682,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2665,
       "step": 103870
     },
     {
       "epoch": 1.6877061298760379,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.23,
       "step": 103880
     },
     {
       "epoch": 1.6878685967734075,
-      "grad_norm": 3.25,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.1983,
       "step": 103890
     },
     {
       "epoch": 1.6880310636707772,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2127,
       "step": 103900
     },
     {
       "epoch": 1.6881935305681468,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.216,
       "step": 103910
     },
     {
       "epoch": 1.6883559974655165,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2153,
       "step": 103920
     },
     {
       "epoch": 1.688518464362886,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.369,
+      "loss": 1.222,
       "step": 103930
     },
     {
       "epoch": 1.6886809312602558,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2481,
       "step": 103940
     },
     {
       "epoch": 1.6888433981576254,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.2472,
       "step": 103950
     },
     {
       "epoch": 1.689005865054995,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2132,
       "step": 103960
     },
     {
       "epoch": 1.6891683319523647,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2417,
       "step": 103970
     },
     {
       "epoch": 1.6893307988497344,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.251,
       "step": 103980
     },
     {
       "epoch": 1.689493265747104,
-      "grad_norm": 2.0625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2046,
       "step": 103990
     },
     {
       "epoch": 1.6896557326444737,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.2596,
       "step": 104000
     },
     {
       "epoch": 1.6898181995418433,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2055,
       "step": 104010
     },
     {
       "epoch": 1.689980666439213,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.249,
       "step": 104020
     },
     {
       "epoch": 1.6901431333365826,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.1988,
       "step": 104030
     },
     {
       "epoch": 1.6903056002339523,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3664,
+      "loss": 1.195,
       "step": 104040
     },
     {
       "epoch": 1.690468067131322,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2113,
       "step": 104050
     },
     {
       "epoch": 1.6906305340286916,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.226,
       "step": 104060
     },
     {
       "epoch": 1.6907930009260612,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3733,
+      "loss": 1.1935,
       "step": 104070
     },
     {
       "epoch": 1.6909554678234309,
-      "grad_norm": 2.71875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.242,
       "step": 104080
     },
     {
       "epoch": 1.6911179347208005,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.2267,
       "step": 104090
     },
     {
       "epoch": 1.6912804016181702,
-      "grad_norm": 3.75,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2529,
       "step": 104100
     },
     {
       "epoch": 1.6914428685155398,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1829,
       "step": 104110
     },
     {
       "epoch": 1.6916053354129095,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.1996,
       "step": 104120
     },
     {
       "epoch": 1.6917678023102791,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3677,
+      "loss": 1.2049,
       "step": 104130
     },
     {
       "epoch": 1.691930269207649,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2215,
       "step": 104140
     },
     {
       "epoch": 1.6920927361050186,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2111,
       "step": 104150
     },
     {
       "epoch": 1.6922552030023883,
-      "grad_norm": 3.375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3627,
+      "loss": 1.22,
       "step": 104160
     },
     {
       "epoch": 1.692417669899758,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2029,
       "step": 104170
     },
     {
       "epoch": 1.6925801367971276,
-      "grad_norm": 3.25,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.1967,
       "step": 104180
     },
     {
       "epoch": 1.6927426036944972,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2047,
       "step": 104190
     },
     {
       "epoch": 1.692905070591867,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.1946,
       "step": 104200
     },
     {
       "epoch": 1.6930675374892366,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2758,
       "step": 104210
     },
     {
       "epoch": 1.6932300043866062,
-      "grad_norm": 4.21875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.37,
+      "loss": 1.1952,
       "step": 104220
     },
     {
       "epoch": 1.6933924712839759,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2611,
       "step": 104230
     },
     {
       "epoch": 1.6935549381813455,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3194,
+      "loss": 1.1899,
       "step": 104240
     },
     {
       "epoch": 1.6937174050787154,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1808,
       "step": 104250
     },
     {
       "epoch": 1.693879871976085,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2284,
       "step": 104260
     },
     {
       "epoch": 1.6940423388734547,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2653,
       "step": 104270
     },
     {
       "epoch": 1.6942048057708243,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2174,
       "step": 104280
     },
     {
       "epoch": 1.694367272668194,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2434,
       "step": 104290
     },
     {
       "epoch": 1.6945297395655636,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2243,
       "step": 104300
     },
     {
       "epoch": 1.6946922064629333,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.2138,
       "step": 104310
     },
     {
       "epoch": 1.694854673360303,
-      "grad_norm": 2.625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2282,
       "step": 104320
     },
     {
       "epoch": 1.6950171402576726,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2252,
       "step": 104330
     },
     {
       "epoch": 1.6951796071550422,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.221,
       "step": 104340
     },
     {
       "epoch": 1.6953420740524119,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3582,
+      "loss": 1.2729,
       "step": 104350
     },
     {
       "epoch": 1.6955045409497815,
-      "grad_norm": 3.125,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3191,
+      "loss": 1.1936,
       "step": 104360
     },
     {
       "epoch": 1.6956670078471512,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2272,
       "step": 104370
     },
     {
       "epoch": 1.6958294747445208,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2082,
       "step": 104380
     },
     {
       "epoch": 1.6959919416418905,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.1951,
       "step": 104390
     },
     {
       "epoch": 1.6961544085392601,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.1776,
       "step": 104400
     },
     {
       "epoch": 1.6963168754366298,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3558,
+      "loss": 1.2205,
       "step": 104410
     },
     {
       "epoch": 1.6964793423339994,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.2074,
       "step": 104420
     },
     {
       "epoch": 1.696641809231369,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3241,
+      "loss": 1.2479,
       "step": 104430
     },
     {
       "epoch": 1.6968042761287387,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2361,
       "step": 104440
     },
     {
       "epoch": 1.6969667430261084,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3153,
+      "loss": 1.1768,
       "step": 104450
     },
     {
       "epoch": 1.697129209923478,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3579,
+      "loss": 1.1931,
       "step": 104460
     },
     {
       "epoch": 1.6972916768208477,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.261,
       "step": 104470
     },
     {
       "epoch": 1.6974541437182173,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2346,
       "step": 104480
     },
     {
       "epoch": 1.697616610615587,
-      "grad_norm": 5.0625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2197,
       "step": 104490
     },
     {
       "epoch": 1.6977790775129566,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.1953,
       "step": 104500
     },
     {
       "epoch": 1.6979415444103263,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2229,
       "step": 104510
     },
     {
       "epoch": 1.698104011307696,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.1826,
       "step": 104520
     },
     {
       "epoch": 1.6982664782050656,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2076,
       "step": 104530
     },
     {
       "epoch": 1.6984289451024353,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.243,
       "step": 104540
     },
     {
       "epoch": 1.698591411999805,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.1842,
       "step": 104550
     },
     {
       "epoch": 1.6987538788971746,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.1831,
       "step": 104560
     },
     {
       "epoch": 1.6989163457945442,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2281,
       "step": 104570
     },
     {
       "epoch": 1.699078812691914,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3718,
+      "loss": 1.2021,
       "step": 104580
     },
     {
       "epoch": 1.6992412795892837,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.1906,
       "step": 104590
     },
     {
       "epoch": 1.6994037464866534,
-      "grad_norm": 2.25,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2015,
       "step": 104600
     },
     {
       "epoch": 1.699566213384023,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.1675,
       "step": 104610
     },
     {
       "epoch": 1.6997286802813927,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2225,
       "step": 104620
     },
     {
       "epoch": 1.6998911471787623,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.2706,
       "step": 104630
     },
     {
       "epoch": 1.700053614076132,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.214,
       "step": 104640
     },
     {
       "epoch": 1.7002160809735016,
-      "grad_norm": 3.375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2541,
       "step": 104650
     },
     {
       "epoch": 1.7003785478708713,
-      "grad_norm": 3.609375,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2099,
       "step": 104660
     },
     {
       "epoch": 1.700541014768241,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.2228,
       "step": 104670
     },
     {
       "epoch": 1.7007034816656106,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2395,
       "step": 104680
     },
     {
       "epoch": 1.7008659485629805,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2489,
       "step": 104690
     },
     {
       "epoch": 1.70102841546035,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.1895,
       "step": 104700
     },
     {
       "epoch": 1.7011908823577198,
-      "grad_norm": 4.21875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.236,
       "step": 104710
     },
     {
       "epoch": 1.7013533492550894,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1883,
       "step": 104720
     },
     {
       "epoch": 1.701515816152459,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2496,
       "step": 104730
     },
     {
       "epoch": 1.7016782830498287,
-      "grad_norm": 2.875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.2499,
       "step": 104740
     },
     {
       "epoch": 1.7018407499471984,
-      "grad_norm": 2.5,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1682,
       "step": 104750
     },
     {
       "epoch": 1.702003216844568,
-      "grad_norm": 3.953125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.316,
+      "loss": 1.2089,
       "step": 104760
     },
     {
       "epoch": 1.7021656837419377,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.2143,
       "step": 104770
     },
     {
       "epoch": 1.7023281506393073,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.1714,
       "step": 104780
     },
     {
       "epoch": 1.702490617536677,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1998,
       "step": 104790
     },
     {
       "epoch": 1.7026530844340466,
-      "grad_norm": 4.375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.216,
       "step": 104800
     },
     {
       "epoch": 1.7028155513314163,
-      "grad_norm": 2.75,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2365,
       "step": 104810
     },
     {
       "epoch": 1.702978018228786,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2463,
       "step": 104820
     },
     {
       "epoch": 1.7031404851261556,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2707,
       "step": 104830
     },
     {
       "epoch": 1.7033029520235252,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.1868,
       "step": 104840
     },
     {
       "epoch": 1.7034654189208949,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3713,
+      "loss": 1.198,
       "step": 104850
     },
     {
       "epoch": 1.7036278858182645,
-      "grad_norm": 3.5,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.1973,
       "step": 104860
     },
     {
       "epoch": 1.7037903527156342,
-      "grad_norm": 4.1875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2011,
       "step": 104870
     },
     {
       "epoch": 1.7039528196130038,
-      "grad_norm": 2.21875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2613,
       "step": 104880
     },
     {
       "epoch": 1.7041152865103735,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2212,
       "step": 104890
     },
     {
       "epoch": 1.7042777534077431,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2159,
       "step": 104900
     },
     {
       "epoch": 1.7044402203051128,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2232,
       "step": 104910
     },
     {
       "epoch": 1.7046026872024824,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3235,
+      "loss": 1.2589,
       "step": 104920
     },
     {
       "epoch": 1.704765154099852,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2126,
       "step": 104930
     },
     {
       "epoch": 1.7049276209972217,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2127,
       "step": 104940
     },
     {
       "epoch": 1.7050900878945914,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2289,
       "step": 104950
     },
     {
       "epoch": 1.705252554791961,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.2418,
       "step": 104960
     },
     {
       "epoch": 1.7054150216893307,
-      "grad_norm": 2.875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2289,
       "step": 104970
     },
     {
       "epoch": 1.7055774885867003,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2279,
       "step": 104980
     },
     {
       "epoch": 1.70573995548407,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2176,
       "step": 104990
     },
     {
       "epoch": 1.7059024223814396,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2484,
       "step": 105000
     },
     {
       "epoch": 1.7060648892788093,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2198,
       "step": 105010
     },
     {
       "epoch": 1.7062273561761792,
-      "grad_norm": 4.0,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2121,
       "step": 105020
     },
     {
       "epoch": 1.7063898230735488,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3732,
+      "loss": 1.2232,
       "step": 105030
     },
     {
       "epoch": 1.7065522899709185,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.2494,
       "step": 105040
     },
     {
       "epoch": 1.706714756868288,
-      "grad_norm": 2.5,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.374,
+      "loss": 1.2072,
       "step": 105050
     },
     {
       "epoch": 1.7068772237656578,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.2189,
       "step": 105060
     },
     {
       "epoch": 1.7070396906630274,
-      "grad_norm": 3.875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2158,
       "step": 105070
     },
     {
       "epoch": 1.707202157560397,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2403,
       "step": 105080
     },
     {
       "epoch": 1.7073646244577667,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2284,
       "step": 105090
     },
     {
       "epoch": 1.7075270913551364,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.1901,
       "step": 105100
     },
     {
       "epoch": 1.707689558252506,
-      "grad_norm": 2.296875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3161,
+      "loss": 1.2336,
       "step": 105110
     },
     {
       "epoch": 1.7078520251498757,
-      "grad_norm": 2.625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2075,
       "step": 105120
     },
     {
       "epoch": 1.7080144920472455,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2439,
       "step": 105130
     },
     {
       "epoch": 1.7081769589446152,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2039,
       "step": 105140
     },
     {
       "epoch": 1.7083394258419848,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2324,
       "step": 105150
     },
     {
       "epoch": 1.7085018927393545,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.1978,
       "step": 105160
     },
     {
       "epoch": 1.7086643596367241,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2193,
       "step": 105170
     },
     {
       "epoch": 1.7088268265340938,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2044,
       "step": 105180
     },
     {
       "epoch": 1.7089892934314634,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.1996,
       "step": 105190
     },
     {
       "epoch": 1.709151760328833,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.226,
       "step": 105200
     },
     {
       "epoch": 1.7093142272262027,
-      "grad_norm": 4.34375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2405,
       "step": 105210
     },
     {
       "epoch": 1.7094766941235724,
-      "grad_norm": 3.125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2321,
       "step": 105220
     },
     {
       "epoch": 1.709639161020942,
-      "grad_norm": 2.875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2532,
       "step": 105230
     },
     {
       "epoch": 1.7098016279183117,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2359,
       "step": 105240
     },
     {
       "epoch": 1.7099640948156813,
-      "grad_norm": 2.28125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2279,
       "step": 105250
     },
     {
       "epoch": 1.710126561713051,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2435,
       "step": 105260
     },
     {
       "epoch": 1.7102890286104206,
-      "grad_norm": 3.828125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2169,
       "step": 105270
     },
     {
       "epoch": 1.7104514955077903,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2035,
       "step": 105280
     },
     {
       "epoch": 1.71061396240516,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2162,
       "step": 105290
     },
     {
       "epoch": 1.7107764293025296,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2164,
       "step": 105300
     },
     {
       "epoch": 1.7109388961998993,
-      "grad_norm": 3.53125,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.24,
       "step": 105310
     },
     {
       "epoch": 1.711101363097269,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.364,
+      "loss": 1.1787,
       "step": 105320
     },
     {
       "epoch": 1.7112638299946386,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2373,
       "step": 105330
     },
     {
       "epoch": 1.7114262968920082,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.1689,
       "step": 105340
     },
     {
       "epoch": 1.7115887637893779,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2407,
       "step": 105350
     },
     {
       "epoch": 1.7117512306867475,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2374,
       "step": 105360
     },
     {
       "epoch": 1.7119136975841172,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2114,
       "step": 105370
     },
     {
       "epoch": 1.7120761644814868,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2015,
       "step": 105380
     },
     {
       "epoch": 1.7122386313788565,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2194,
       "step": 105390
     },
     {
       "epoch": 1.712401098276226,
-      "grad_norm": 3.71875,
+      "grad_norm": 15.375,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2165,
       "step": 105400
     },
     {
       "epoch": 1.7125635651735958,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2411,
       "step": 105410
     },
     {
       "epoch": 1.7127260320709654,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2215,
       "step": 105420
     },
     {
       "epoch": 1.712888498968335,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2329,
       "step": 105430
     },
     {
       "epoch": 1.7130509658657047,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.206,
       "step": 105440
     },
     {
       "epoch": 1.7132134327630744,
-      "grad_norm": 2.125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2289,
       "step": 105450
     },
     {
       "epoch": 1.7133758996604442,
-      "grad_norm": 3.03125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2014,
       "step": 105460
     },
     {
       "epoch": 1.7135383665578139,
-      "grad_norm": 3.5,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2614,
       "step": 105470
     },
     {
       "epoch": 1.7137008334551835,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.202,
       "step": 105480
     },
     {
       "epoch": 1.7138633003525532,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.1815,
       "step": 105490
     },
     {
       "epoch": 1.7140257672499228,
-      "grad_norm": 2.625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2368,
       "step": 105500
     },
     {
       "epoch": 1.7141882341472925,
-      "grad_norm": 2.75,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3732,
+      "loss": 1.2401,
       "step": 105510
     },
     {
       "epoch": 1.7143507010446621,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2216,
       "step": 105520
     },
     {
       "epoch": 1.7145131679420318,
-      "grad_norm": 3.703125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 1.2479,
       "step": 105530
     },
     {
       "epoch": 1.7146756348394014,
-      "grad_norm": 3.640625,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.1647,
       "step": 105540
     },
     {
       "epoch": 1.714838101736771,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2468,
       "step": 105550
     },
     {
       "epoch": 1.7150005686341407,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.1713,
       "step": 105560
     },
     {
       "epoch": 1.7151630355315106,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.188,
       "step": 105570
     },
     {
       "epoch": 1.7153255024288803,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2178,
       "step": 105580
     },
     {
       "epoch": 1.71548796932625,
-      "grad_norm": 3.734375,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3622,
+      "loss": 1.1991,
       "step": 105590
     },
     {
       "epoch": 1.7156504362236196,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.236,
       "step": 105600
     },
     {
       "epoch": 1.7158129031209892,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2284,
       "step": 105610
     },
     {
       "epoch": 1.7159753700183589,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2015,
       "step": 105620
     },
     {
       "epoch": 1.7161378369157285,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2499,
       "step": 105630
     },
     {
       "epoch": 1.7163003038130982,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.2781,
       "step": 105640
     },
     {
       "epoch": 1.7164627707104678,
-      "grad_norm": 2.5,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2315,
       "step": 105650
     },
     {
       "epoch": 1.7166252376078375,
-      "grad_norm": 2.625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1917,
       "step": 105660
     },
     {
       "epoch": 1.7167877045052071,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2393,
       "step": 105670
     },
     {
       "epoch": 1.7169501714025768,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2522,
       "step": 105680
     },
     {
       "epoch": 1.7171126382999464,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1968,
       "step": 105690
     },
     {
       "epoch": 1.717275105197316,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.359,
+      "loss": 1.1795,
       "step": 105700
     },
     {
       "epoch": 1.7174375720946857,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2316,
       "step": 105710
     },
     {
       "epoch": 1.7176000389920554,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2112,
       "step": 105720
     },
     {
       "epoch": 1.717762505889425,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2632,
       "step": 105730
     },
     {
       "epoch": 1.7179249727867947,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3153,
+      "loss": 1.224,
       "step": 105740
     },
     {
       "epoch": 1.7180874396841643,
-      "grad_norm": 3.703125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.231,
       "step": 105750
     },
     {
       "epoch": 1.718249906581534,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2234,
       "step": 105760
     },
     {
       "epoch": 1.7184123734789036,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2217,
       "step": 105770
     },
     {
       "epoch": 1.7185748403762733,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2171,
       "step": 105780
     },
     {
       "epoch": 1.718737307273643,
-      "grad_norm": 2.375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.1914,
       "step": 105790
     },
     {
       "epoch": 1.7188997741710126,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.201,
       "step": 105800
     },
     {
       "epoch": 1.7190622410683822,
-      "grad_norm": 3.625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2102,
       "step": 105810
     },
     {
       "epoch": 1.7192247079657519,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2346,
       "step": 105820
     },
     {
       "epoch": 1.7193871748631215,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3651,
+      "loss": 1.2055,
       "step": 105830
     },
     {
       "epoch": 1.7195496417604912,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.2381,
       "step": 105840
     },
     {
       "epoch": 1.7197121086578608,
-      "grad_norm": 6.03125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2511,
       "step": 105850
     },
     {
       "epoch": 1.7198745755552305,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.1933,
       "step": 105860
     },
     {
       "epoch": 1.7200370424526001,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2452,
       "step": 105870
     },
     {
       "epoch": 1.7201995093499698,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.1937,
       "step": 105880
     },
     {
       "epoch": 1.7203619762473394,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.2286,
       "step": 105890
     },
     {
       "epoch": 1.7205244431447093,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2269,
       "step": 105900
     },
     {
       "epoch": 1.720686910042079,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.1991,
       "step": 105910
     },
     {
       "epoch": 1.7208493769394486,
-      "grad_norm": 3.359375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.258,
       "step": 105920
     },
     {
       "epoch": 1.7210118438368183,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2271,
       "step": 105930
     },
     {
       "epoch": 1.721174310734188,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.2027,
       "step": 105940
     },
     {
       "epoch": 1.7213367776315576,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2341,
       "step": 105950
     },
     {
       "epoch": 1.7214992445289272,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3598,
+      "loss": 1.2362,
       "step": 105960
     },
     {
       "epoch": 1.7216617114262969,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1897,
       "step": 105970
     },
     {
       "epoch": 1.7218241783236665,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2239,
       "step": 105980
     },
     {
       "epoch": 1.7219866452210362,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2462,
       "step": 105990
     },
     {
       "epoch": 1.7221491121184058,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.1844,
       "step": 106000
     },
     {
       "epoch": 1.7223115790157757,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2395,
       "step": 106010
     },
     {
       "epoch": 1.7224740459131453,
-      "grad_norm": 2.46875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.182,
       "step": 106020
     },
     {
       "epoch": 1.722636512810515,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.1964,
       "step": 106030
     },
     {
       "epoch": 1.7227989797078846,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2536,
       "step": 106040
     },
     {
       "epoch": 1.7229614466052543,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2256,
       "step": 106050
     },
     {
       "epoch": 1.723123913502624,
-      "grad_norm": 3.296875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.1809,
       "step": 106060
     },
     {
       "epoch": 1.7232863803999936,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.1868,
       "step": 106070
     },
     {
       "epoch": 1.7234488472973633,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3208,
+      "loss": 1.1549,
       "step": 106080
     },
     {
       "epoch": 1.723611314194733,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2443,
       "step": 106090
     },
     {
       "epoch": 1.7237737810921026,
-      "grad_norm": 2.84375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2321,
       "step": 106100
     },
     {
       "epoch": 1.7239362479894722,
-      "grad_norm": 3.6875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2235,
       "step": 106110
     },
     {
       "epoch": 1.7240987148868419,
-      "grad_norm": 4.8125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2515,
       "step": 106120
     },
     {
       "epoch": 1.7242611817842115,
-      "grad_norm": 4.625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.2482,
       "step": 106130
     },
     {
       "epoch": 1.7244236486815812,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3527,
+      "loss": 1.2143,
       "step": 106140
     },
     {
       "epoch": 1.7245861155789508,
-      "grad_norm": 2.875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.233,
       "step": 106150
     },
     {
       "epoch": 1.7247485824763205,
-      "grad_norm": 2.25,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3196,
+      "loss": 1.2107,
       "step": 106160
     },
     {
       "epoch": 1.72491104937369,
-      "grad_norm": 3.625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.1884,
       "step": 106170
     },
     {
       "epoch": 1.7250735162710598,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3793,
+      "loss": 1.2605,
       "step": 106180
     },
     {
       "epoch": 1.7252359831684294,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.239,
       "step": 106190
     },
     {
       "epoch": 1.725398450065799,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2269,
       "step": 106200
     },
     {
       "epoch": 1.7255609169631687,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2845,
       "step": 106210
     },
     {
       "epoch": 1.7257233838605384,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2292,
       "step": 106220
     },
     {
       "epoch": 1.725885850757908,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2249,
       "step": 106230
     },
     {
       "epoch": 1.7260483176552777,
-      "grad_norm": 2.671875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2351,
       "step": 106240
     },
     {
       "epoch": 1.7262107845526473,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.1949,
       "step": 106250
     },
     {
       "epoch": 1.726373251450017,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2197,
       "step": 106260
     },
     {
       "epoch": 1.7265357183473866,
-      "grad_norm": 2.375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.223,
       "step": 106270
     },
     {
       "epoch": 1.7266981852447563,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.1747,
       "step": 106280
     },
     {
       "epoch": 1.726860652142126,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2387,
       "step": 106290
     },
     {
       "epoch": 1.7270231190394956,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2281,
       "step": 106300
     },
     {
       "epoch": 1.7271855859368652,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2385,
       "step": 106310
     },
     {
       "epoch": 1.7273480528342349,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2464,
       "step": 106320
     },
     {
       "epoch": 1.7275105197316045,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3164,
+      "loss": 1.1904,
       "step": 106330
     },
     {
       "epoch": 1.7276729866289744,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2362,
       "step": 106340
     },
     {
       "epoch": 1.727835453526344,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2053,
       "step": 106350
     },
     {
       "epoch": 1.7279979204237137,
-      "grad_norm": 3.125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2126,
       "step": 106360
     },
     {
       "epoch": 1.7281603873210833,
-      "grad_norm": 2.875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3614,
+      "loss": 1.1971,
       "step": 106370
     },
     {
       "epoch": 1.728322854218453,
-      "grad_norm": 3.0,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.1912,
       "step": 106380
     },
     {
       "epoch": 1.7284853211158226,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.1844,
       "step": 106390
     },
     {
       "epoch": 1.7286477880131923,
-      "grad_norm": 2.8125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.1872,
       "step": 106400
     },
     {
       "epoch": 1.728810254910562,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.1841,
       "step": 106410
     },
     {
       "epoch": 1.7289727218079316,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.221,
       "step": 106420
     },
     {
       "epoch": 1.7291351887053013,
-      "grad_norm": 3.375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2124,
       "step": 106430
     },
     {
       "epoch": 1.729297655602671,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3667,
+      "loss": 1.2182,
       "step": 106440
     },
     {
       "epoch": 1.7294601225000408,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3593,
+      "loss": 1.249,
       "step": 106450
     },
     {
       "epoch": 1.7296225893974104,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1956,
       "step": 106460
     },
     {
       "epoch": 1.72978505629478,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.207,
       "step": 106470
     },
     {
       "epoch": 1.7299475231921497,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.239,
       "step": 106480
     },
     {
       "epoch": 1.7301099900895194,
-      "grad_norm": 3.5,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2246,
       "step": 106490
     },
     {
       "epoch": 1.730272456986889,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.2139,
       "step": 106500
     },
     {
       "epoch": 1.7304349238842587,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.1842,
       "step": 106510
     },
     {
       "epoch": 1.7305973907816283,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2482,
       "step": 106520
     },
     {
       "epoch": 1.730759857678998,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3596,
+      "loss": 1.2643,
       "step": 106530
     },
     {
       "epoch": 1.7309223245763676,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.22,
       "step": 106540
     },
     {
       "epoch": 1.7310847914737373,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1957,
       "step": 106550
     },
     {
       "epoch": 1.731247258371107,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.2227,
       "step": 106560
     },
     {
       "epoch": 1.7314097252684766,
-      "grad_norm": 3.765625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.2123,
       "step": 106570
     },
     {
       "epoch": 1.7315721921658462,
-      "grad_norm": 3.71875,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3838,
+      "loss": 1.2022,
       "step": 106580
     },
     {
       "epoch": 1.7317346590632159,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.174,
       "step": 106590
     },
     {
       "epoch": 1.7318971259605855,
-      "grad_norm": 2.75,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2112,
       "step": 106600
     },
     {
       "epoch": 1.7320595928579552,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3659,
+      "loss": 1.2043,
       "step": 106610
     },
     {
       "epoch": 1.7322220597553248,
-      "grad_norm": 2.578125,
+      "grad_norm": 13.625,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2473,
       "step": 106620
     },
     {
       "epoch": 1.7323845266526945,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3668,
+      "loss": 1.1987,
       "step": 106630
     },
     {
       "epoch": 1.7325469935500641,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3654,
+      "loss": 1.2177,
       "step": 106640
     },
     {
       "epoch": 1.7327094604474338,
-      "grad_norm": 1.7890625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2147,
       "step": 106650
     },
     {
       "epoch": 1.7328719273448034,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3134,
+      "loss": 1.1936,
       "step": 106660
     },
     {
       "epoch": 1.733034394242173,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3282,
+      "loss": 1.2032,
       "step": 106670
     },
     {
       "epoch": 1.7331968611395427,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2483,
       "step": 106680
     },
     {
       "epoch": 1.7333593280369124,
-      "grad_norm": 3.78125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2158,
       "step": 106690
     },
     {
       "epoch": 1.733521794934282,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2118,
       "step": 106700
     },
     {
       "epoch": 1.7336842618316517,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2492,
       "step": 106710
     },
     {
       "epoch": 1.7338467287290213,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2139,
       "step": 106720
     },
     {
       "epoch": 1.734009195626391,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.367,
+      "loss": 1.2114,
       "step": 106730
     },
     {
       "epoch": 1.7341716625237606,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2351,
       "step": 106740
     },
     {
       "epoch": 1.7343341294211303,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2587,
       "step": 106750
     },
     {
       "epoch": 1.7344965963185,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.239,
       "step": 106760
     },
     {
       "epoch": 1.7346590632158696,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2007,
       "step": 106770
     },
     {
       "epoch": 1.7348215301132395,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2195,
       "step": 106780
     },
     {
       "epoch": 1.7349839970106091,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.1841,
       "step": 106790
     },
     {
       "epoch": 1.7351464639079788,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.2506,
       "step": 106800
     },
     {
       "epoch": 1.7353089308053484,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2144,
       "step": 106810
     },
     {
       "epoch": 1.735471397702718,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3274,
+      "loss": 1.257,
       "step": 106820
     },
     {
       "epoch": 1.7356338646000877,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2462,
       "step": 106830
     },
     {
       "epoch": 1.7357963314974574,
-      "grad_norm": 2.5,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.1856,
       "step": 106840
     },
     {
       "epoch": 1.735958798394827,
-      "grad_norm": 3.25,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2061,
       "step": 106850
     },
     {
       "epoch": 1.7361212652921967,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.1907,
       "step": 106860
     },
     {
       "epoch": 1.7362837321895663,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.2175,
       "step": 106870
     },
     {
       "epoch": 1.736446199086936,
-      "grad_norm": 2.375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1991,
       "step": 106880
     },
     {
       "epoch": 1.7366086659843059,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.23,
       "step": 106890
     },
     {
       "epoch": 1.7367711328816755,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2168,
       "step": 106900
     },
     {
       "epoch": 1.7369335997790452,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2612,
       "step": 106910
     },
     {
       "epoch": 1.7370960666764148,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.1973,
       "step": 106920
     },
     {
       "epoch": 1.7372585335737845,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.174,
       "step": 106930
     },
     {
       "epoch": 1.737421000471154,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2557,
       "step": 106940
     },
     {
       "epoch": 1.7375834673685238,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.1797,
       "step": 106950
     },
     {
       "epoch": 1.7377459342658934,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2237,
       "step": 106960
     },
     {
       "epoch": 1.737908401163263,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3774,
+      "loss": 1.2103,
       "step": 106970
     },
     {
       "epoch": 1.7380708680606327,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3205,
+      "loss": 1.187,
       "step": 106980
     },
     {
       "epoch": 1.7382333349580024,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2347,
       "step": 106990
     },
     {
       "epoch": 1.738395801855372,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2261,
       "step": 107000
     },
     {
       "epoch": 1.7385582687527417,
-      "grad_norm": 2.625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2165,
       "step": 107010
     },
     {
       "epoch": 1.7387207356501113,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 1.2486,
       "step": 107020
     },
     {
       "epoch": 1.738883202547481,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.1873,
       "step": 107030
     },
     {
       "epoch": 1.7390456694448506,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3281,
+      "loss": 1.2191,
       "step": 107040
     },
     {
       "epoch": 1.7392081363422203,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3587,
+      "loss": 1.2144,
       "step": 107050
     },
     {
       "epoch": 1.73937060323959,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2038,
       "step": 107060
     },
     {
       "epoch": 1.7395330701369596,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3595,
+      "loss": 1.2298,
       "step": 107070
     },
     {
       "epoch": 1.7396955370343292,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2281,
       "step": 107080
     },
     {
       "epoch": 1.7398580039316989,
-      "grad_norm": 1.8984375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2004,
       "step": 107090
     },
     {
       "epoch": 1.7400204708290685,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2155,
       "step": 107100
     },
     {
       "epoch": 1.7401829377264382,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.226,
       "step": 107110
     },
     {
       "epoch": 1.7403454046238078,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.1866,
       "step": 107120
     },
     {
       "epoch": 1.7405078715211775,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2082,
       "step": 107130
     },
     {
       "epoch": 1.7406703384185471,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.227,
       "step": 107140
     },
     {
       "epoch": 1.7408328053159168,
-      "grad_norm": 2.75,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2231,
       "step": 107150
     },
     {
       "epoch": 1.7409952722132864,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2076,
       "step": 107160
     },
     {
       "epoch": 1.741157739110656,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3657,
+      "loss": 1.1886,
       "step": 107170
     },
     {
       "epoch": 1.7413202060080257,
-      "grad_norm": 2.546875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2383,
       "step": 107180
     },
     {
       "epoch": 1.7414826729053954,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3215,
+      "loss": 1.224,
       "step": 107190
     },
     {
       "epoch": 1.741645139802765,
-      "grad_norm": 2.421875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2369,
       "step": 107200
     },
     {
       "epoch": 1.741807606700135,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.2369,
       "step": 107210
     },
     {
       "epoch": 1.7419700735975046,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.2357,
       "step": 107220
     },
     {
       "epoch": 1.7421325404948742,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3681,
+      "loss": 1.2031,
       "step": 107230
     },
     {
       "epoch": 1.7422950073922439,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2041,
       "step": 107240
     },
     {
       "epoch": 1.7424574742896135,
-      "grad_norm": 3.515625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2282,
       "step": 107250
     },
     {
       "epoch": 1.7426199411869832,
-      "grad_norm": 3.890625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.1689,
       "step": 107260
     },
     {
       "epoch": 1.7427824080843528,
-      "grad_norm": 2.625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.1813,
       "step": 107270
     },
     {
       "epoch": 1.7429448749817225,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.258,
       "step": 107280
     },
     {
       "epoch": 1.743107341879092,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3144,
+      "loss": 1.2172,
       "step": 107290
     },
     {
       "epoch": 1.7432698087764618,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3183,
+      "loss": 1.2093,
       "step": 107300
     },
     {
       "epoch": 1.7434322756738314,
-      "grad_norm": 2.0,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2185,
       "step": 107310
     },
     {
       "epoch": 1.7435947425712013,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2471,
       "step": 107320
     },
     {
       "epoch": 1.743757209468571,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2487,
       "step": 107330
     },
     {
       "epoch": 1.7439196763659406,
-      "grad_norm": 3.125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.243,
       "step": 107340
     },
     {
       "epoch": 1.7440821432633102,
-      "grad_norm": 2.625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2666,
       "step": 107350
     },
     {
       "epoch": 1.7442446101606799,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.3161,
       "step": 107360
     },
     {
       "epoch": 1.7444070770580495,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2339,
       "step": 107370
     },
     {
       "epoch": 1.7445695439554192,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3212,
+      "loss": 1.1947,
       "step": 107380
     },
     {
       "epoch": 1.7447320108527888,
-      "grad_norm": 4.375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2284,
       "step": 107390
     },
     {
       "epoch": 1.7448944777501585,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2342,
       "step": 107400
     },
     {
       "epoch": 1.7450569446475281,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2275,
       "step": 107410
     },
     {
       "epoch": 1.7452194115448978,
-      "grad_norm": 1.9921875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.22,
       "step": 107420
     },
     {
       "epoch": 1.7453818784422674,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3162,
+      "loss": 1.2482,
       "step": 107430
     },
     {
       "epoch": 1.745544345339637,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2907,
       "step": 107440
     },
     {
       "epoch": 1.7457068122370067,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2482,
       "step": 107450
     },
     {
       "epoch": 1.7458692791343764,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1844,
       "step": 107460
     },
     {
       "epoch": 1.746031746031746,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2575,
       "step": 107470
     },
     {
       "epoch": 1.7461942129291157,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.1982,
       "step": 107480
     },
     {
       "epoch": 1.7463566798264853,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.2104,
       "step": 107490
     },
     {
       "epoch": 1.746519146723855,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2542,
       "step": 107500
     },
     {
       "epoch": 1.7466816136212246,
-      "grad_norm": 2.25,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2669,
       "step": 107510
     },
     {
       "epoch": 1.7468440805185943,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.1979,
       "step": 107520
     },
     {
       "epoch": 1.747006547415964,
-      "grad_norm": 2.625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.223,
       "step": 107530
     },
     {
       "epoch": 1.7471690143133336,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2116,
       "step": 107540
     },
     {
       "epoch": 1.7473314812107033,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.1991,
       "step": 107550
     },
     {
       "epoch": 1.747493948108073,
-      "grad_norm": 2.625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2597,
       "step": 107560
     },
     {
       "epoch": 1.7476564150054426,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2193,
       "step": 107570
     },
     {
       "epoch": 1.7478188819028122,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1861,
       "step": 107580
     },
     {
       "epoch": 1.7479813488001819,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2433,
       "step": 107590
     },
     {
       "epoch": 1.7481438156975515,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3679,
+      "loss": 1.2232,
       "step": 107600
     },
     {
       "epoch": 1.7483062825949212,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2219,
       "step": 107610
     },
     {
       "epoch": 1.7484687494922908,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2433,
       "step": 107620
     },
     {
       "epoch": 1.7486312163896605,
-      "grad_norm": 4.21875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2107,
       "step": 107630
     },
     {
       "epoch": 1.74879368328703,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.2051,
       "step": 107640
     },
     {
       "epoch": 1.7489561501844,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3564,
+      "loss": 1.22,
       "step": 107650
     },
     {
       "epoch": 1.7491186170817696,
-      "grad_norm": 2.25,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3162,
+      "loss": 1.2009,
       "step": 107660
     },
     {
       "epoch": 1.7492810839791393,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2167,
       "step": 107670
     },
     {
       "epoch": 1.749443550876509,
-      "grad_norm": 2.890625,
+      "grad_norm": 5.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.1636,
       "step": 107680
     },
     {
       "epoch": 1.7496060177738786,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.33,
+      "loss": 1.2421,
       "step": 107690
     },
     {
       "epoch": 1.7497684846712482,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2066,
       "step": 107700
     },
     {
       "epoch": 1.7499309515686179,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2313,
       "step": 107710
     },
     {
       "epoch": 1.7500934184659875,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.2132,
       "step": 107720
     },
     {
       "epoch": 1.7502558853633572,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.1937,
       "step": 107730
     },
     {
       "epoch": 1.7504183522607268,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3667,
+      "loss": 1.2286,
       "step": 107740
     },
     {
       "epoch": 1.7505808191580965,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2059,
       "step": 107750
     },
     {
       "epoch": 1.7507432860554664,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2507,
       "step": 107760
     },
     {
       "epoch": 1.750905752952836,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.2473,
       "step": 107770
     },
     {
       "epoch": 1.7510682198502057,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2352,
       "step": 107780
     },
     {
       "epoch": 1.7512306867475753,
-      "grad_norm": 2.625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.2568,
       "step": 107790
     },
     {
       "epoch": 1.751393153644945,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.1853,
       "step": 107800
     },
     {
       "epoch": 1.7515556205423146,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3643,
+      "loss": 1.259,
       "step": 107810
     },
     {
       "epoch": 1.7517180874396843,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3614,
+      "loss": 1.2351,
       "step": 107820
     },
     {
       "epoch": 1.751880554337054,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2448,
       "step": 107830
     },
     {
       "epoch": 1.7520430212344236,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3305,
+      "loss": 1.2232,
       "step": 107840
     },
     {
       "epoch": 1.7522054881317932,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.1812,
       "step": 107850
     },
     {
       "epoch": 1.7523679550291629,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3386,
+      "loss": 1.2402,
       "step": 107860
     },
     {
       "epoch": 1.7525304219265325,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2077,
       "step": 107870
     },
     {
       "epoch": 1.7526928888239022,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3139,
+      "loss": 1.2093,
       "step": 107880
     },
     {
       "epoch": 1.7528553557212718,
-      "grad_norm": 3.796875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2186,
       "step": 107890
     },
     {
       "epoch": 1.7530178226186415,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2125,
       "step": 107900
     },
     {
       "epoch": 1.7531802895160111,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.1745,
       "step": 107910
     },
     {
       "epoch": 1.7533427564133808,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2108,
       "step": 107920
     },
     {
       "epoch": 1.7535052233107504,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3629,
+      "loss": 1.1857,
       "step": 107930
     },
     {
       "epoch": 1.75366769020812,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.2543,
       "step": 107940
     },
     {
       "epoch": 1.7538301571054897,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.2219,
       "step": 107950
     },
     {
       "epoch": 1.7539926240028594,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2091,
       "step": 107960
     },
     {
       "epoch": 1.754155090900229,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2212,
       "step": 107970
     },
     {
       "epoch": 1.7543175577975987,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2237,
       "step": 107980
     },
     {
       "epoch": 1.7544800246949683,
-      "grad_norm": 2.375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3129,
+      "loss": 1.2394,
       "step": 107990
     },
     {
       "epoch": 1.754642491592338,
-      "grad_norm": 2.578125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.1978,
       "step": 108000
     },
     {
       "epoch": 1.7548049584897076,
-      "grad_norm": 2.015625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2419,
       "step": 108010
     },
     {
       "epoch": 1.7549674253870773,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2579,
       "step": 108020
     },
     {
       "epoch": 1.755129892284447,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2115,
       "step": 108030
     },
     {
       "epoch": 1.7552923591818166,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.2693,
       "step": 108040
     },
     {
       "epoch": 1.7554548260791862,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2235,
       "step": 108050
     },
     {
       "epoch": 1.7556172929765559,
-      "grad_norm": 3.765625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.1854,
       "step": 108060
     },
     {
       "epoch": 1.7557797598739255,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2305,
       "step": 108070
     },
     {
       "epoch": 1.7559422267712952,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.274,
       "step": 108080
     },
     {
       "epoch": 1.756104693668665,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.2149,
       "step": 108090
     },
     {
       "epoch": 1.7562671605660347,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.1992,
       "step": 108100
     },
     {
       "epoch": 1.7564296274634044,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2016,
       "step": 108110
     },
     {
       "epoch": 1.756592094360774,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2388,
       "step": 108120
     },
     {
       "epoch": 1.7567545612581437,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.231,
       "step": 108130
     },
     {
       "epoch": 1.7569170281555133,
-      "grad_norm": 2.875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.1882,
       "step": 108140
     },
     {
       "epoch": 1.757079495052883,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3185,
+      "loss": 1.2124,
       "step": 108150
     },
     {
       "epoch": 1.7572419619502526,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2146,
       "step": 108160
     },
     {
       "epoch": 1.7574044288476223,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.1997,
       "step": 108170
     },
     {
       "epoch": 1.757566895744992,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3704,
+      "loss": 1.2433,
       "step": 108180
     },
     {
       "epoch": 1.7577293626423616,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.2101,
       "step": 108190
     },
     {
       "epoch": 1.7578918295397314,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.2106,
       "step": 108200
     },
     {
       "epoch": 1.758054296437101,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3125,
+      "loss": 1.1367,
       "step": 108210
     },
     {
       "epoch": 1.7582167633344707,
-      "grad_norm": 3.640625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3609,
+      "loss": 1.215,
       "step": 108220
     },
     {
       "epoch": 1.7583792302318404,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2053,
       "step": 108230
     },
     {
       "epoch": 1.75854169712921,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.1979,
       "step": 108240
     },
     {
       "epoch": 1.7587041640265797,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3268,
+      "loss": 1.2258,
       "step": 108250
     },
     {
       "epoch": 1.7588666309239493,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2552,
       "step": 108260
     },
     {
       "epoch": 1.759029097821319,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2304,
       "step": 108270
     },
     {
       "epoch": 1.7591915647186886,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2294,
       "step": 108280
     },
     {
       "epoch": 1.7593540316160583,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3375,
+      "loss": 1.2029,
       "step": 108290
     },
     {
       "epoch": 1.759516498513428,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2089,
       "step": 108300
     },
     {
       "epoch": 1.7596789654107976,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2135,
       "step": 108310
     },
     {
       "epoch": 1.7598414323081673,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2359,
       "step": 108320
     },
     {
       "epoch": 1.760003899205537,
-      "grad_norm": 2.09375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2348,
       "step": 108330
     },
     {
       "epoch": 1.7601663661029066,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.1662,
       "step": 108340
     },
     {
       "epoch": 1.7603288330002762,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2596,
       "step": 108350
     },
     {
       "epoch": 1.7604912998976459,
-      "grad_norm": 2.25,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2035,
       "step": 108360
     },
     {
       "epoch": 1.7606537667950155,
-      "grad_norm": 2.625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2317,
       "step": 108370
     },
     {
       "epoch": 1.7608162336923852,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3224,
+      "loss": 1.2276,
       "step": 108380
     },
     {
       "epoch": 1.7609787005897548,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.1888,
       "step": 108390
     },
     {
       "epoch": 1.7611411674871245,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3367,
+      "loss": 1.2115,
       "step": 108400
     },
     {
       "epoch": 1.761303634384494,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.1936,
       "step": 108410
     },
     {
       "epoch": 1.7614661012818638,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2413,
       "step": 108420
     },
     {
       "epoch": 1.7616285681792334,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.221,
       "step": 108430
     },
     {
       "epoch": 1.761791035076603,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2117,
       "step": 108440
     },
     {
       "epoch": 1.7619535019739727,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2549,
       "step": 108450
     },
     {
       "epoch": 1.7621159688713424,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2103,
       "step": 108460
     },
     {
       "epoch": 1.762278435768712,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.2185,
       "step": 108470
     },
     {
       "epoch": 1.7624409026660817,
-      "grad_norm": 3.34375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.2385,
       "step": 108480
     },
     {
       "epoch": 1.7626033695634513,
-      "grad_norm": 5.0,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.1978,
       "step": 108490
     },
     {
       "epoch": 1.762765836460821,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.1927,
       "step": 108500
     },
     {
       "epoch": 1.7629283033581906,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.1964,
       "step": 108510
     },
     {
       "epoch": 1.7630907702555603,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2079,
       "step": 108520
     },
     {
       "epoch": 1.7632532371529301,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.1948,
       "step": 108530
     },
     {
       "epoch": 1.7634157040502998,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2342,
       "step": 108540
     },
     {
       "epoch": 1.7635781709476694,
-      "grad_norm": 2.25,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3224,
+      "loss": 1.2161,
       "step": 108550
     },
     {
       "epoch": 1.763740637845039,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.206,
       "step": 108560
     },
     {
       "epoch": 1.7639031047424087,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3625,
+      "loss": 1.2183,
       "step": 108570
     },
     {
       "epoch": 1.7640655716397784,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1876,
       "step": 108580
     },
     {
       "epoch": 1.764228038537148,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2248,
       "step": 108590
     },
     {
       "epoch": 1.7643905054345177,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2249,
       "step": 108600
     },
     {
       "epoch": 1.7645529723318873,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2309,
       "step": 108610
     },
     {
       "epoch": 1.764715439229257,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3271,
+      "loss": 1.2436,
       "step": 108620
     },
     {
       "epoch": 1.7648779061266266,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1649,
       "step": 108630
     },
     {
       "epoch": 1.7650403730239965,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3237,
+      "loss": 1.2307,
       "step": 108640
     },
     {
       "epoch": 1.7652028399213662,
-      "grad_norm": 3.625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2497,
       "step": 108650
     },
     {
       "epoch": 1.7653653068187358,
-      "grad_norm": 1.9921875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1988,
       "step": 108660
     },
     {
       "epoch": 1.7655277737161055,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.206,
       "step": 108670
     },
     {
       "epoch": 1.7656902406134751,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.2227,
       "step": 108680
     },
     {
       "epoch": 1.7658527075108448,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2059,
       "step": 108690
     },
     {
       "epoch": 1.7660151744082144,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.1865,
       "step": 108700
     },
     {
       "epoch": 1.766177641305584,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.2184,
       "step": 108710
     },
     {
       "epoch": 1.7663401082029537,
-      "grad_norm": 2.046875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 1.2544,
       "step": 108720
     },
     {
       "epoch": 1.7665025751003234,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2151,
       "step": 108730
     },
     {
       "epoch": 1.766665041997693,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.1943,
       "step": 108740
     },
     {
       "epoch": 1.7668275088950627,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1979,
       "step": 108750
     },
     {
       "epoch": 1.7669899757924323,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3238,
+      "loss": 1.1972,
       "step": 108760
     },
     {
       "epoch": 1.767152442689802,
-      "grad_norm": 3.125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.1524,
       "step": 108770
     },
     {
       "epoch": 1.7673149095871716,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.1934,
       "step": 108780
     },
     {
       "epoch": 1.7674773764845413,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.1707,
       "step": 108790
     },
     {
       "epoch": 1.767639843381911,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2023,
       "step": 108800
     },
     {
       "epoch": 1.7678023102792806,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.2361,
       "step": 108810
     },
     {
       "epoch": 1.7679647771766502,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.253,
       "step": 108820
     },
     {
       "epoch": 1.7681272440740199,
-      "grad_norm": 2.5,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2178,
       "step": 108830
     },
     {
       "epoch": 1.7682897109713895,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.2461,
       "step": 108840
     },
     {
       "epoch": 1.7684521778687592,
-      "grad_norm": 2.34375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2258,
       "step": 108850
     },
     {
       "epoch": 1.7686146447661288,
-      "grad_norm": 3.5,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2289,
       "step": 108860
     },
     {
       "epoch": 1.7687771116634985,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.1956,
       "step": 108870
     },
     {
       "epoch": 1.7689395785608681,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3619,
+      "loss": 1.2472,
       "step": 108880
     },
     {
       "epoch": 1.7691020454582378,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2193,
       "step": 108890
     },
     {
       "epoch": 1.7692645123556074,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3232,
+      "loss": 1.2416,
       "step": 108900
     },
     {
       "epoch": 1.769426979252977,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.214,
       "step": 108910
     },
     {
       "epoch": 1.7695894461503467,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.1999,
       "step": 108920
     },
     {
       "epoch": 1.7697519130477164,
-      "grad_norm": 5.3125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.1686,
       "step": 108930
     },
     {
       "epoch": 1.769914379945086,
-      "grad_norm": 2.078125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2291,
       "step": 108940
     },
     {
       "epoch": 1.7700768468424557,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.2658,
       "step": 108950
     },
     {
       "epoch": 1.7702393137398253,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.1915,
       "step": 108960
     },
     {
       "epoch": 1.7704017806371952,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.375,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2062,
       "step": 108970
     },
     {
       "epoch": 1.7705642475345649,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.2595,
       "step": 108980
     },
     {
       "epoch": 1.7707267144319345,
-      "grad_norm": 3.25,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2197,
       "step": 108990
     },
     {
       "epoch": 1.7708891813293042,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2075,
       "step": 109000
     },
     {
       "epoch": 1.7710516482266738,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2494,
       "step": 109010
     },
     {
       "epoch": 1.7712141151240435,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2004,
       "step": 109020
     },
     {
       "epoch": 1.7713765820214131,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2758,
       "step": 109030
     },
     {
       "epoch": 1.7715390489187828,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3075,
+      "loss": 1.2813,
       "step": 109040
     },
     {
       "epoch": 1.7717015158161524,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3441,
+      "loss": 1.2071,
       "step": 109050
     },
     {
       "epoch": 1.771863982713522,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2226,
       "step": 109060
     },
     {
       "epoch": 1.7720264496108917,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.225,
       "step": 109070
     },
     {
       "epoch": 1.7721889165082616,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.1702,
       "step": 109080
     },
     {
       "epoch": 1.7723513834056313,
-      "grad_norm": 3.125,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3741,
+      "loss": 1.1931,
       "step": 109090
     },
     {
       "epoch": 1.772513850303001,
-      "grad_norm": 2.75,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2065,
       "step": 109100
     },
     {
       "epoch": 1.7726763172003706,
-      "grad_norm": 2.375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2156,
       "step": 109110
     },
     {
       "epoch": 1.7728387840977402,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2107,
       "step": 109120
     },
     {
       "epoch": 1.7730012509951099,
-      "grad_norm": 3.125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2136,
       "step": 109130
     },
     {
       "epoch": 1.7731637178924795,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2174,
       "step": 109140
     },
     {
       "epoch": 1.7733261847898492,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.1926,
       "step": 109150
     },
     {
       "epoch": 1.7734886516872188,
-      "grad_norm": 3.0,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.1943,
       "step": 109160
     },
     {
       "epoch": 1.7736511185845885,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2229,
       "step": 109170
     },
     {
       "epoch": 1.773813585481958,
-      "grad_norm": 3.5,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2533,
       "step": 109180
     },
     {
       "epoch": 1.7739760523793278,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2617,
       "step": 109190
     },
     {
       "epoch": 1.7741385192766974,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2758,
       "step": 109200
     },
     {
       "epoch": 1.774300986174067,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1724,
       "step": 109210
     },
     {
       "epoch": 1.7744634530714367,
-      "grad_norm": 2.59375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2611,
       "step": 109220
     },
     {
       "epoch": 1.7746259199688064,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1655,
       "step": 109230
     },
     {
       "epoch": 1.774788386866176,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3629,
+      "loss": 1.1891,
       "step": 109240
     },
     {
       "epoch": 1.7749508537635457,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2336,
       "step": 109250
     },
     {
       "epoch": 1.7751133206609153,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3599,
+      "loss": 1.2303,
       "step": 109260
     },
     {
       "epoch": 1.775275787558285,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.3652,
+      "loss": 1.1714,
       "step": 109270
     },
     {
       "epoch": 1.7754382544556546,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.1696,
       "step": 109280
     },
     {
       "epoch": 1.7756007213530243,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2338,
       "step": 109290
     },
     {
       "epoch": 1.775763188250394,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2099,
       "step": 109300
     },
     {
       "epoch": 1.7759256551477636,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2092,
       "step": 109310
     },
     {
       "epoch": 1.7760881220451332,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.1923,
       "step": 109320
     },
     {
       "epoch": 1.7762505889425029,
-      "grad_norm": 3.765625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.333,
+      "loss": 1.2105,
       "step": 109330
     },
     {
       "epoch": 1.7764130558398725,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3161,
+      "loss": 1.2207,
       "step": 109340
     },
     {
       "epoch": 1.7765755227372422,
-      "grad_norm": 3.9375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.1742,
       "step": 109350
     },
     {
       "epoch": 1.7767379896346118,
-      "grad_norm": 2.375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.1529,
       "step": 109360
     },
     {
       "epoch": 1.7769004565319815,
-      "grad_norm": 2.625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2015,
       "step": 109370
     },
     {
       "epoch": 1.7770629234293511,
-      "grad_norm": 3.0625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3771,
+      "loss": 1.2525,
       "step": 109380
     },
     {
       "epoch": 1.7772253903267208,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.183,
       "step": 109390
     },
     {
       "epoch": 1.7773878572240904,
-      "grad_norm": 2.5625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2251,
       "step": 109400
     },
     {
       "epoch": 1.7775503241214603,
-      "grad_norm": 2.5,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2401,
       "step": 109410
     },
     {
       "epoch": 1.77771279101883,
-      "grad_norm": 1.9296875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.2193,
       "step": 109420
     },
     {
       "epoch": 1.7778752579161996,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2427,
       "step": 109430
     },
     {
       "epoch": 1.7780377248135693,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2402,
       "step": 109440
     },
     {
       "epoch": 1.778200191710939,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2325,
       "step": 109450
     },
     {
       "epoch": 1.7783626586083086,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2145,
       "step": 109460
     },
     {
       "epoch": 1.7785251255056782,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.1909,
       "step": 109470
     },
     {
       "epoch": 1.7786875924030479,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3239,
+      "loss": 1.2534,
       "step": 109480
     },
     {
       "epoch": 1.7788500593004175,
-      "grad_norm": 4.3125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2172,
       "step": 109490
     },
     {
       "epoch": 1.7790125261977872,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3254,
+      "loss": 1.2374,
       "step": 109500
     },
     {
       "epoch": 1.7791749930951568,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2079,
       "step": 109510
     },
     {
       "epoch": 1.7793374599925267,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2322,
       "step": 109520
     },
     {
       "epoch": 1.7794999268898963,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.1839,
       "step": 109530
     },
     {
       "epoch": 1.779662393787266,
-      "grad_norm": 3.375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2233,
       "step": 109540
     },
     {
       "epoch": 1.7798248606846356,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.213,
       "step": 109550
     },
     {
       "epoch": 1.7799873275820053,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2324,
       "step": 109560
     },
     {
       "epoch": 1.780149794479375,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2626,
       "step": 109570
     },
     {
       "epoch": 1.7803122613767446,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.1906,
       "step": 109580
     },
     {
       "epoch": 1.7804747282741142,
-      "grad_norm": 4.28125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.196,
       "step": 109590
     },
     {
       "epoch": 1.7806371951714839,
-      "grad_norm": 2.234375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2637,
       "step": 109600
     },
     {
       "epoch": 1.7807996620688535,
-      "grad_norm": 2.125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.2135,
       "step": 109610
     },
     {
       "epoch": 1.7809621289662232,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.222,
       "step": 109620
     },
     {
       "epoch": 1.7811245958635928,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.329,
+      "loss": 1.2229,
       "step": 109630
     },
     {
       "epoch": 1.7812870627609625,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2165,
       "step": 109640
     },
     {
       "epoch": 1.7814495296583321,
-      "grad_norm": 2.640625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.1865,
       "step": 109650
     },
     {
       "epoch": 1.7816119965557018,
-      "grad_norm": 2.625,
+      "grad_norm": 12.75,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2355,
       "step": 109660
     },
     {
       "epoch": 1.7817744634530714,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2041,
       "step": 109670
     },
     {
       "epoch": 1.781936930350441,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.1972,
       "step": 109680
     },
     {
       "epoch": 1.7820993972478107,
-      "grad_norm": 2.03125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2226,
       "step": 109690
     },
     {
       "epoch": 1.7822618641451804,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2487,
       "step": 109700
     },
     {
       "epoch": 1.78242433104255,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3128,
+      "loss": 1.1508,
       "step": 109710
     },
     {
       "epoch": 1.7825867979399197,
-      "grad_norm": 2.75,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.212,
       "step": 109720
     },
     {
       "epoch": 1.7827492648372893,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1859,
       "step": 109730
     },
     {
       "epoch": 1.782911731734659,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2103,
       "step": 109740
     },
     {
       "epoch": 1.7830741986320287,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3481,
+      "loss": 1.2238,
       "step": 109750
     },
     {
       "epoch": 1.7832366655293983,
-      "grad_norm": 3.46875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.2155,
       "step": 109760
     },
     {
       "epoch": 1.783399132426768,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2207,
       "step": 109770
     },
     {
       "epoch": 1.7835615993241376,
-      "grad_norm": 2.1875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.1901,
       "step": 109780
     },
     {
       "epoch": 1.7837240662215073,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2091,
       "step": 109790
     },
     {
       "epoch": 1.783886533118877,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.342,
+      "loss": 1.2426,
       "step": 109800
     },
     {
       "epoch": 1.7840490000162466,
-      "grad_norm": 4.40625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2252,
       "step": 109810
     },
     {
       "epoch": 1.7842114669136162,
-      "grad_norm": 4.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.2754,
       "step": 109820
     },
     {
       "epoch": 1.7843739338109859,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2014,
       "step": 109830
     },
     {
       "epoch": 1.7845364007083555,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.241,
       "step": 109840
     },
     {
       "epoch": 1.7846988676057254,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.1909,
       "step": 109850
     },
     {
       "epoch": 1.784861334503095,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2293,
       "step": 109860
     },
     {
       "epoch": 1.7850238014004647,
-      "grad_norm": 4.34375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.229,
       "step": 109870
     },
     {
       "epoch": 1.7851862682978343,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.219,
       "step": 109880
     },
     {
       "epoch": 1.785348735195204,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2301,
       "step": 109890
     },
     {
       "epoch": 1.7855112020925736,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.2252,
       "step": 109900
     },
     {
       "epoch": 1.7856736689899433,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2169,
       "step": 109910
     },
     {
       "epoch": 1.785836135887313,
-      "grad_norm": 4.21875,
+      "grad_norm": 13.75,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2267,
       "step": 109920
     },
     {
       "epoch": 1.7859986027846826,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.2123,
       "step": 109930
     },
     {
       "epoch": 1.7861610696820522,
-      "grad_norm": 2.375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3201,
+      "loss": 1.2241,
       "step": 109940
     },
     {
       "epoch": 1.7863235365794219,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2065,
       "step": 109950
     },
     {
       "epoch": 1.7864860034767918,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3281,
+      "loss": 1.2526,
       "step": 109960
     },
     {
       "epoch": 1.7866484703741614,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.368,
+      "loss": 1.2643,
       "step": 109970
     },
     {
       "epoch": 1.786810937271531,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2249,
       "step": 109980
     },
     {
       "epoch": 1.7869734041689007,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3683,
+      "loss": 1.228,
       "step": 109990
     },
     {
       "epoch": 1.7871358710662704,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2131,
       "step": 110000
     },
     {
       "epoch": 1.78729833796364,
-      "grad_norm": 3.125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.185,
       "step": 110010
     },
     {
       "epoch": 1.7874608048610097,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2091,
       "step": 110020
     },
     {
       "epoch": 1.7876232717583793,
-      "grad_norm": 3.125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.246,
       "step": 110030
     },
     {
       "epoch": 1.787785738655749,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2156,
       "step": 110040
     },
     {
       "epoch": 1.7879482055531186,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.2057,
       "step": 110050
     },
     {
       "epoch": 1.7881106724504883,
-      "grad_norm": 2.984375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2373,
       "step": 110060
     },
     {
       "epoch": 1.788273139347858,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.1728,
       "step": 110070
     },
     {
       "epoch": 1.7884356062452276,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3272,
+      "loss": 1.2329,
       "step": 110080
     },
     {
       "epoch": 1.7885980731425972,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2286,
       "step": 110090
     },
     {
       "epoch": 1.7887605400399669,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3577,
+      "loss": 1.1986,
       "step": 110100
     },
     {
       "epoch": 1.7889230069373365,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.198,
       "step": 110110
     },
     {
       "epoch": 1.7890854738347062,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2353,
       "step": 110120
     },
     {
       "epoch": 1.7892479407320758,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2014,
       "step": 110130
     },
     {
       "epoch": 1.7894104076294455,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2071,
       "step": 110140
     },
     {
       "epoch": 1.7895728745268151,
-      "grad_norm": 2.15625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3189,
+      "loss": 1.2222,
       "step": 110150
     },
     {
       "epoch": 1.7897353414241848,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2461,
       "step": 110160
     },
     {
       "epoch": 1.7898978083215544,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2559,
       "step": 110170
     },
     {
       "epoch": 1.790060275218924,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2243,
       "step": 110180
     },
     {
       "epoch": 1.7902227421162937,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3582,
+      "loss": 1.239,
       "step": 110190
     },
     {
       "epoch": 1.7903852090136634,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1862,
       "step": 110200
     },
     {
       "epoch": 1.790547675911033,
-      "grad_norm": 2.25,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.2729,
       "step": 110210
     },
     {
       "epoch": 1.7907101428084027,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.204,
       "step": 110220
     },
     {
       "epoch": 1.7908726097057723,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3185,
+      "loss": 1.1889,
       "step": 110230
     },
     {
       "epoch": 1.791035076603142,
-      "grad_norm": 2.625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.213,
       "step": 110240
     },
     {
       "epoch": 1.7911975435005116,
-      "grad_norm": 2.5,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.317,
+      "loss": 1.2038,
       "step": 110250
     },
     {
       "epoch": 1.7913600103978813,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.2073,
       "step": 110260
     },
     {
       "epoch": 1.791522477295251,
-      "grad_norm": 3.796875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2335,
       "step": 110270
     },
     {
       "epoch": 1.7916849441926206,
-      "grad_norm": 2.203125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2644,
       "step": 110280
     },
     {
       "epoch": 1.7918474110899905,
-      "grad_norm": 2.859375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.1904,
       "step": 110290
     },
     {
       "epoch": 1.7920098779873601,
-      "grad_norm": 2.203125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2437,
       "step": 110300
     },
     {
       "epoch": 1.7921723448847298,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.2503,
       "step": 110310
     },
     {
       "epoch": 1.7923348117820994,
-      "grad_norm": 2.375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2036,
       "step": 110320
     },
     {
       "epoch": 1.792497278679469,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2628,
       "step": 110330
     },
     {
       "epoch": 1.7926597455768387,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3219,
+      "loss": 1.2227,
       "step": 110340
     },
     {
       "epoch": 1.7928222124742084,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.2571,
       "step": 110350
     },
     {
       "epoch": 1.792984679371578,
-      "grad_norm": 2.625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.194,
       "step": 110360
     },
     {
       "epoch": 1.7931471462689477,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2404,
       "step": 110370
     },
     {
       "epoch": 1.7933096131663173,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.2166,
       "step": 110380
     },
     {
       "epoch": 1.793472080063687,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2319,
       "step": 110390
     },
     {
       "epoch": 1.7936345469610568,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.219,
       "step": 110400
     },
     {
       "epoch": 1.7937970138584265,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2304,
       "step": 110410
     },
     {
       "epoch": 1.7939594807557961,
-      "grad_norm": 4.53125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2444,
       "step": 110420
     },
     {
       "epoch": 1.7941219476531658,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2424,
       "step": 110430
     },
     {
       "epoch": 1.7942844145505354,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2108,
       "step": 110440
     },
     {
       "epoch": 1.794446881447905,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2321,
       "step": 110450
     },
     {
       "epoch": 1.7946093483452747,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2193,
       "step": 110460
     },
     {
       "epoch": 1.7947718152426444,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2547,
       "step": 110470
     },
     {
       "epoch": 1.794934282140014,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2547,
       "step": 110480
     },
     {
       "epoch": 1.7950967490373837,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.1947,
       "step": 110490
     },
     {
       "epoch": 1.7952592159347533,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2319,
       "step": 110500
     },
     {
       "epoch": 1.795421682832123,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3663,
+      "loss": 1.1937,
       "step": 110510
     },
     {
       "epoch": 1.7955841497294927,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2041,
       "step": 110520
     },
     {
       "epoch": 1.7957466166268623,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2555,
       "step": 110530
     },
     {
       "epoch": 1.795909083524232,
-      "grad_norm": 3.953125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3593,
+      "loss": 1.2728,
       "step": 110540
     },
     {
       "epoch": 1.7960715504216016,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3555,
+      "loss": 1.2167,
       "step": 110550
     },
     {
       "epoch": 1.7962340173189713,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3177,
+      "loss": 1.2244,
       "step": 110560
     },
     {
       "epoch": 1.796396484216341,
-      "grad_norm": 2.875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2032,
       "step": 110570
     },
     {
       "epoch": 1.7965589511137106,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3721,
+      "loss": 1.1836,
       "step": 110580
     },
     {
       "epoch": 1.7967214180110802,
-      "grad_norm": 2.203125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2548,
       "step": 110590
     },
     {
       "epoch": 1.7968838849084499,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1835,
       "step": 110600
     },
     {
       "epoch": 1.7970463518058195,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1761,
       "step": 110610
     },
     {
       "epoch": 1.7972088187031892,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.1805,
       "step": 110620
     },
     {
       "epoch": 1.7973712856005588,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2407,
       "step": 110630
     },
     {
       "epoch": 1.7975337524979285,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.198,
       "step": 110640
     },
     {
       "epoch": 1.7976962193952981,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.1991,
       "step": 110650
     },
     {
       "epoch": 1.7978586862926678,
-      "grad_norm": 2.875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2065,
       "step": 110660
     },
     {
       "epoch": 1.7980211531900374,
-      "grad_norm": 2.75,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2505,
       "step": 110670
     },
     {
       "epoch": 1.798183620087407,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2283,
       "step": 110680
     },
     {
       "epoch": 1.7983460869847767,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2164,
       "step": 110690
     },
     {
       "epoch": 1.7985085538821464,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2655,
       "step": 110700
     },
     {
       "epoch": 1.798671020779516,
-      "grad_norm": 3.125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2655,
       "step": 110710
     },
     {
       "epoch": 1.798833487676886,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.1707,
       "step": 110720
     },
     {
       "epoch": 1.7989959545742555,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2163,
       "step": 110730
     },
     {
       "epoch": 1.7991584214716252,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2138,
       "step": 110740
     },
     {
       "epoch": 1.7993208883689948,
-      "grad_norm": 2.5,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.1836,
       "step": 110750
     },
     {
       "epoch": 1.7994833552663645,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.1959,
       "step": 110760
     },
     {
       "epoch": 1.7996458221637341,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2204,
       "step": 110770
     },
     {
       "epoch": 1.7998082890611038,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3506,
+      "loss": 1.209,
       "step": 110780
     },
     {
       "epoch": 1.7999707559584734,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2281,
       "step": 110790
     },
     {
       "epoch": 1.800133222855843,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3252,
+      "loss": 1.1751,
       "step": 110800
     },
     {
       "epoch": 1.8002956897532127,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.262,
       "step": 110810
     },
     {
       "epoch": 1.8004581566505824,
-      "grad_norm": 4.25,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2194,
       "step": 110820
     },
     {
       "epoch": 1.8006206235479523,
-      "grad_norm": 2.125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.1982,
       "step": 110830
     },
     {
       "epoch": 1.800783090445322,
-      "grad_norm": 2.25,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3354,
+      "loss": 1.1861,
       "step": 110840
     },
     {
       "epoch": 1.8009455573426916,
-      "grad_norm": 3.375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2039,
       "step": 110850
     },
     {
       "epoch": 1.8011080242400612,
-      "grad_norm": 3.5,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3163,
+      "loss": 1.2277,
       "step": 110860
     },
     {
       "epoch": 1.8012704911374309,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.1987,
       "step": 110870
     },
     {
       "epoch": 1.8014329580348005,
-      "grad_norm": 2.609375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2145,
       "step": 110880
     },
     {
       "epoch": 1.8015954249321702,
-      "grad_norm": 2.5,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.319,
+      "loss": 1.2142,
       "step": 110890
     },
     {
       "epoch": 1.8017578918295398,
-      "grad_norm": 3.25,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.1934,
       "step": 110900
     },
     {
       "epoch": 1.8019203587269095,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.2225,
       "step": 110910
     },
     {
       "epoch": 1.8020828256242791,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.183,
       "step": 110920
     },
     {
       "epoch": 1.8022452925216488,
-      "grad_norm": 2.6875,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.1921,
       "step": 110930
     },
     {
       "epoch": 1.8024077594190184,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2054,
       "step": 110940
     },
     {
       "epoch": 1.802570226316388,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2459,
       "step": 110950
     },
     {
       "epoch": 1.8027326932137577,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3155,
+      "loss": 1.2242,
       "step": 110960
     },
     {
       "epoch": 1.8028951601111274,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.206,
       "step": 110970
     },
     {
       "epoch": 1.803057627008497,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.2513,
       "step": 110980
     },
     {
       "epoch": 1.8032200939058667,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.1856,
       "step": 110990
     },
     {
       "epoch": 1.8033825608032363,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.2256,
       "step": 111000
     },
     {
       "epoch": 1.803545027700606,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2352,
       "step": 111010
     },
     {
       "epoch": 1.8037074945979756,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.1975,
       "step": 111020
     },
     {
       "epoch": 1.8038699614953453,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2158,
       "step": 111030
     },
     {
       "epoch": 1.804032428392715,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3211,
+      "loss": 1.2024,
       "step": 111040
     },
     {
       "epoch": 1.8041948952900846,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.1886,
       "step": 111050
     },
     {
       "epoch": 1.8043573621874542,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2413,
       "step": 111060
     },
     {
       "epoch": 1.804519829084824,
-      "grad_norm": 2.875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3184,
+      "loss": 1.2371,
       "step": 111070
     },
     {
       "epoch": 1.8046822959821935,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2465,
       "step": 111080
     },
     {
       "epoch": 1.8048447628795632,
-      "grad_norm": 2.875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2593,
       "step": 111090
     },
     {
       "epoch": 1.8050072297769328,
-      "grad_norm": 2.0625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.1696,
       "step": 111100
     },
     {
       "epoch": 1.8051696966743025,
-      "grad_norm": 2.375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.2046,
       "step": 111110
     },
     {
       "epoch": 1.8053321635716721,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.1982,
       "step": 111120
     },
     {
       "epoch": 1.8054946304690418,
-      "grad_norm": 2.375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.235,
       "step": 111130
     },
     {
       "epoch": 1.8056570973664114,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3239,
+      "loss": 1.1935,
       "step": 111140
     },
     {
       "epoch": 1.805819564263781,
-      "grad_norm": 3.640625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.381,
+      "loss": 1.2473,
       "step": 111150
     },
     {
       "epoch": 1.805982031161151,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.199,
       "step": 111160
     },
     {
       "epoch": 1.8061444980585206,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.1871,
       "step": 111170
     },
     {
       "epoch": 1.8063069649558903,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2021,
       "step": 111180
     },
     {
       "epoch": 1.80646943185326,
-      "grad_norm": 2.5,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.227,
       "step": 111190
     },
     {
       "epoch": 1.8066318987506296,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2316,
       "step": 111200
     },
     {
       "epoch": 1.8067943656479992,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.1897,
       "step": 111210
     },
     {
       "epoch": 1.8069568325453689,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2114,
       "step": 111220
     },
     {
       "epoch": 1.8071192994427385,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.2113,
       "step": 111230
     },
     {
       "epoch": 1.8072817663401082,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2392,
       "step": 111240
     },
     {
       "epoch": 1.8074442332374778,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.1972,
       "step": 111250
     },
     {
       "epoch": 1.8076067001348475,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2514,
       "step": 111260
     },
     {
       "epoch": 1.8077691670322174,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.1968,
       "step": 111270
     },
     {
       "epoch": 1.807931633929587,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2296,
       "step": 111280
     },
     {
       "epoch": 1.8080941008269567,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2165,
       "step": 111290
     },
     {
       "epoch": 1.8082565677243263,
-      "grad_norm": 4.03125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.2259,
       "step": 111300
     },
     {
       "epoch": 1.808419034621696,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2162,
       "step": 111310
     },
     {
       "epoch": 1.8085815015190656,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.1948,
       "step": 111320
     },
     {
       "epoch": 1.8087439684164353,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.214,
       "step": 111330
     },
     {
       "epoch": 1.808906435313805,
-      "grad_norm": 2.375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2501,
       "step": 111340
     },
     {
       "epoch": 1.8090689022111746,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2313,
       "step": 111350
     },
     {
       "epoch": 1.8092313691085442,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.2144,
       "step": 111360
     },
     {
       "epoch": 1.8093938360059139,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2058,
       "step": 111370
     },
     {
       "epoch": 1.8095563029032835,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.1964,
       "step": 111380
     },
     {
       "epoch": 1.8097187698006532,
-      "grad_norm": 3.15625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3632,
+      "loss": 1.2475,
       "step": 111390
     },
     {
       "epoch": 1.8098812366980228,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.1708,
       "step": 111400
     },
     {
       "epoch": 1.8100437035953925,
-      "grad_norm": 3.0625,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.1778,
       "step": 111410
     },
     {
       "epoch": 1.8102061704927621,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3222,
+      "loss": 1.2415,
       "step": 111420
     },
     {
       "epoch": 1.8103686373901318,
-      "grad_norm": 4.0,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3171,
+      "loss": 1.1912,
       "step": 111430
     },
     {
       "epoch": 1.8105311042875014,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2232,
       "step": 111440
     },
     {
       "epoch": 1.810693571184871,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.198,
       "step": 111450
     },
     {
       "epoch": 1.8108560380822407,
-      "grad_norm": 2.875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.1858,
       "step": 111460
     },
     {
       "epoch": 1.8110185049796104,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2321,
       "step": 111470
     },
     {
       "epoch": 1.81118097187698,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.2515,
       "step": 111480
     },
     {
       "epoch": 1.8113434387743497,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3641,
+      "loss": 1.2125,
       "step": 111490
     },
     {
       "epoch": 1.8115059056717193,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2041,
       "step": 111500
     },
     {
       "epoch": 1.811668372569089,
-      "grad_norm": 2.375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3736,
+      "loss": 1.228,
       "step": 111510
     },
     {
       "epoch": 1.8118308394664586,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.2332,
       "step": 111520
     },
     {
       "epoch": 1.8119933063638283,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2196,
       "step": 111530
     },
     {
       "epoch": 1.812155773261198,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1907,
       "step": 111540
     },
     {
       "epoch": 1.8123182401585676,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.1688,
       "step": 111550
     },
     {
       "epoch": 1.8124807070559372,
-      "grad_norm": 4.53125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2187,
       "step": 111560
     },
     {
       "epoch": 1.8126431739533069,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2432,
       "step": 111570
     },
     {
       "epoch": 1.8128056408506765,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.1935,
       "step": 111580
     },
     {
       "epoch": 1.8129681077480462,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2228,
       "step": 111590
     },
     {
       "epoch": 1.813130574645416,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.19,
       "step": 111600
     },
     {
       "epoch": 1.8132930415427857,
-      "grad_norm": 2.578125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3635,
+      "loss": 1.2526,
       "step": 111610
     },
     {
       "epoch": 1.8134555084401554,
-      "grad_norm": 4.0,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2021,
       "step": 111620
     },
     {
       "epoch": 1.813617975337525,
-      "grad_norm": 2.25,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.1662,
       "step": 111630
     },
     {
       "epoch": 1.8137804422348947,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1919,
       "step": 111640
     },
     {
       "epoch": 1.8139429091322643,
-      "grad_norm": 3.203125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.204,
       "step": 111650
     },
     {
       "epoch": 1.814105376029634,
-      "grad_norm": 2.25,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2167,
       "step": 111660
     },
     {
       "epoch": 1.8142678429270036,
-      "grad_norm": 2.75,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1988,
       "step": 111670
     },
     {
       "epoch": 1.8144303098243733,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2215,
       "step": 111680
     },
     {
       "epoch": 1.814592776721743,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2157,
       "step": 111690
     },
     {
       "epoch": 1.8147552436191126,
-      "grad_norm": 2.921875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2321,
       "step": 111700
     },
     {
       "epoch": 1.8149177105164824,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.1913,
       "step": 111710
     },
     {
       "epoch": 1.815080177413852,
-      "grad_norm": 4.21875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2306,
       "step": 111720
     },
     {
       "epoch": 1.8152426443112217,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1976,
       "step": 111730
     },
     {
       "epoch": 1.8154051112085914,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.232,
       "step": 111740
     },
     {
       "epoch": 1.815567578105961,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2537,
       "step": 111750
     },
     {
       "epoch": 1.8157300450033307,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2406,
       "step": 111760
     },
     {
       "epoch": 1.8158925119007003,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.322,
+      "loss": 1.1959,
       "step": 111770
     },
     {
       "epoch": 1.81605497879807,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3527,
+      "loss": 1.2666,
       "step": 111780
     },
     {
       "epoch": 1.8162174456954396,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.2067,
       "step": 111790
     },
     {
       "epoch": 1.8163799125928093,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3629,
+      "loss": 1.259,
       "step": 111800
     },
     {
       "epoch": 1.816542379490179,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2129,
       "step": 111810
     },
     {
       "epoch": 1.8167048463875486,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.198,
       "step": 111820
     },
     {
       "epoch": 1.8168673132849182,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2267,
       "step": 111830
     },
     {
       "epoch": 1.817029780182288,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2166,
       "step": 111840
     },
     {
       "epoch": 1.8171922470796575,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2328,
       "step": 111850
     },
     {
       "epoch": 1.8173547139770272,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.2394,
       "step": 111860
     },
     {
       "epoch": 1.8175171808743968,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3194,
+      "loss": 1.2035,
       "step": 111870
     },
     {
       "epoch": 1.8176796477717665,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2468,
       "step": 111880
     },
     {
       "epoch": 1.8178421146691361,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3609,
+      "loss": 1.2161,
       "step": 111890
     },
     {
       "epoch": 1.8180045815665058,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3624,
+      "loss": 1.1941,
       "step": 111900
     },
     {
       "epoch": 1.8181670484638754,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.204,
       "step": 111910
     },
     {
       "epoch": 1.818329515361245,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.1954,
       "step": 111920
     },
     {
       "epoch": 1.8184919822586147,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.319,
+      "loss": 1.2002,
       "step": 111930
     },
     {
       "epoch": 1.8186544491559844,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.2039,
       "step": 111940
     },
     {
       "epoch": 1.818816916053354,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2489,
       "step": 111950
     },
     {
       "epoch": 1.8189793829507237,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3374,
+      "loss": 1.22,
       "step": 111960
     },
     {
       "epoch": 1.8191418498480934,
-      "grad_norm": 3.40625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3534,
+      "loss": 1.1938,
       "step": 111970
     },
     {
       "epoch": 1.819304316745463,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2374,
       "step": 111980
     },
     {
       "epoch": 1.8194667836428327,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.75,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.228,
       "step": 111990
     },
     {
       "epoch": 1.8196292505402023,
-      "grad_norm": 2.171875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.2153,
       "step": 112000
     },
     {
       "epoch": 1.819791717437572,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.2268,
       "step": 112010
     },
     {
       "epoch": 1.8199541843349416,
-      "grad_norm": 3.9375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2132,
       "step": 112020
     },
     {
       "epoch": 1.8201166512323113,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2723,
       "step": 112030
     },
     {
       "epoch": 1.8202791181296811,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.2317,
       "step": 112040
     },
     {
       "epoch": 1.8204415850270508,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2269,
       "step": 112050
     },
     {
       "epoch": 1.8206040519244204,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2071,
       "step": 112060
     },
     {
       "epoch": 1.82076651882179,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.2359,
       "step": 112070
     },
     {
       "epoch": 1.8209289857191597,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2453,
       "step": 112080
     },
     {
       "epoch": 1.8210914526165294,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2142,
       "step": 112090
     },
     {
       "epoch": 1.821253919513899,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2194,
       "step": 112100
     },
     {
       "epoch": 1.8214163864112687,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.1956,
       "step": 112110
     },
     {
       "epoch": 1.8215788533086383,
-      "grad_norm": 1.9765625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.309,
+      "loss": 1.1707,
       "step": 112120
     },
     {
       "epoch": 1.821741320206008,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2268,
       "step": 112130
     },
     {
       "epoch": 1.8219037871033776,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2051,
       "step": 112140
     },
     {
       "epoch": 1.8220662540007475,
-      "grad_norm": 3.0,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.186,
       "step": 112150
     },
     {
       "epoch": 1.8222287208981172,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2407,
       "step": 112160
     },
     {
       "epoch": 1.8223911877954868,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.2111,
       "step": 112170
     },
     {
       "epoch": 1.8225536546928565,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3547,
+      "loss": 1.1905,
       "step": 112180
     },
     {
       "epoch": 1.8227161215902261,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1838,
       "step": 112190
     },
     {
       "epoch": 1.8228785884875958,
-      "grad_norm": 3.625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3315,
+      "loss": 1.2929,
       "step": 112200
     },
     {
       "epoch": 1.8230410553849654,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.1949,
       "step": 112210
     },
     {
       "epoch": 1.823203522282335,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2799,
       "step": 112220
     },
     {
       "epoch": 1.8233659891797047,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2298,
       "step": 112230
     },
     {
       "epoch": 1.8235284560770744,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2044,
       "step": 112240
     },
     {
       "epoch": 1.823690922974444,
-      "grad_norm": 3.0,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.1985,
       "step": 112250
     },
     {
       "epoch": 1.8238533898718137,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2107,
       "step": 112260
     },
     {
       "epoch": 1.8240158567691833,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3682,
+      "loss": 1.1825,
       "step": 112270
     },
     {
       "epoch": 1.824178323666553,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.1932,
       "step": 112280
     },
     {
       "epoch": 1.8243407905639226,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2227,
       "step": 112290
     },
     {
       "epoch": 1.8245032574612923,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2207,
       "step": 112300
     },
     {
       "epoch": 1.824665724358662,
-      "grad_norm": 2.625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2428,
       "step": 112310
     },
     {
       "epoch": 1.8248281912560316,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.1786,
       "step": 112320
     },
     {
       "epoch": 1.8249906581534012,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2051,
       "step": 112330
     },
     {
       "epoch": 1.8251531250507709,
-      "grad_norm": 3.375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.2455,
       "step": 112340
     },
     {
       "epoch": 1.8253155919481405,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.1828,
       "step": 112350
     },
     {
       "epoch": 1.8254780588455102,
-      "grad_norm": 2.5,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.1847,
       "step": 112360
     },
     {
       "epoch": 1.8256405257428798,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.1779,
       "step": 112370
     },
     {
       "epoch": 1.8258029926402495,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3154,
+      "loss": 1.2732,
       "step": 112380
     },
     {
       "epoch": 1.8259654595376191,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1855,
       "step": 112390
     },
     {
       "epoch": 1.8261279264349888,
-      "grad_norm": 4.40625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3218,
+      "loss": 1.2491,
       "step": 112400
     },
     {
       "epoch": 1.8262903933323584,
-      "grad_norm": 3.953125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3558,
+      "loss": 1.1795,
       "step": 112410
     },
     {
       "epoch": 1.826452860229728,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.1791,
       "step": 112420
     },
     {
       "epoch": 1.8266153271270977,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.1956,
       "step": 112430
     },
     {
       "epoch": 1.8267777940244674,
-      "grad_norm": 3.34375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1849,
       "step": 112440
     },
     {
       "epoch": 1.826940260921837,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2079,
       "step": 112450
     },
     {
       "epoch": 1.8271027278192067,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2364,
       "step": 112460
     },
     {
       "epoch": 1.8272651947165763,
-      "grad_norm": 3.0,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2177,
       "step": 112470
     },
     {
       "epoch": 1.8274276616139462,
-      "grad_norm": 2.078125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2061,
       "step": 112480
     },
     {
       "epoch": 1.8275901285113159,
-      "grad_norm": 3.984375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.1858,
       "step": 112490
     },
     {
       "epoch": 1.8277525954086855,
-      "grad_norm": 3.90625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.1676,
       "step": 112500
     },
     {
       "epoch": 1.8279150623060552,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2222,
       "step": 112510
     },
     {
       "epoch": 1.8280775292034248,
-      "grad_norm": 2.875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.1934,
       "step": 112520
     },
     {
       "epoch": 1.8282399961007945,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2341,
       "step": 112530
     },
     {
       "epoch": 1.8284024629981641,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.2014,
       "step": 112540
     },
     {
       "epoch": 1.8285649298955338,
-      "grad_norm": 3.546875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2293,
       "step": 112550
     },
     {
       "epoch": 1.8287273967929034,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2066,
       "step": 112560
     },
     {
       "epoch": 1.828889863690273,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2217,
       "step": 112570
     },
     {
       "epoch": 1.8290523305876427,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2133,
       "step": 112580
     },
     {
       "epoch": 1.8292147974850126,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2537,
       "step": 112590
     },
     {
       "epoch": 1.8293772643823822,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2178,
       "step": 112600
     },
     {
       "epoch": 1.829539731279752,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2264,
       "step": 112610
     },
     {
       "epoch": 1.8297021981771215,
-      "grad_norm": 3.0,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3648,
+      "loss": 1.2227,
       "step": 112620
     },
     {
       "epoch": 1.8298646650744912,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.1768,
       "step": 112630
     },
     {
       "epoch": 1.8300271319718608,
-      "grad_norm": 6.25,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.195,
       "step": 112640
     },
     {
       "epoch": 1.8301895988692305,
-      "grad_norm": 3.953125,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.2936,
       "step": 112650
     },
     {
       "epoch": 1.8303520657666001,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3246,
+      "loss": 1.2432,
       "step": 112660
     },
     {
       "epoch": 1.8305145326639698,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.213,
       "step": 112670
     },
     {
       "epoch": 1.8306769995613394,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2228,
       "step": 112680
     },
     {
       "epoch": 1.830839466458709,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2327,
       "step": 112690
     },
     {
       "epoch": 1.8310019333560787,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.1778,
       "step": 112700
     },
     {
       "epoch": 1.8311644002534484,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2132,
       "step": 112710
     },
     {
       "epoch": 1.831326867150818,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2379,
       "step": 112720
     },
     {
       "epoch": 1.8314893340481877,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2536,
       "step": 112730
     },
     {
       "epoch": 1.8316518009455574,
-      "grad_norm": 5.3125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2452,
       "step": 112740
     },
     {
       "epoch": 1.831814267842927,
-      "grad_norm": 3.0,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.227,
       "step": 112750
     },
     {
       "epoch": 1.8319767347402967,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2494,
       "step": 112760
     },
     {
       "epoch": 1.8321392016376663,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.236,
       "step": 112770
     },
     {
       "epoch": 1.832301668535036,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2147,
       "step": 112780
     },
     {
       "epoch": 1.8324641354324056,
-      "grad_norm": 3.0,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.2072,
       "step": 112790
     },
     {
       "epoch": 1.8326266023297753,
-      "grad_norm": 2.25,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2288,
       "step": 112800
     },
     {
       "epoch": 1.832789069227145,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2351,
       "step": 112810
     },
     {
       "epoch": 1.8329515361245146,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.1767,
       "step": 112820
     },
     {
       "epoch": 1.8331140030218842,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3714,
+      "loss": 1.1918,
       "step": 112830
     },
     {
       "epoch": 1.8332764699192539,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2325,
       "step": 112840
     },
     {
       "epoch": 1.8334389368166235,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2055,
       "step": 112850
     },
     {
       "epoch": 1.8336014037139932,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2238,
       "step": 112860
     },
     {
       "epoch": 1.8337638706113628,
-      "grad_norm": 3.21875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.314,
+      "loss": 1.2706,
       "step": 112870
     },
     {
       "epoch": 1.8339263375087325,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2176,
       "step": 112880
     },
     {
       "epoch": 1.8340888044061021,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2198,
       "step": 112890
     },
     {
       "epoch": 1.8342512713034718,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.1817,
       "step": 112900
     },
     {
       "epoch": 1.8344137382008414,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2127,
       "step": 112910
     },
     {
       "epoch": 1.8345762050982113,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2184,
       "step": 112920
     },
     {
       "epoch": 1.834738671995581,
-      "grad_norm": 3.15625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2489,
       "step": 112930
     },
     {
       "epoch": 1.8349011388929506,
-      "grad_norm": 3.125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2222,
       "step": 112940
     },
     {
       "epoch": 1.8350636057903202,
-      "grad_norm": 3.25,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3624,
+      "loss": 1.2065,
       "step": 112950
     },
     {
       "epoch": 1.83522607268769,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2641,
       "step": 112960
     },
     {
       "epoch": 1.8353885395850595,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2635,
       "step": 112970
     },
     {
       "epoch": 1.8355510064824292,
-      "grad_norm": 2.75,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.279,
       "step": 112980
     },
     {
       "epoch": 1.8357134733797988,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.1595,
       "step": 112990
     },
     {
       "epoch": 1.8358759402771685,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.2228,
       "step": 113000
     },
     {
       "epoch": 1.8360384071745381,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2052,
       "step": 113010
     },
     {
       "epoch": 1.8362008740719078,
-      "grad_norm": 2.75,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3309,
+      "loss": 1.2,
       "step": 113020
     },
     {
       "epoch": 1.8363633409692777,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3283,
+      "loss": 1.2455,
       "step": 113030
     },
     {
       "epoch": 1.8365258078666473,
-      "grad_norm": 3.75,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3647,
+      "loss": 1.1992,
       "step": 113040
     },
     {
       "epoch": 1.836688274764017,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.2682,
       "step": 113050
     },
     {
       "epoch": 1.8368507416613866,
-      "grad_norm": 4.03125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2245,
       "step": 113060
     },
     {
       "epoch": 1.8370132085587563,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2249,
       "step": 113070
     },
     {
       "epoch": 1.837175675456126,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2338,
       "step": 113080
     },
     {
       "epoch": 1.8373381423534956,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2159,
       "step": 113090
     },
     {
       "epoch": 1.8375006092508652,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2313,
       "step": 113100
     },
     {
       "epoch": 1.8376630761482349,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2366,
       "step": 113110
     },
     {
       "epoch": 1.8378255430456045,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3641,
+      "loss": 1.2147,
       "step": 113120
     },
     {
       "epoch": 1.8379880099429742,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2392,
       "step": 113130
     },
     {
       "epoch": 1.8381504768403438,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2184,
       "step": 113140
     },
     {
       "epoch": 1.8383129437377135,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3754,
+      "loss": 1.2475,
       "step": 113150
     },
     {
       "epoch": 1.8384754106350831,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.2507,
       "step": 113160
     },
     {
       "epoch": 1.8386378775324528,
-      "grad_norm": 3.671875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.1943,
       "step": 113170
     },
     {
       "epoch": 1.8388003444298224,
-      "grad_norm": 1.6640625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2411,
       "step": 113180
     },
     {
       "epoch": 1.838962811327192,
-      "grad_norm": 1.875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.1774,
       "step": 113190
     },
     {
       "epoch": 1.8391252782245617,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.1866,
       "step": 113200
     },
     {
       "epoch": 1.8392877451219314,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3564,
+      "loss": 1.2071,
       "step": 113210
     },
     {
       "epoch": 1.839450212019301,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3493,
+      "loss": 1.1936,
       "step": 113220
     },
     {
       "epoch": 1.8396126789166707,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2098,
       "step": 113230
     },
     {
       "epoch": 1.8397751458140403,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.2831,
       "step": 113240
     },
     {
       "epoch": 1.83993761271141,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2376,
       "step": 113250
     },
     {
       "epoch": 1.8401000796087796,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2429,
       "step": 113260
     },
     {
       "epoch": 1.8402625465061493,
-      "grad_norm": 4.375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.1799,
       "step": 113270
     },
     {
       "epoch": 1.840425013403519,
-      "grad_norm": 3.484375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2041,
       "step": 113280
     },
     {
       "epoch": 1.8405874803008886,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2073,
       "step": 113290
     },
     {
       "epoch": 1.8407499471982582,
-      "grad_norm": 1.8984375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2198,
       "step": 113300
     },
     {
       "epoch": 1.840912414095628,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2612,
       "step": 113310
     },
     {
       "epoch": 1.8410748809929975,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2139,
       "step": 113320
     },
     {
       "epoch": 1.8412373478903672,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2262,
       "step": 113330
     },
     {
       "epoch": 1.8413998147877368,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2463,
       "step": 113340
     },
     {
       "epoch": 1.8415622816851065,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2074,
       "step": 113350
     },
     {
       "epoch": 1.8417247485824764,
-      "grad_norm": 2.171875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2688,
       "step": 113360
     },
     {
       "epoch": 1.841887215479846,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.175,
       "step": 113370
     },
     {
       "epoch": 1.8420496823772157,
-      "grad_norm": 3.140625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2429,
       "step": 113380
     },
     {
       "epoch": 1.8422121492745853,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2028,
       "step": 113390
     },
     {
       "epoch": 1.842374616171955,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2801,
       "step": 113400
     },
     {
       "epoch": 1.8425370830693246,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2681,
       "step": 113410
     },
     {
       "epoch": 1.8426995499666943,
-      "grad_norm": 3.375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.2178,
       "step": 113420
     },
     {
       "epoch": 1.842862016864064,
-      "grad_norm": 2.421875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2016,
       "step": 113430
     },
     {
       "epoch": 1.8430244837614336,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2246,
       "step": 113440
     },
     {
       "epoch": 1.8431869506588032,
-      "grad_norm": 3.515625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3248,
+      "loss": 1.2196,
       "step": 113450
     },
     {
       "epoch": 1.8433494175561729,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2327,
       "step": 113460
     },
     {
       "epoch": 1.8435118844535427,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2386,
       "step": 113470
     },
     {
       "epoch": 1.8436743513509124,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.1943,
       "step": 113480
     },
     {
       "epoch": 1.843836818248282,
-      "grad_norm": 4.5,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2241,
       "step": 113490
     },
     {
       "epoch": 1.8439992851456517,
-      "grad_norm": 3.75,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3266,
+      "loss": 1.2054,
       "step": 113500
     },
     {
       "epoch": 1.8441617520430214,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2088,
       "step": 113510
     },
     {
       "epoch": 1.844324218940391,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2228,
       "step": 113520
     },
     {
       "epoch": 1.8444866858377607,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2105,
       "step": 113530
     },
     {
       "epoch": 1.8446491527351303,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2133,
       "step": 113540
     },
     {
       "epoch": 1.8448116196325,
-      "grad_norm": 3.25,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2236,
       "step": 113550
     },
     {
       "epoch": 1.8449740865298696,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2042,
       "step": 113560
     },
     {
       "epoch": 1.8451365534272393,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2273,
       "step": 113570
     },
     {
       "epoch": 1.845299020324609,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2261,
       "step": 113580
     },
     {
       "epoch": 1.8454614872219786,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2374,
       "step": 113590
     },
     {
       "epoch": 1.8456239541193482,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.1953,
       "step": 113600
     },
     {
       "epoch": 1.8457864210167179,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2482,
       "step": 113610
     },
     {
       "epoch": 1.8459488879140875,
-      "grad_norm": 3.25,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1766,
       "step": 113620
     },
     {
       "epoch": 1.8461113548114572,
-      "grad_norm": 3.0,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.2044,
       "step": 113630
     },
     {
       "epoch": 1.8462738217088268,
-      "grad_norm": 3.84375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2135,
       "step": 113640
     },
     {
       "epoch": 1.8464362886061965,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.2164,
       "step": 113650
     },
     {
       "epoch": 1.8465987555035661,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2357,
       "step": 113660
     },
     {
       "epoch": 1.8467612224009358,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.222,
       "step": 113670
     },
     {
       "epoch": 1.8469236892983054,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3525,
+      "loss": 1.2279,
       "step": 113680
     },
     {
       "epoch": 1.847086156195675,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3516,
+      "loss": 1.2042,
       "step": 113690
     },
     {
       "epoch": 1.8472486230930447,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.363,
+      "loss": 1.231,
       "step": 113700
     },
     {
       "epoch": 1.8474110899904144,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2154,
       "step": 113710
     },
     {
       "epoch": 1.847573556887784,
-      "grad_norm": 2.609375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3165,
+      "loss": 1.2602,
       "step": 113720
     },
     {
       "epoch": 1.8477360237851537,
-      "grad_norm": 5.375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.1959,
       "step": 113730
     },
     {
       "epoch": 1.8478984906825233,
-      "grad_norm": 3.0,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.2172,
       "step": 113740
     },
     {
       "epoch": 1.848060957579893,
-      "grad_norm": 2.328125,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2261,
       "step": 113750
     },
     {
       "epoch": 1.8482234244772626,
-      "grad_norm": 3.265625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2426,
       "step": 113760
     },
     {
       "epoch": 1.8483858913746323,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3264,
+      "loss": 1.2027,
       "step": 113770
     },
     {
       "epoch": 1.848548358272002,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.1919,
       "step": 113780
     },
     {
       "epoch": 1.8487108251693716,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.25,
       "step": 113790
     },
     {
       "epoch": 1.8488732920667414,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.1899,
       "step": 113800
     },
     {
       "epoch": 1.849035758964111,
-      "grad_norm": 2.625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.24,
       "step": 113810
     },
     {
       "epoch": 1.8491982258614807,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2299,
       "step": 113820
     },
     {
       "epoch": 1.8493606927588504,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2254,
       "step": 113830
     },
     {
       "epoch": 1.84952315965622,
-      "grad_norm": 4.34375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.224,
       "step": 113840
     },
     {
       "epoch": 1.8496856265535897,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.2122,
       "step": 113850
     },
     {
       "epoch": 1.8498480934509594,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1997,
       "step": 113860
     },
     {
       "epoch": 1.850010560348329,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.1798,
       "step": 113870
     },
     {
       "epoch": 1.8501730272456987,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2161,
       "step": 113880
     },
     {
       "epoch": 1.8503354941430683,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.2514,
       "step": 113890
     },
     {
       "epoch": 1.850497961040438,
-      "grad_norm": 2.234375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.323,
+      "loss": 1.2171,
       "step": 113900
     },
     {
       "epoch": 1.8506604279378078,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3094,
+      "loss": 1.2209,
       "step": 113910
     },
     {
       "epoch": 1.8508228948351775,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2177,
       "step": 113920
     },
     {
       "epoch": 1.8509853617325471,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2238,
       "step": 113930
     },
     {
       "epoch": 1.8511478286299168,
-      "grad_norm": 4.03125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2061,
       "step": 113940
     },
     {
       "epoch": 1.8513102955272864,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2006,
       "step": 113950
     },
     {
       "epoch": 1.851472762424656,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2233,
       "step": 113960
     },
     {
       "epoch": 1.8516352293220257,
-      "grad_norm": 2.015625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2206,
       "step": 113970
     },
     {
       "epoch": 1.8517976962193954,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.1855,
       "step": 113980
     },
     {
       "epoch": 1.851960163116765,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.2552,
       "step": 113990
     },
     {
       "epoch": 1.8521226300141347,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3634,
+      "loss": 1.1953,
       "step": 114000
     },
     {
       "epoch": 1.8522850969115043,
-      "grad_norm": 2.515625,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2559,
       "step": 114010
     },
     {
       "epoch": 1.852447563808874,
-      "grad_norm": 1.9375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.188,
       "step": 114020
     },
     {
       "epoch": 1.8526100307062436,
-      "grad_norm": 2.109375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3322,
+      "loss": 1.2165,
       "step": 114030
     },
     {
       "epoch": 1.8527724976036133,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2145,
       "step": 114040
     },
     {
       "epoch": 1.852934964500983,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.2514,
       "step": 114050
     },
     {
       "epoch": 1.8530974313983526,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2236,
       "step": 114060
     },
     {
       "epoch": 1.8532598982957222,
-      "grad_norm": 2.234375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2214,
       "step": 114070
     },
     {
       "epoch": 1.853422365193092,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3319,
+      "loss": 1.2354,
       "step": 114080
     },
     {
       "epoch": 1.8535848320904615,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.2294,
       "step": 114090
     },
     {
       "epoch": 1.8537472989878312,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2032,
       "step": 114100
     },
     {
       "epoch": 1.8539097658852008,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3561,
+      "loss": 1.2404,
       "step": 114110
     },
     {
       "epoch": 1.8540722327825705,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2129,
       "step": 114120
     },
     {
       "epoch": 1.8542346996799401,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.196,
       "step": 114130
     },
     {
       "epoch": 1.8543971665773098,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2459,
       "step": 114140
     },
     {
       "epoch": 1.8545596334746794,
-      "grad_norm": 2.296875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2427,
       "step": 114150
     },
     {
       "epoch": 1.854722100372049,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2583,
       "step": 114160
     },
     {
       "epoch": 1.8548845672694187,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.1905,
       "step": 114170
     },
     {
       "epoch": 1.8550470341667884,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.239,
       "step": 114180
     },
     {
       "epoch": 1.855209501064158,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3541,
+      "loss": 1.229,
       "step": 114190
     },
     {
       "epoch": 1.8553719679615277,
-      "grad_norm": 3.59375,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2279,
       "step": 114200
     },
     {
       "epoch": 1.8555344348588974,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2254,
       "step": 114210
     },
     {
       "epoch": 1.855696901756267,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3768,
+      "loss": 1.2157,
       "step": 114220
     },
     {
       "epoch": 1.8558593686536369,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.2292,
       "step": 114230
     },
     {
       "epoch": 1.8560218355510065,
-      "grad_norm": 2.25,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2288,
       "step": 114240
     },
     {
       "epoch": 1.8561843024483762,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.2176,
       "step": 114250
     },
     {
       "epoch": 1.8563467693457458,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2025,
       "step": 114260
     },
     {
       "epoch": 1.8565092362431155,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.19,
       "step": 114270
     },
     {
       "epoch": 1.8566717031404851,
-      "grad_norm": 2.4375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2292,
       "step": 114280
     },
     {
       "epoch": 1.8568341700378548,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2294,
       "step": 114290
     },
     {
       "epoch": 1.8569966369352244,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2338,
       "step": 114300
     },
     {
       "epoch": 1.857159103832594,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3138,
+      "loss": 1.2221,
       "step": 114310
     },
     {
       "epoch": 1.8573215707299637,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3405,
+      "loss": 1.2103,
       "step": 114320
     },
     {
       "epoch": 1.8574840376273334,
-      "grad_norm": 3.25,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2603,
       "step": 114330
     },
     {
       "epoch": 1.8576465045247033,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2055,
       "step": 114340
     },
     {
       "epoch": 1.857808971422073,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.1976,
       "step": 114350
     },
     {
       "epoch": 1.8579714383194426,
-      "grad_norm": 4.375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2624,
       "step": 114360
     },
     {
       "epoch": 1.8581339052168122,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2359,
       "step": 114370
     },
     {
       "epoch": 1.8582963721141819,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2445,
       "step": 114380
     },
     {
       "epoch": 1.8584588390115515,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3614,
+      "loss": 1.2156,
       "step": 114390
     },
     {
       "epoch": 1.8586213059089212,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3626,
+      "loss": 1.2256,
       "step": 114400
     },
     {
       "epoch": 1.8587837728062908,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3487,
+      "loss": 1.2413,
       "step": 114410
     },
     {
       "epoch": 1.8589462397036605,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.1876,
       "step": 114420
     },
     {
       "epoch": 1.8591087066010301,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3171,
+      "loss": 1.2488,
       "step": 114430
     },
     {
       "epoch": 1.8592711734983998,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.245,
       "step": 114440
     },
     {
       "epoch": 1.8594336403957694,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3414,
+      "loss": 1.2335,
       "step": 114450
     },
     {
       "epoch": 1.859596107293139,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2415,
       "step": 114460
     },
     {
       "epoch": 1.8597585741905087,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3166,
+      "loss": 1.1994,
       "step": 114470
     },
     {
       "epoch": 1.8599210410878784,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3419,
+      "loss": 1.2094,
       "step": 114480
     },
     {
       "epoch": 1.860083507985248,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2076,
       "step": 114490
     },
     {
       "epoch": 1.8602459748826177,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2196,
       "step": 114500
     },
     {
       "epoch": 1.8604084417799873,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2469,
       "step": 114510
     },
     {
       "epoch": 1.860570908677357,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.23,
       "step": 114520
     },
     {
       "epoch": 1.8607333755747266,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3617,
+      "loss": 1.2408,
       "step": 114530
     },
     {
       "epoch": 1.8608958424720963,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3433,
+      "loss": 1.1939,
       "step": 114540
     },
     {
       "epoch": 1.861058309369466,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.1894,
       "step": 114550
     },
     {
       "epoch": 1.8612207762668356,
-      "grad_norm": 3.25,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3295,
+      "loss": 1.1981,
       "step": 114560
     },
     {
       "epoch": 1.8613832431642052,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.1974,
       "step": 114570
     },
     {
       "epoch": 1.8615457100615749,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.2502,
       "step": 114580
     },
     {
       "epoch": 1.8617081769589445,
-      "grad_norm": 3.625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2157,
       "step": 114590
     },
     {
       "epoch": 1.8618706438563142,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2138,
       "step": 114600
     },
     {
       "epoch": 1.8620331107536838,
-      "grad_norm": 3.171875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3612,
+      "loss": 1.1645,
       "step": 114610
     },
     {
       "epoch": 1.8621955776510535,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.2383,
       "step": 114620
     },
     {
       "epoch": 1.8623580445484231,
-      "grad_norm": 3.90625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2742,
       "step": 114630
     },
     {
       "epoch": 1.8625205114457928,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.1853,
       "step": 114640
     },
     {
       "epoch": 1.8626829783431624,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2004,
       "step": 114650
     },
     {
       "epoch": 1.862845445240532,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.218,
       "step": 114660
     },
     {
       "epoch": 1.863007912137902,
-      "grad_norm": 3.0,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2196,
       "step": 114670
     },
     {
       "epoch": 1.8631703790352716,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.217,
       "step": 114680
     },
     {
       "epoch": 1.8633328459326413,
-      "grad_norm": 4.375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.231,
       "step": 114690
     },
     {
       "epoch": 1.863495312830011,
-      "grad_norm": 3.125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.1526,
       "step": 114700
     },
     {
       "epoch": 1.8636577797273806,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.1999,
       "step": 114710
     },
     {
       "epoch": 1.8638202466247502,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2179,
       "step": 114720
     },
     {
       "epoch": 1.8639827135221199,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3536,
+      "loss": 1.186,
       "step": 114730
     },
     {
       "epoch": 1.8641451804194895,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2454,
       "step": 114740
     },
     {
       "epoch": 1.8643076473168592,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.309,
+      "loss": 1.1909,
       "step": 114750
     },
     {
       "epoch": 1.8644701142142288,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.229,
       "step": 114760
     },
     {
       "epoch": 1.8646325811115985,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.2167,
       "step": 114770
     },
     {
       "epoch": 1.8647950480089683,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2075,
       "step": 114780
     },
     {
       "epoch": 1.864957514906338,
-      "grad_norm": 3.984375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.2378,
       "step": 114790
     },
     {
       "epoch": 1.8651199818037076,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2055,
       "step": 114800
     },
     {
       "epoch": 1.8652824487010773,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.356,
+      "loss": 1.2114,
       "step": 114810
     },
     {
       "epoch": 1.865444915598447,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2226,
       "step": 114820
     },
     {
       "epoch": 1.8656073824958166,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3473,
+      "loss": 1.2101,
       "step": 114830
     },
     {
       "epoch": 1.8657698493931862,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3634,
+      "loss": 1.1878,
       "step": 114840
     },
     {
       "epoch": 1.865932316290556,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2118,
       "step": 114850
     },
     {
       "epoch": 1.8660947831879255,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.1806,
       "step": 114860
     },
     {
       "epoch": 1.8662572500852952,
-      "grad_norm": 3.375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.236,
       "step": 114870
     },
     {
       "epoch": 1.8664197169826648,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3601,
+      "loss": 1.1763,
       "step": 114880
     },
     {
       "epoch": 1.8665821838800345,
-      "grad_norm": 3.25,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2036,
       "step": 114890
     },
     {
       "epoch": 1.8667446507774041,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.2136,
       "step": 114900
     },
     {
       "epoch": 1.8669071176747738,
-      "grad_norm": 4.59375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3446,
+      "loss": 1.215,
       "step": 114910
     },
     {
       "epoch": 1.8670695845721434,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.248,
       "step": 114920
     },
     {
       "epoch": 1.867232051469513,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.2218,
       "step": 114930
     },
     {
       "epoch": 1.8673945183668827,
-      "grad_norm": 3.4375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.2584,
       "step": 114940
     },
     {
       "epoch": 1.8675569852642524,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3689,
+      "loss": 1.2202,
       "step": 114950
     },
     {
       "epoch": 1.867719452161622,
-      "grad_norm": 4.53125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3318,
+      "loss": 1.2,
       "step": 114960
     },
     {
       "epoch": 1.8678819190589917,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3683,
+      "loss": 1.2223,
       "step": 114970
     },
     {
       "epoch": 1.8680443859563614,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.1927,
       "step": 114980
     },
     {
       "epoch": 1.868206852853731,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3553,
+      "loss": 1.1983,
       "step": 114990
     },
     {
       "epoch": 1.8683693197511007,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3224,
+      "loss": 1.2455,
       "step": 115000
     },
     {
       "epoch": 1.8685317866484703,
-      "grad_norm": 2.359375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3301,
+      "loss": 1.2073,
       "step": 115010
     },
     {
       "epoch": 1.86869425354584,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2349,
       "step": 115020
     },
     {
       "epoch": 1.8688567204432096,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2846,
       "step": 115030
     },
     {
       "epoch": 1.8690191873405793,
-      "grad_norm": 2.4375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.314,
+      "loss": 1.1762,
       "step": 115040
     },
     {
       "epoch": 1.869181654237949,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3277,
+      "loss": 1.1787,
       "step": 115050
     },
     {
       "epoch": 1.8693441211353186,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2187,
       "step": 115060
     },
     {
       "epoch": 1.8695065880326882,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3537,
+      "loss": 1.1896,
       "step": 115070
     },
     {
       "epoch": 1.8696690549300579,
-      "grad_norm": 2.484375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2272,
       "step": 115080
     },
     {
       "epoch": 1.8698315218274275,
-      "grad_norm": 3.015625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2078,
       "step": 115090
     },
     {
       "epoch": 1.8699939887247972,
-      "grad_norm": 3.5,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.2566,
       "step": 115100
     },
     {
       "epoch": 1.870156455622167,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.1795,
       "step": 115110
     },
     {
       "epoch": 1.8703189225195367,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.212,
       "step": 115120
     },
     {
       "epoch": 1.8704813894169063,
-      "grad_norm": 3.953125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2154,
       "step": 115130
     },
     {
       "epoch": 1.870643856314276,
-      "grad_norm": 3.09375,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.234,
       "step": 115140
     },
     {
       "epoch": 1.8708063232116456,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3655,
+      "loss": 1.1688,
       "step": 115150
     },
     {
       "epoch": 1.8709687901090153,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2125,
       "step": 115160
     },
     {
       "epoch": 1.871131257006385,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3328,
+      "loss": 1.2296,
       "step": 115170
     },
     {
       "epoch": 1.8712937239037546,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2199,
       "step": 115180
     },
     {
       "epoch": 1.8714561908011242,
-      "grad_norm": 3.125,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3425,
+      "loss": 1.2303,
       "step": 115190
     },
     {
       "epoch": 1.871618657698494,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2172,
       "step": 115200
     },
     {
       "epoch": 1.8717811245958635,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2318,
       "step": 115210
     },
     {
       "epoch": 1.8719435914932334,
-      "grad_norm": 2.625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2058,
       "step": 115220
     },
     {
       "epoch": 1.872106058390603,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2349,
       "step": 115230
     },
     {
       "epoch": 1.8722685252879727,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3562,
+      "loss": 1.2047,
       "step": 115240
     },
     {
       "epoch": 1.8724309921853424,
-      "grad_norm": 2.875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2025,
       "step": 115250
     },
     {
       "epoch": 1.872593459082712,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.2444,
       "step": 115260
     },
     {
       "epoch": 1.8727559259800817,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.2264,
       "step": 115270
     },
     {
       "epoch": 1.8729183928774513,
-      "grad_norm": 4.46875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2242,
       "step": 115280
     },
     {
       "epoch": 1.873080859774821,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.245,
       "step": 115290
     },
     {
       "epoch": 1.8732433266721906,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2194,
       "step": 115300
     },
     {
       "epoch": 1.8734057935695603,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2147,
       "step": 115310
     },
     {
       "epoch": 1.87356826046693,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2729,
       "step": 115320
     },
     {
       "epoch": 1.8737307273642996,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.244,
       "step": 115330
     },
     {
       "epoch": 1.8738931942616692,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.1891,
       "step": 115340
     },
     {
       "epoch": 1.8740556611590389,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.194,
       "step": 115350
     },
     {
       "epoch": 1.8742181280564085,
-      "grad_norm": 2.984375,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3498,
+      "loss": 1.2306,
       "step": 115360
     },
     {
       "epoch": 1.8743805949537782,
-      "grad_norm": 3.59375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.2301,
       "step": 115370
     },
     {
       "epoch": 1.8745430618511478,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2245,
       "step": 115380
     },
     {
       "epoch": 1.8747055287485175,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 1.2731,
       "step": 115390
     },
     {
       "epoch": 1.8748679956458871,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3644,
+      "loss": 1.2346,
       "step": 115400
     },
     {
       "epoch": 1.8750304625432568,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2196,
       "step": 115410
     },
     {
       "epoch": 1.8751929294406264,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2076,
       "step": 115420
     },
     {
       "epoch": 1.875355396337996,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.261,
       "step": 115430
     },
     {
       "epoch": 1.8755178632353657,
-      "grad_norm": 2.46875,
+      "grad_norm": 5.5,
       "learning_rate": 5e-05,
-      "loss": 0.3289,
+      "loss": 1.2253,
       "step": 115440
     },
     {
       "epoch": 1.8756803301327354,
-      "grad_norm": 2.75,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.373,
+      "loss": 1.272,
       "step": 115450
     },
     {
       "epoch": 1.875842797030105,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2219,
       "step": 115460
     },
     {
       "epoch": 1.8760052639274747,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2063,
       "step": 115470
     },
     {
       "epoch": 1.8761677308248443,
-      "grad_norm": 4.4375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2379,
       "step": 115480
     },
     {
       "epoch": 1.876330197722214,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3457,
+      "loss": 1.2222,
       "step": 115490
     },
     {
       "epoch": 1.8764926646195836,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3449,
+      "loss": 1.2547,
       "step": 115500
     },
     {
       "epoch": 1.8766551315169533,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2495,
       "step": 115510
     },
     {
       "epoch": 1.876817598414323,
-      "grad_norm": 3.0,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2247,
       "step": 115520
     },
     {
       "epoch": 1.8769800653116926,
-      "grad_norm": 2.625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3352,
+      "loss": 1.2322,
       "step": 115530
     },
     {
       "epoch": 1.8771425322090622,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.2214,
       "step": 115540
     },
     {
       "epoch": 1.8773049991064321,
-      "grad_norm": 3.375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 1.2533,
       "step": 115550
     },
     {
       "epoch": 1.8774674660038018,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3279,
+      "loss": 1.191,
       "step": 115560
     },
     {
       "epoch": 1.8776299329011714,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2002,
       "step": 115570
     },
     {
       "epoch": 1.877792399798541,
-      "grad_norm": 1.78125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3613,
+      "loss": 1.201,
       "step": 115580
     },
     {
       "epoch": 1.8779548666959107,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1892,
       "step": 115590
     },
     {
       "epoch": 1.8781173335932804,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 1.2076,
       "step": 115600
     },
     {
       "epoch": 1.87827980049065,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.2271,
       "step": 115610
     },
     {
       "epoch": 1.8784422673880197,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.1865,
       "step": 115620
     },
     {
       "epoch": 1.8786047342853893,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2069,
       "step": 115630
     },
     {
       "epoch": 1.878767201182759,
-      "grad_norm": 3.25,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2156,
       "step": 115640
     },
     {
       "epoch": 1.8789296680801286,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3615,
+      "loss": 1.1675,
       "step": 115650
     },
     {
       "epoch": 1.8790921349774985,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.2617,
       "step": 115660
     },
     {
       "epoch": 1.8792546018748681,
-      "grad_norm": 2.625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.2046,
       "step": 115670
     },
     {
       "epoch": 1.8794170687722378,
-      "grad_norm": 2.78125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2356,
       "step": 115680
     },
     {
       "epoch": 1.8795795356696074,
-      "grad_norm": 2.5,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.1937,
       "step": 115690
     },
     {
       "epoch": 1.879742002566977,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.2289,
       "step": 115700
     },
     {
       "epoch": 1.8799044694643468,
-      "grad_norm": 3.25,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3245,
+      "loss": 1.2104,
       "step": 115710
     },
     {
       "epoch": 1.8800669363617164,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.2381,
       "step": 115720
     },
     {
       "epoch": 1.880229403259086,
-      "grad_norm": 2.65625,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2473,
       "step": 115730
     },
     {
       "epoch": 1.8803918701564557,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3499,
+      "loss": 1.2174,
       "step": 115740
     },
     {
       "epoch": 1.8805543370538254,
-      "grad_norm": 3.375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.2547,
       "step": 115750
     },
     {
       "epoch": 1.880716803951195,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2452,
       "step": 115760
     },
     {
       "epoch": 1.8808792708485647,
-      "grad_norm": 3.65625,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1968,
       "step": 115770
     },
     {
       "epoch": 1.8810417377459343,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1999,
       "step": 115780
     },
     {
       "epoch": 1.881204204643304,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.229,
       "step": 115790
     },
     {
       "epoch": 1.8813666715406736,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2334,
       "step": 115800
     },
     {
       "epoch": 1.8815291384380433,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3285,
+      "loss": 1.2068,
       "step": 115810
     },
     {
       "epoch": 1.881691605335413,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3233,
+      "loss": 1.2289,
       "step": 115820
     },
     {
       "epoch": 1.8818540722327826,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.1792,
       "step": 115830
     },
     {
       "epoch": 1.8820165391301522,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2342,
       "step": 115840
     },
     {
       "epoch": 1.8821790060275219,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.2176,
       "step": 115850
     },
     {
       "epoch": 1.8823414729248915,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2384,
       "step": 115860
     },
     {
       "epoch": 1.8825039398222612,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2294,
       "step": 115870
     },
     {
       "epoch": 1.8826664067196308,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.2465,
       "step": 115880
     },
     {
       "epoch": 1.8828288736170005,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2144,
       "step": 115890
     },
     {
       "epoch": 1.8829913405143701,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3183,
+      "loss": 1.2392,
       "step": 115900
     },
     {
       "epoch": 1.8831538074117398,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2518,
       "step": 115910
     },
     {
       "epoch": 1.8833162743091094,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2171,
       "step": 115920
     },
     {
       "epoch": 1.883478741206479,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2485,
       "step": 115930
     },
     {
       "epoch": 1.8836412081038487,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.189,
       "step": 115940
     },
     {
       "epoch": 1.8838036750012184,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2194,
       "step": 115950
     },
     {
       "epoch": 1.883966141898588,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.2136,
       "step": 115960
     },
     {
       "epoch": 1.8841286087959577,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2352,
       "step": 115970
     },
     {
       "epoch": 1.8842910756933273,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2221,
       "step": 115980
     },
     {
       "epoch": 1.8844535425906972,
-      "grad_norm": 4.03125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2065,
       "step": 115990
     },
     {
       "epoch": 1.8846160094880668,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2469,
       "step": 116000
     },
     {
       "epoch": 1.8847784763854365,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2157,
       "step": 116010
     },
     {
       "epoch": 1.8849409432828061,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3338,
+      "loss": 1.2057,
       "step": 116020
     },
     {
       "epoch": 1.8851034101801758,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3268,
+      "loss": 1.174,
       "step": 116030
     },
     {
       "epoch": 1.8852658770775454,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.315,
+      "loss": 1.2333,
       "step": 116040
     },
     {
       "epoch": 1.885428343974915,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.2441,
       "step": 116050
     },
     {
       "epoch": 1.8855908108722848,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3592,
+      "loss": 1.179,
       "step": 116060
     },
     {
       "epoch": 1.8857532777696544,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2204,
       "step": 116070
     },
     {
       "epoch": 1.885915744667024,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2327,
       "step": 116080
     },
     {
       "epoch": 1.8860782115643937,
-      "grad_norm": 2.375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2079,
       "step": 116090
     },
     {
       "epoch": 1.8862406784617636,
-      "grad_norm": 3.515625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2416,
       "step": 116100
     },
     {
       "epoch": 1.8864031453591332,
-      "grad_norm": 4.25,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3495,
+      "loss": 1.1641,
       "step": 116110
     },
     {
       "epoch": 1.8865656122565029,
-      "grad_norm": 4.0,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3264,
+      "loss": 1.2352,
       "step": 116120
     },
     {
       "epoch": 1.8867280791538725,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2077,
       "step": 116130
     },
     {
       "epoch": 1.8868905460512422,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3066,
+      "loss": 1.1916,
       "step": 116140
     },
     {
       "epoch": 1.8870530129486118,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.199,
       "step": 116150
     },
     {
       "epoch": 1.8872154798459815,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.1678,
       "step": 116160
     },
     {
       "epoch": 1.8873779467433511,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.1964,
       "step": 116170
     },
     {
       "epoch": 1.8875404136407208,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.23,
       "step": 116180
     },
     {
       "epoch": 1.8877028805380904,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2463,
       "step": 116190
     },
     {
       "epoch": 1.88786534743546,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.1968,
       "step": 116200
     },
     {
       "epoch": 1.8880278143328297,
-      "grad_norm": 2.21875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2165,
       "step": 116210
     },
     {
       "epoch": 1.8881902812301994,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.211,
       "step": 116220
     },
     {
       "epoch": 1.888352748127569,
-      "grad_norm": 2.375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3324,
+      "loss": 1.2431,
       "step": 116230
     },
     {
       "epoch": 1.8885152150249387,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3187,
+      "loss": 1.2168,
       "step": 116240
     },
     {
       "epoch": 1.8886776819223083,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2162,
       "step": 116250
     },
     {
       "epoch": 1.888840148819678,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2162,
       "step": 116260
     },
     {
       "epoch": 1.8890026157170476,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3251,
+      "loss": 1.2248,
       "step": 116270
     },
     {
       "epoch": 1.8891650826144173,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.1764,
       "step": 116280
     },
     {
       "epoch": 1.889327549511787,
-      "grad_norm": 2.625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2105,
       "step": 116290
     },
     {
       "epoch": 1.8894900164091566,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2276,
       "step": 116300
     },
     {
       "epoch": 1.8896524833065262,
-      "grad_norm": 3.71875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2,
       "step": 116310
     },
     {
       "epoch": 1.889814950203896,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3167,
+      "loss": 1.209,
       "step": 116320
     },
     {
       "epoch": 1.8899774171012655,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2288,
       "step": 116330
     },
     {
       "epoch": 1.8901398839986352,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3209,
+      "loss": 1.2176,
       "step": 116340
     },
     {
       "epoch": 1.8903023508960048,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.1696,
       "step": 116350
     },
     {
       "epoch": 1.8904648177933745,
-      "grad_norm": 2.546875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3335,
+      "loss": 1.2019,
       "step": 116360
     },
     {
       "epoch": 1.8906272846907441,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3694,
+      "loss": 1.2062,
       "step": 116370
     },
     {
       "epoch": 1.8907897515881138,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1792,
       "step": 116380
     },
     {
       "epoch": 1.8909522184854834,
-      "grad_norm": 2.15625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2187,
       "step": 116390
     },
     {
       "epoch": 1.891114685382853,
-      "grad_norm": 1.8671875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2718,
       "step": 116400
     },
     {
       "epoch": 1.8912771522802228,
-      "grad_norm": 2.375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2146,
       "step": 116410
     },
     {
       "epoch": 1.8914396191775924,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2216,
       "step": 116420
     },
     {
       "epoch": 1.8916020860749623,
-      "grad_norm": 3.71875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.1882,
       "step": 116430
     },
     {
       "epoch": 1.891764552972332,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2415,
       "step": 116440
     },
     {
       "epoch": 1.8919270198697016,
-      "grad_norm": 3.421875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3503,
+      "loss": 1.1858,
       "step": 116450
     },
     {
       "epoch": 1.8920894867670712,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3461,
+      "loss": 1.2545,
       "step": 116460
     },
     {
       "epoch": 1.8922519536644409,
-      "grad_norm": 2.0625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.1985,
       "step": 116470
     },
     {
       "epoch": 1.8924144205618105,
-      "grad_norm": 3.0,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3359,
+      "loss": 1.2248,
       "step": 116480
     },
     {
       "epoch": 1.8925768874591802,
-      "grad_norm": 3.984375,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3642,
+      "loss": 1.2268,
       "step": 116490
     },
     {
       "epoch": 1.8927393543565498,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3652,
+      "loss": 1.2248,
       "step": 116500
     },
     {
       "epoch": 1.8929018212539195,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3155,
+      "loss": 1.2049,
       "step": 116510
     },
     {
       "epoch": 1.8930642881512891,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2219,
       "step": 116520
     },
     {
       "epoch": 1.8932267550486588,
-      "grad_norm": 4.28125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2395,
       "step": 116530
     },
     {
       "epoch": 1.8933892219460287,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3653,
+      "loss": 1.2419,
       "step": 116540
     },
     {
       "epoch": 1.8935516888433983,
-      "grad_norm": 3.671875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.1923,
       "step": 116550
     },
     {
       "epoch": 1.893714155740768,
-      "grad_norm": 2.59375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2451,
       "step": 116560
     },
     {
       "epoch": 1.8938766226381376,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2052,
       "step": 116570
     },
     {
       "epoch": 1.8940390895355073,
-      "grad_norm": 4.25,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3605,
+      "loss": 1.2108,
       "step": 116580
     },
     {
       "epoch": 1.894201556432877,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2191,
       "step": 116590
     },
     {
       "epoch": 1.8943640233302466,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.197,
       "step": 116600
     },
     {
       "epoch": 1.8945264902276162,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2637,
       "step": 116610
     },
     {
       "epoch": 1.8946889571249859,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3558,
+      "loss": 1.1971,
       "step": 116620
     },
     {
       "epoch": 1.8948514240223555,
-      "grad_norm": 2.78125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.2247,
       "step": 116630
     },
     {
       "epoch": 1.8950138909197252,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3527,
+      "loss": 1.2252,
       "step": 116640
     },
     {
       "epoch": 1.8951763578170948,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3217,
+      "loss": 1.1762,
       "step": 116650
     },
     {
       "epoch": 1.8953388247144645,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2454,
       "step": 116660
     },
     {
       "epoch": 1.8955012916118341,
-      "grad_norm": 2.796875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.2657,
       "step": 116670
     },
     {
       "epoch": 1.8956637585092038,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.2017,
       "step": 116680
     },
     {
       "epoch": 1.8958262254065734,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2445,
       "step": 116690
     },
     {
       "epoch": 1.895988692303943,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3491,
+      "loss": 1.2442,
       "step": 116700
     },
     {
       "epoch": 1.8961511592013127,
-      "grad_norm": 2.5,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2172,
       "step": 116710
     },
     {
       "epoch": 1.8963136260986824,
-      "grad_norm": 3.265625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2379,
       "step": 116720
     },
     {
       "epoch": 1.896476092996052,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2449,
       "step": 116730
     },
     {
       "epoch": 1.8966385598934217,
-      "grad_norm": 3.75,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.205,
       "step": 116740
     },
     {
       "epoch": 1.8968010267907913,
-      "grad_norm": 4.1875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.191,
       "step": 116750
     },
     {
       "epoch": 1.896963493688161,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2489,
       "step": 116760
     },
     {
       "epoch": 1.8971259605855306,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2195,
       "step": 116770
     },
     {
       "epoch": 1.8972884274829003,
-      "grad_norm": 3.125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.2676,
       "step": 116780
     },
     {
       "epoch": 1.89745089438027,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2067,
       "step": 116790
     },
     {
       "epoch": 1.8976133612776396,
-      "grad_norm": 2.546875,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3549,
+      "loss": 1.2001,
       "step": 116800
     },
     {
       "epoch": 1.8977758281750092,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2488,
       "step": 116810
     },
     {
       "epoch": 1.8979382950723789,
-      "grad_norm": 3.125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3447,
+      "loss": 1.2006,
       "step": 116820
     },
     {
       "epoch": 1.8981007619697485,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.2071,
       "step": 116830
     },
     {
       "epoch": 1.8982632288671182,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3228,
+      "loss": 1.2112,
       "step": 116840
     },
     {
       "epoch": 1.8984256957644878,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3538,
+      "loss": 1.2021,
       "step": 116850
     },
     {
       "epoch": 1.8985881626618575,
-      "grad_norm": 3.0,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.1866,
       "step": 116860
     },
     {
       "epoch": 1.8987506295592274,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3256,
+      "loss": 1.2246,
       "step": 116870
     },
     {
       "epoch": 1.898913096456597,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3341,
+      "loss": 1.2064,
       "step": 116880
     },
     {
       "epoch": 1.8990755633539667,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3243,
+      "loss": 1.2183,
       "step": 116890
     },
     {
       "epoch": 1.8992380302513363,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2389,
       "step": 116900
     },
     {
       "epoch": 1.899400497148706,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3275,
+      "loss": 1.2308,
       "step": 116910
     },
     {
       "epoch": 1.8995629640460756,
-      "grad_norm": 4.59375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3521,
+      "loss": 1.2402,
       "step": 116920
     },
     {
       "epoch": 1.8997254309434453,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2219,
       "step": 116930
     },
     {
       "epoch": 1.899887897840815,
-      "grad_norm": 4.78125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2179,
       "step": 116940
     },
     {
       "epoch": 1.9000503647381846,
-      "grad_norm": 2.25,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.1862,
       "step": 116950
     },
     {
       "epoch": 1.9002128316355542,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3149,
+      "loss": 1.2259,
       "step": 116960
     },
     {
       "epoch": 1.9003752985329239,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.226,
       "step": 116970
     },
     {
       "epoch": 1.9005377654302937,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.2071,
       "step": 116980
     },
     {
       "epoch": 1.9007002323276634,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2255,
       "step": 116990
     },
     {
       "epoch": 1.900862699225033,
-      "grad_norm": 3.859375,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2564,
       "step": 117000
     },
     {
       "epoch": 1.9010251661224027,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2296,
       "step": 117010
     },
     {
       "epoch": 1.9011876330197723,
-      "grad_norm": 3.59375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.1823,
       "step": 117020
     },
     {
       "epoch": 1.901350099917142,
-      "grad_norm": 2.359375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3345,
+      "loss": 1.2335,
       "step": 117030
     },
     {
       "epoch": 1.9015125668145116,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2386,
       "step": 117040
     },
     {
       "epoch": 1.9016750337118813,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.1932,
       "step": 117050
     },
     {
       "epoch": 1.901837500609251,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3131,
+      "loss": 1.1901,
       "step": 117060
     },
     {
       "epoch": 1.9019999675066206,
-      "grad_norm": 2.5625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2006,
       "step": 117070
     },
     {
       "epoch": 1.9021624344039902,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2673,
       "step": 117080
     },
     {
       "epoch": 1.90232490130136,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3373,
+      "loss": 1.1823,
       "step": 117090
     },
     {
       "epoch": 1.9024873681987295,
-      "grad_norm": 3.65625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2569,
       "step": 117100
     },
     {
       "epoch": 1.9026498350960992,
-      "grad_norm": 4.625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3589,
+      "loss": 1.2442,
       "step": 117110
     },
     {
       "epoch": 1.9028123019934688,
-      "grad_norm": 3.359375,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.2347,
       "step": 117120
     },
     {
       "epoch": 1.9029747688908385,
-      "grad_norm": 2.203125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3092,
+      "loss": 1.1672,
       "step": 117130
     },
     {
       "epoch": 1.9031372357882081,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.2217,
       "step": 117140
     },
     {
       "epoch": 1.9032997026855778,
-      "grad_norm": 3.375,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2078,
       "step": 117150
     },
     {
       "epoch": 1.9034621695829474,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 1.2376,
       "step": 117160
     },
     {
       "epoch": 1.903624636480317,
-      "grad_norm": 1.3828125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3183,
+      "loss": 1.241,
       "step": 117170
     },
     {
       "epoch": 1.9037871033776868,
-      "grad_norm": 2.859375,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2151,
       "step": 117180
     },
     {
       "epoch": 1.9039495702750564,
-      "grad_norm": 2.640625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2268,
       "step": 117190
     },
     {
       "epoch": 1.904112037172426,
-      "grad_norm": 2.515625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2617,
       "step": 117200
     },
     {
       "epoch": 1.9042745040697957,
-      "grad_norm": 2.625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2495,
       "step": 117210
     },
     {
       "epoch": 1.9044369709671654,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3265,
+      "loss": 1.2307,
       "step": 117220
     },
     {
       "epoch": 1.904599437864535,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.1948,
       "step": 117230
     },
     {
       "epoch": 1.9047619047619047,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2064,
       "step": 117240
     },
     {
       "epoch": 1.9049243716592743,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.1942,
       "step": 117250
     },
     {
       "epoch": 1.905086838556644,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2287,
       "step": 117260
     },
     {
       "epoch": 1.9052493054540136,
-      "grad_norm": 3.484375,
+      "grad_norm": 5.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.2201,
       "step": 117270
     },
     {
       "epoch": 1.9054117723513833,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2087,
       "step": 117280
     },
     {
       "epoch": 1.905574239248753,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3636,
+      "loss": 1.1981,
       "step": 117290
     },
     {
       "epoch": 1.9057367061461226,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2423,
       "step": 117300
     },
     {
       "epoch": 1.9058991730434924,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3458,
+      "loss": 1.2062,
       "step": 117310
     },
     {
       "epoch": 1.906061639940862,
-      "grad_norm": 3.4375,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.2246,
       "step": 117320
     },
     {
       "epoch": 1.9062241068382317,
-      "grad_norm": 4.4375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2289,
       "step": 117330
     },
     {
       "epoch": 1.9063865737356014,
-      "grad_norm": 2.828125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2041,
       "step": 117340
     },
     {
       "epoch": 1.906549040632971,
-      "grad_norm": 3.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.2505,
       "step": 117350
     },
     {
       "epoch": 1.9067115075303407,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2206,
       "step": 117360
     },
     {
       "epoch": 1.9068739744277103,
-      "grad_norm": 2.453125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3197,
+      "loss": 1.2449,
       "step": 117370
     },
     {
       "epoch": 1.90703644132508,
-      "grad_norm": 1.9765625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.2386,
       "step": 117380
     },
     {
       "epoch": 1.9071989082224496,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3288,
+      "loss": 1.2328,
       "step": 117390
     },
     {
       "epoch": 1.9073613751198193,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3336,
+      "loss": 1.2179,
       "step": 117400
     },
     {
       "epoch": 1.9075238420171892,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.251,
       "step": 117410
     },
     {
       "epoch": 1.9076863089145588,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.18,
       "step": 117420
     },
     {
       "epoch": 1.9078487758119285,
-      "grad_norm": 3.109375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.2034,
       "step": 117430
     },
     {
       "epoch": 1.9080112427092981,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.336,
+      "loss": 1.1705,
       "step": 117440
     },
     {
       "epoch": 1.9081737096066678,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2364,
       "step": 117450
     },
     {
       "epoch": 1.9083361765040374,
-      "grad_norm": 2.015625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1935,
       "step": 117460
     },
     {
       "epoch": 1.908498643401407,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.1952,
       "step": 117470
     },
     {
       "epoch": 1.9086611102987767,
-      "grad_norm": 2.078125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3551,
+      "loss": 1.2285,
       "step": 117480
     },
     {
       "epoch": 1.9088235771961464,
-      "grad_norm": 3.09375,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3636,
+      "loss": 1.2156,
       "step": 117490
     },
     {
       "epoch": 1.908986044093516,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.1954,
       "step": 117500
     },
     {
       "epoch": 1.9091485109908857,
-      "grad_norm": 3.578125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2158,
       "step": 117510
     },
     {
       "epoch": 1.9093109778882553,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.1772,
       "step": 117520
     },
     {
       "epoch": 1.909473444785625,
-      "grad_norm": 1.96875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.261,
       "step": 117530
     },
     {
       "epoch": 1.9096359116829946,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3486,
+      "loss": 1.2206,
       "step": 117540
     },
     {
       "epoch": 1.9097983785803643,
-      "grad_norm": 3.0625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3627,
+      "loss": 1.2361,
       "step": 117550
     },
     {
       "epoch": 1.909960845477734,
-      "grad_norm": 3.5,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.1549,
       "step": 117560
     },
     {
       "epoch": 1.9101233123751036,
-      "grad_norm": 2.84375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2149,
       "step": 117570
     },
     {
       "epoch": 1.9102857792724732,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2131,
       "step": 117580
     },
     {
       "epoch": 1.9104482461698429,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.358,
+      "loss": 1.2191,
       "step": 117590
     },
     {
       "epoch": 1.9106107130672125,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.313,
+      "loss": 1.2196,
       "step": 117600
     },
     {
       "epoch": 1.9107731799645822,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2318,
       "step": 117610
     },
     {
       "epoch": 1.9109356468619518,
-      "grad_norm": 2.53125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2594,
       "step": 117620
     },
     {
       "epoch": 1.9110981137593215,
-      "grad_norm": 2.875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3607,
+      "loss": 1.2027,
       "step": 117630
     },
     {
       "epoch": 1.9112605806566911,
-      "grad_norm": 4.1875,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.1595,
       "step": 117640
     },
     {
       "epoch": 1.9114230475540608,
-      "grad_norm": 3.25,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2239,
       "step": 117650
     },
     {
       "epoch": 1.9115855144514304,
-      "grad_norm": 3.109375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2375,
       "step": 117660
     },
     {
       "epoch": 1.9117479813488,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3294,
+      "loss": 1.2389,
       "step": 117670
     },
     {
       "epoch": 1.9119104482461697,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2386,
       "step": 117680
     },
     {
       "epoch": 1.9120729151435394,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.2273,
       "step": 117690
     },
     {
       "epoch": 1.912235382040909,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.2412,
       "step": 117700
     },
     {
       "epoch": 1.9123978489382787,
-      "grad_norm": 2.75,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3617,
+      "loss": 1.2053,
       "step": 117710
     },
     {
       "epoch": 1.9125603158356483,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3387,
+      "loss": 1.225,
       "step": 117720
     },
     {
       "epoch": 1.912722782733018,
-      "grad_norm": 2.5,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.1944,
       "step": 117730
     },
     {
       "epoch": 1.9128852496303879,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3284,
+      "loss": 1.2395,
       "step": 117740
     },
     {
       "epoch": 1.9130477165277575,
-      "grad_norm": 3.0,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3181,
+      "loss": 1.2021,
       "step": 117750
     },
     {
       "epoch": 1.9132101834251272,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.233,
       "step": 117760
     },
     {
       "epoch": 1.9133726503224968,
-      "grad_norm": 3.25,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2323,
       "step": 117770
     },
     {
       "epoch": 1.9135351172198665,
-      "grad_norm": 3.8125,
+      "grad_norm": 5.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.1901,
       "step": 117780
     },
     {
       "epoch": 1.9136975841172361,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3568,
+      "loss": 1.2077,
       "step": 117790
     },
     {
       "epoch": 1.9138600510146058,
-      "grad_norm": 3.203125,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2015,
       "step": 117800
     },
     {
       "epoch": 1.9140225179119754,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3464,
+      "loss": 1.2363,
       "step": 117810
     },
     {
       "epoch": 1.914184984809345,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2349,
       "step": 117820
     },
     {
       "epoch": 1.9143474517067147,
-      "grad_norm": 3.03125,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2267,
       "step": 117830
     },
     {
       "epoch": 1.9145099186040844,
-      "grad_norm": 3.203125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2083,
       "step": 117840
     },
     {
       "epoch": 1.9146723855014542,
-      "grad_norm": 2.546875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3321,
+      "loss": 1.2056,
       "step": 117850
     },
     {
       "epoch": 1.914834852398824,
-      "grad_norm": 3.5,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.2673,
       "step": 117860
     },
     {
       "epoch": 1.9149973192961935,
-      "grad_norm": 2.65625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3298,
+      "loss": 1.2238,
       "step": 117870
     },
     {
       "epoch": 1.9151597861935632,
-      "grad_norm": 4.71875,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.1959,
       "step": 117880
     },
     {
       "epoch": 1.9153222530909328,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3523,
+      "loss": 1.1897,
       "step": 117890
     },
     {
       "epoch": 1.9154847199883025,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.331,
+      "loss": 1.2717,
       "step": 117900
     },
     {
       "epoch": 1.9156471868856721,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3133,
+      "loss": 1.2362,
       "step": 117910
     },
     {
       "epoch": 1.9158096537830418,
-      "grad_norm": 3.5,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2394,
       "step": 117920
     },
     {
       "epoch": 1.9159721206804115,
-      "grad_norm": 3.234375,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3619,
+      "loss": 1.171,
       "step": 117930
     },
     {
       "epoch": 1.916134587577781,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2426,
       "step": 117940
     },
     {
       "epoch": 1.9162970544751508,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3306,
+      "loss": 1.2534,
       "step": 117950
     },
     {
       "epoch": 1.9164595213725204,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.2117,
       "step": 117960
     },
     {
       "epoch": 1.91662198826989,
-      "grad_norm": 3.375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.2278,
       "step": 117970
     },
     {
       "epoch": 1.9167844551672597,
-      "grad_norm": 3.0,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2381,
       "step": 117980
     },
     {
       "epoch": 1.9169469220646294,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.2127,
       "step": 117990
     },
     {
       "epoch": 1.917109388961999,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2266,
       "step": 118000
     },
     {
       "epoch": 1.9172718558593687,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2022,
       "step": 118010
     },
     {
       "epoch": 1.9174343227567383,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3545,
+      "loss": 1.2234,
       "step": 118020
     },
     {
       "epoch": 1.917596789654108,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3287,
+      "loss": 1.2022,
       "step": 118030
     },
     {
       "epoch": 1.9177592565514776,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2461,
       "step": 118040
     },
     {
       "epoch": 1.9179217234488473,
-      "grad_norm": 2.75,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2014,
       "step": 118050
     },
     {
       "epoch": 1.918084190346217,
-      "grad_norm": 2.625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.218,
       "step": 118060
     },
     {
       "epoch": 1.9182466572435866,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2503,
       "step": 118070
     },
     {
       "epoch": 1.9184091241409562,
-      "grad_norm": 7.28125,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.1937,
       "step": 118080
     },
     {
       "epoch": 1.9185715910383259,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.1734,
       "step": 118090
     },
     {
       "epoch": 1.9187340579356955,
-      "grad_norm": 2.09375,
+      "grad_norm": 5.9375,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.1555,
       "step": 118100
     },
     {
       "epoch": 1.9188965248330652,
-      "grad_norm": 3.625,
+      "grad_norm": 5.25,
       "learning_rate": 5e-05,
-      "loss": 0.3255,
+      "loss": 1.2005,
       "step": 118110
     },
     {
       "epoch": 1.9190589917304348,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3524,
+      "loss": 1.2267,
       "step": 118120
     },
     {
       "epoch": 1.9192214586278045,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.31,
+      "loss": 1.2184,
       "step": 118130
     },
     {
       "epoch": 1.9193839255251741,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3557,
+      "loss": 1.2047,
       "step": 118140
     },
     {
       "epoch": 1.9195463924225438,
-      "grad_norm": 2.1875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.2256,
       "step": 118150
     },
     {
       "epoch": 1.9197088593199134,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2334,
       "step": 118160
     },
     {
       "epoch": 1.919871326217283,
-      "grad_norm": 2.8125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3468,
+      "loss": 1.1861,
       "step": 118170
     },
     {
       "epoch": 1.920033793114653,
-      "grad_norm": 3.53125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.228,
       "step": 118180
     },
     {
       "epoch": 1.9201962600120226,
-      "grad_norm": 4.28125,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2018,
       "step": 118190
     },
     {
       "epoch": 1.9203587269093922,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2364,
       "step": 118200
     },
     {
       "epoch": 1.920521193806762,
-      "grad_norm": 2.625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2422,
       "step": 118210
     },
     {
       "epoch": 1.9206836607041315,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3637,
+      "loss": 1.2552,
       "step": 118220
     },
     {
       "epoch": 1.9208461276015012,
-      "grad_norm": 3.28125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2034,
       "step": 118230
     },
     {
       "epoch": 1.9210085944988708,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3518,
+      "loss": 1.2017,
       "step": 118240
     },
     {
       "epoch": 1.9211710613962405,
-      "grad_norm": 2.265625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3116,
+      "loss": 1.219,
       "step": 118250
     },
     {
       "epoch": 1.9213335282936101,
-      "grad_norm": 3.765625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.361,
+      "loss": 1.2377,
       "step": 118260
     },
     {
       "epoch": 1.9214959951909798,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2223,
       "step": 118270
     },
     {
       "epoch": 1.9216584620883495,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2058,
       "step": 118280
     },
     {
       "epoch": 1.9218209289857193,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3268,
+      "loss": 1.2468,
       "step": 118290
     },
     {
       "epoch": 1.921983395883089,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2384,
       "step": 118300
     },
     {
       "epoch": 1.9221458627804586,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3365,
+      "loss": 1.2505,
       "step": 118310
     },
     {
       "epoch": 1.9223083296778283,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3668,
+      "loss": 1.2539,
       "step": 118320
     },
     {
       "epoch": 1.922470796575198,
-      "grad_norm": 3.171875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.352,
+      "loss": 1.2444,
       "step": 118330
     },
     {
       "epoch": 1.9226332634725676,
-      "grad_norm": 3.296875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.362,
+      "loss": 1.2106,
       "step": 118340
     },
     {
       "epoch": 1.9227957303699372,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.2261,
       "step": 118350
     },
     {
       "epoch": 1.9229581972673069,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.198,
       "step": 118360
     },
     {
       "epoch": 1.9231206641646765,
-      "grad_norm": 2.046875,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 1.2603,
       "step": 118370
     },
     {
       "epoch": 1.9232831310620462,
-      "grad_norm": 3.25,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3535,
+      "loss": 1.2281,
       "step": 118380
     },
     {
       "epoch": 1.9234455979594158,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.1905,
       "step": 118390
     },
     {
       "epoch": 1.9236080648567855,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2119,
       "step": 118400
     },
     {
       "epoch": 1.9237705317541551,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3453,
+      "loss": 1.2452,
       "step": 118410
     },
     {
       "epoch": 1.9239329986515248,
-      "grad_norm": 2.96875,
+      "grad_norm": 5.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2125,
       "step": 118420
     },
     {
       "epoch": 1.9240954655488944,
-      "grad_norm": 3.0,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3377,
+      "loss": 1.2416,
       "step": 118430
     },
     {
       "epoch": 1.924257932446264,
-      "grad_norm": 3.046875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2265,
       "step": 118440
     },
     {
       "epoch": 1.9244203993436337,
-      "grad_norm": 3.921875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3519,
+      "loss": 1.2032,
       "step": 118450
     },
     {
       "epoch": 1.9245828662410034,
-      "grad_norm": 4.8125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.2102,
       "step": 118460
     },
     {
       "epoch": 1.924745333138373,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2233,
       "step": 118470
     },
     {
       "epoch": 1.9249078000357427,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2386,
       "step": 118480
     },
     {
       "epoch": 1.9250702669331123,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3366,
+      "loss": 1.2392,
       "step": 118490
     },
     {
       "epoch": 1.925232733830482,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2124,
       "step": 118500
     },
     {
       "epoch": 1.9253952007278516,
-      "grad_norm": 3.03125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.2086,
       "step": 118510
     },
     {
       "epoch": 1.9255576676252213,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2221,
       "step": 118520
     },
     {
       "epoch": 1.925720134522591,
-      "grad_norm": 2.921875,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3504,
+      "loss": 1.2056,
       "step": 118530
     },
     {
       "epoch": 1.9258826014199606,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2236,
       "step": 118540
     },
     {
       "epoch": 1.9260450683173302,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.2494,
       "step": 118550
     },
     {
       "epoch": 1.9262075352147,
-      "grad_norm": 2.234375,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3271,
+      "loss": 1.2027,
       "step": 118560
     },
     {
       "epoch": 1.9263700021120695,
-      "grad_norm": 3.703125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.1779,
       "step": 118570
     },
     {
       "epoch": 1.9265324690094392,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.2452,
       "step": 118580
     },
     {
       "epoch": 1.9266949359068088,
-      "grad_norm": 3.859375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2361,
       "step": 118590
     },
     {
       "epoch": 1.9268574028041785,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3757,
+      "loss": 1.1804,
       "step": 118600
     },
     {
       "epoch": 1.9270198697015481,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.1994,
       "step": 118610
     },
     {
       "epoch": 1.927182336598918,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3411,
+      "loss": 1.1964,
       "step": 118620
     },
     {
       "epoch": 1.9273448034962877,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3126,
+      "loss": 1.2385,
       "step": 118630
     },
     {
       "epoch": 1.9275072703936573,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.2151,
       "step": 118640
     },
     {
       "epoch": 1.927669737291027,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3395,
+      "loss": 1.1808,
       "step": 118650
     },
     {
       "epoch": 1.9278322041883966,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.1993,
       "step": 118660
     },
     {
       "epoch": 1.9279946710857663,
-      "grad_norm": 3.078125,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3598,
+      "loss": 1.244,
       "step": 118670
     },
     {
       "epoch": 1.928157137983136,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3684,
+      "loss": 1.2202,
       "step": 118680
     },
     {
       "epoch": 1.9283196048805056,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2215,
       "step": 118690
     },
     {
       "epoch": 1.9284820717778752,
-      "grad_norm": 3.21875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2266,
       "step": 118700
     },
     {
       "epoch": 1.9286445386752449,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3647,
+      "loss": 1.2151,
       "step": 118710
     },
     {
       "epoch": 1.9288070055726145,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2389,
       "step": 118720
     },
     {
       "epoch": 1.9289694724699844,
-      "grad_norm": 5.9375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.245,
       "step": 118730
     },
     {
       "epoch": 1.929131939367354,
-      "grad_norm": 2.5,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3263,
+      "loss": 1.2122,
       "step": 118740
     },
     {
       "epoch": 1.9292944062647237,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.1999,
       "step": 118750
     },
     {
       "epoch": 1.9294568731620934,
-      "grad_norm": 2.921875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3664,
+      "loss": 1.2439,
       "step": 118760
     },
     {
       "epoch": 1.929619340059463,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3507,
+      "loss": 1.2453,
       "step": 118770
     },
     {
       "epoch": 1.9297818069568327,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3477,
+      "loss": 1.2344,
       "step": 118780
     },
     {
       "epoch": 1.9299442738542023,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3183,
+      "loss": 1.237,
       "step": 118790
     },
     {
       "epoch": 1.930106740751572,
-      "grad_norm": 2.765625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.1866,
       "step": 118800
     },
     {
       "epoch": 1.9302692076489416,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2492,
       "step": 118810
     },
     {
       "epoch": 1.9304316745463113,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2017,
       "step": 118820
     },
     {
       "epoch": 1.930594141443681,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.2371,
       "step": 118830
     },
     {
       "epoch": 1.9307566083410506,
-      "grad_norm": 3.75,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3478,
+      "loss": 1.1777,
       "step": 118840
     },
     {
       "epoch": 1.9309190752384202,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2196,
       "step": 118850
     },
     {
       "epoch": 1.9310815421357899,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3408,
+      "loss": 1.239,
       "step": 118860
     },
     {
       "epoch": 1.9312440090331595,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2142,
       "step": 118870
     },
     {
       "epoch": 1.9314064759305292,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.2085,
       "step": 118880
     },
     {
       "epoch": 1.9315689428278988,
-      "grad_norm": 3.140625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2088,
       "step": 118890
     },
     {
       "epoch": 1.9317314097252685,
-      "grad_norm": 4.21875,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2244,
       "step": 118900
     },
     {
       "epoch": 1.9318938766226381,
-      "grad_norm": 3.21875,
+      "grad_norm": 5.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2549,
       "step": 118910
     },
     {
       "epoch": 1.9320563435200078,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3581,
+      "loss": 1.2238,
       "step": 118920
     },
     {
       "epoch": 1.9322188104173774,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.2421,
       "step": 118930
     },
     {
       "epoch": 1.932381277314747,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.1962,
       "step": 118940
     },
     {
       "epoch": 1.9325437442121167,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.2537,
       "step": 118950
     },
     {
       "epoch": 1.9327062111094864,
-      "grad_norm": 2.671875,
+      "grad_norm": 5.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3681,
+      "loss": 1.2315,
       "step": 118960
     },
     {
       "epoch": 1.932868678006856,
-      "grad_norm": 2.453125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2189,
       "step": 118970
     },
     {
       "epoch": 1.9330311449042257,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2004,
       "step": 118980
     },
     {
       "epoch": 1.9331936118015953,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3332,
+      "loss": 1.2021,
       "step": 118990
     },
     {
       "epoch": 1.933356078698965,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2151,
       "step": 119000
     },
     {
       "epoch": 1.9335185455963346,
-      "grad_norm": 2.40625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3385,
+      "loss": 1.2582,
       "step": 119010
     },
     {
       "epoch": 1.9336810124937043,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3456,
+      "loss": 1.2251,
       "step": 119020
     },
     {
       "epoch": 1.933843479391074,
-      "grad_norm": 2.875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.1977,
       "step": 119030
     },
     {
       "epoch": 1.9340059462884436,
-      "grad_norm": 2.890625,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2321,
       "step": 119040
     },
     {
       "epoch": 1.9341684131858132,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.2088,
       "step": 119050
     },
     {
       "epoch": 1.934330880083183,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3348,
+      "loss": 1.221,
       "step": 119060
     },
     {
       "epoch": 1.9344933469805528,
-      "grad_norm": 3.84375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3353,
+      "loss": 1.216,
       "step": 119070
     },
     {
       "epoch": 1.9346558138779224,
-      "grad_norm": 4.5,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2145,
       "step": 119080
     },
     {
       "epoch": 1.934818280775292,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2293,
       "step": 119090
     },
     {
       "epoch": 1.9349807476726617,
-      "grad_norm": 2.328125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2718,
       "step": 119100
     },
     {
       "epoch": 1.9351432145700314,
-      "grad_norm": 3.875,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3651,
+      "loss": 1.2276,
       "step": 119110
     },
     {
       "epoch": 1.935305681467401,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3668,
+      "loss": 1.2188,
       "step": 119120
     },
     {
       "epoch": 1.9354681483647707,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2101,
       "step": 119130
     },
     {
       "epoch": 1.9356306152621403,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2265,
       "step": 119140
     },
     {
       "epoch": 1.93579308215951,
-      "grad_norm": 1.84375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2633,
       "step": 119150
     },
     {
       "epoch": 1.9359555490568796,
-      "grad_norm": 2.875,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3415,
+      "loss": 1.2592,
       "step": 119160
     },
     {
       "epoch": 1.9361180159542495,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.1967,
       "step": 119170
     },
     {
       "epoch": 1.9362804828516191,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2309,
       "step": 119180
     },
     {
       "epoch": 1.9364429497489888,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3544,
+      "loss": 1.2054,
       "step": 119190
     },
     {
       "epoch": 1.9366054166463584,
-      "grad_norm": 2.390625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2339,
       "step": 119200
     },
     {
       "epoch": 1.936767883543728,
-      "grad_norm": 2.03125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2217,
       "step": 119210
     },
     {
       "epoch": 1.9369303504410977,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3454,
+      "loss": 1.2197,
       "step": 119220
     },
     {
       "epoch": 1.9370928173384674,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.1994,
       "step": 119230
     },
     {
       "epoch": 1.937255284235837,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3413,
+      "loss": 1.252,
       "step": 119240
     },
     {
       "epoch": 1.9374177511332067,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2497,
       "step": 119250
     },
     {
       "epoch": 1.9375802180305763,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.1874,
       "step": 119260
     },
     {
       "epoch": 1.937742684927946,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3733,
+      "loss": 1.2342,
       "step": 119270
     },
     {
       "epoch": 1.9379051518253156,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3737,
+      "loss": 1.169,
       "step": 119280
     },
     {
       "epoch": 1.9380676187226853,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2396,
       "step": 119290
     },
     {
       "epoch": 1.938230085620055,
-      "grad_norm": 3.75,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3601,
+      "loss": 1.2304,
       "step": 119300
     },
     {
       "epoch": 1.9383925525174246,
-      "grad_norm": 5.09375,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2251,
       "step": 119310
     },
     {
       "epoch": 1.9385550194147942,
-      "grad_norm": 1.875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3226,
+      "loss": 1.2282,
       "step": 119320
     },
     {
       "epoch": 1.938717486312164,
-      "grad_norm": 3.1875,
+      "grad_norm": 5.625,
       "learning_rate": 5e-05,
-      "loss": 0.3247,
+      "loss": 1.2096,
       "step": 119330
     },
     {
       "epoch": 1.9388799532095335,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2092,
       "step": 119340
     },
     {
       "epoch": 1.9390424201069032,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2445,
       "step": 119350
     },
     {
       "epoch": 1.9392048870042728,
-      "grad_norm": 1.90625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2306,
       "step": 119360
     },
     {
       "epoch": 1.9393673539016425,
-      "grad_norm": 3.71875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3455,
+      "loss": 1.1841,
       "step": 119370
     },
     {
       "epoch": 1.9395298207990121,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3156,
+      "loss": 1.2311,
       "step": 119380
     },
     {
       "epoch": 1.9396922876963818,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2027,
       "step": 119390
     },
     {
       "epoch": 1.9398547545937515,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.2259,
       "step": 119400
     },
     {
       "epoch": 1.940017221491121,
-      "grad_norm": 2.28125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.1978,
       "step": 119410
     },
     {
       "epoch": 1.9401796883884908,
-      "grad_norm": 2.21875,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.1717,
       "step": 119420
     },
     {
       "epoch": 1.9403421552858604,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3179,
+      "loss": 1.2464,
       "step": 119430
     },
     {
       "epoch": 1.94050462218323,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2405,
       "step": 119440
     },
     {
       "epoch": 1.9406670890805997,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3196,
+      "loss": 1.2354,
       "step": 119450
     },
     {
       "epoch": 1.9408295559779694,
-      "grad_norm": 4.125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2076,
       "step": 119460
     },
     {
       "epoch": 1.940992022875339,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2178,
       "step": 119470
     },
     {
       "epoch": 1.9411544897727087,
-      "grad_norm": 2.875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.2319,
       "step": 119480
     },
     {
       "epoch": 1.9413169566700783,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.2149,
       "step": 119490
     },
     {
       "epoch": 1.9414794235674482,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3528,
+      "loss": 1.1966,
       "step": 119500
     },
     {
       "epoch": 1.9416418904648178,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2079,
       "step": 119510
     },
     {
       "epoch": 1.9418043573621875,
-      "grad_norm": 3.03125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3531,
+      "loss": 1.2044,
       "step": 119520
     },
     {
       "epoch": 1.9419668242595571,
-      "grad_norm": 3.265625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2029,
       "step": 119530
     },
     {
       "epoch": 1.9421292911569268,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3662,
+      "loss": 1.1831,
       "step": 119540
     },
     {
       "epoch": 1.9422917580542964,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2333,
       "step": 119550
     },
     {
       "epoch": 1.942454224951666,
-      "grad_norm": 4.0625,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.207,
       "step": 119560
     },
     {
       "epoch": 1.9426166918490357,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2219,
       "step": 119570
     },
     {
       "epoch": 1.9427791587464054,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.2576,
       "step": 119580
     },
     {
       "epoch": 1.942941625643775,
-      "grad_norm": 3.0,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3729,
+      "loss": 1.191,
       "step": 119590
     },
     {
       "epoch": 1.9431040925411447,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.2484,
       "step": 119600
     },
     {
       "epoch": 1.9432665594385146,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3418,
+      "loss": 1.185,
       "step": 119610
     },
     {
       "epoch": 1.9434290263358842,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3253,
+      "loss": 1.2338,
       "step": 119620
     },
     {
       "epoch": 1.9435914932332539,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2182,
       "step": 119630
     },
     {
       "epoch": 1.9437539601306235,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1909,
       "step": 119640
     },
     {
       "epoch": 1.9439164270279932,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.367,
+      "loss": 1.2378,
       "step": 119650
     },
     {
       "epoch": 1.9440788939253628,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2194,
       "step": 119660
     },
     {
       "epoch": 1.9442413608227325,
-      "grad_norm": 4.0,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2575,
       "step": 119670
     },
     {
       "epoch": 1.9444038277201021,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3231,
+      "loss": 1.2597,
       "step": 119680
     },
     {
       "epoch": 1.9445662946174718,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2169,
       "step": 119690
     },
     {
       "epoch": 1.9447287615148414,
-      "grad_norm": 3.0,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3286,
+      "loss": 1.2261,
       "step": 119700
     },
     {
       "epoch": 1.944891228412211,
-      "grad_norm": 3.828125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 1.1945,
       "step": 119710
     },
     {
       "epoch": 1.9450536953095807,
-      "grad_norm": 2.296875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3554,
+      "loss": 1.1789,
       "step": 119720
     },
     {
       "epoch": 1.9452161622069504,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3576,
+      "loss": 1.2054,
       "step": 119730
     },
     {
       "epoch": 1.94537862910432,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.2359,
       "step": 119740
     },
     {
       "epoch": 1.9455410960016897,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2042,
       "step": 119750
     },
     {
       "epoch": 1.9457035628990593,
-      "grad_norm": 1.921875,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2175,
       "step": 119760
     },
     {
       "epoch": 1.945866029796429,
-      "grad_norm": 2.921875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.2085,
       "step": 119770
     },
     {
       "epoch": 1.9460284966937986,
-      "grad_norm": 3.53125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3391,
+      "loss": 1.1967,
       "step": 119780
     },
     {
       "epoch": 1.9461909635911683,
-      "grad_norm": 2.5625,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3296,
+      "loss": 1.2625,
       "step": 119790
     },
     {
       "epoch": 1.946353430488538,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.346,
+      "loss": 1.2196,
       "step": 119800
     },
     {
       "epoch": 1.9465158973859076,
-      "grad_norm": 3.90625,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3428,
+      "loss": 1.2415,
       "step": 119810
     },
     {
       "epoch": 1.9466783642832772,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.2091,
       "step": 119820
     },
     {
       "epoch": 1.9468408311806469,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3252,
+      "loss": 1.2062,
       "step": 119830
     },
     {
       "epoch": 1.9470032980780165,
-      "grad_norm": 3.203125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2327,
       "step": 119840
     },
     {
       "epoch": 1.9471657649753862,
-      "grad_norm": 3.125,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3174,
+      "loss": 1.2208,
       "step": 119850
     },
     {
       "epoch": 1.9473282318727558,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2367,
       "step": 119860
     },
     {
       "epoch": 1.9474906987701255,
-      "grad_norm": 3.125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3422,
+      "loss": 1.2227,
       "step": 119870
     },
     {
       "epoch": 1.9476531656674951,
-      "grad_norm": 3.625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.2013,
       "step": 119880
     },
     {
       "epoch": 1.9478156325648648,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3611,
+      "loss": 1.2655,
       "step": 119890
     },
     {
       "epoch": 1.9479780994622344,
-      "grad_norm": 3.0,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3563,
+      "loss": 1.192,
       "step": 119900
     },
     {
       "epoch": 1.948140566359604,
-      "grad_norm": 2.234375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3407,
+      "loss": 1.2253,
       "step": 119910
     },
     {
       "epoch": 1.9483030332569737,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3406,
+      "loss": 1.225,
       "step": 119920
     },
     {
       "epoch": 1.9484655001543434,
-      "grad_norm": 3.8125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.2168,
       "step": 119930
     },
     {
       "epoch": 1.9486279670517133,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.2257,
       "step": 119940
     },
     {
       "epoch": 1.948790433949083,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 1.2066,
       "step": 119950
     },
     {
       "epoch": 1.9489529008464526,
-      "grad_norm": 3.125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.2158,
       "step": 119960
     },
     {
       "epoch": 1.9491153677438222,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.1953,
       "step": 119970
     },
     {
       "epoch": 1.9492778346411919,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3401,
+      "loss": 1.2012,
       "step": 119980
     },
     {
       "epoch": 1.9494403015385615,
-      "grad_norm": 2.3125,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.1925,
       "step": 119990
     },
     {
       "epoch": 1.9496027684359312,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.215,
       "step": 120000
     },
     {
       "epoch": 1.9497652353333008,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3631,
+      "loss": 1.1841,
       "step": 120010
     },
     {
       "epoch": 1.9499277022306705,
-      "grad_norm": 2.46875,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2038,
       "step": 120020
     },
     {
       "epoch": 1.9500901691280401,
-      "grad_norm": 2.484375,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.1948,
       "step": 120030
     },
     {
       "epoch": 1.9502526360254098,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3588,
+      "loss": 1.2184,
       "step": 120040
     },
     {
       "epoch": 1.9504151029227796,
-      "grad_norm": 3.0,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2245,
       "step": 120050
     },
     {
       "epoch": 1.9505775698201493,
-      "grad_norm": 2.90625,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3636,
+      "loss": 1.2546,
       "step": 120060
     },
     {
       "epoch": 1.950740036717519,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.3597,
+      "loss": 1.1758,
       "step": 120070
     },
     {
       "epoch": 1.9509025036148886,
-      "grad_norm": 2.453125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3267,
+      "loss": 1.2199,
       "step": 120080
     },
     {
       "epoch": 1.9510649705122582,
-      "grad_norm": 3.5,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3476,
+      "loss": 1.1898,
       "step": 120090
     },
     {
       "epoch": 1.951227437409628,
-      "grad_norm": 3.5,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2166,
       "step": 120100
     },
     {
       "epoch": 1.9513899043069975,
-      "grad_norm": 3.375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.2153,
       "step": 120110
     },
     {
       "epoch": 1.9515523712043672,
-      "grad_norm": 4.03125,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3785,
+      "loss": 1.2446,
       "step": 120120
     },
     {
       "epoch": 1.9517148381017368,
-      "grad_norm": 4.59375,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3515,
+      "loss": 1.2035,
       "step": 120130
     },
     {
       "epoch": 1.9518773049991065,
-      "grad_norm": 2.984375,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2191,
       "step": 120140
     },
     {
       "epoch": 1.9520397718964762,
-      "grad_norm": 2.625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.181,
       "step": 120150
     },
     {
       "epoch": 1.9522022387938458,
-      "grad_norm": 3.0,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3438,
+      "loss": 1.1968,
       "step": 120160
     },
     {
       "epoch": 1.9523647056912155,
-      "grad_norm": 2.625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3459,
+      "loss": 1.2303,
       "step": 120170
     },
     {
       "epoch": 1.952527172588585,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3432,
+      "loss": 1.1847,
       "step": 120180
     },
     {
       "epoch": 1.9526896394859548,
-      "grad_norm": 2.28125,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3567,
+      "loss": 1.2422,
       "step": 120190
     },
     {
       "epoch": 1.9528521063833244,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3261,
+      "loss": 1.2417,
       "step": 120200
     },
     {
       "epoch": 1.953014573280694,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.1985,
       "step": 120210
     },
     {
       "epoch": 1.9531770401780637,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3259,
+      "loss": 1.2533,
       "step": 120220
     },
     {
       "epoch": 1.9533395070754334,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.351,
+      "loss": 1.1915,
       "step": 120230
     },
     {
       "epoch": 1.953501973972803,
-      "grad_norm": 3.3125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3337,
+      "loss": 1.2485,
       "step": 120240
     },
     {
       "epoch": 1.9536644408701727,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2168,
       "step": 120250
     },
     {
       "epoch": 1.9538269077675423,
-      "grad_norm": 3.453125,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3213,
+      "loss": 1.2047,
       "step": 120260
     },
     {
       "epoch": 1.953989374664912,
-      "grad_norm": 3.015625,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3092,
+      "loss": 1.2595,
       "step": 120270
     },
     {
       "epoch": 1.9541518415622816,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2682,
       "step": 120280
     },
     {
       "epoch": 1.9543143084596513,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2063,
       "step": 120290
     },
     {
       "epoch": 1.954476775357021,
-      "grad_norm": 2.875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2089,
       "step": 120300
     },
     {
       "epoch": 1.9546392422543906,
-      "grad_norm": 2.8125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2518,
       "step": 120310
     },
     {
       "epoch": 1.9548017091517602,
-      "grad_norm": 3.5,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3316,
+      "loss": 1.2304,
       "step": 120320
     },
     {
       "epoch": 1.9549641760491299,
-      "grad_norm": 3.5,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3663,
+      "loss": 1.2673,
       "step": 120330
     },
     {
       "epoch": 1.9551266429464995,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2236,
       "step": 120340
     },
     {
       "epoch": 1.9552891098438692,
-      "grad_norm": 2.421875,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3633,
+      "loss": 1.2213,
       "step": 120350
     },
     {
       "epoch": 1.9554515767412388,
-      "grad_norm": 2.75,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2064,
       "step": 120360
     },
     {
       "epoch": 1.9556140436386085,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.178,
       "step": 120370
     },
     {
       "epoch": 1.9557765105359783,
-      "grad_norm": 4.03125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3402,
+      "loss": 1.2259,
       "step": 120380
     },
     {
       "epoch": 1.955938977433348,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.2401,
       "step": 120390
     },
     {
       "epoch": 1.9561014443307176,
-      "grad_norm": 3.234375,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.183,
       "step": 120400
     },
     {
       "epoch": 1.9562639112280873,
-      "grad_norm": 2.3125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3623,
+      "loss": 1.2269,
       "step": 120410
     },
     {
       "epoch": 1.956426378125457,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3526,
+      "loss": 1.2417,
       "step": 120420
     },
     {
       "epoch": 1.9565888450228266,
-      "grad_norm": 2.375,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2597,
       "step": 120430
     },
     {
       "epoch": 1.9567513119201962,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.1822,
       "step": 120440
     },
     {
       "epoch": 1.956913778817566,
-      "grad_norm": 2.109375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.1578,
       "step": 120450
     },
     {
       "epoch": 1.9570762457149355,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3293,
+      "loss": 1.1908,
       "step": 120460
     },
     {
       "epoch": 1.9572387126123052,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3257,
+      "loss": 1.2344,
       "step": 120470
     },
     {
       "epoch": 1.9574011795096748,
-      "grad_norm": 3.40625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2343,
       "step": 120480
     },
     {
       "epoch": 1.9575636464070447,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.1893,
       "step": 120490
     },
     {
       "epoch": 1.9577261133044144,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.2198,
       "step": 120500
     },
     {
       "epoch": 1.957888580201784,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3517,
+      "loss": 1.2,
       "step": 120510
     },
     {
       "epoch": 1.9580510470991537,
-      "grad_norm": 2.578125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2318,
       "step": 120520
     },
     {
       "epoch": 1.9582135139965233,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3417,
+      "loss": 1.2122,
       "step": 120530
     },
     {
       "epoch": 1.958375980893893,
-      "grad_norm": 3.3125,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.341,
+      "loss": 1.2156,
       "step": 120540
     },
     {
       "epoch": 1.9585384477912626,
-      "grad_norm": 5.15625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3585,
+      "loss": 1.243,
       "step": 120550
     },
     {
       "epoch": 1.9587009146886323,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3424,
+      "loss": 1.1836,
       "step": 120560
     },
     {
       "epoch": 1.958863381586002,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.2128,
       "step": 120570
     },
     {
       "epoch": 1.9590258484833716,
-      "grad_norm": 3.34375,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.3559,
+      "loss": 1.2365,
       "step": 120580
     },
     {
       "epoch": 1.9591883153807412,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.1986,
       "step": 120590
     },
     {
       "epoch": 1.9593507822781109,
-      "grad_norm": 3.09375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2459,
       "step": 120600
     },
     {
       "epoch": 1.9595132491754805,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2233,
       "step": 120610
     },
     {
       "epoch": 1.9596757160728502,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.1983,
       "step": 120620
     },
     {
       "epoch": 1.9598381829702198,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2107,
       "step": 120630
     },
     {
       "epoch": 1.9600006498675895,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2184,
       "step": 120640
     },
     {
       "epoch": 1.9601631167649591,
-      "grad_norm": 3.5625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2384,
       "step": 120650
     },
     {
       "epoch": 1.9603255836623288,
-      "grad_norm": 2.484375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2036,
       "step": 120660
     },
     {
       "epoch": 1.9604880505596984,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3618,
+      "loss": 1.2006,
       "step": 120670
     },
     {
       "epoch": 1.960650517457068,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3694,
+      "loss": 1.2034,
       "step": 120680
     },
     {
       "epoch": 1.9608129843544377,
-      "grad_norm": 2.6875,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.202,
       "step": 120690
     },
     {
       "epoch": 1.9609754512518074,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.2548,
       "step": 120700
     },
     {
       "epoch": 1.961137918149177,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.1965,
       "step": 120710
     },
     {
       "epoch": 1.9613003850465467,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3326,
+      "loss": 1.1671,
       "step": 120720
     },
     {
       "epoch": 1.9614628519439163,
-      "grad_norm": 3.390625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3533,
+      "loss": 1.1915,
       "step": 120730
     },
     {
       "epoch": 1.961625318841286,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.327,
+      "loss": 1.2196,
       "step": 120740
     },
     {
       "epoch": 1.9617877857386556,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3637,
+      "loss": 1.2108,
       "step": 120750
     },
     {
       "epoch": 1.9619502526360253,
-      "grad_norm": 2.75,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3505,
+      "loss": 1.275,
       "step": 120760
     },
     {
       "epoch": 1.962112719533395,
-      "grad_norm": 2.671875,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2049,
       "step": 120770
     },
     {
       "epoch": 1.9622751864307646,
-      "grad_norm": 4.3125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3398,
+      "loss": 1.2119,
       "step": 120780
     },
     {
       "epoch": 1.9624376533281342,
-      "grad_norm": 2.84375,
+      "grad_norm": 7.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.1865,
       "step": 120790
     },
     {
       "epoch": 1.962600120225504,
-      "grad_norm": 2.703125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3369,
+      "loss": 1.1673,
       "step": 120800
     },
     {
       "epoch": 1.9627625871228738,
-      "grad_norm": 3.265625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3469,
+      "loss": 1.1912,
       "step": 120810
     },
     {
       "epoch": 1.9629250540202434,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3186,
+      "loss": 1.2434,
       "step": 120820
     },
     {
       "epoch": 1.963087520917613,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3448,
+      "loss": 1.1976,
       "step": 120830
     },
     {
       "epoch": 1.9632499878149827,
-      "grad_norm": 3.015625,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3593,
+      "loss": 1.2173,
       "step": 120840
     },
     {
       "epoch": 1.9634124547123524,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.1891,
       "step": 120850
     },
     {
       "epoch": 1.963574921609722,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3393,
+      "loss": 1.2079,
       "step": 120860
     },
     {
       "epoch": 1.9637373885070917,
-      "grad_norm": 2.875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3227,
+      "loss": 1.1705,
       "step": 120870
     },
     {
       "epoch": 1.9638998554044613,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3342,
+      "loss": 1.2343,
       "step": 120880
     },
     {
       "epoch": 1.964062322301831,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 1.2383,
       "step": 120890
     },
     {
       "epoch": 1.9642247891992006,
-      "grad_norm": 2.25,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3376,
+      "loss": 1.2156,
       "step": 120900
     },
     {
       "epoch": 1.9643872560965703,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3639,
+      "loss": 1.2094,
       "step": 120910
     },
     {
       "epoch": 1.9645497229939402,
-      "grad_norm": 2.28125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2277,
       "step": 120920
     },
     {
       "epoch": 1.9647121898913098,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3584,
+      "loss": 1.1652,
       "step": 120930
     },
     {
       "epoch": 1.9648746567886795,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2345,
       "step": 120940
     },
     {
       "epoch": 1.965037123686049,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.1952,
       "step": 120950
     },
     {
       "epoch": 1.9651995905834188,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3194,
+      "loss": 1.1642,
       "step": 120960
     },
     {
       "epoch": 1.9653620574807884,
-      "grad_norm": 3.9375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2486,
       "step": 120970
     },
     {
       "epoch": 1.965524524378158,
-      "grad_norm": 3.90625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.2275,
       "step": 120980
     },
     {
       "epoch": 1.9656869912755277,
-      "grad_norm": 3.296875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.3512,
+      "loss": 1.2295,
       "step": 120990
     },
     {
       "epoch": 1.9658494581728974,
-      "grad_norm": 2.75,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2266,
       "step": 121000
     },
     {
       "epoch": 1.966011925070267,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3271,
+      "loss": 1.1772,
       "step": 121010
     },
     {
       "epoch": 1.9661743919676367,
-      "grad_norm": 2.703125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.326,
+      "loss": 1.2036,
       "step": 121020
     },
     {
       "epoch": 1.9663368588650063,
-      "grad_norm": 3.5625,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2287,
       "step": 121030
     },
     {
       "epoch": 1.966499325762376,
-      "grad_norm": 3.390625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3329,
+      "loss": 1.2141,
       "step": 121040
     },
     {
       "epoch": 1.9666617926597456,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3276,
+      "loss": 1.2057,
       "step": 121050
     },
     {
       "epoch": 1.9668242595571153,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3463,
+      "loss": 1.2406,
       "step": 121060
     },
     {
       "epoch": 1.966986726454485,
-      "grad_norm": 2.734375,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3442,
+      "loss": 1.2552,
       "step": 121070
     },
     {
       "epoch": 1.9671491933518546,
-      "grad_norm": 3.0,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3552,
+      "loss": 1.2158,
       "step": 121080
     },
     {
       "epoch": 1.9673116602492242,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.2047,
       "step": 121090
     },
     {
       "epoch": 1.9674741271465939,
-      "grad_norm": 2.53125,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.1959,
       "step": 121100
     },
     {
       "epoch": 1.9676365940439635,
-      "grad_norm": 3.125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2282,
       "step": 121110
     },
     {
       "epoch": 1.9677990609413332,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3134,
+      "loss": 1.2547,
       "step": 121120
     },
     {
       "epoch": 1.9679615278387028,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.332,
+      "loss": 1.2312,
       "step": 121130
     },
     {
       "epoch": 1.9681239947360725,
-      "grad_norm": 3.328125,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3586,
+      "loss": 1.1746,
       "step": 121140
     },
     {
       "epoch": 1.9682864616334421,
-      "grad_norm": 2.5625,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3396,
+      "loss": 1.1882,
       "step": 121150
     },
     {
       "epoch": 1.9684489285308118,
-      "grad_norm": 3.5,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.2403,
       "step": 121160
     },
     {
       "epoch": 1.9686113954281814,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3372,
+      "loss": 1.2195,
       "step": 121170
     },
     {
       "epoch": 1.968773862325551,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.1921,
       "step": 121180
     },
     {
       "epoch": 1.9689363292229207,
-      "grad_norm": 4.09375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2118,
       "step": 121190
     },
     {
       "epoch": 1.9690987961202904,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.2506,
       "step": 121200
     },
     {
       "epoch": 1.96926126301766,
-      "grad_norm": 1.890625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3303,
+      "loss": 1.224,
       "step": 121210
     },
     {
       "epoch": 1.9694237299150297,
-      "grad_norm": 2.1875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3494,
+      "loss": 1.2059,
       "step": 121220
     },
     {
       "epoch": 1.9695861968123993,
-      "grad_norm": 3.96875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.1974,
       "step": 121230
     },
     {
       "epoch": 1.969748663709769,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3368,
+      "loss": 1.229,
       "step": 121240
     },
     {
       "epoch": 1.9699111306071388,
-      "grad_norm": 2.640625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3451,
+      "loss": 1.238,
       "step": 121250
     },
     {
       "epoch": 1.9700735975045085,
-      "grad_norm": 4.46875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2346,
       "step": 121260
     },
     {
       "epoch": 1.9702360644018782,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3349,
+      "loss": 1.1835,
       "step": 121270
     },
     {
       "epoch": 1.9703985312992478,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.34,
+      "loss": 1.2302,
       "step": 121280
     },
     {
       "epoch": 1.9705609981966175,
-      "grad_norm": 2.71875,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3273,
+      "loss": 1.2247,
       "step": 121290
     },
     {
       "epoch": 1.970723465093987,
-      "grad_norm": 3.0,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.1889,
       "step": 121300
     },
     {
       "epoch": 1.9708859319913568,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3445,
+      "loss": 1.2279,
       "step": 121310
     },
     {
       "epoch": 1.9710483988887264,
-      "grad_norm": 2.828125,
+      "grad_norm": 5.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3371,
+      "loss": 1.2042,
       "step": 121320
     },
     {
       "epoch": 1.971210865786096,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2649,
       "step": 121330
     },
     {
       "epoch": 1.9713733326834657,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.328,
+      "loss": 1.2408,
       "step": 121340
     },
     {
       "epoch": 1.9715357995808354,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.3513,
+      "loss": 1.2399,
       "step": 121350
     },
     {
       "epoch": 1.9716982664782052,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.353,
+      "loss": 1.2008,
       "step": 121360
     },
     {
       "epoch": 1.9718607333755749,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3077,
+      "loss": 1.1821,
       "step": 121370
     },
     {
       "epoch": 1.9720232002729445,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3578,
+      "loss": 1.1988,
       "step": 121380
     },
     {
       "epoch": 1.9721856671703142,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3308,
+      "loss": 1.195,
       "step": 121390
     },
     {
       "epoch": 1.9723481340676838,
-      "grad_norm": 2.390625,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2134,
       "step": 121400
     },
     {
       "epoch": 1.9725106009650535,
-      "grad_norm": 2.359375,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2063,
       "step": 121410
     },
     {
       "epoch": 1.9726730678624231,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3317,
+      "loss": 1.2162,
       "step": 121420
     },
     {
       "epoch": 1.9728355347597928,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.1841,
       "step": 121430
     },
     {
       "epoch": 1.9729980016571624,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3344,
+      "loss": 1.2652,
       "step": 121440
     },
     {
       "epoch": 1.973160468554532,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.2559,
       "step": 121450
     },
     {
       "epoch": 1.9733229354519017,
-      "grad_norm": 2.890625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.344,
+      "loss": 1.1958,
       "step": 121460
     },
     {
       "epoch": 1.9734854023492714,
-      "grad_norm": 4.46875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.2547,
       "step": 121470
     },
     {
       "epoch": 1.973647869246641,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3312,
+      "loss": 1.2564,
       "step": 121480
     },
     {
       "epoch": 1.9738103361440107,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3416,
+      "loss": 1.2103,
       "step": 121490
     },
     {
       "epoch": 1.9739728030413803,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 1.2024,
       "step": 121500
     },
     {
       "epoch": 1.97413526993875,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.347,
+      "loss": 1.2262,
       "step": 121510
     },
     {
       "epoch": 1.9742977368361196,
-      "grad_norm": 2.78125,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2207,
       "step": 121520
     },
     {
       "epoch": 1.9744602037334893,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.325,
+      "loss": 1.2595,
       "step": 121530
     },
     {
       "epoch": 1.974622670630859,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3509,
+      "loss": 1.241,
       "step": 121540
     },
     {
       "epoch": 1.9747851375282286,
-      "grad_norm": 2.703125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.2791,
       "step": 121550
     },
     {
       "epoch": 1.9749476044255982,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3351,
+      "loss": 1.2342,
       "step": 121560
     },
     {
       "epoch": 1.975110071322968,
-      "grad_norm": 2.609375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.2232,
       "step": 121570
     },
     {
       "epoch": 1.9752725382203375,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3302,
+      "loss": 1.2282,
       "step": 121580
     },
     {
       "epoch": 1.9754350051177072,
-      "grad_norm": 2.125,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3427,
+      "loss": 1.2111,
       "step": 121590
     },
     {
       "epoch": 1.9755974720150768,
-      "grad_norm": 3.046875,
+      "grad_norm": 7.5,
       "learning_rate": 5e-05,
-      "loss": 0.3673,
+      "loss": 1.2528,
       "step": 121600
     },
     {
       "epoch": 1.9757599389124465,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3392,
+      "loss": 1.2263,
       "step": 121610
     },
     {
       "epoch": 1.9759224058098162,
-      "grad_norm": 2.25,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.1856,
       "step": 121620
     },
     {
       "epoch": 1.9760848727071858,
-      "grad_norm": 4.21875,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2086,
       "step": 121630
     },
     {
       "epoch": 1.9762473396045555,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3443,
+      "loss": 1.2686,
       "step": 121640
     },
     {
       "epoch": 1.976409806501925,
-      "grad_norm": 4.9375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3546,
+      "loss": 1.1999,
       "step": 121650
     },
     {
       "epoch": 1.9765722733992948,
-      "grad_norm": 2.828125,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3436,
+      "loss": 1.2994,
       "step": 121660
     },
     {
       "epoch": 1.9767347402966644,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3431,
+      "loss": 1.2207,
       "step": 121670
     },
     {
       "epoch": 1.976897207194034,
-      "grad_norm": 3.78125,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3602,
+      "loss": 1.2006,
       "step": 121680
     },
     {
       "epoch": 1.977059674091404,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3389,
+      "loss": 1.2196,
       "step": 121690
     },
     {
       "epoch": 1.9772221409887736,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3564,
+      "loss": 1.2185,
       "step": 121700
     },
     {
       "epoch": 1.9773846078861432,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3311,
+      "loss": 1.2033,
       "step": 121710
     },
     {
       "epoch": 1.9775470747835129,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.75,
       "learning_rate": 5e-05,
-      "loss": 0.337,
+      "loss": 1.2117,
       "step": 121720
     },
     {
       "epoch": 1.9777095416808825,
-      "grad_norm": 3.46875,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2117,
       "step": 121730
     },
     {
       "epoch": 1.9778720085782522,
-      "grad_norm": 2.875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.3502,
+      "loss": 1.2319,
       "step": 121740
     },
     {
       "epoch": 1.9780344754756218,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.2193,
       "step": 121750
     },
     {
       "epoch": 1.9781969423729915,
-      "grad_norm": 2.59375,
+      "grad_norm": 6.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3334,
+      "loss": 1.2621,
       "step": 121760
     },
     {
       "epoch": 1.9783594092703611,
-      "grad_norm": 2.15625,
+      "grad_norm": 6.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2328,
       "step": 121770
     },
     {
       "epoch": 1.9785218761677308,
-      "grad_norm": 2.34375,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.355,
+      "loss": 1.2366,
       "step": 121780
     },
     {
       "epoch": 1.9786843430651004,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.625,
       "learning_rate": 5e-05,
-      "loss": 0.3207,
+      "loss": 1.2194,
       "step": 121790
     },
     {
       "epoch": 1.9788468099624703,
-      "grad_norm": 2.546875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3566,
+      "loss": 1.2361,
       "step": 121800
     },
     {
       "epoch": 1.97900927685984,
-      "grad_norm": 3.640625,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2237,
       "step": 121810
     },
     {
       "epoch": 1.9791717437572096,
-      "grad_norm": 2.71875,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.2149,
       "step": 121820
     },
     {
       "epoch": 1.9793342106545793,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3146,
+      "loss": 1.221,
       "step": 121830
     },
     {
       "epoch": 1.979496677551949,
-      "grad_norm": 2.515625,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.338,
+      "loss": 1.2133,
       "step": 121840
     },
     {
       "epoch": 1.9796591444493186,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3307,
+      "loss": 1.2024,
       "step": 121850
     },
     {
       "epoch": 1.9798216113466882,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.3575,
+      "loss": 1.1574,
       "step": 121860
     },
     {
       "epoch": 1.9799840782440579,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3343,
+      "loss": 1.2301,
       "step": 121870
     },
     {
       "epoch": 1.9801465451414275,
-      "grad_norm": 4.625,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3466,
+      "loss": 1.2288,
       "step": 121880
     },
     {
       "epoch": 1.9803090120387972,
-      "grad_norm": 3.171875,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3323,
+      "loss": 1.2093,
       "step": 121890
     },
     {
       "epoch": 1.9804714789361668,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3522,
+      "loss": 1.1769,
       "step": 121900
     },
     {
       "epoch": 1.9806339458335365,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3492,
+      "loss": 1.1973,
       "step": 121910
     },
     {
       "epoch": 1.9807964127309061,
-      "grad_norm": 2.90625,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3571,
+      "loss": 1.2306,
       "step": 121920
     },
     {
       "epoch": 1.9809588796282758,
-      "grad_norm": 3.359375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2414,
       "step": 121930
     },
     {
       "epoch": 1.9811213465256454,
-      "grad_norm": 2.515625,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.36,
+      "loss": 1.2249,
       "step": 121940
     },
     {
       "epoch": 1.981283813423015,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.1615,
       "step": 121950
     },
     {
       "epoch": 1.9814462803203847,
-      "grad_norm": 3.53125,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3327,
+      "loss": 1.2262,
       "step": 121960
     },
     {
       "epoch": 1.9816087472177544,
-      "grad_norm": 3.78125,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3339,
+      "loss": 1.2185,
       "step": 121970
     },
     {
       "epoch": 1.981771214115124,
-      "grad_norm": 2.171875,
+      "grad_norm": 6.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3435,
+      "loss": 1.2208,
       "step": 121980
     },
     {
       "epoch": 1.9819336810124937,
-      "grad_norm": 3.359375,
+      "grad_norm": 5.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3539,
+      "loss": 1.2407,
       "step": 121990
     },
     {
       "epoch": 1.9820961479098633,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3394,
+      "loss": 1.2259,
       "step": 122000
     },
     {
       "epoch": 1.982258614807233,
-      "grad_norm": 3.015625,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.1911,
       "step": 122010
     },
     {
       "epoch": 1.9824210817046026,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.354,
+      "loss": 1.2451,
       "step": 122020
     },
     {
       "epoch": 1.9825835486019723,
-      "grad_norm": 2.34375,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3437,
+      "loss": 1.2235,
       "step": 122030
     },
     {
       "epoch": 1.982746015499342,
-      "grad_norm": 2.75,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3278,
+      "loss": 1.1983,
       "step": 122040
     },
     {
       "epoch": 1.9829084823967116,
-      "grad_norm": 3.609375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3269,
+      "loss": 1.2439,
       "step": 122050
     },
     {
       "epoch": 1.9830709492940812,
-      "grad_norm": 2.5625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3331,
+      "loss": 1.1988,
       "step": 122060
     },
     {
       "epoch": 1.9832334161914509,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2123,
       "step": 122070
     },
     {
       "epoch": 1.9833958830888205,
-      "grad_norm": 2.796875,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3572,
+      "loss": 1.2116,
       "step": 122080
     },
     {
       "epoch": 1.9835583499861902,
-      "grad_norm": 2.6875,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3621,
+      "loss": 1.2345,
       "step": 122090
     },
     {
       "epoch": 1.9837208168835598,
-      "grad_norm": 3.453125,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.238,
       "step": 122100
     },
     {
       "epoch": 1.9838832837809295,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3429,
+      "loss": 1.2444,
       "step": 122110
     },
     {
       "epoch": 1.9840457506782991,
-      "grad_norm": 3.234375,
+      "grad_norm": 6.75,
       "learning_rate": 5e-05,
-      "loss": 0.3363,
+      "loss": 1.2177,
       "step": 122120
     },
     {
       "epoch": 1.984208217575669,
-      "grad_norm": 2.265625,
+      "grad_norm": 7.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.1979,
       "step": 122130
     },
     {
       "epoch": 1.9843706844730387,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3472,
+      "loss": 1.1753,
       "step": 122140
     },
     {
       "epoch": 1.9845331513704083,
-      "grad_norm": 2.109375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3444,
+      "loss": 1.2241,
       "step": 122150
     },
     {
       "epoch": 1.984695618267778,
-      "grad_norm": 2.40625,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3548,
+      "loss": 1.2376,
       "step": 122160
     },
     {
       "epoch": 1.9848580851651476,
-      "grad_norm": 2.78125,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3475,
+      "loss": 1.2383,
       "step": 122170
     },
     {
       "epoch": 1.9850205520625173,
-      "grad_norm": 2.4375,
+      "grad_norm": 6.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3236,
+      "loss": 1.1975,
       "step": 122180
     },
     {
       "epoch": 1.985183018959887,
-      "grad_norm": 2.453125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3119,
+      "loss": 1.2089,
       "step": 122190
     },
     {
       "epoch": 1.9853454858572566,
-      "grad_norm": 2.8125,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3355,
+      "loss": 1.2473,
       "step": 122200
     },
     {
       "epoch": 1.9855079527546262,
-      "grad_norm": 3.34375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3489,
+      "loss": 1.1707,
       "step": 122210
     },
     {
       "epoch": 1.9856704196519959,
-      "grad_norm": 2.0625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2106,
       "step": 122220
     },
     {
       "epoch": 1.9858328865493655,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3399,
+      "loss": 1.2078,
       "step": 122230
     },
     {
       "epoch": 1.9859953534467354,
-      "grad_norm": 3.5625,
+      "grad_norm": 6.09375,
       "learning_rate": 5e-05,
-      "loss": 0.357,
+      "loss": 1.198,
       "step": 122240
     },
     {
       "epoch": 1.986157820344105,
-      "grad_norm": 3.109375,
+      "grad_norm": 6.875,
       "learning_rate": 5e-05,
-      "loss": 0.3292,
+      "loss": 1.2272,
       "step": 122250
     },
     {
       "epoch": 1.9863202872414747,
-      "grad_norm": 3.65625,
+      "grad_norm": 7.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3485,
+      "loss": 1.195,
       "step": 122260
     },
     {
       "epoch": 1.9864827541388443,
-      "grad_norm": 2.40625,
+      "grad_norm": 5.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3333,
+      "loss": 1.2633,
       "step": 122270
     },
     {
       "epoch": 1.986645221036214,
-      "grad_norm": 3.8125,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3529,
+      "loss": 1.1986,
       "step": 122280
     },
     {
       "epoch": 1.9868076879335836,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 1.2446,
       "step": 122290
     },
     {
       "epoch": 1.9869701548309533,
-      "grad_norm": 2.828125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3423,
+      "loss": 1.1957,
       "step": 122300
     },
     {
       "epoch": 1.987132621728323,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.349,
+      "loss": 1.2498,
       "step": 122310
     },
     {
       "epoch": 1.9872950886256926,
-      "grad_norm": 3.21875,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 1.2333,
       "step": 122320
     },
     {
       "epoch": 1.9874575555230622,
-      "grad_norm": 2.21875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3325,
+      "loss": 1.2545,
       "step": 122330
     },
     {
       "epoch": 1.987620022420432,
-      "grad_norm": 2.4375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.335,
+      "loss": 1.1767,
       "step": 122340
     },
     {
       "epoch": 1.9877824893178015,
-      "grad_norm": 3.125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3097,
+      "loss": 1.2208,
       "step": 122350
     },
     {
       "epoch": 1.9879449562151712,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3404,
+      "loss": 1.1891,
       "step": 122360
     },
     {
       "epoch": 1.9881074231125409,
-      "grad_norm": 3.0625,
+      "grad_norm": 7.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3242,
+      "loss": 1.2148,
       "step": 122370
     },
     {
       "epoch": 1.9882698900099105,
-      "grad_norm": 2.3125,
+      "grad_norm": 6.375,
       "learning_rate": 5e-05,
-      "loss": 0.3412,
+      "loss": 1.2126,
       "step": 122380
     },
     {
       "epoch": 1.9884323569072802,
-      "grad_norm": 2.703125,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3483,
+      "loss": 1.2259,
       "step": 122390
     },
     {
       "epoch": 1.9885948238046498,
-      "grad_norm": 2.765625,
+      "grad_norm": 6.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3382,
+      "loss": 1.1892,
       "step": 122400
     },
     {
       "epoch": 1.9887572907020195,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2419,
       "step": 122410
     },
     {
       "epoch": 1.988919757599389,
-      "grad_norm": 3.734375,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.223,
       "step": 122420
     },
     {
       "epoch": 1.9890822244967588,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.2211,
       "step": 122430
     },
     {
       "epoch": 1.9892446913941284,
-      "grad_norm": 2.953125,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2233,
       "step": 122440
     },
     {
       "epoch": 1.989407158291498,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.25,
       "learning_rate": 5e-05,
-      "loss": 0.3474,
+      "loss": 1.2247,
       "step": 122450
     },
     {
       "epoch": 1.9895696251888677,
-      "grad_norm": 2.75,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3384,
+      "loss": 1.2475,
       "step": 122460
     },
     {
       "epoch": 1.9897320920862374,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3409,
+      "loss": 1.2278,
       "step": 122470
     },
     {
       "epoch": 1.989894558983607,
-      "grad_norm": 2.53125,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3434,
+      "loss": 1.1921,
       "step": 122480
     },
     {
       "epoch": 1.9900570258809767,
-      "grad_norm": 2.75,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3543,
+      "loss": 1.1955,
       "step": 122490
     },
     {
       "epoch": 1.9902194927783463,
-      "grad_norm": 2.8125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3362,
+      "loss": 1.2003,
       "step": 122500
     },
     {
       "epoch": 1.990381959675716,
-      "grad_norm": 2.3125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3496,
+      "loss": 1.1926,
       "step": 122510
     },
     {
       "epoch": 1.9905444265730856,
-      "grad_norm": 3.140625,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3608,
+      "loss": 1.2621,
       "step": 122520
     },
     {
       "epoch": 1.9907068934704553,
-      "grad_norm": 4.0,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3356,
+      "loss": 1.1969,
       "step": 122530
     },
     {
       "epoch": 1.990869360367825,
-      "grad_norm": 3.15625,
+      "grad_norm": 6.34375,
       "learning_rate": 5e-05,
-      "loss": 0.348,
+      "loss": 1.2446,
       "step": 122540
     },
     {
       "epoch": 1.9910318272651946,
-      "grad_norm": 2.15625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3542,
+      "loss": 1.2546,
       "step": 122550
     },
     {
       "epoch": 1.9911942941625642,
-      "grad_norm": 3.390625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2517,
       "step": 122560
     },
     {
       "epoch": 1.991356761059934,
-      "grad_norm": 3.328125,
+      "grad_norm": 7.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3314,
+      "loss": 1.209,
       "step": 122570
     },
     {
       "epoch": 1.9915192279573037,
-      "grad_norm": 2.90625,
+      "grad_norm": 7.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2607,
       "step": 122580
     },
     {
       "epoch": 1.9916816948546734,
-      "grad_norm": 3.734375,
+      "grad_norm": 6.0,
       "learning_rate": 5e-05,
-      "loss": 0.3234,
+      "loss": 1.2297,
       "step": 122590
     },
     {
       "epoch": 1.991844161752043,
-      "grad_norm": 4.25,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3462,
+      "loss": 1.2048,
       "step": 122600
     },
     {
       "epoch": 1.9920066286494127,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3604,
+      "loss": 1.2044,
       "step": 122610
     },
     {
       "epoch": 1.9921690955467823,
-      "grad_norm": 2.796875,
+      "grad_norm": 5.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3383,
+      "loss": 1.2228,
       "step": 122620
     },
     {
       "epoch": 1.992331562444152,
-      "grad_norm": 3.40625,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3671,
+      "loss": 1.2055,
       "step": 122630
     },
     {
       "epoch": 1.9924940293415216,
-      "grad_norm": 3.1875,
+      "grad_norm": 6.65625,
       "learning_rate": 5e-05,
-      "loss": 0.3514,
+      "loss": 1.1977,
       "step": 122640
     },
     {
       "epoch": 1.9926564962388913,
-      "grad_norm": 2.984375,
+      "grad_norm": 7.84375,
       "learning_rate": 5e-05,
-      "loss": 0.3397,
+      "loss": 1.2364,
       "step": 122650
     },
     {
       "epoch": 1.992818963136261,
-      "grad_norm": 4.4375,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3403,
+      "loss": 1.2513,
       "step": 122660
     },
     {
       "epoch": 1.9929814300336306,
-      "grad_norm": 2.9375,
+      "grad_norm": 6.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3304,
+      "loss": 1.2278,
       "step": 122670
     },
     {
       "epoch": 1.9931438969310005,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.343,
+      "loss": 1.1892,
       "step": 122680
     },
     {
       "epoch": 1.9933063638283701,
-      "grad_norm": 2.953125,
+      "grad_norm": 7.09375,
       "learning_rate": 5e-05,
-      "loss": 0.3291,
+      "loss": 1.2261,
       "step": 122690
     },
     {
       "epoch": 1.9934688307257398,
-      "grad_norm": 2.609375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3599,
+      "loss": 1.2187,
       "step": 122700
     },
     {
       "epoch": 1.9936312976231094,
-      "grad_norm": 2.328125,
+      "grad_norm": 5.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3467,
+      "loss": 1.271,
       "step": 122710
     },
     {
       "epoch": 1.993793764520479,
-      "grad_norm": 4.15625,
+      "grad_norm": 6.59375,
       "learning_rate": 5e-05,
-      "loss": 0.345,
+      "loss": 1.2405,
       "step": 122720
     },
     {
       "epoch": 1.9939562314178487,
-      "grad_norm": 3.0625,
+      "grad_norm": 6.78125,
       "learning_rate": 5e-05,
-      "loss": 0.35,
+      "loss": 1.2277,
       "step": 122730
     },
     {
       "epoch": 1.9941186983152184,
-      "grad_norm": 2.859375,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3471,
+      "loss": 1.1957,
       "step": 122740
     },
     {
       "epoch": 1.994281165212588,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3381,
+      "loss": 1.2531,
       "step": 122750
     },
     {
       "epoch": 1.9944436321099577,
-      "grad_norm": 2.8125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3244,
+      "loss": 1.1705,
       "step": 122760
     },
     {
       "epoch": 1.9946060990073273,
-      "grad_norm": 3.359375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3378,
+      "loss": 1.2464,
       "step": 122770
     },
     {
       "epoch": 1.994768565904697,
-      "grad_norm": 2.625,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3583,
+      "loss": 1.2069,
       "step": 122780
     },
     {
       "epoch": 1.9949310328020666,
-      "grad_norm": 2.96875,
+      "grad_norm": 6.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3532,
+      "loss": 1.258,
       "step": 122790
     },
     {
       "epoch": 1.9950934996994363,
-      "grad_norm": 3.28125,
+      "grad_norm": 6.21875,
       "learning_rate": 5e-05,
-      "loss": 0.3667,
+      "loss": 1.2394,
       "step": 122800
     },
     {
       "epoch": 1.995255966596806,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3735,
+      "loss": 1.2167,
       "step": 122810
     },
     {
       "epoch": 1.9954184334941756,
-      "grad_norm": 2.875,
+      "grad_norm": 5.875,
       "learning_rate": 5e-05,
-      "loss": 0.3225,
+      "loss": 1.2018,
       "step": 122820
     },
     {
       "epoch": 1.9955809003915452,
-      "grad_norm": 3.640625,
+      "grad_norm": 6.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3511,
+      "loss": 1.2283,
       "step": 122830
     },
     {
       "epoch": 1.9957433672889149,
-      "grad_norm": 3.25,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3488,
+      "loss": 1.2304,
       "step": 122840
     },
     {
       "epoch": 1.9959058341862845,
-      "grad_norm": 3.0,
+      "grad_norm": 6.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3452,
+      "loss": 1.2181,
       "step": 122850
     },
     {
       "epoch": 1.9960683010836542,
-      "grad_norm": 2.734375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3347,
+      "loss": 1.2175,
       "step": 122860
     },
     {
       "epoch": 1.9962307679810238,
-      "grad_norm": 2.46875,
+      "grad_norm": 7.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3176,
+      "loss": 1.241,
       "step": 122870
     },
     {
       "epoch": 1.9963932348783935,
-      "grad_norm": 3.671875,
+      "grad_norm": 6.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3569,
+      "loss": 1.191,
       "step": 122880
     },
     {
       "epoch": 1.9965557017757631,
-      "grad_norm": 2.625,
+      "grad_norm": 6.71875,
       "learning_rate": 5e-05,
-      "loss": 0.3297,
+      "loss": 1.1856,
       "step": 122890
     },
     {
       "epoch": 1.9967181686731328,
-      "grad_norm": 2.640625,
+      "grad_norm": 7.375,
       "learning_rate": 5e-05,
-      "loss": 0.3388,
+      "loss": 1.2115,
       "step": 122900
     },
     {
       "epoch": 1.9968806355705024,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.0,
       "learning_rate": 5e-05,
-      "loss": 0.3501,
+      "loss": 1.235,
       "step": 122910
     },
     {
       "epoch": 1.997043102467872,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.1994,
       "step": 122920
     },
     {
       "epoch": 1.9972055693652417,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3775,
+      "loss": 1.1917,
       "step": 122930
     },
     {
       "epoch": 1.9973680362626114,
-      "grad_norm": 2.421875,
+      "grad_norm": 6.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3591,
+      "loss": 1.2106,
       "step": 122940
     },
     {
       "epoch": 1.997530503159981,
-      "grad_norm": 3.5,
+      "grad_norm": 7.40625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.1974,
       "step": 122950
     },
     {
       "epoch": 1.9976929700573507,
-      "grad_norm": 3.25,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3508,
+      "loss": 1.2205,
       "step": 122960
     },
     {
       "epoch": 1.9978554369547203,
-      "grad_norm": 2.25,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3497,
+      "loss": 1.1866,
       "step": 122970
     },
     {
       "epoch": 1.99801790385209,
-      "grad_norm": 2.9375,
+      "grad_norm": 7.28125,
       "learning_rate": 5e-05,
-      "loss": 0.3479,
+      "loss": 1.2389,
       "step": 122980
     },
     {
       "epoch": 1.9981803707494596,
-      "grad_norm": 3.125,
+      "grad_norm": 7.125,
       "learning_rate": 5e-05,
-      "loss": 0.3573,
+      "loss": 1.2122,
       "step": 122990
     },
     {
       "epoch": 1.9983428376468293,
-      "grad_norm": 2.953125,
+      "grad_norm": 5.78125,
       "learning_rate": 5e-05,
-      "loss": 0.3426,
+      "loss": 1.2094,
       "step": 123000
     },
     {
       "epoch": 1.9985053045441992,
-      "grad_norm": 3.34375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3465,
+      "loss": 1.1867,
       "step": 123010
     },
     {
       "epoch": 1.9986677714415688,
-      "grad_norm": 2.671875,
+      "grad_norm": 6.5,
       "learning_rate": 5e-05,
-      "loss": 0.3421,
+      "loss": 1.2047,
       "step": 123020
     },
     {
       "epoch": 1.9988302383389385,
-      "grad_norm": 3.046875,
+      "grad_norm": 5.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3379,
+      "loss": 1.2293,
       "step": 123030
     },
     {
       "epoch": 1.9989927052363081,
-      "grad_norm": 2.59375,
+      "grad_norm": 7.03125,
       "learning_rate": 5e-05,
-      "loss": 0.3122,
+      "loss": 1.2448,
       "step": 123040
     },
     {
       "epoch": 1.9991551721336778,
-      "grad_norm": 2.390625,
+      "grad_norm": 5.96875,
       "learning_rate": 5e-05,
-      "loss": 0.3556,
+      "loss": 1.1972,
       "step": 123050
     },
     {
       "epoch": 1.9993176390310474,
-      "grad_norm": 2.578125,
+      "grad_norm": 6.125,
       "learning_rate": 5e-05,
-      "loss": 0.3482,
+      "loss": 1.1707,
       "step": 123060
     },
     {
       "epoch": 1.999480105928417,
-      "grad_norm": 1.96875,
+      "grad_norm": 6.15625,
       "learning_rate": 5e-05,
-      "loss": 0.3358,
+      "loss": 1.2293,
       "step": 123070
     },
     {
       "epoch": 1.9996425728257867,
-      "grad_norm": 2.546875,
+      "grad_norm": 6.53125,
       "learning_rate": 5e-05,
-      "loss": 0.3484,
+      "loss": 1.2156,
       "step": 123080
     },
     {
       "epoch": 1.9998050397231564,
-      "grad_norm": 2.65625,
+      "grad_norm": 6.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3439,
+      "loss": 1.2181,
       "step": 123090
     },
     {
       "epoch": 1.999967506620526,
-      "grad_norm": 3.078125,
+      "grad_norm": 6.90625,
       "learning_rate": 5e-05,
-      "loss": 0.3313,
+      "loss": 1.1862,
       "step": 123100
     },
     {
       "epoch": 2.000129973517896,
-      "grad_norm": 2.8125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3299,
+      "loss": 0.986,
       "step": 123110
     },
     {
       "epoch": 2.0002924404152655,
-      "grad_norm": 2.859375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.3191,
+      "loss": 0.9926,
       "step": 123120
     },
     {
       "epoch": 2.000454907312635,
-      "grad_norm": 2.296875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3357,
+      "loss": 0.9411,
       "step": 123130
     },
     {
       "epoch": 2.000617374210005,
-      "grad_norm": 2.421875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3156,
+      "loss": 0.9683,
       "step": 123140
     },
     {
       "epoch": 2.0007798411073745,
-      "grad_norm": 4.5,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.3131,
+      "loss": 0.9599,
       "step": 123150
     },
     {
       "epoch": 2.000942308004744,
-      "grad_norm": 2.359375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3128,
+      "loss": 0.9473,
       "step": 123160
     },
     {
       "epoch": 2.001104774902114,
-      "grad_norm": 2.859375,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.309,
+      "loss": 0.9799,
       "step": 123170
     },
     {
       "epoch": 2.0012672417994835,
-      "grad_norm": 2.71875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.3032,
+      "loss": 0.9407,
       "step": 123180
     },
     {
       "epoch": 2.001429708696853,
-      "grad_norm": 2.65625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3095,
+      "loss": 0.9405,
       "step": 123190
     },
     {
       "epoch": 2.0015921755942228,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3203,
+      "loss": 0.9215,
       "step": 123200
     },
     {
       "epoch": 2.0017546424915924,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3156,
+      "loss": 0.9208,
       "step": 123210
     },
     {
       "epoch": 2.001917109388962,
-      "grad_norm": 1.9765625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3167,
+      "loss": 0.9159,
       "step": 123220
     },
     {
       "epoch": 2.0020795762863317,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3146,
+      "loss": 0.9272,
       "step": 123230
     },
     {
       "epoch": 2.0022420431837014,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.3262,
+      "loss": 0.9459,
       "step": 123240
     },
     {
       "epoch": 2.002404510081071,
-      "grad_norm": 2.625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3192,
+      "loss": 0.9214,
       "step": 123250
     },
     {
       "epoch": 2.0025669769784407,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.3103,
+      "loss": 0.9385,
       "step": 123260
     },
     {
       "epoch": 2.0027294438758103,
-      "grad_norm": 2.125,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.308,
+      "loss": 0.9579,
       "step": 123270
     },
     {
       "epoch": 2.00289191077318,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3361,
+      "loss": 0.9141,
       "step": 123280
     },
     {
       "epoch": 2.0030543776705496,
-      "grad_norm": 2.078125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2984,
+      "loss": 0.9319,
       "step": 123290
     },
     {
       "epoch": 2.0032168445679193,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3364,
+      "loss": 0.9356,
       "step": 123300
     },
     {
       "epoch": 2.003379311465289,
-      "grad_norm": 2.171875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3147,
+      "loss": 0.9472,
       "step": 123310
     },
     {
       "epoch": 2.0035417783626586,
-      "grad_norm": 3.890625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3059,
+      "loss": 0.9419,
       "step": 123320
     },
     {
       "epoch": 2.003704245260028,
-      "grad_norm": 2.546875,
+      "grad_norm": 7.46875,
       "learning_rate": 5e-05,
-      "loss": 0.3042,
+      "loss": 0.9418,
       "step": 123330
     },
     {
       "epoch": 2.003866712157398,
-      "grad_norm": 2.390625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3254,
+      "loss": 0.9682,
       "step": 123340
     },
     {
       "epoch": 2.0040291790547675,
-      "grad_norm": 3.171875,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3162,
+      "loss": 0.9258,
       "step": 123350
     },
     {
       "epoch": 2.004191645952137,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.339,
+      "loss": 0.9513,
       "step": 123360
     },
     {
       "epoch": 2.004354112849507,
-      "grad_norm": 2.71875,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3196,
+      "loss": 0.963,
       "step": 123370
     },
     {
       "epoch": 2.0045165797468765,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.3159,
+      "loss": 0.9603,
       "step": 123380
     },
     {
       "epoch": 2.004679046644246,
-      "grad_norm": 2.984375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3346,
+      "loss": 0.9533,
       "step": 123390
     },
     {
       "epoch": 2.0048415135416158,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3216,
+      "loss": 0.9507,
       "step": 123400
     },
     {
       "epoch": 2.0050039804389854,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.334,
+      "loss": 0.9099,
       "step": 123410
     },
     {
       "epoch": 2.005166447336355,
-      "grad_norm": 2.609375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3137,
+      "loss": 0.9324,
       "step": 123420
     },
     {
       "epoch": 2.0053289142337247,
-      "grad_norm": 2.953125,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2952,
+      "loss": 0.9388,
       "step": 123430
     },
     {
       "epoch": 2.0054913811310944,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3074,
+      "loss": 0.9252,
       "step": 123440
     },
     {
       "epoch": 2.005653848028464,
-      "grad_norm": 4.34375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.3127,
+      "loss": 0.9024,
       "step": 123450
     },
     {
       "epoch": 2.0058163149258337,
-      "grad_norm": 2.296875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3124,
+      "loss": 0.9313,
       "step": 123460
     },
     {
       "epoch": 2.0059787818232033,
-      "grad_norm": 2.265625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.3026,
+      "loss": 0.9566,
       "step": 123470
     },
     {
       "epoch": 2.006141248720573,
-      "grad_norm": 3.265625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2997,
+      "loss": 0.9374,
       "step": 123480
     },
     {
       "epoch": 2.0063037156179426,
-      "grad_norm": 2.34375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3024,
+      "loss": 0.9359,
       "step": 123490
     },
     {
       "epoch": 2.0064661825153127,
-      "grad_norm": 3.421875,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2939,
+      "loss": 0.968,
       "step": 123500
     },
     {
       "epoch": 2.0066286494126824,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3108,
+      "loss": 0.9294,
       "step": 123510
     },
     {
       "epoch": 2.006791116310052,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2964,
+      "loss": 0.905,
       "step": 123520
     },
     {
       "epoch": 2.0069535832074217,
-      "grad_norm": 3.140625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2902,
+      "loss": 0.9403,
       "step": 123530
     },
     {
       "epoch": 2.0071160501047913,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.3189,
+      "loss": 0.9552,
       "step": 123540
     },
     {
       "epoch": 2.007278517002161,
-      "grad_norm": 2.265625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2875,
+      "loss": 0.9386,
       "step": 123550
     },
     {
       "epoch": 2.0074409838995306,
-      "grad_norm": 3.25,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2925,
+      "loss": 0.9315,
       "step": 123560
     },
     {
       "epoch": 2.0076034507969003,
-      "grad_norm": 3.03125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2989,
+      "loss": 0.953,
       "step": 123570
     },
     {
       "epoch": 2.00776591769427,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.3059,
+      "loss": 0.9392,
       "step": 123580
     },
     {
       "epoch": 2.0079283845916396,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3139,
+      "loss": 0.912,
       "step": 123590
     },
     {
       "epoch": 2.0080908514890092,
-      "grad_norm": 2.6875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3157,
+      "loss": 0.946,
       "step": 123600
     },
     {
       "epoch": 2.008253318386379,
-      "grad_norm": 2.328125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2993,
+      "loss": 0.9296,
       "step": 123610
     },
     {
       "epoch": 2.0084157852837485,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2989,
+      "loss": 0.9502,
       "step": 123620
     },
     {
       "epoch": 2.008578252181118,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.3149,
+      "loss": 0.9346,
       "step": 123630
     },
     {
       "epoch": 2.008740719078488,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.3052,
+      "loss": 0.9797,
       "step": 123640
     },
     {
       "epoch": 2.0089031859758575,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3073,
+      "loss": 0.9549,
       "step": 123650
     },
     {
       "epoch": 2.009065652873227,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2916,
+      "loss": 0.9562,
       "step": 123660
     },
     {
       "epoch": 2.009228119770597,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2985,
+      "loss": 0.943,
       "step": 123670
     },
     {
       "epoch": 2.0093905866679664,
-      "grad_norm": 2.671875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2887,
+      "loss": 0.925,
       "step": 123680
     },
     {
       "epoch": 2.009553053565336,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2973,
+      "loss": 0.9134,
       "step": 123690
     },
     {
       "epoch": 2.0097155204627057,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2887,
+      "loss": 0.9487,
       "step": 123700
     },
     {
       "epoch": 2.0098779873600754,
-      "grad_norm": 2.609375,
+      "grad_norm": 13.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3144,
+      "loss": 0.9218,
       "step": 123710
     },
     {
       "epoch": 2.010040454257445,
-      "grad_norm": 2.78125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.3067,
+      "loss": 0.9427,
       "step": 123720
     },
     {
       "epoch": 2.0102029211548147,
-      "grad_norm": 2.171875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.294,
+      "loss": 0.9803,
       "step": 123730
     },
     {
       "epoch": 2.0103653880521843,
-      "grad_norm": 3.875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2796,
+      "loss": 0.9455,
       "step": 123740
     },
     {
       "epoch": 2.010527854949554,
-      "grad_norm": 3.25,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.29,
+      "loss": 0.9207,
       "step": 123750
     },
     {
       "epoch": 2.0106903218469236,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2941,
+      "loss": 0.9113,
       "step": 123760
     },
     {
       "epoch": 2.0108527887442933,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.284,
+      "loss": 0.9579,
       "step": 123770
     },
     {
       "epoch": 2.011015255641663,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.3093,
+      "loss": 0.9253,
       "step": 123780
     },
     {
       "epoch": 2.0111777225390326,
-      "grad_norm": 2.296875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2877,
+      "loss": 0.9316,
       "step": 123790
     },
     {
       "epoch": 2.0113401894364022,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.3088,
+      "loss": 0.9314,
       "step": 123800
     },
     {
       "epoch": 2.011502656333772,
-      "grad_norm": 3.25,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2772,
+      "loss": 0.9032,
       "step": 123810
     },
     {
       "epoch": 2.0116651232311415,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2872,
+      "loss": 0.9352,
       "step": 123820
     },
     {
       "epoch": 2.011827590128511,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2809,
+      "loss": 0.9062,
       "step": 123830
     },
     {
       "epoch": 2.011990057025881,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2942,
+      "loss": 0.9138,
       "step": 123840
     },
     {
       "epoch": 2.0121525239232505,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2833,
+      "loss": 0.9327,
       "step": 123850
     },
     {
       "epoch": 2.01231499082062,
-      "grad_norm": 3.484375,
+      "grad_norm": 12.875,
       "learning_rate": 5e-05,
-      "loss": 0.2961,
+      "loss": 0.9206,
       "step": 123860
     },
     {
       "epoch": 2.01247745771799,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 0.944,
       "step": 123870
     },
     {
       "epoch": 2.0126399246153595,
-      "grad_norm": 4.0,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2945,
+      "loss": 0.9064,
       "step": 123880
     },
     {
       "epoch": 2.012802391512729,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2907,
+      "loss": 0.9302,
       "step": 123890
     },
     {
       "epoch": 2.0129648584100988,
-      "grad_norm": 2.046875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2887,
+      "loss": 0.9179,
       "step": 123900
     },
     {
       "epoch": 2.0131273253074684,
-      "grad_norm": 3.25,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3028,
+      "loss": 0.9374,
       "step": 123910
     },
     {
       "epoch": 2.013289792204838,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2844,
+      "loss": 0.978,
       "step": 123920
     },
     {
       "epoch": 2.0134522591022077,
-      "grad_norm": 2.34375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2903,
+      "loss": 0.9215,
       "step": 123930
     },
     {
       "epoch": 2.013614725999578,
-      "grad_norm": 3.109375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2982,
+      "loss": 0.9802,
       "step": 123940
     },
     {
       "epoch": 2.0137771928969475,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2877,
+      "loss": 0.956,
       "step": 123950
     },
     {
       "epoch": 2.013939659794317,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2998,
+      "loss": 0.9251,
       "step": 123960
     },
     {
       "epoch": 2.0141021266916868,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.3091,
+      "loss": 0.9086,
       "step": 123970
     },
     {
       "epoch": 2.0142645935890564,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.299,
+      "loss": 0.9043,
       "step": 123980
     },
     {
       "epoch": 2.014427060486426,
-      "grad_norm": 2.453125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2922,
+      "loss": 0.9311,
       "step": 123990
     },
     {
       "epoch": 2.0145895273837957,
-      "grad_norm": 2.40625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.282,
+      "loss": 0.9512,
       "step": 124000
     },
     {
       "epoch": 2.0147519942811654,
-      "grad_norm": 3.1875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2934,
+      "loss": 0.9645,
       "step": 124010
     },
     {
       "epoch": 2.014914461178535,
-      "grad_norm": 2.75,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2961,
+      "loss": 0.9554,
       "step": 124020
     },
     {
       "epoch": 2.0150769280759047,
-      "grad_norm": 2.921875,
+      "grad_norm": 7.25,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9343,
       "step": 124030
     },
     {
       "epoch": 2.0152393949732743,
-      "grad_norm": 2.859375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9608,
       "step": 124040
     },
     {
       "epoch": 2.015401861870644,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2728,
+      "loss": 0.9325,
       "step": 124050
     },
     {
       "epoch": 2.0155643287680136,
-      "grad_norm": 2.859375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2871,
+      "loss": 0.9304,
       "step": 124060
     },
     {
       "epoch": 2.0157267956653833,
-      "grad_norm": 2.796875,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2802,
+      "loss": 0.9257,
       "step": 124070
     },
     {
       "epoch": 2.015889262562753,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2955,
+      "loss": 0.9423,
       "step": 124080
     },
     {
       "epoch": 2.0160517294601226,
-      "grad_norm": 2.59375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2889,
+      "loss": 0.9111,
       "step": 124090
     },
     {
       "epoch": 2.016214196357492,
-      "grad_norm": 2.46875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9441,
       "step": 124100
     },
     {
       "epoch": 2.016376663254862,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.285,
+      "loss": 0.9547,
       "step": 124110
     },
     {
       "epoch": 2.0165391301522315,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.964,
       "step": 124120
     },
     {
       "epoch": 2.016701597049601,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3027,
+      "loss": 0.9522,
       "step": 124130
     },
     {
       "epoch": 2.016864063946971,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3204,
+      "loss": 0.9462,
       "step": 124140
     },
     {
       "epoch": 2.0170265308443405,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2969,
+      "loss": 0.9452,
       "step": 124150
     },
     {
       "epoch": 2.01718899774171,
-      "grad_norm": 3.3125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2759,
+      "loss": 0.9268,
       "step": 124160
     },
     {
       "epoch": 2.0173514646390798,
-      "grad_norm": 2.890625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.9464,
       "step": 124170
     },
     {
       "epoch": 2.0175139315364494,
-      "grad_norm": 3.625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2532,
+      "loss": 0.9276,
       "step": 124180
     },
     {
       "epoch": 2.017676398433819,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2907,
+      "loss": 0.981,
       "step": 124190
     },
     {
       "epoch": 2.0178388653311887,
-      "grad_norm": 2.21875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2778,
+      "loss": 0.9362,
       "step": 124200
     },
     {
       "epoch": 2.0180013322285584,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.9508,
       "step": 124210
     },
     {
       "epoch": 2.018163799125928,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2948,
+      "loss": 0.9025,
       "step": 124220
     },
     {
       "epoch": 2.0183262660232977,
-      "grad_norm": 3.625,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2794,
+      "loss": 0.9718,
       "step": 124230
     },
     {
       "epoch": 2.0184887329206673,
-      "grad_norm": 3.3125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.272,
+      "loss": 0.9597,
       "step": 124240
     },
     {
       "epoch": 2.018651199818037,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 0.9083,
       "step": 124250
     },
     {
       "epoch": 2.0188136667154066,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2663,
+      "loss": 0.9449,
       "step": 124260
     },
     {
       "epoch": 2.0189761336127763,
-      "grad_norm": 2.953125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2711,
+      "loss": 0.9203,
       "step": 124270
     },
     {
       "epoch": 2.019138600510146,
-      "grad_norm": 2.515625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2818,
+      "loss": 0.9009,
       "step": 124280
     },
     {
       "epoch": 2.0193010674075156,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2829,
+      "loss": 0.9321,
       "step": 124290
     },
     {
       "epoch": 2.0194635343048852,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.3042,
+      "loss": 0.9343,
       "step": 124300
     },
     {
       "epoch": 2.019626001202255,
-      "grad_norm": 2.609375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2763,
+      "loss": 0.9428,
       "step": 124310
     },
     {
       "epoch": 2.0197884680996245,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2938,
+      "loss": 0.9614,
       "step": 124320
     },
     {
       "epoch": 2.019950934996994,
-      "grad_norm": 4.78125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2725,
+      "loss": 0.9246,
       "step": 124330
     },
     {
       "epoch": 2.020113401894364,
-      "grad_norm": 2.734375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2705,
+      "loss": 0.9401,
       "step": 124340
     },
     {
       "epoch": 2.0202758687917335,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2791,
+      "loss": 0.9273,
       "step": 124350
     },
     {
       "epoch": 2.020438335689103,
-      "grad_norm": 2.984375,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2841,
+      "loss": 0.8878,
       "step": 124360
     },
     {
       "epoch": 2.020600802586473,
-      "grad_norm": 3.109375,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2786,
+      "loss": 0.9582,
       "step": 124370
     },
     {
       "epoch": 2.020763269483843,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.28,
+      "loss": 0.9457,
       "step": 124380
     },
     {
       "epoch": 2.0209257363812125,
-      "grad_norm": 2.15625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.282,
+      "loss": 0.9292,
       "step": 124390
     },
     {
       "epoch": 2.021088203278582,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2732,
+      "loss": 0.9092,
       "step": 124400
     },
     {
       "epoch": 2.021250670175952,
-      "grad_norm": 2.984375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2631,
+      "loss": 0.9321,
       "step": 124410
     },
     {
       "epoch": 2.0214131370733215,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2821,
+      "loss": 0.9219,
       "step": 124420
     },
     {
       "epoch": 2.021575603970691,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2716,
+      "loss": 0.9441,
       "step": 124430
     },
     {
       "epoch": 2.021738070868061,
-      "grad_norm": 2.734375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2888,
+      "loss": 0.9103,
       "step": 124440
     },
     {
       "epoch": 2.0219005377654304,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2897,
+      "loss": 0.9059,
       "step": 124450
     },
     {
       "epoch": 2.0220630046628,
-      "grad_norm": 2.984375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2788,
+      "loss": 0.9293,
       "step": 124460
     },
     {
       "epoch": 2.0222254715601697,
-      "grad_norm": 3.78125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2754,
+      "loss": 0.9455,
       "step": 124470
     },
     {
       "epoch": 2.0223879384575394,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2766,
+      "loss": 0.9589,
       "step": 124480
     },
     {
       "epoch": 2.022550405354909,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2861,
+      "loss": 0.9079,
       "step": 124490
     },
     {
       "epoch": 2.0227128722522787,
-      "grad_norm": 2.46875,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.268,
+      "loss": 0.8911,
       "step": 124500
     },
     {
       "epoch": 2.0228753391496483,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2885,
+      "loss": 0.9272,
       "step": 124510
     },
     {
       "epoch": 2.023037806047018,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2834,
+      "loss": 0.9032,
       "step": 124520
     },
     {
       "epoch": 2.0232002729443876,
-      "grad_norm": 2.890625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2785,
+      "loss": 0.9266,
       "step": 124530
     },
     {
       "epoch": 2.0233627398417573,
-      "grad_norm": 3.28125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2881,
+      "loss": 0.9478,
       "step": 124540
     },
     {
       "epoch": 2.023525206739127,
-      "grad_norm": 3.375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2836,
+      "loss": 0.9096,
       "step": 124550
     },
     {
       "epoch": 2.0236876736364966,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9408,
       "step": 124560
     },
     {
       "epoch": 2.0238501405338662,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9327,
       "step": 124570
     },
     {
       "epoch": 2.024012607431236,
-      "grad_norm": 3.796875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2815,
+      "loss": 0.907,
       "step": 124580
     },
     {
       "epoch": 2.0241750743286056,
-      "grad_norm": 3.078125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2769,
+      "loss": 0.9545,
       "step": 124590
     },
     {
       "epoch": 2.024337541225975,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2737,
+      "loss": 0.9183,
       "step": 124600
     },
     {
       "epoch": 2.024500008123345,
-      "grad_norm": 3.25,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2917,
+      "loss": 0.9046,
       "step": 124610
     },
     {
       "epoch": 2.0246624750207145,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2759,
+      "loss": 0.9118,
       "step": 124620
     },
     {
       "epoch": 2.024824941918084,
-      "grad_norm": 3.78125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2785,
+      "loss": 0.8942,
       "step": 124630
     },
     {
       "epoch": 2.024987408815454,
-      "grad_norm": 2.484375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2527,
+      "loss": 0.9489,
       "step": 124640
     },
     {
       "epoch": 2.0251498757128235,
-      "grad_norm": 2.734375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9189,
       "step": 124650
     },
     {
       "epoch": 2.025312342610193,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2632,
+      "loss": 0.9339,
       "step": 124660
     },
     {
       "epoch": 2.0254748095075628,
-      "grad_norm": 3.84375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2801,
+      "loss": 0.9348,
       "step": 124670
     },
     {
       "epoch": 2.0256372764049324,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9323,
       "step": 124680
     },
     {
       "epoch": 2.025799743302302,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2741,
+      "loss": 0.9556,
       "step": 124690
     },
     {
       "epoch": 2.0259622101996717,
-      "grad_norm": 2.484375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2711,
+      "loss": 0.9424,
       "step": 124700
     },
     {
       "epoch": 2.0261246770970414,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2842,
+      "loss": 0.9313,
       "step": 124710
     },
     {
       "epoch": 2.026287143994411,
-      "grad_norm": 3.375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2736,
+      "loss": 0.9089,
       "step": 124720
     },
     {
       "epoch": 2.0264496108917807,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2882,
+      "loss": 0.9246,
       "step": 124730
     },
     {
       "epoch": 2.0266120777891503,
-      "grad_norm": 2.796875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.262,
+      "loss": 0.9449,
       "step": 124740
     },
     {
       "epoch": 2.02677454468652,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.268,
+      "loss": 0.915,
       "step": 124750
     },
     {
       "epoch": 2.0269370115838896,
-      "grad_norm": 3.859375,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2694,
+      "loss": 0.9457,
       "step": 124760
     },
     {
       "epoch": 2.0270994784812593,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.8894,
       "step": 124770
     },
     {
       "epoch": 2.027261945378629,
-      "grad_norm": 2.75,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2702,
+      "loss": 0.9378,
       "step": 124780
     },
     {
       "epoch": 2.0274244122759986,
-      "grad_norm": 2.40625,
+      "grad_norm": 12.75,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 0.9294,
       "step": 124790
     },
     {
       "epoch": 2.027586879173368,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2677,
+      "loss": 0.9389,
       "step": 124800
     },
     {
       "epoch": 2.027749346070738,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2665,
+      "loss": 0.9382,
       "step": 124810
     },
     {
       "epoch": 2.027911812968108,
-      "grad_norm": 3.8125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9359,
       "step": 124820
     },
     {
       "epoch": 2.0280742798654776,
-      "grad_norm": 2.6875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2428,
+      "loss": 0.9059,
       "step": 124830
     },
     {
       "epoch": 2.0282367467628473,
-      "grad_norm": 3.1875,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2571,
+      "loss": 0.9208,
       "step": 124840
     },
     {
       "epoch": 2.028399213660217,
-      "grad_norm": 2.875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2351,
+      "loss": 0.9264,
       "step": 124850
     },
     {
       "epoch": 2.0285616805575866,
-      "grad_norm": 2.53125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9461,
       "step": 124860
     },
     {
       "epoch": 2.028724147454956,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9115,
       "step": 124870
     },
     {
       "epoch": 2.028886614352326,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2456,
+      "loss": 0.9073,
       "step": 124880
     },
     {
       "epoch": 2.0290490812496955,
-      "grad_norm": 2.34375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 0.9052,
       "step": 124890
     },
     {
       "epoch": 2.029211548147065,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2546,
+      "loss": 0.9053,
       "step": 124900
     },
     {
       "epoch": 2.029374015044435,
-      "grad_norm": 3.703125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2767,
+      "loss": 0.9262,
       "step": 124910
     },
     {
       "epoch": 2.0295364819418045,
-      "grad_norm": 3.546875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2503,
+      "loss": 0.9321,
       "step": 124920
     },
     {
       "epoch": 2.029698948839174,
-      "grad_norm": 3.015625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 0.9264,
       "step": 124930
     },
     {
       "epoch": 2.0298614157365438,
-      "grad_norm": 3.109375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 0.9575,
       "step": 124940
     },
     {
       "epoch": 2.0300238826339134,
-      "grad_norm": 3.140625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2785,
+      "loss": 0.9307,
       "step": 124950
     },
     {
       "epoch": 2.030186349531283,
-      "grad_norm": 3.265625,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2892,
+      "loss": 0.9178,
       "step": 124960
     },
     {
       "epoch": 2.0303488164286527,
-      "grad_norm": 3.578125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 0.9221,
       "step": 124970
     },
     {
       "epoch": 2.0305112833260224,
-      "grad_norm": 2.875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2761,
+      "loss": 0.9686,
       "step": 124980
     },
     {
       "epoch": 2.030673750223392,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2778,
+      "loss": 0.9579,
       "step": 124990
     },
     {
       "epoch": 2.0308362171207617,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2746,
+      "loss": 0.9491,
       "step": 125000
     },
     {
       "epoch": 2.0309986840181313,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2841,
+      "loss": 0.941,
       "step": 125010
     },
     {
       "epoch": 2.031161150915501,
-      "grad_norm": 2.734375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2839,
+      "loss": 0.9966,
       "step": 125020
     },
     {
       "epoch": 2.0313236178128706,
-      "grad_norm": 2.5625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2557,
+      "loss": 0.9307,
       "step": 125030
     },
     {
       "epoch": 2.0314860847102403,
-      "grad_norm": 3.140625,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9492,
       "step": 125040
     },
     {
       "epoch": 2.03164855160761,
-      "grad_norm": 3.0,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2777,
+      "loss": 0.9582,
       "step": 125050
     },
     {
       "epoch": 2.0318110185049796,
-      "grad_norm": 3.25,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2745,
+      "loss": 0.9521,
       "step": 125060
     },
     {
       "epoch": 2.0319734854023492,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2489,
+      "loss": 0.9502,
       "step": 125070
     },
     {
       "epoch": 2.032135952299719,
-      "grad_norm": 3.375,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 0.9454,
       "step": 125080
     },
     {
       "epoch": 2.0322984191970885,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9388,
       "step": 125090
     },
     {
       "epoch": 2.032460886094458,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2824,
+      "loss": 0.9382,
       "step": 125100
     },
     {
       "epoch": 2.032623352991828,
-      "grad_norm": 3.6875,
+      "grad_norm": 12.125,
       "learning_rate": 5e-05,
-      "loss": 0.2856,
+      "loss": 0.9051,
       "step": 125110
     },
     {
       "epoch": 2.0327858198891975,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2855,
+      "loss": 0.9704,
       "step": 125120
     },
     {
       "epoch": 2.032948286786567,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9197,
       "step": 125130
     },
     {
       "epoch": 2.033110753683937,
-      "grad_norm": 3.140625,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2799,
+      "loss": 0.9451,
       "step": 125140
     },
     {
       "epoch": 2.0332732205813064,
-      "grad_norm": 3.984375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2637,
+      "loss": 0.9151,
       "step": 125150
     },
     {
       "epoch": 2.033435687478676,
-      "grad_norm": 4.625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2749,
+      "loss": 0.9162,
       "step": 125160
     },
     {
       "epoch": 2.0335981543760457,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.2622,
+      "loss": 0.8917,
       "step": 125170
     },
     {
       "epoch": 2.0337606212734154,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2713,
+      "loss": 0.9045,
       "step": 125180
     },
     {
       "epoch": 2.033923088170785,
-      "grad_norm": 2.65625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2709,
+      "loss": 0.9417,
       "step": 125190
     },
     {
       "epoch": 2.0340855550681547,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2465,
+      "loss": 0.9525,
       "step": 125200
     },
     {
       "epoch": 2.0342480219655243,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2862,
+      "loss": 0.9314,
       "step": 125210
     },
     {
       "epoch": 2.034410488862894,
-      "grad_norm": 2.6875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2593,
+      "loss": 0.9269,
       "step": 125220
     },
     {
       "epoch": 2.0345729557602636,
-      "grad_norm": 2.78125,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9224,
       "step": 125230
     },
     {
       "epoch": 2.0347354226576333,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2752,
+      "loss": 0.9551,
       "step": 125240
     },
     {
       "epoch": 2.034897889555003,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 0.9308,
       "step": 125250
     },
     {
       "epoch": 2.035060356452373,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9748,
       "step": 125260
     },
     {
       "epoch": 2.0352228233497427,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2603,
+      "loss": 0.9414,
       "step": 125270
     },
     {
       "epoch": 2.0353852902471123,
-      "grad_norm": 5.9375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2506,
+      "loss": 0.9368,
       "step": 125280
     },
     {
       "epoch": 2.035547757144482,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2672,
+      "loss": 0.935,
       "step": 125290
     },
     {
       "epoch": 2.0357102240418516,
-      "grad_norm": 3.671875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2423,
+      "loss": 0.917,
       "step": 125300
     },
     {
       "epoch": 2.0358726909392213,
-      "grad_norm": 3.25,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2523,
+      "loss": 0.9355,
       "step": 125310
     },
     {
       "epoch": 2.036035157836591,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2725,
+      "loss": 0.9292,
       "step": 125320
     },
     {
       "epoch": 2.0361976247339606,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2838,
+      "loss": 0.9388,
       "step": 125330
     },
     {
       "epoch": 2.0363600916313302,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9354,
       "step": 125340
     },
     {
       "epoch": 2.0365225585287,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2725,
+      "loss": 0.9295,
       "step": 125350
     },
     {
       "epoch": 2.0366850254260696,
-      "grad_norm": 3.28125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2693,
+      "loss": 0.9397,
       "step": 125360
     },
     {
       "epoch": 2.036847492323439,
-      "grad_norm": 3.15625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.263,
+      "loss": 0.9378,
       "step": 125370
     },
     {
       "epoch": 2.037009959220809,
-      "grad_norm": 3.4375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2816,
+      "loss": 0.9336,
       "step": 125380
     },
     {
       "epoch": 2.0371724261181785,
-      "grad_norm": 2.578125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2372,
+      "loss": 0.9514,
       "step": 125390
     },
     {
       "epoch": 2.037334893015548,
-      "grad_norm": 3.828125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9074,
       "step": 125400
     },
     {
       "epoch": 2.037497359912918,
-      "grad_norm": 3.828125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9383,
       "step": 125410
     },
     {
       "epoch": 2.0376598268102875,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2625,
+      "loss": 0.9231,
       "step": 125420
     },
     {
       "epoch": 2.037822293707657,
-      "grad_norm": 3.125,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2472,
+      "loss": 0.9298,
       "step": 125430
     },
     {
       "epoch": 2.0379847606050268,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2739,
+      "loss": 0.929,
       "step": 125440
     },
     {
       "epoch": 2.0381472275023964,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.272,
+      "loss": 0.96,
       "step": 125450
     },
     {
       "epoch": 2.038309694399766,
-      "grad_norm": 2.453125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9457,
       "step": 125460
     },
     {
       "epoch": 2.0384721612971357,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2565,
+      "loss": 0.9252,
       "step": 125470
     },
     {
       "epoch": 2.0386346281945054,
-      "grad_norm": 3.734375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.297,
+      "loss": 0.9371,
       "step": 125480
     },
     {
       "epoch": 2.038797095091875,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2881,
+      "loss": 0.9546,
       "step": 125490
     },
     {
       "epoch": 2.0389595619892447,
-      "grad_norm": 2.6875,
+      "grad_norm": 12.0,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.9441,
       "step": 125500
     },
     {
       "epoch": 2.0391220288866143,
-      "grad_norm": 3.625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2764,
+      "loss": 0.9562,
       "step": 125510
     },
     {
       "epoch": 2.039284495783984,
-      "grad_norm": 2.75,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2833,
+      "loss": 0.915,
       "step": 125520
     },
     {
       "epoch": 2.0394469626813536,
-      "grad_norm": 3.8125,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2798,
+      "loss": 0.9033,
       "step": 125530
     },
     {
       "epoch": 2.0396094295787233,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.9506,
       "step": 125540
     },
     {
       "epoch": 2.039771896476093,
-      "grad_norm": 3.078125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 0.9624,
       "step": 125550
     },
     {
       "epoch": 2.0399343633734626,
-      "grad_norm": 3.03125,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2818,
+      "loss": 0.9473,
       "step": 125560
     },
     {
       "epoch": 2.040096830270832,
-      "grad_norm": 4.125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2619,
+      "loss": 0.9809,
       "step": 125570
     },
     {
       "epoch": 2.040259297168202,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 0.9461,
       "step": 125580
     },
     {
       "epoch": 2.0404217640655715,
-      "grad_norm": 2.53125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9388,
       "step": 125590
     },
     {
       "epoch": 2.040584230962941,
-      "grad_norm": 3.296875,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2745,
+      "loss": 0.9145,
       "step": 125600
     },
     {
       "epoch": 2.040746697860311,
-      "grad_norm": 3.5,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.267,
+      "loss": 0.9108,
       "step": 125610
     },
     {
       "epoch": 2.0409091647576805,
-      "grad_norm": 2.65625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.276,
+      "loss": 0.9264,
       "step": 125620
     },
     {
       "epoch": 2.04107163165505,
-      "grad_norm": 3.265625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9353,
       "step": 125630
     },
     {
       "epoch": 2.0412340985524198,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 0.9324,
       "step": 125640
     },
     {
       "epoch": 2.0413965654497894,
-      "grad_norm": 3.15625,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2728,
+      "loss": 0.9541,
       "step": 125650
     },
     {
       "epoch": 2.041559032347159,
-      "grad_norm": 3.734375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2509,
+      "loss": 0.9373,
       "step": 125660
     },
     {
       "epoch": 2.0417214992445287,
-      "grad_norm": 2.953125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2797,
+      "loss": 0.9258,
       "step": 125670
     },
     {
       "epoch": 2.0418839661418984,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2537,
+      "loss": 0.9232,
       "step": 125680
     },
     {
       "epoch": 2.0420464330392685,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9084,
       "step": 125690
     },
     {
       "epoch": 2.042208899936638,
-      "grad_norm": 4.5,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2732,
+      "loss": 0.8976,
       "step": 125700
     },
     {
       "epoch": 2.0423713668340078,
-      "grad_norm": 3.0,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.921,
       "step": 125710
     },
     {
       "epoch": 2.0425338337313774,
-      "grad_norm": 4.0,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2708,
+      "loss": 0.9724,
       "step": 125720
     },
     {
       "epoch": 2.042696300628747,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.269,
+      "loss": 0.9211,
       "step": 125730
     },
     {
       "epoch": 2.0428587675261167,
-      "grad_norm": 3.1875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2841,
+      "loss": 0.9398,
       "step": 125740
     },
     {
       "epoch": 2.0430212344234864,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2471,
+      "loss": 0.9266,
       "step": 125750
     },
     {
       "epoch": 2.043183701320856,
-      "grad_norm": 2.640625,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2606,
+      "loss": 0.953,
       "step": 125760
     },
     {
       "epoch": 2.0433461682182257,
-      "grad_norm": 2.8125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2641,
+      "loss": 0.9644,
       "step": 125770
     },
     {
       "epoch": 2.0435086351155953,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2818,
+      "loss": 0.9507,
       "step": 125780
     },
     {
       "epoch": 2.043671102012965,
-      "grad_norm": 3.25,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2712,
+      "loss": 0.9035,
       "step": 125790
     },
     {
       "epoch": 2.0438335689103346,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2791,
+      "loss": 0.9023,
       "step": 125800
     },
     {
       "epoch": 2.0439960358077043,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2557,
+      "loss": 0.9631,
       "step": 125810
     },
     {
       "epoch": 2.044158502705074,
-      "grad_norm": 3.375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9751,
       "step": 125820
     },
     {
       "epoch": 2.0443209696024436,
-      "grad_norm": 2.265625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2548,
+      "loss": 0.9235,
       "step": 125830
     },
     {
       "epoch": 2.0444834364998132,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.267,
+      "loss": 0.9604,
       "step": 125840
     },
     {
       "epoch": 2.044645903397183,
-      "grad_norm": 3.375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2554,
+      "loss": 0.9764,
       "step": 125850
     },
     {
       "epoch": 2.0448083702945525,
-      "grad_norm": 3.140625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2622,
+      "loss": 0.8844,
       "step": 125860
     },
     {
       "epoch": 2.044970837191922,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2396,
+      "loss": 0.9213,
       "step": 125870
     },
     {
       "epoch": 2.045133304089292,
-      "grad_norm": 2.90625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2563,
+      "loss": 0.9215,
       "step": 125880
     },
     {
       "epoch": 2.0452957709866615,
-      "grad_norm": 4.21875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2839,
+      "loss": 0.9695,
       "step": 125890
     },
     {
       "epoch": 2.045458237884031,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2568,
+      "loss": 0.914,
       "step": 125900
     },
     {
       "epoch": 2.045620704781401,
-      "grad_norm": 3.359375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2589,
+      "loss": 0.9206,
       "step": 125910
     },
     {
       "epoch": 2.0457831716787704,
-      "grad_norm": 3.234375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2667,
+      "loss": 0.9522,
       "step": 125920
     },
     {
       "epoch": 2.04594563857614,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.3008,
+      "loss": 0.9152,
       "step": 125930
     },
     {
       "epoch": 2.0461081054735097,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2781,
+      "loss": 0.9311,
       "step": 125940
     },
     {
       "epoch": 2.0462705723708794,
-      "grad_norm": 2.734375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2597,
+      "loss": 0.9276,
       "step": 125950
     },
     {
       "epoch": 2.046433039268249,
-      "grad_norm": 3.25,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2704,
+      "loss": 0.9124,
       "step": 125960
     },
     {
       "epoch": 2.0465955061656187,
-      "grad_norm": 3.125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2617,
+      "loss": 0.9632,
       "step": 125970
     },
     {
       "epoch": 2.0467579730629883,
-      "grad_norm": 2.765625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.953,
       "step": 125980
     },
     {
       "epoch": 2.046920439960358,
-      "grad_norm": 2.84375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2695,
+      "loss": 0.9466,
       "step": 125990
     },
     {
       "epoch": 2.0470829068577276,
-      "grad_norm": 1.8828125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2417,
+      "loss": 0.9447,
       "step": 126000
     },
     {
       "epoch": 2.0472453737550973,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2771,
+      "loss": 0.9386,
       "step": 126010
     },
     {
       "epoch": 2.047407840652467,
-      "grad_norm": 3.203125,
+      "grad_norm": 12.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2456,
+      "loss": 0.9249,
       "step": 126020
     },
     {
       "epoch": 2.0475703075498366,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2342,
+      "loss": 0.9034,
       "step": 126030
     },
     {
       "epoch": 2.0477327744472062,
-      "grad_norm": 3.625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2745,
+      "loss": 0.9485,
       "step": 126040
     },
     {
       "epoch": 2.047895241344576,
-      "grad_norm": 2.75,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9653,
       "step": 126050
     },
     {
       "epoch": 2.0480577082419456,
-      "grad_norm": 3.234375,
+      "grad_norm": 12.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2872,
+      "loss": 0.9359,
       "step": 126060
     },
     {
       "epoch": 2.048220175139315,
-      "grad_norm": 2.78125,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2754,
+      "loss": 0.9269,
       "step": 126070
     },
     {
       "epoch": 2.048382642036685,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2867,
+      "loss": 0.9735,
       "step": 126080
     },
     {
       "epoch": 2.0485451089340545,
-      "grad_norm": 2.90625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 0.9304,
       "step": 126090
     },
     {
       "epoch": 2.048707575831424,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.277,
+      "loss": 0.9429,
       "step": 126100
     },
     {
       "epoch": 2.048870042728794,
-      "grad_norm": 3.375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2684,
+      "loss": 0.9526,
       "step": 126110
     },
     {
       "epoch": 2.0490325096261635,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2683,
+      "loss": 0.9732,
       "step": 126120
     },
     {
       "epoch": 2.049194976523533,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2344,
+      "loss": 0.9464,
       "step": 126130
     },
     {
       "epoch": 2.049357443420903,
-      "grad_norm": 3.078125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2548,
+      "loss": 0.9342,
       "step": 126140
     },
     {
       "epoch": 2.049519910318273,
-      "grad_norm": 2.703125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2357,
+      "loss": 0.9482,
       "step": 126150
     },
     {
       "epoch": 2.0496823772156425,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9658,
       "step": 126160
     },
     {
       "epoch": 2.049844844113012,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.255,
+      "loss": 0.9669,
       "step": 126170
     },
     {
       "epoch": 2.050007311010382,
-      "grad_norm": 3.140625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2413,
+      "loss": 0.931,
       "step": 126180
     },
     {
       "epoch": 2.0501697779077515,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2615,
+      "loss": 0.9357,
       "step": 126190
     },
     {
       "epoch": 2.050332244805121,
-      "grad_norm": 2.484375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2526,
+      "loss": 0.9715,
       "step": 126200
     },
     {
       "epoch": 2.0504947117024908,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2525,
+      "loss": 0.9234,
       "step": 126210
     },
     {
       "epoch": 2.0506571785998604,
-      "grad_norm": 3.078125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2616,
+      "loss": 0.9836,
       "step": 126220
     },
     {
       "epoch": 2.05081964549723,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9404,
       "step": 126230
     },
     {
       "epoch": 2.0509821123945997,
-      "grad_norm": 5.15625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2657,
+      "loss": 0.9366,
       "step": 126240
     },
     {
       "epoch": 2.0511445792919694,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.234,
+      "loss": 0.9219,
       "step": 126250
     },
     {
       "epoch": 2.051307046189339,
-      "grad_norm": 3.125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9549,
       "step": 126260
     },
     {
       "epoch": 2.0514695130867087,
-      "grad_norm": 2.953125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2604,
+      "loss": 0.9626,
       "step": 126270
     },
     {
       "epoch": 2.0516319799840783,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2814,
+      "loss": 0.9436,
       "step": 126280
     },
     {
       "epoch": 2.051794446881448,
-      "grad_norm": 3.375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2741,
+      "loss": 0.9211,
       "step": 126290
     },
     {
       "epoch": 2.0519569137788176,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2551,
+      "loss": 0.9141,
       "step": 126300
     },
     {
       "epoch": 2.0521193806761873,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2341,
+      "loss": 0.9957,
       "step": 126310
     },
     {
       "epoch": 2.052281847573557,
-      "grad_norm": 3.0,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2589,
+      "loss": 0.9188,
       "step": 126320
     },
     {
       "epoch": 2.0524443144709266,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.252,
+      "loss": 0.9504,
       "step": 126330
     },
     {
       "epoch": 2.052606781368296,
-      "grad_norm": 3.28125,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2803,
+      "loss": 0.9843,
       "step": 126340
     },
     {
       "epoch": 2.052769248265666,
-      "grad_norm": 5.03125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.268,
+      "loss": 0.9177,
       "step": 126350
     },
     {
       "epoch": 2.0529317151630355,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.276,
+      "loss": 0.919,
       "step": 126360
     },
     {
       "epoch": 2.053094182060405,
-      "grad_norm": 2.78125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9352,
       "step": 126370
     },
     {
       "epoch": 2.053256648957775,
-      "grad_norm": 3.1875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2615,
+      "loss": 0.9491,
       "step": 126380
     },
     {
       "epoch": 2.0534191158551445,
-      "grad_norm": 4.09375,
+      "grad_norm": 15.25,
       "learning_rate": 5e-05,
-      "loss": 0.2568,
+      "loss": 0.9317,
       "step": 126390
     },
     {
       "epoch": 2.053581582752514,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 0.9354,
       "step": 126400
     },
     {
       "epoch": 2.0537440496498838,
-      "grad_norm": 3.171875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2553,
+      "loss": 0.9457,
       "step": 126410
     },
     {
       "epoch": 2.0539065165472534,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2719,
+      "loss": 0.9589,
       "step": 126420
     },
     {
       "epoch": 2.054068983444623,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9609,
       "step": 126430
     },
     {
       "epoch": 2.0542314503419927,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2652,
+      "loss": 0.9275,
       "step": 126440
     },
     {
       "epoch": 2.0543939172393624,
-      "grad_norm": 3.421875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2616,
+      "loss": 0.9581,
       "step": 126450
     },
     {
       "epoch": 2.054556384136732,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2837,
+      "loss": 0.9477,
       "step": 126460
     },
     {
       "epoch": 2.0547188510341017,
-      "grad_norm": 3.96875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2436,
+      "loss": 0.9361,
       "step": 126470
     },
     {
       "epoch": 2.0548813179314713,
-      "grad_norm": 2.828125,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2642,
+      "loss": 0.9686,
       "step": 126480
     },
     {
       "epoch": 2.055043784828841,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2546,
+      "loss": 0.923,
       "step": 126490
     },
     {
       "epoch": 2.0552062517262106,
-      "grad_norm": 3.96875,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9536,
       "step": 126500
     },
     {
       "epoch": 2.0553687186235803,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.285,
+      "loss": 0.9829,
       "step": 126510
     },
     {
       "epoch": 2.05553118552095,
-      "grad_norm": 3.453125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2714,
+      "loss": 0.9339,
       "step": 126520
     },
     {
       "epoch": 2.0556936524183196,
-      "grad_norm": 3.25,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.278,
+      "loss": 0.9028,
       "step": 126530
     },
     {
       "epoch": 2.0558561193156892,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2616,
+      "loss": 0.948,
       "step": 126540
     },
     {
       "epoch": 2.056018586213059,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2431,
+      "loss": 0.9468,
       "step": 126550
     },
     {
       "epoch": 2.0561810531104285,
-      "grad_norm": 3.59375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2386,
+      "loss": 0.9414,
       "step": 126560
     },
     {
       "epoch": 2.0563435200077986,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9278,
       "step": 126570
     },
     {
       "epoch": 2.0565059869051683,
-      "grad_norm": 4.875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9537,
       "step": 126580
     },
     {
       "epoch": 2.056668453802538,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9628,
       "step": 126590
     },
     {
       "epoch": 2.0568309206999076,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2332,
+      "loss": 1.0092,
       "step": 126600
     },
     {
       "epoch": 2.0569933875972772,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2495,
+      "loss": 0.9227,
       "step": 126610
     },
     {
       "epoch": 2.057155854494647,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2641,
+      "loss": 0.9463,
       "step": 126620
     },
     {
       "epoch": 2.0573183213920165,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2471,
+      "loss": 0.962,
       "step": 126630
     },
     {
       "epoch": 2.057480788289386,
-      "grad_norm": 2.875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2528,
+      "loss": 0.9387,
       "step": 126640
     },
     {
       "epoch": 2.057643255186756,
-      "grad_norm": 3.78125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9431,
       "step": 126650
     },
     {
       "epoch": 2.0578057220841255,
-      "grad_norm": 3.609375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2528,
+      "loss": 0.9678,
       "step": 126660
     },
     {
       "epoch": 2.057968188981495,
-      "grad_norm": 2.546875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9572,
       "step": 126670
     },
     {
       "epoch": 2.058130655878865,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2389,
+      "loss": 0.9645,
       "step": 126680
     },
     {
       "epoch": 2.0582931227762344,
-      "grad_norm": 3.75,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2312,
+      "loss": 0.9188,
       "step": 126690
     },
     {
       "epoch": 2.058455589673604,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.9347,
       "step": 126700
     },
     {
       "epoch": 2.0586180565709737,
-      "grad_norm": 2.953125,
+      "grad_norm": 12.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2508,
+      "loss": 0.9308,
       "step": 126710
     },
     {
       "epoch": 2.0587805234683434,
-      "grad_norm": 2.875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2411,
+      "loss": 0.9758,
       "step": 126720
     },
     {
       "epoch": 2.058942990365713,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2517,
+      "loss": 0.9193,
       "step": 126730
     },
     {
       "epoch": 2.0591054572630827,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9412,
       "step": 126740
     },
     {
       "epoch": 2.0592679241604523,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.287,
+      "loss": 0.9117,
       "step": 126750
     },
     {
       "epoch": 2.059430391057822,
-      "grad_norm": 2.765625,
+      "grad_norm": 7.71875,
       "learning_rate": 5e-05,
-      "loss": 0.2453,
+      "loss": 0.9477,
       "step": 126760
     },
     {
       "epoch": 2.0595928579551916,
-      "grad_norm": 2.890625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9661,
       "step": 126770
     },
     {
       "epoch": 2.0597553248525613,
-      "grad_norm": 2.609375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2834,
+      "loss": 0.9573,
       "step": 126780
     },
     {
       "epoch": 2.059917791749931,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2497,
+      "loss": 0.9847,
       "step": 126790
     },
     {
       "epoch": 2.0600802586473006,
-      "grad_norm": 3.125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2342,
+      "loss": 0.9213,
       "step": 126800
     },
     {
       "epoch": 2.0602427255446703,
-      "grad_norm": 4.0625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2352,
+      "loss": 0.9562,
       "step": 126810
     },
     {
       "epoch": 2.06040519244204,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2514,
+      "loss": 0.9138,
       "step": 126820
     },
     {
       "epoch": 2.0605676593394096,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 1.0095,
       "step": 126830
     },
     {
       "epoch": 2.060730126236779,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2415,
+      "loss": 0.9188,
       "step": 126840
     },
     {
       "epoch": 2.060892593134149,
-      "grad_norm": 3.25,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2478,
+      "loss": 0.9776,
       "step": 126850
     },
     {
       "epoch": 2.0610550600315185,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2571,
+      "loss": 0.9426,
       "step": 126860
     },
     {
       "epoch": 2.061217526928888,
-      "grad_norm": 2.546875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 0.9488,
       "step": 126870
     },
     {
       "epoch": 2.061379993826258,
-      "grad_norm": 3.140625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9388,
       "step": 126880
     },
     {
       "epoch": 2.0615424607236275,
-      "grad_norm": 3.53125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 0.9812,
       "step": 126890
     },
     {
       "epoch": 2.061704927620997,
-      "grad_norm": 2.921875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2363,
+      "loss": 0.9411,
       "step": 126900
     },
     {
       "epoch": 2.0618673945183668,
-      "grad_norm": 3.546875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9231,
       "step": 126910
     },
     {
       "epoch": 2.0620298614157364,
-      "grad_norm": 2.6875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9195,
       "step": 126920
     },
     {
       "epoch": 2.062192328313106,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2514,
+      "loss": 0.959,
       "step": 126930
     },
     {
       "epoch": 2.0623547952104757,
-      "grad_norm": 3.6875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 0.9373,
       "step": 126940
     },
     {
       "epoch": 2.0625172621078454,
-      "grad_norm": 4.46875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2546,
+      "loss": 0.9198,
       "step": 126950
     },
     {
       "epoch": 2.062679729005215,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2624,
+      "loss": 0.9412,
       "step": 126960
     },
     {
       "epoch": 2.0628421959025847,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2351,
+      "loss": 0.9258,
       "step": 126970
     },
     {
       "epoch": 2.0630046627999543,
-      "grad_norm": 2.90625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2468,
+      "loss": 0.9184,
       "step": 126980
     },
     {
       "epoch": 2.063167129697324,
-      "grad_norm": 3.375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2225,
+      "loss": 0.9142,
       "step": 126990
     },
     {
       "epoch": 2.063329596594694,
-      "grad_norm": 3.078125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2389,
+      "loss": 0.9312,
       "step": 127000
     },
     {
       "epoch": 2.0634920634920633,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2474,
+      "loss": 0.8984,
       "step": 127010
     },
     {
       "epoch": 2.0636545303894334,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2486,
+      "loss": 0.9374,
       "step": 127020
     },
     {
       "epoch": 2.063816997286803,
-      "grad_norm": 3.625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2798,
+      "loss": 0.9689,
       "step": 127030
     },
     {
       "epoch": 2.0639794641841727,
-      "grad_norm": 3.203125,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.272,
+      "loss": 0.9296,
       "step": 127040
     },
     {
       "epoch": 2.0641419310815423,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2836,
+      "loss": 0.9536,
       "step": 127050
     },
     {
       "epoch": 2.064304397978912,
-      "grad_norm": 4.59375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2637,
+      "loss": 0.9588,
       "step": 127060
     },
     {
       "epoch": 2.0644668648762816,
-      "grad_norm": 2.390625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2649,
+      "loss": 0.9213,
       "step": 127070
     },
     {
       "epoch": 2.0646293317736513,
-      "grad_norm": 3.734375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2549,
+      "loss": 0.9552,
       "step": 127080
     },
     {
       "epoch": 2.064791798671021,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2668,
+      "loss": 0.9469,
       "step": 127090
     },
     {
       "epoch": 2.0649542655683906,
-      "grad_norm": 2.9375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2604,
+      "loss": 0.9289,
       "step": 127100
     },
     {
       "epoch": 2.06511673246576,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2452,
+      "loss": 0.9257,
       "step": 127110
     },
     {
       "epoch": 2.06527919936313,
-      "grad_norm": 2.765625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2496,
+      "loss": 0.9432,
       "step": 127120
     },
     {
       "epoch": 2.0654416662604995,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2436,
+      "loss": 0.9832,
       "step": 127130
     },
     {
       "epoch": 2.065604133157869,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2525,
+      "loss": 0.9528,
       "step": 127140
     },
     {
       "epoch": 2.065766600055239,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2744,
+      "loss": 0.9467,
       "step": 127150
     },
     {
       "epoch": 2.0659290669526085,
-      "grad_norm": 3.65625,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2456,
+      "loss": 0.9169,
       "step": 127160
     },
     {
       "epoch": 2.066091533849978,
-      "grad_norm": 5.03125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2746,
+      "loss": 0.9601,
       "step": 127170
     },
     {
       "epoch": 2.0662540007473478,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2775,
+      "loss": 0.949,
       "step": 127180
     },
     {
       "epoch": 2.0664164676447174,
-      "grad_norm": 5.09375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 0.967,
       "step": 127190
     },
     {
       "epoch": 2.066578934542087,
-      "grad_norm": 3.671875,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2766,
+      "loss": 0.938,
       "step": 127200
     },
     {
       "epoch": 2.0667414014394567,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2426,
+      "loss": 0.961,
       "step": 127210
     },
     {
       "epoch": 2.0669038683368264,
-      "grad_norm": 3.625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9428,
       "step": 127220
     },
     {
       "epoch": 2.067066335234196,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2537,
+      "loss": 0.9504,
       "step": 127230
     },
     {
       "epoch": 2.0672288021315657,
-      "grad_norm": 3.359375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.3026,
+      "loss": 0.9274,
       "step": 127240
     },
     {
       "epoch": 2.0673912690289353,
-      "grad_norm": 3.3125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.941,
       "step": 127250
     },
     {
       "epoch": 2.067553735926305,
-      "grad_norm": 3.328125,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.2423,
+      "loss": 0.9013,
       "step": 127260
     },
     {
       "epoch": 2.0677162028236746,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2393,
+      "loss": 0.93,
       "step": 127270
     },
     {
       "epoch": 2.0678786697210443,
-      "grad_norm": 3.359375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2364,
+      "loss": 0.9368,
       "step": 127280
     },
     {
       "epoch": 2.068041136618414,
-      "grad_norm": 3.90625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2757,
+      "loss": 0.9123,
       "step": 127290
     },
     {
       "epoch": 2.0682036035157836,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2907,
+      "loss": 0.9851,
       "step": 127300
     },
     {
       "epoch": 2.0683660704131532,
-      "grad_norm": 2.875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 0.9381,
       "step": 127310
     },
     {
       "epoch": 2.068528537310523,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2629,
+      "loss": 0.9305,
       "step": 127320
     },
     {
       "epoch": 2.0686910042078925,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2615,
+      "loss": 0.8981,
       "step": 127330
     },
     {
       "epoch": 2.068853471105262,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2513,
+      "loss": 0.9649,
       "step": 127340
     },
     {
       "epoch": 2.069015938002632,
-      "grad_norm": 2.6875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2426,
+      "loss": 0.9418,
       "step": 127350
     },
     {
       "epoch": 2.0691784049000015,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2431,
+      "loss": 0.9417,
       "step": 127360
     },
     {
       "epoch": 2.069340871797371,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9396,
       "step": 127370
     },
     {
       "epoch": 2.069503338694741,
-      "grad_norm": 4.25,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2587,
+      "loss": 0.9201,
       "step": 127380
     },
     {
       "epoch": 2.0696658055921104,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9588,
       "step": 127390
     },
     {
       "epoch": 2.06982827248948,
-      "grad_norm": 3.25,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2522,
+      "loss": 0.9505,
       "step": 127400
     },
     {
       "epoch": 2.0699907393868497,
-      "grad_norm": 3.640625,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2399,
+      "loss": 0.9274,
       "step": 127410
     },
     {
       "epoch": 2.0701532062842194,
-      "grad_norm": 3.84375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2576,
+      "loss": 0.9182,
       "step": 127420
     },
     {
       "epoch": 2.070315673181589,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9274,
       "step": 127430
     },
     {
       "epoch": 2.0704781400789587,
-      "grad_norm": 2.8125,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.2777,
+      "loss": 0.9601,
       "step": 127440
     },
     {
       "epoch": 2.070640606976329,
-      "grad_norm": 3.859375,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2589,
+      "loss": 0.9454,
       "step": 127450
     },
     {
       "epoch": 2.0708030738736984,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2657,
+      "loss": 0.9577,
       "step": 127460
     },
     {
       "epoch": 2.070965540771068,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2557,
+      "loss": 0.949,
       "step": 127470
     },
     {
       "epoch": 2.0711280076684377,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.262,
+      "loss": 0.9711,
       "step": 127480
     },
     {
       "epoch": 2.0712904745658074,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.249,
+      "loss": 0.9575,
       "step": 127490
     },
     {
       "epoch": 2.071452941463177,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9184,
       "step": 127500
     },
     {
       "epoch": 2.0716154083605467,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2366,
+      "loss": 0.9757,
       "step": 127510
     },
     {
       "epoch": 2.0717778752579163,
-      "grad_norm": 3.03125,
+      "grad_norm": 12.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2405,
+      "loss": 0.9269,
       "step": 127520
     },
     {
       "epoch": 2.071940342155286,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2466,
+      "loss": 0.9435,
       "step": 127530
     },
     {
       "epoch": 2.0721028090526556,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9306,
       "step": 127540
     },
     {
       "epoch": 2.0722652759500253,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9305,
       "step": 127550
     },
     {
       "epoch": 2.072427742847395,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9367,
       "step": 127560
     },
     {
       "epoch": 2.0725902097447646,
-      "grad_norm": 3.171875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.259,
+      "loss": 0.9367,
       "step": 127570
     },
     {
       "epoch": 2.0727526766421343,
-      "grad_norm": 3.09375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2482,
+      "loss": 0.9557,
       "step": 127580
     },
     {
       "epoch": 2.072915143539504,
-      "grad_norm": 2.734375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2484,
+      "loss": 0.9282,
       "step": 127590
     },
     {
       "epoch": 2.0730776104368736,
-      "grad_norm": 2.84375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2497,
+      "loss": 0.9644,
       "step": 127600
     },
     {
       "epoch": 2.073240077334243,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2536,
+      "loss": 0.9601,
       "step": 127610
     },
     {
       "epoch": 2.073402544231613,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2495,
+      "loss": 0.9443,
       "step": 127620
     },
     {
       "epoch": 2.0735650111289825,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.241,
+      "loss": 0.9427,
       "step": 127630
     },
     {
       "epoch": 2.073727478026352,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2738,
+      "loss": 0.9616,
       "step": 127640
     },
     {
       "epoch": 2.073889944923722,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2537,
+      "loss": 0.9109,
       "step": 127650
     },
     {
       "epoch": 2.0740524118210915,
-      "grad_norm": 5.03125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2643,
+      "loss": 0.9835,
       "step": 127660
     },
     {
       "epoch": 2.074214878718461,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9384,
       "step": 127670
     },
     {
       "epoch": 2.0743773456158308,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2422,
+      "loss": 0.954,
       "step": 127680
     },
     {
       "epoch": 2.0745398125132004,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2567,
+      "loss": 0.9917,
       "step": 127690
     },
     {
       "epoch": 2.07470227941057,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2477,
+      "loss": 0.978,
       "step": 127700
     },
     {
       "epoch": 2.0748647463079397,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2555,
+      "loss": 0.9762,
       "step": 127710
     },
     {
       "epoch": 2.0750272132053094,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9306,
       "step": 127720
     },
     {
       "epoch": 2.075189680102679,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2529,
+      "loss": 0.9607,
       "step": 127730
     },
     {
       "epoch": 2.0753521470000487,
-      "grad_norm": 3.390625,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9141,
       "step": 127740
     },
     {
       "epoch": 2.0755146138974183,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2447,
+      "loss": 0.9447,
       "step": 127750
     },
     {
       "epoch": 2.075677080794788,
-      "grad_norm": 3.984375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9822,
       "step": 127760
     },
     {
       "epoch": 2.0758395476921576,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2432,
+      "loss": 0.9582,
       "step": 127770
     },
     {
       "epoch": 2.0760020145895273,
-      "grad_norm": 4.46875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2512,
+      "loss": 0.949,
       "step": 127780
     },
     {
       "epoch": 2.076164481486897,
-      "grad_norm": 4.375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.277,
+      "loss": 0.9717,
       "step": 127790
     },
     {
       "epoch": 2.0763269483842666,
-      "grad_norm": 3.125,
+      "grad_norm": 12.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2358,
+      "loss": 0.9739,
       "step": 127800
     },
     {
       "epoch": 2.076489415281636,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.238,
+      "loss": 0.9547,
       "step": 127810
     },
     {
       "epoch": 2.076651882179006,
-      "grad_norm": 3.390625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2421,
+      "loss": 0.9488,
       "step": 127820
     },
     {
       "epoch": 2.0768143490763755,
-      "grad_norm": 3.015625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2616,
+      "loss": 0.9293,
       "step": 127830
     },
     {
       "epoch": 2.076976815973745,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9376,
       "step": 127840
     },
     {
       "epoch": 2.077139282871115,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.275,
+      "loss": 0.9485,
       "step": 127850
     },
     {
       "epoch": 2.0773017497684845,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.268,
+      "loss": 0.9729,
       "step": 127860
     },
     {
       "epoch": 2.077464216665854,
-      "grad_norm": 3.09375,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2518,
+      "loss": 0.9543,
       "step": 127870
     },
     {
       "epoch": 2.077626683563224,
-      "grad_norm": 4.875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.9653,
       "step": 127880
     },
     {
       "epoch": 2.0777891504605934,
-      "grad_norm": 3.234375,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2532,
+      "loss": 0.9955,
       "step": 127890
     },
     {
       "epoch": 2.0779516173579635,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2623,
+      "loss": 0.9311,
       "step": 127900
     },
     {
       "epoch": 2.078114084255333,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2659,
+      "loss": 0.987,
       "step": 127910
     },
     {
       "epoch": 2.078276551152703,
-      "grad_norm": 4.53125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2665,
+      "loss": 0.9153,
       "step": 127920
     },
     {
       "epoch": 2.0784390180500725,
-      "grad_norm": 4.15625,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2823,
+      "loss": 0.9591,
       "step": 127930
     },
     {
       "epoch": 2.078601484947442,
-      "grad_norm": 2.921875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2943,
+      "loss": 0.9251,
       "step": 127940
     },
     {
       "epoch": 2.0787639518448118,
-      "grad_norm": 3.03125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2806,
+      "loss": 0.9466,
       "step": 127950
     },
     {
       "epoch": 2.0789264187421814,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2441,
+      "loss": 0.9606,
       "step": 127960
     },
     {
       "epoch": 2.079088885639551,
-      "grad_norm": 5.90625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2786,
+      "loss": 0.9557,
       "step": 127970
     },
     {
       "epoch": 2.0792513525369207,
-      "grad_norm": 3.359375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2539,
+      "loss": 0.9188,
       "step": 127980
     },
     {
       "epoch": 2.0794138194342904,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2736,
+      "loss": 0.9509,
       "step": 127990
     },
     {
       "epoch": 2.07957628633166,
-      "grad_norm": 4.1875,
+      "grad_norm": 12.125,
       "learning_rate": 5e-05,
-      "loss": 0.2811,
+      "loss": 0.9744,
       "step": 128000
     },
     {
       "epoch": 2.0797387532290297,
-      "grad_norm": 2.625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2595,
+      "loss": 0.9636,
       "step": 128010
     },
     {
       "epoch": 2.0799012201263993,
-      "grad_norm": 3.0,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 0.9472,
       "step": 128020
     },
     {
       "epoch": 2.080063687023769,
-      "grad_norm": 3.984375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9779,
       "step": 128030
     },
     {
       "epoch": 2.0802261539211386,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2404,
+      "loss": 0.9289,
       "step": 128040
     },
     {
       "epoch": 2.0803886208185083,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2772,
+      "loss": 0.9428,
       "step": 128050
     },
     {
       "epoch": 2.080551087715878,
-      "grad_norm": 3.265625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2635,
+      "loss": 0.9449,
       "step": 128060
     },
     {
       "epoch": 2.0807135546132476,
-      "grad_norm": 3.046875,
+      "grad_norm": 12.125,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 0.9316,
       "step": 128070
     },
     {
       "epoch": 2.0808760215106172,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2652,
+      "loss": 0.9616,
       "step": 128080
     },
     {
       "epoch": 2.081038488407987,
-      "grad_norm": 3.890625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.237,
+      "loss": 0.9387,
       "step": 128090
     },
     {
       "epoch": 2.0812009553053565,
-      "grad_norm": 3.265625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2436,
+      "loss": 0.994,
       "step": 128100
     },
     {
       "epoch": 2.081363422202726,
-      "grad_norm": 2.484375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 0.9579,
       "step": 128110
     },
     {
       "epoch": 2.081525889100096,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2412,
+      "loss": 0.9699,
       "step": 128120
     },
     {
       "epoch": 2.0816883559974655,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2538,
+      "loss": 0.9364,
       "step": 128130
     },
     {
       "epoch": 2.081850822894835,
-      "grad_norm": 4.84375,
+      "grad_norm": 13.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2488,
+      "loss": 0.977,
       "step": 128140
     },
     {
       "epoch": 2.082013289792205,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2536,
+      "loss": 0.9572,
       "step": 128150
     },
     {
       "epoch": 2.0821757566895744,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2399,
+      "loss": 0.9488,
       "step": 128160
     },
     {
       "epoch": 2.082338223586944,
-      "grad_norm": 3.5,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.2194,
+      "loss": 0.983,
       "step": 128170
     },
     {
       "epoch": 2.0825006904843137,
-      "grad_norm": 3.59375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2442,
+      "loss": 0.9254,
       "step": 128180
     },
     {
       "epoch": 2.0826631573816834,
-      "grad_norm": 3.109375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9366,
       "step": 128190
     },
     {
       "epoch": 2.082825624279053,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2551,
+      "loss": 0.9372,
       "step": 128200
     },
     {
       "epoch": 2.0829880911764227,
-      "grad_norm": 2.796875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9164,
       "step": 128210
     },
     {
       "epoch": 2.0831505580737923,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9836,
       "step": 128220
     },
     {
       "epoch": 2.083313024971162,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2566,
+      "loss": 0.9551,
       "step": 128230
     },
     {
       "epoch": 2.0834754918685316,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9554,
       "step": 128240
     },
     {
       "epoch": 2.0836379587659013,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2417,
+      "loss": 0.9606,
       "step": 128250
     },
     {
       "epoch": 2.083800425663271,
-      "grad_norm": 2.734375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2515,
+      "loss": 0.9574,
       "step": 128260
     },
     {
       "epoch": 2.0839628925606406,
-      "grad_norm": 3.125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2509,
+      "loss": 0.9836,
       "step": 128270
     },
     {
       "epoch": 2.0841253594580103,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2598,
+      "loss": 0.9351,
       "step": 128280
     },
     {
       "epoch": 2.08428782635538,
-      "grad_norm": 4.125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2382,
+      "loss": 0.939,
       "step": 128290
     },
     {
       "epoch": 2.0844502932527496,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2258,
+      "loss": 0.9555,
       "step": 128300
     },
     {
       "epoch": 2.084612760150119,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2373,
+      "loss": 0.9308,
       "step": 128310
     },
     {
       "epoch": 2.084775227047489,
-      "grad_norm": 4.5,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2502,
+      "loss": 0.9397,
       "step": 128320
     },
     {
       "epoch": 2.084937693944859,
-      "grad_norm": 4.46875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2755,
+      "loss": 0.9707,
       "step": 128330
     },
     {
       "epoch": 2.0851001608422286,
-      "grad_norm": 3.734375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2617,
+      "loss": 0.9489,
       "step": 128340
     },
     {
       "epoch": 2.0852626277395983,
-      "grad_norm": 3.828125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9254,
       "step": 128350
     },
     {
       "epoch": 2.085425094636968,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2522,
+      "loss": 0.9701,
       "step": 128360
     },
     {
       "epoch": 2.0855875615343376,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2731,
+      "loss": 0.9635,
       "step": 128370
     },
     {
       "epoch": 2.085750028431707,
-      "grad_norm": 5.53125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2672,
+      "loss": 0.9233,
       "step": 128380
     },
     {
       "epoch": 2.085912495329077,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2634,
+      "loss": 0.9358,
       "step": 128390
     },
     {
       "epoch": 2.0860749622264465,
-      "grad_norm": 2.6875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2491,
+      "loss": 0.9724,
       "step": 128400
     },
     {
       "epoch": 2.086237429123816,
-      "grad_norm": 3.96875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2551,
+      "loss": 0.9457,
       "step": 128410
     },
     {
       "epoch": 2.086399896021186,
-      "grad_norm": 3.734375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9284,
       "step": 128420
     },
     {
       "epoch": 2.0865623629185555,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.275,
+      "loss": 0.9407,
       "step": 128430
     },
     {
       "epoch": 2.086724829815925,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2775,
+      "loss": 0.958,
       "step": 128440
     },
     {
       "epoch": 2.0868872967132948,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2747,
+      "loss": 0.964,
       "step": 128450
     },
     {
       "epoch": 2.0870497636106644,
-      "grad_norm": 3.4375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2717,
+      "loss": 0.9312,
       "step": 128460
     },
     {
       "epoch": 2.087212230508034,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2773,
+      "loss": 0.96,
       "step": 128470
     },
     {
       "epoch": 2.0873746974054037,
-      "grad_norm": 2.09375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9687,
       "step": 128480
     },
     {
       "epoch": 2.0875371643027734,
-      "grad_norm": 3.5,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2591,
+      "loss": 0.9486,
       "step": 128490
     },
     {
       "epoch": 2.087699631200143,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2396,
+      "loss": 0.9575,
       "step": 128500
     },
     {
       "epoch": 2.0878620980975127,
-      "grad_norm": 4.125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2423,
+      "loss": 0.9619,
       "step": 128510
     },
     {
       "epoch": 2.0880245649948823,
-      "grad_norm": 2.671875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2527,
+      "loss": 0.9855,
       "step": 128520
     },
     {
       "epoch": 2.088187031892252,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2614,
+      "loss": 0.9634,
       "step": 128530
     },
     {
       "epoch": 2.0883494987896216,
-      "grad_norm": 2.703125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9546,
       "step": 128540
     },
     {
       "epoch": 2.0885119656869913,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2438,
+      "loss": 0.9425,
       "step": 128550
     },
     {
       "epoch": 2.088674432584361,
-      "grad_norm": 3.625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.9404,
       "step": 128560
     },
     {
       "epoch": 2.0888368994817306,
-      "grad_norm": 3.984375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.285,
+      "loss": 0.9434,
       "step": 128570
     },
     {
       "epoch": 2.0889993663791,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2718,
+      "loss": 0.9503,
       "step": 128580
     },
     {
       "epoch": 2.08916183327647,
-      "grad_norm": 3.625,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2809,
+      "loss": 0.9638,
       "step": 128590
     },
     {
       "epoch": 2.0893243001738395,
-      "grad_norm": 3.109375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2509,
+      "loss": 0.9087,
       "step": 128600
     },
     {
       "epoch": 2.089486767071209,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2772,
+      "loss": 0.9768,
       "step": 128610
     },
     {
       "epoch": 2.089649233968579,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2496,
+      "loss": 0.9287,
       "step": 128620
     },
     {
       "epoch": 2.0898117008659485,
-      "grad_norm": 3.921875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2569,
+      "loss": 0.9516,
       "step": 128630
     },
     {
       "epoch": 2.089974167763318,
-      "grad_norm": 3.171875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2347,
+      "loss": 0.9732,
       "step": 128640
     },
     {
       "epoch": 2.0901366346606878,
-      "grad_norm": 3.984375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2385,
+      "loss": 0.9685,
       "step": 128650
     },
     {
       "epoch": 2.0902991015580574,
-      "grad_norm": 3.265625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2394,
+      "loss": 0.9491,
       "step": 128660
     },
     {
       "epoch": 2.090461568455427,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2368,
+      "loss": 0.9202,
       "step": 128670
     },
     {
       "epoch": 2.0906240353527967,
-      "grad_norm": 5.46875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2469,
+      "loss": 0.9611,
       "step": 128680
     },
     {
       "epoch": 2.0907865022501664,
-      "grad_norm": 2.765625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2862,
+      "loss": 0.9531,
       "step": 128690
     },
     {
       "epoch": 2.090948969147536,
-      "grad_norm": 4.28125,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 0.9445,
       "step": 128700
     },
     {
       "epoch": 2.0911114360449057,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2694,
+      "loss": 0.9504,
       "step": 128710
     },
     {
       "epoch": 2.0912739029422753,
-      "grad_norm": 2.859375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9655,
       "step": 128720
     },
     {
       "epoch": 2.091436369839645,
-      "grad_norm": 3.625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2934,
+      "loss": 0.9673,
       "step": 128730
     },
     {
       "epoch": 2.0915988367370146,
-      "grad_norm": 3.484375,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.9103,
       "step": 128740
     },
     {
       "epoch": 2.0917613036343843,
-      "grad_norm": 4.6875,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2675,
+      "loss": 0.9589,
       "step": 128750
     },
     {
       "epoch": 2.0919237705317544,
-      "grad_norm": 4.28125,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2719,
+      "loss": 0.9252,
       "step": 128760
     },
     {
       "epoch": 2.092086237429124,
-      "grad_norm": 3.875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9697,
       "step": 128770
     },
     {
       "epoch": 2.0922487043264937,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2781,
+      "loss": 0.9356,
       "step": 128780
     },
     {
       "epoch": 2.0924111712238633,
-      "grad_norm": 3.875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.252,
+      "loss": 0.929,
       "step": 128790
     },
     {
       "epoch": 2.092573638121233,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2379,
+      "loss": 0.9421,
       "step": 128800
     },
     {
       "epoch": 2.0927361050186026,
-      "grad_norm": 2.90625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2357,
+      "loss": 0.9276,
       "step": 128810
     },
     {
       "epoch": 2.0928985719159723,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2112,
+      "loss": 0.9651,
       "step": 128820
     },
     {
       "epoch": 2.093061038813342,
-      "grad_norm": 3.4375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.236,
+      "loss": 0.9757,
       "step": 128830
     },
     {
       "epoch": 2.0932235057107116,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 0.919,
       "step": 128840
     },
     {
       "epoch": 2.0933859726080812,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2708,
+      "loss": 0.9525,
       "step": 128850
     },
     {
       "epoch": 2.093548439505451,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2346,
+      "loss": 0.9368,
       "step": 128860
     },
     {
       "epoch": 2.0937109064028205,
-      "grad_norm": 4.3125,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2522,
+      "loss": 0.9509,
       "step": 128870
     },
     {
       "epoch": 2.09387337330019,
-      "grad_norm": 4.34375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.251,
+      "loss": 0.9635,
       "step": 128880
     },
     {
       "epoch": 2.09403584019756,
-      "grad_norm": 2.40625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9617,
       "step": 128890
     },
     {
       "epoch": 2.0941983070949295,
-      "grad_norm": 2.78125,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.9691,
       "step": 128900
     },
     {
       "epoch": 2.094360773992299,
-      "grad_norm": 5.375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.263,
+      "loss": 0.9959,
       "step": 128910
     },
     {
       "epoch": 2.094523240889669,
-      "grad_norm": 3.1875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2539,
+      "loss": 0.9858,
       "step": 128920
     },
     {
       "epoch": 2.0946857077870384,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 0.9334,
       "step": 128930
     },
     {
       "epoch": 2.094848174684408,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 0.976,
       "step": 128940
     },
     {
       "epoch": 2.0950106415817777,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2944,
+      "loss": 0.9275,
       "step": 128950
     },
     {
       "epoch": 2.0951731084791474,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2736,
+      "loss": 0.985,
       "step": 128960
     },
     {
       "epoch": 2.095335575376517,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2677,
+      "loss": 0.9631,
       "step": 128970
     },
     {
       "epoch": 2.0954980422738867,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2677,
+      "loss": 0.9435,
       "step": 128980
     },
     {
       "epoch": 2.0956605091712563,
-      "grad_norm": 3.4375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2552,
+      "loss": 0.9471,
       "step": 128990
     },
     {
       "epoch": 2.095822976068626,
-      "grad_norm": 3.5,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2414,
+      "loss": 0.9594,
       "step": 129000
     },
     {
       "epoch": 2.0959854429659956,
-      "grad_norm": 3.546875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9578,
       "step": 129010
     },
     {
       "epoch": 2.0961479098633653,
-      "grad_norm": 3.15625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9501,
       "step": 129020
     },
     {
       "epoch": 2.096310376760735,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.267,
+      "loss": 0.9466,
       "step": 129030
     },
     {
       "epoch": 2.0964728436581046,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2425,
+      "loss": 0.9414,
       "step": 129040
     },
     {
       "epoch": 2.0966353105554743,
-      "grad_norm": 2.875,
+      "grad_norm": 12.0,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 1.0048,
       "step": 129050
     },
     {
       "epoch": 2.096797777452844,
-      "grad_norm": 3.328125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2465,
+      "loss": 0.9503,
       "step": 129060
     },
     {
       "epoch": 2.0969602443502136,
-      "grad_norm": 3.875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2298,
+      "loss": 0.9903,
       "step": 129070
     },
     {
       "epoch": 2.097122711247583,
-      "grad_norm": 3.59375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2383,
+      "loss": 0.9586,
       "step": 129080
     },
     {
       "epoch": 2.097285178144953,
-      "grad_norm": 3.96875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.265,
+      "loss": 0.9667,
       "step": 129090
     },
     {
       "epoch": 2.0974476450423225,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.9455,
       "step": 129100
     },
     {
       "epoch": 2.097610111939692,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2717,
+      "loss": 0.9502,
       "step": 129110
     },
     {
       "epoch": 2.097772578837062,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2394,
+      "loss": 0.9499,
       "step": 129120
     },
     {
       "epoch": 2.0979350457344315,
-      "grad_norm": 3.84375,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2769,
+      "loss": 0.9295,
       "step": 129130
     },
     {
       "epoch": 2.098097512631801,
-      "grad_norm": 2.734375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2401,
+      "loss": 0.9712,
       "step": 129140
     },
     {
       "epoch": 2.0982599795291708,
-      "grad_norm": 3.25,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 0.9512,
       "step": 129150
     },
     {
       "epoch": 2.0984224464265404,
-      "grad_norm": 4.625,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2459,
+      "loss": 0.9582,
       "step": 129160
     },
     {
       "epoch": 2.09858491332391,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2389,
+      "loss": 0.9488,
       "step": 129170
     },
     {
       "epoch": 2.0987473802212797,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9478,
       "step": 129180
     },
     {
       "epoch": 2.0989098471186494,
-      "grad_norm": 3.578125,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2424,
+      "loss": 0.9591,
       "step": 129190
     },
     {
       "epoch": 2.099072314016019,
-      "grad_norm": 4.25,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9081,
       "step": 129200
     },
     {
       "epoch": 2.099234780913389,
-      "grad_norm": 4.59375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 0.9621,
       "step": 129210
     },
     {
       "epoch": 2.0993972478107588,
-      "grad_norm": 4.78125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 0.9483,
       "step": 129220
     },
     {
       "epoch": 2.0995597147081284,
-      "grad_norm": 4.84375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2542,
+      "loss": 0.9315,
       "step": 129230
     },
     {
       "epoch": 2.099722181605498,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.24,
+      "loss": 0.9974,
       "step": 129240
     },
     {
       "epoch": 2.0998846485028677,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9417,
       "step": 129250
     },
     {
       "epoch": 2.1000471154002374,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2868,
+      "loss": 0.9424,
       "step": 129260
     },
     {
       "epoch": 2.100209582297607,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2714,
+      "loss": 0.9445,
       "step": 129270
     },
     {
       "epoch": 2.1003720491949767,
-      "grad_norm": 2.921875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2617,
+      "loss": 0.9509,
       "step": 129280
     },
     {
       "epoch": 2.1005345160923463,
-      "grad_norm": 4.0625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2868,
+      "loss": 0.9578,
       "step": 129290
     },
     {
       "epoch": 2.100696982989716,
-      "grad_norm": 3.640625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 0.9653,
       "step": 129300
     },
     {
       "epoch": 2.1008594498870856,
-      "grad_norm": 3.953125,
+      "grad_norm": 12.375,
       "learning_rate": 5e-05,
-      "loss": 0.2654,
+      "loss": 0.9831,
       "step": 129310
     },
     {
       "epoch": 2.1010219167844553,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.251,
+      "loss": 0.9605,
       "step": 129320
     },
     {
       "epoch": 2.101184383681825,
-      "grad_norm": 3.25,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2432,
+      "loss": 0.9578,
       "step": 129330
     },
     {
       "epoch": 2.1013468505791946,
-      "grad_norm": 4.59375,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2668,
+      "loss": 0.9551,
       "step": 129340
     },
     {
       "epoch": 2.101509317476564,
-      "grad_norm": 2.609375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.269,
+      "loss": 0.9413,
       "step": 129350
     },
     {
       "epoch": 2.101671784373934,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2642,
+      "loss": 0.9645,
       "step": 129360
     },
     {
       "epoch": 2.1018342512713035,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2464,
+      "loss": 0.92,
       "step": 129370
     },
     {
       "epoch": 2.101996718168673,
-      "grad_norm": 3.25,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2529,
+      "loss": 0.9781,
       "step": 129380
     },
     {
       "epoch": 2.102159185066043,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9145,
       "step": 129390
     },
     {
       "epoch": 2.1023216519634125,
-      "grad_norm": 2.78125,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2451,
+      "loss": 0.9905,
       "step": 129400
     },
     {
       "epoch": 2.102484118860782,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2828,
+      "loss": 0.986,
       "step": 129410
     },
     {
       "epoch": 2.1026465857581518,
-      "grad_norm": 3.703125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2702,
+      "loss": 0.9158,
       "step": 129420
     },
     {
       "epoch": 2.1028090526555214,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2803,
+      "loss": 0.9589,
       "step": 129430
     },
     {
       "epoch": 2.102971519552891,
-      "grad_norm": 4.125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2632,
+      "loss": 0.9801,
       "step": 129440
     },
     {
       "epoch": 2.1031339864502607,
-      "grad_norm": 2.890625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2532,
+      "loss": 0.975,
       "step": 129450
     },
     {
       "epoch": 2.1032964533476304,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2477,
+      "loss": 0.9748,
       "step": 129460
     },
     {
       "epoch": 2.103458920245,
-      "grad_norm": 2.21875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 0.9694,
       "step": 129470
     },
     {
       "epoch": 2.1036213871423697,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2858,
+      "loss": 0.9748,
       "step": 129480
     },
     {
       "epoch": 2.1037838540397393,
-      "grad_norm": 2.90625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2644,
+      "loss": 0.9097,
       "step": 129490
     },
     {
       "epoch": 2.103946320937109,
-      "grad_norm": 4.46875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2424,
+      "loss": 0.947,
       "step": 129500
     },
     {
       "epoch": 2.1041087878344786,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.271,
+      "loss": 0.9754,
       "step": 129510
     },
     {
       "epoch": 2.1042712547318483,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2708,
+      "loss": 0.9513,
       "step": 129520
     },
     {
       "epoch": 2.104433721629218,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2568,
+      "loss": 0.9439,
       "step": 129530
     },
     {
       "epoch": 2.1045961885265876,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2551,
+      "loss": 0.946,
       "step": 129540
     },
     {
       "epoch": 2.1047586554239572,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.263,
+      "loss": 0.9447,
       "step": 129550
     },
     {
       "epoch": 2.104921122321327,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2485,
+      "loss": 0.939,
       "step": 129560
     },
     {
       "epoch": 2.1050835892186965,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2445,
+      "loss": 0.9654,
       "step": 129570
     },
     {
       "epoch": 2.105246056116066,
-      "grad_norm": 4.125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9586,
       "step": 129580
     },
     {
       "epoch": 2.105408523013436,
-      "grad_norm": 3.875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2568,
+      "loss": 0.9247,
       "step": 129590
     },
     {
       "epoch": 2.1055709899108055,
-      "grad_norm": 3.75,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2507,
+      "loss": 0.9601,
       "step": 129600
     },
     {
       "epoch": 2.105733456808175,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.9629,
       "step": 129610
     },
     {
       "epoch": 2.105895923705545,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2526,
+      "loss": 0.9267,
       "step": 129620
     },
     {
       "epoch": 2.1060583906029144,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2724,
+      "loss": 0.9601,
       "step": 129630
     },
     {
       "epoch": 2.1062208575002845,
-      "grad_norm": 2.75,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2454,
+      "loss": 0.9379,
       "step": 129640
     },
     {
       "epoch": 2.106383324397654,
-      "grad_norm": 4.03125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2451,
+      "loss": 0.9431,
       "step": 129650
     },
     {
       "epoch": 2.106545791295024,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2619,
+      "loss": 0.9483,
       "step": 129660
     },
     {
       "epoch": 2.1067082581923935,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9643,
       "step": 129670
     },
     {
       "epoch": 2.106870725089763,
-      "grad_norm": 4.21875,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2572,
+      "loss": 0.9033,
       "step": 129680
     },
     {
       "epoch": 2.107033191987133,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 0.9701,
       "step": 129690
     },
     {
       "epoch": 2.1071956588845024,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.237,
+      "loss": 0.9555,
       "step": 129700
     },
     {
       "epoch": 2.107358125781872,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2427,
+      "loss": 0.941,
       "step": 129710
     },
     {
       "epoch": 2.1075205926792417,
-      "grad_norm": 3.3125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2505,
+      "loss": 0.9399,
       "step": 129720
     },
     {
       "epoch": 2.1076830595766114,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2699,
+      "loss": 0.9253,
       "step": 129730
     },
     {
       "epoch": 2.107845526473981,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9759,
       "step": 129740
     },
     {
       "epoch": 2.1080079933713507,
-      "grad_norm": 3.25,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2636,
+      "loss": 0.9666,
       "step": 129750
     },
     {
       "epoch": 2.1081704602687203,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 0.9399,
       "step": 129760
     },
     {
       "epoch": 2.10833292716609,
-      "grad_norm": 3.9375,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2591,
+      "loss": 0.9356,
       "step": 129770
     },
     {
       "epoch": 2.1084953940634596,
-      "grad_norm": 2.734375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2479,
+      "loss": 0.9279,
       "step": 129780
     },
     {
       "epoch": 2.1086578609608293,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.226,
+      "loss": 0.9348,
       "step": 129790
     },
     {
       "epoch": 2.108820327858199,
-      "grad_norm": 3.984375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9444,
       "step": 129800
     },
     {
       "epoch": 2.1089827947555686,
-      "grad_norm": 3.75,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9247,
       "step": 129810
     },
     {
       "epoch": 2.1091452616529383,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2692,
+      "loss": 0.9741,
       "step": 129820
     },
     {
       "epoch": 2.109307728550308,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2752,
+      "loss": 0.9664,
       "step": 129830
     },
     {
       "epoch": 2.1094701954476776,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2557,
+      "loss": 0.9464,
       "step": 129840
     },
     {
       "epoch": 2.109632662345047,
-      "grad_norm": 2.75,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2561,
+      "loss": 0.9571,
       "step": 129850
     },
     {
       "epoch": 2.109795129242417,
-      "grad_norm": 3.140625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2501,
+      "loss": 0.9539,
       "step": 129860
     },
     {
       "epoch": 2.1099575961397865,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2756,
+      "loss": 0.9853,
       "step": 129870
     },
     {
       "epoch": 2.110120063037156,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2446,
+      "loss": 0.9744,
       "step": 129880
     },
     {
       "epoch": 2.110282529934526,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2452,
+      "loss": 0.9486,
       "step": 129890
     },
     {
       "epoch": 2.1104449968318955,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2816,
+      "loss": 0.9801,
       "step": 129900
     },
     {
       "epoch": 2.110607463729265,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9353,
       "step": 129910
     },
     {
       "epoch": 2.1107699306266348,
-      "grad_norm": 3.234375,
+      "grad_norm": 12.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.9682,
       "step": 129920
     },
     {
       "epoch": 2.1109323975240044,
-      "grad_norm": 3.625,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9549,
       "step": 129930
     },
     {
       "epoch": 2.111094864421374,
-      "grad_norm": 3.875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2472,
+      "loss": 0.9561,
       "step": 129940
     },
     {
       "epoch": 2.1112573313187437,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.9368,
       "step": 129950
     },
     {
       "epoch": 2.1114197982161134,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2464,
+      "loss": 0.9667,
       "step": 129960
     },
     {
       "epoch": 2.111582265113483,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2491,
+      "loss": 0.93,
       "step": 129970
     },
     {
       "epoch": 2.1117447320108527,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2583,
+      "loss": 0.9543,
       "step": 129980
     },
     {
       "epoch": 2.1119071989082223,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.251,
+      "loss": 0.9772,
       "step": 129990
     },
     {
       "epoch": 2.112069665805592,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2678,
+      "loss": 0.9421,
       "step": 130000
     },
     {
       "epoch": 2.1122321327029616,
-      "grad_norm": 3.484375,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9546,
       "step": 130010
     },
     {
       "epoch": 2.1123945996003313,
-      "grad_norm": 3.84375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.255,
+      "loss": 0.9528,
       "step": 130020
     },
     {
       "epoch": 2.112557066497701,
-      "grad_norm": 3.828125,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2622,
+      "loss": 0.9828,
       "step": 130030
     },
     {
       "epoch": 2.1127195333950706,
-      "grad_norm": 2.875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2599,
+      "loss": 0.9519,
       "step": 130040
     },
     {
       "epoch": 2.11288200029244,
-      "grad_norm": 4.3125,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2738,
+      "loss": 0.9904,
       "step": 130050
     },
     {
       "epoch": 2.11304446718981,
-      "grad_norm": 3.40625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9902,
       "step": 130060
     },
     {
       "epoch": 2.1132069340871795,
-      "grad_norm": 3.859375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2884,
+      "loss": 0.944,
       "step": 130070
     },
     {
       "epoch": 2.113369400984549,
-      "grad_norm": 3.25,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.266,
+      "loss": 0.9638,
       "step": 130080
     },
     {
       "epoch": 2.1135318678819193,
-      "grad_norm": 3.515625,
+      "grad_norm": 12.375,
       "learning_rate": 5e-05,
-      "loss": 0.2699,
+      "loss": 0.9558,
       "step": 130090
     },
     {
       "epoch": 2.113694334779289,
-      "grad_norm": 3.8125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.269,
+      "loss": 0.9623,
       "step": 130100
     },
     {
       "epoch": 2.1138568016766586,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9426,
       "step": 130110
     },
     {
       "epoch": 2.114019268574028,
-      "grad_norm": 4.5,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2417,
+      "loss": 0.9509,
       "step": 130120
     },
     {
       "epoch": 2.114181735471398,
-      "grad_norm": 4.125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9105,
       "step": 130130
     },
     {
       "epoch": 2.1143442023687675,
-      "grad_norm": 3.453125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2823,
+      "loss": 0.937,
       "step": 130140
     },
     {
       "epoch": 2.114506669266137,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 0.9584,
       "step": 130150
     },
     {
       "epoch": 2.114669136163507,
-      "grad_norm": 2.84375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2544,
+      "loss": 0.9461,
       "step": 130160
     },
     {
       "epoch": 2.1148316030608765,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2437,
+      "loss": 0.8891,
       "step": 130170
     },
     {
       "epoch": 2.114994069958246,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2679,
+      "loss": 0.9868,
       "step": 130180
     },
     {
       "epoch": 2.1151565368556158,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9355,
       "step": 130190
     },
     {
       "epoch": 2.1153190037529854,
-      "grad_norm": 4.375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2537,
+      "loss": 0.9656,
       "step": 130200
     },
     {
       "epoch": 2.115481470650355,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2292,
+      "loss": 0.9776,
       "step": 130210
     },
     {
       "epoch": 2.1156439375477247,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2295,
+      "loss": 0.9593,
       "step": 130220
     },
     {
       "epoch": 2.1158064044450944,
-      "grad_norm": 3.53125,
+      "grad_norm": 12.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2515,
+      "loss": 0.9514,
       "step": 130230
     },
     {
       "epoch": 2.115968871342464,
-      "grad_norm": 3.109375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2411,
+      "loss": 0.9978,
       "step": 130240
     },
     {
       "epoch": 2.1161313382398337,
-      "grad_norm": 3.625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.9098,
       "step": 130250
     },
     {
       "epoch": 2.1162938051372033,
-      "grad_norm": 3.25,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2497,
+      "loss": 0.9686,
       "step": 130260
     },
     {
       "epoch": 2.116456272034573,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 0.9562,
       "step": 130270
     },
     {
       "epoch": 2.1166187389319426,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9629,
       "step": 130280
     },
     {
       "epoch": 2.1167812058293123,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2363,
+      "loss": 0.9715,
       "step": 130290
     },
     {
       "epoch": 2.116943672726682,
-      "grad_norm": 3.15625,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2531,
+      "loss": 0.9502,
       "step": 130300
     },
     {
       "epoch": 2.1171061396240516,
-      "grad_norm": 4.25,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2538,
+      "loss": 0.9733,
       "step": 130310
     },
     {
       "epoch": 2.1172686065214212,
-      "grad_norm": 3.671875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9476,
       "step": 130320
     },
     {
       "epoch": 2.117431073418791,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2673,
+      "loss": 0.9688,
       "step": 130330
     },
     {
       "epoch": 2.1175935403161605,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2449,
+      "loss": 0.9228,
       "step": 130340
     },
     {
       "epoch": 2.11775600721353,
-      "grad_norm": 3.234375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.261,
+      "loss": 0.9632,
       "step": 130350
     },
     {
       "epoch": 2.1179184741109,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2685,
+      "loss": 0.9329,
       "step": 130360
     },
     {
       "epoch": 2.1180809410082695,
-      "grad_norm": 2.71875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2495,
+      "loss": 0.9907,
       "step": 130370
     },
     {
       "epoch": 2.118243407905639,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2831,
+      "loss": 0.9373,
       "step": 130380
     },
     {
       "epoch": 2.118405874803009,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2832,
+      "loss": 0.9832,
       "step": 130390
     },
     {
       "epoch": 2.1185683417003784,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.9544,
       "step": 130400
     },
     {
       "epoch": 2.118730808597748,
-      "grad_norm": 3.375,
+      "grad_norm": 12.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2595,
+      "loss": 0.9616,
       "step": 130410
     },
     {
       "epoch": 2.1188932754951177,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2654,
+      "loss": 0.9192,
       "step": 130420
     },
     {
       "epoch": 2.1190557423924874,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2976,
+      "loss": 0.9573,
       "step": 130430
     },
     {
       "epoch": 2.119218209289857,
-      "grad_norm": 3.984375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2816,
+      "loss": 0.9852,
       "step": 130440
     },
     {
       "epoch": 2.1193806761872267,
-      "grad_norm": 2.921875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2397,
+      "loss": 0.9299,
       "step": 130450
     },
     {
       "epoch": 2.1195431430845963,
-      "grad_norm": 3.546875,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2508,
+      "loss": 0.9596,
       "step": 130460
     },
     {
       "epoch": 2.119705609981966,
-      "grad_norm": 3.359375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2755,
+      "loss": 0.9405,
       "step": 130470
     },
     {
       "epoch": 2.1198680768793356,
-      "grad_norm": 3.953125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2455,
+      "loss": 0.9458,
       "step": 130480
     },
     {
       "epoch": 2.1200305437767053,
-      "grad_norm": 3.640625,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2306,
+      "loss": 0.9777,
       "step": 130490
     },
     {
       "epoch": 2.120193010674075,
-      "grad_norm": 3.78125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2513,
+      "loss": 0.9749,
       "step": 130500
     },
     {
       "epoch": 2.1203554775714446,
-      "grad_norm": 4.0,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2431,
+      "loss": 0.9889,
       "step": 130510
     },
     {
       "epoch": 2.1205179444688147,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2274,
+      "loss": 0.9481,
       "step": 130520
     },
     {
       "epoch": 2.1206804113661843,
-      "grad_norm": 4.5,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2254,
+      "loss": 0.9523,
       "step": 130530
     },
     {
       "epoch": 2.120842878263554,
-      "grad_norm": 3.859375,
+      "grad_norm": 7.90625,
       "learning_rate": 5e-05,
-      "loss": 0.2298,
+      "loss": 0.9365,
       "step": 130540
     },
     {
       "epoch": 2.1210053451609236,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2503,
+      "loss": 0.9824,
       "step": 130550
     },
     {
       "epoch": 2.1211678120582933,
-      "grad_norm": 3.1875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2466,
+      "loss": 0.9654,
       "step": 130560
     },
     {
       "epoch": 2.121330278955663,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.244,
+      "loss": 0.9724,
       "step": 130570
     },
     {
       "epoch": 2.1214927458530326,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2769,
+      "loss": 0.9506,
       "step": 130580
     },
     {
       "epoch": 2.1216552127504023,
-      "grad_norm": 3.234375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2591,
+      "loss": 0.9938,
       "step": 130590
     },
     {
       "epoch": 2.121817679647772,
-      "grad_norm": 3.34375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2713,
+      "loss": 0.9381,
       "step": 130600
     },
     {
       "epoch": 2.1219801465451416,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2834,
+      "loss": 0.9488,
       "step": 130610
     },
     {
       "epoch": 2.122142613442511,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2624,
+      "loss": 0.962,
       "step": 130620
     },
     {
       "epoch": 2.122305080339881,
-      "grad_norm": 4.125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2467,
+      "loss": 0.9854,
       "step": 130630
     },
     {
       "epoch": 2.1224675472372505,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2627,
+      "loss": 0.9668,
       "step": 130640
     },
     {
       "epoch": 2.12263001413462,
-      "grad_norm": 3.9375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2452,
+      "loss": 0.964,
       "step": 130650
     },
     {
       "epoch": 2.12279248103199,
-      "grad_norm": 6.0625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2546,
+      "loss": 0.9565,
       "step": 130660
     },
     {
       "epoch": 2.1229549479293595,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2511,
+      "loss": 0.9686,
       "step": 130670
     },
     {
       "epoch": 2.123117414826729,
-      "grad_norm": 3.0625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2622,
+      "loss": 0.9988,
       "step": 130680
     },
     {
       "epoch": 2.1232798817240988,
-      "grad_norm": 3.625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2733,
+      "loss": 0.9477,
       "step": 130690
     },
     {
       "epoch": 2.1234423486214684,
-      "grad_norm": 5.125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2564,
+      "loss": 0.9501,
       "step": 130700
     },
     {
       "epoch": 2.123604815518838,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9615,
       "step": 130710
     },
     {
       "epoch": 2.1237672824162077,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2502,
+      "loss": 0.9459,
       "step": 130720
     },
     {
       "epoch": 2.1239297493135774,
-      "grad_norm": 4.375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2846,
+      "loss": 0.9644,
       "step": 130730
     },
     {
       "epoch": 2.124092216210947,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2669,
+      "loss": 0.9516,
       "step": 130740
     },
     {
       "epoch": 2.1242546831083167,
-      "grad_norm": 3.625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2769,
+      "loss": 0.9463,
       "step": 130750
     },
     {
       "epoch": 2.1244171500056863,
-      "grad_norm": 4.0,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2553,
+      "loss": 0.9876,
       "step": 130760
     },
     {
       "epoch": 2.124579616903056,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2565,
+      "loss": 0.9122,
       "step": 130770
     },
     {
       "epoch": 2.1247420838004256,
-      "grad_norm": 4.90625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2588,
+      "loss": 0.912,
       "step": 130780
     },
     {
       "epoch": 2.1249045506977953,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.259,
+      "loss": 0.9367,
       "step": 130790
     },
     {
       "epoch": 2.125067017595165,
-      "grad_norm": 3.875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.274,
+      "loss": 0.9364,
       "step": 130800
     },
     {
       "epoch": 2.1252294844925346,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2568,
+      "loss": 0.9527,
       "step": 130810
     },
     {
       "epoch": 2.125391951389904,
-      "grad_norm": 3.90625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2659,
+      "loss": 0.9229,
       "step": 130820
     },
     {
       "epoch": 2.125554418287274,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2836,
+      "loss": 0.9522,
       "step": 130830
     },
     {
       "epoch": 2.1257168851846435,
-      "grad_norm": 4.625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2764,
+      "loss": 0.9623,
       "step": 130840
     },
     {
       "epoch": 2.125879352082013,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2711,
+      "loss": 0.9605,
       "step": 130850
     },
     {
       "epoch": 2.126041818979383,
-      "grad_norm": 3.875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 0.9372,
       "step": 130860
     },
     {
       "epoch": 2.1262042858767525,
-      "grad_norm": 4.65625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2518,
+      "loss": 0.9546,
       "step": 130870
     },
     {
       "epoch": 2.126366752774122,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2603,
+      "loss": 0.991,
       "step": 130880
     },
     {
       "epoch": 2.1265292196714918,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2543,
+      "loss": 0.9219,
       "step": 130890
     },
     {
       "epoch": 2.1266916865688614,
-      "grad_norm": 2.875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2412,
+      "loss": 0.9647,
       "step": 130900
     },
     {
       "epoch": 2.126854153466231,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2153,
+      "loss": 0.9349,
       "step": 130910
     },
     {
       "epoch": 2.1270166203636007,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2418,
+      "loss": 0.9724,
       "step": 130920
     },
     {
       "epoch": 2.1271790872609704,
-      "grad_norm": 4.125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2502,
+      "loss": 0.9529,
       "step": 130930
     },
     {
       "epoch": 2.12734155415834,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2533,
+      "loss": 0.9703,
       "step": 130940
     },
     {
       "epoch": 2.12750402105571,
-      "grad_norm": 3.546875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9433,
       "step": 130950
     },
     {
       "epoch": 2.1276664879530793,
-      "grad_norm": 2.875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2742,
+      "loss": 0.9471,
       "step": 130960
     },
     {
       "epoch": 2.1278289548504494,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2564,
+      "loss": 0.9594,
       "step": 130970
     },
     {
       "epoch": 2.127991421747819,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.9826,
       "step": 130980
     },
     {
       "epoch": 2.1281538886451887,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2413,
+      "loss": 0.9399,
       "step": 130990
     },
     {
       "epoch": 2.1283163555425584,
-      "grad_norm": 3.703125,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2173,
+      "loss": 0.9564,
       "step": 131000
     },
     {
       "epoch": 2.128478822439928,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2428,
+      "loss": 0.9708,
       "step": 131010
     },
     {
       "epoch": 2.1286412893372977,
-      "grad_norm": 3.328125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2565,
+      "loss": 0.9618,
       "step": 131020
     },
     {
       "epoch": 2.1288037562346673,
-      "grad_norm": 3.46875,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.979,
       "step": 131030
     },
     {
       "epoch": 2.128966223132037,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9678,
       "step": 131040
     },
     {
       "epoch": 2.1291286900294066,
-      "grad_norm": 2.609375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 0.9479,
       "step": 131050
     },
     {
       "epoch": 2.1292911569267763,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2416,
+      "loss": 0.9425,
       "step": 131060
     },
     {
       "epoch": 2.129453623824146,
-      "grad_norm": 3.625,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2337,
+      "loss": 0.9715,
       "step": 131070
     },
     {
       "epoch": 2.1296160907215156,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2435,
+      "loss": 0.9498,
       "step": 131080
     },
     {
       "epoch": 2.1297785576188852,
-      "grad_norm": 3.359375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2481,
+      "loss": 0.9594,
       "step": 131090
     },
     {
       "epoch": 2.129941024516255,
-      "grad_norm": 4.375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2448,
+      "loss": 0.9506,
       "step": 131100
     },
     {
       "epoch": 2.1301034914136245,
-      "grad_norm": 4.78125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2729,
+      "loss": 0.9669,
       "step": 131110
     },
     {
       "epoch": 2.130265958310994,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2589,
+      "loss": 0.9618,
       "step": 131120
     },
     {
       "epoch": 2.130428425208364,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9646,
       "step": 131130
     },
     {
       "epoch": 2.1305908921057335,
-      "grad_norm": 4.65625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2736,
+      "loss": 0.9229,
       "step": 131140
     },
     {
       "epoch": 2.130753359003103,
-      "grad_norm": 2.90625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2494,
+      "loss": 0.9503,
       "step": 131150
     },
     {
       "epoch": 2.130915825900473,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2518,
+      "loss": 0.9404,
       "step": 131160
     },
     {
       "epoch": 2.1310782927978424,
-      "grad_norm": 3.75,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2504,
+      "loss": 0.9724,
       "step": 131170
     },
     {
       "epoch": 2.131240759695212,
-      "grad_norm": 3.71875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2617,
+      "loss": 0.9483,
       "step": 131180
     },
     {
       "epoch": 2.1314032265925817,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2482,
+      "loss": 0.9592,
       "step": 131190
     },
     {
       "epoch": 2.1315656934899514,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.25,
+      "loss": 0.9647,
       "step": 131200
     },
     {
       "epoch": 2.131728160387321,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2227,
+      "loss": 0.9291,
       "step": 131210
     },
     {
       "epoch": 2.1318906272846907,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2438,
+      "loss": 0.9591,
       "step": 131220
     },
     {
       "epoch": 2.1320530941820603,
-      "grad_norm": 5.65625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.9661,
       "step": 131230
     },
     {
       "epoch": 2.13221556107943,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9994,
       "step": 131240
     },
     {
       "epoch": 2.1323780279767997,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9878,
       "step": 131250
     },
     {
       "epoch": 2.1325404948741693,
-      "grad_norm": 3.0625,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2499,
+      "loss": 0.9646,
       "step": 131260
     },
     {
       "epoch": 2.132702961771539,
-      "grad_norm": 3.5,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2575,
+      "loss": 0.9656,
       "step": 131270
     },
     {
       "epoch": 2.1328654286689086,
-      "grad_norm": 3.34375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2594,
+      "loss": 0.982,
       "step": 131280
     },
     {
       "epoch": 2.1330278955662783,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2783,
+      "loss": 0.9596,
       "step": 131290
     },
     {
       "epoch": 2.133190362463648,
-      "grad_norm": 4.125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2841,
+      "loss": 0.946,
       "step": 131300
     },
     {
       "epoch": 2.1333528293610176,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9812,
       "step": 131310
     },
     {
       "epoch": 2.133515296258387,
-      "grad_norm": 4.40625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.266,
+      "loss": 0.9513,
       "step": 131320
     },
     {
       "epoch": 2.133677763155757,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2463,
+      "loss": 0.9751,
       "step": 131330
     },
     {
       "epoch": 2.1338402300531265,
-      "grad_norm": 3.90625,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2593,
+      "loss": 0.9546,
       "step": 131340
     },
     {
       "epoch": 2.134002696950496,
-      "grad_norm": 3.75,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9606,
       "step": 131350
     },
     {
       "epoch": 2.134165163847866,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2597,
+      "loss": 0.93,
       "step": 131360
     },
     {
       "epoch": 2.1343276307452355,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2567,
+      "loss": 0.943,
       "step": 131370
     },
     {
       "epoch": 2.1344900976426056,
-      "grad_norm": 4.125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2736,
+      "loss": 0.9819,
       "step": 131380
     },
     {
       "epoch": 2.1346525645399748,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2767,
+      "loss": 0.9679,
       "step": 131390
     },
     {
       "epoch": 2.134815031437345,
-      "grad_norm": 3.3125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2485,
+      "loss": 0.9631,
       "step": 131400
     },
     {
       "epoch": 2.1349774983347145,
-      "grad_norm": 4.375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2729,
+      "loss": 0.9637,
       "step": 131410
     },
     {
       "epoch": 2.135139965232084,
-      "grad_norm": 3.8125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2676,
+      "loss": 0.9382,
       "step": 131420
     },
     {
       "epoch": 2.135302432129454,
-      "grad_norm": 3.4375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2372,
+      "loss": 0.9586,
       "step": 131430
     },
     {
       "epoch": 2.1354648990268235,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2505,
+      "loss": 0.9743,
       "step": 131440
     },
     {
       "epoch": 2.135627365924193,
-      "grad_norm": 3.28125,
+      "grad_norm": 12.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2437,
+      "loss": 0.9997,
       "step": 131450
     },
     {
       "epoch": 2.1357898328215628,
-      "grad_norm": 3.25,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2358,
+      "loss": 0.9602,
       "step": 131460
     },
     {
       "epoch": 2.1359522997189324,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9338,
       "step": 131470
     },
     {
       "epoch": 2.136114766616302,
-      "grad_norm": 3.625,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 0.9505,
       "step": 131480
     },
     {
       "epoch": 2.1362772335136717,
-      "grad_norm": 2.96875,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 0.9577,
       "step": 131490
     },
     {
       "epoch": 2.1364397004110414,
-      "grad_norm": 3.078125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2457,
+      "loss": 1.0101,
       "step": 131500
     },
     {
       "epoch": 2.136602167308411,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9763,
       "step": 131510
     },
     {
       "epoch": 2.1367646342057807,
-      "grad_norm": 4.75,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.288,
+      "loss": 0.9348,
       "step": 131520
     },
     {
       "epoch": 2.1369271011031503,
-      "grad_norm": 3.890625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2784,
+      "loss": 0.9638,
       "step": 131530
     },
     {
       "epoch": 2.13708956800052,
-      "grad_norm": 4.53125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2844,
+      "loss": 0.9811,
       "step": 131540
     },
     {
       "epoch": 2.1372520348978896,
-      "grad_norm": 4.25,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2925,
+      "loss": 0.9338,
       "step": 131550
     },
     {
       "epoch": 2.1374145017952593,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 1.0065,
       "step": 131560
     },
     {
       "epoch": 2.137576968692629,
-      "grad_norm": 4.125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2775,
+      "loss": 0.9765,
       "step": 131570
     },
     {
       "epoch": 2.1377394355899986,
-      "grad_norm": 4.75,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9903,
       "step": 131580
     },
     {
       "epoch": 2.137901902487368,
-      "grad_norm": 3.875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.272,
+      "loss": 0.988,
       "step": 131590
     },
     {
       "epoch": 2.138064369384738,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9829,
       "step": 131600
     },
     {
       "epoch": 2.1382268362821075,
-      "grad_norm": 5.3125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2662,
+      "loss": 0.979,
       "step": 131610
     },
     {
       "epoch": 2.138389303179477,
-      "grad_norm": 3.96875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2384,
+      "loss": 0.978,
       "step": 131620
     },
     {
       "epoch": 2.138551770076847,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.224,
+      "loss": 0.9655,
       "step": 131630
     },
     {
       "epoch": 2.1387142369742165,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2447,
+      "loss": 0.9538,
       "step": 131640
     },
     {
       "epoch": 2.138876703871586,
-      "grad_norm": 4.125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2668,
+      "loss": 1.0026,
       "step": 131650
     },
     {
       "epoch": 2.1390391707689558,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2634,
+      "loss": 0.9689,
       "step": 131660
     },
     {
       "epoch": 2.1392016376663254,
-      "grad_norm": 4.375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2756,
+      "loss": 0.9496,
       "step": 131670
     },
     {
       "epoch": 2.139364104563695,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.246,
+      "loss": 0.9856,
       "step": 131680
     },
     {
       "epoch": 2.1395265714610647,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9815,
       "step": 131690
     },
     {
       "epoch": 2.1396890383584344,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.282,
+      "loss": 0.9931,
       "step": 131700
     },
     {
       "epoch": 2.139851505255804,
-      "grad_norm": 3.40625,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2846,
+      "loss": 0.9212,
       "step": 131710
     },
     {
       "epoch": 2.1400139721531737,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2419,
+      "loss": 0.9865,
       "step": 131720
     },
     {
       "epoch": 2.1401764390505433,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2669,
+      "loss": 0.9709,
       "step": 131730
     },
     {
       "epoch": 2.140338905947913,
-      "grad_norm": 2.953125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2704,
+      "loss": 0.9451,
       "step": 131740
     },
     {
       "epoch": 2.1405013728452826,
-      "grad_norm": 2.984375,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2405,
+      "loss": 0.969,
       "step": 131750
     },
     {
       "epoch": 2.1406638397426523,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.9634,
       "step": 131760
     },
     {
       "epoch": 2.140826306640022,
-      "grad_norm": 3.546875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2467,
+      "loss": 1.0007,
       "step": 131770
     },
     {
       "epoch": 2.1409887735373916,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2704,
+      "loss": 0.9427,
       "step": 131780
     },
     {
       "epoch": 2.1411512404347612,
-      "grad_norm": 3.59375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2681,
+      "loss": 0.9604,
       "step": 131790
     },
     {
       "epoch": 2.141313707332131,
-      "grad_norm": 3.484375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2655,
+      "loss": 0.9884,
       "step": 131800
     },
     {
       "epoch": 2.1414761742295005,
-      "grad_norm": 4.125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9875,
       "step": 131810
     },
     {
       "epoch": 2.14163864112687,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2897,
+      "loss": 0.9944,
       "step": 131820
     },
     {
       "epoch": 2.1418011080242403,
-      "grad_norm": 3.171875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2638,
+      "loss": 0.9506,
       "step": 131830
     },
     {
       "epoch": 2.1419635749216095,
-      "grad_norm": 4.75,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9493,
       "step": 131840
     },
     {
       "epoch": 2.1421260418189796,
-      "grad_norm": 2.71875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2463,
+      "loss": 0.9457,
       "step": 131850
     },
     {
       "epoch": 2.1422885087163492,
-      "grad_norm": 3.671875,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.251,
+      "loss": 0.9289,
       "step": 131860
     },
     {
       "epoch": 2.142450975613719,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2616,
+      "loss": 0.9835,
       "step": 131870
     },
     {
       "epoch": 2.1426134425110885,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2338,
+      "loss": 0.9495,
       "step": 131880
     },
     {
       "epoch": 2.142775909408458,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2433,
+      "loss": 0.9635,
       "step": 131890
     },
     {
       "epoch": 2.142938376305828,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2785,
+      "loss": 0.9438,
       "step": 131900
     },
     {
       "epoch": 2.1431008432031975,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2221,
+      "loss": 0.958,
       "step": 131910
     },
     {
       "epoch": 2.143263310100567,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2474,
+      "loss": 0.9322,
       "step": 131920
     },
     {
       "epoch": 2.143425776997937,
-      "grad_norm": 4.125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2742,
+      "loss": 0.976,
       "step": 131930
     },
     {
       "epoch": 2.1435882438953064,
-      "grad_norm": 3.984375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.251,
+      "loss": 0.9647,
       "step": 131940
     },
     {
       "epoch": 2.143750710792676,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2559,
+      "loss": 0.9538,
       "step": 131950
     },
     {
       "epoch": 2.1439131776900457,
-      "grad_norm": 3.09375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2437,
+      "loss": 0.9439,
       "step": 131960
     },
     {
       "epoch": 2.1440756445874154,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2594,
+      "loss": 0.9518,
       "step": 131970
     },
     {
       "epoch": 2.144238111484785,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9735,
       "step": 131980
     },
     {
       "epoch": 2.1444005783821547,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2718,
+      "loss": 0.9377,
       "step": 131990
     },
     {
       "epoch": 2.1445630452795243,
-      "grad_norm": 3.625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2813,
+      "loss": 0.9652,
       "step": 132000
     },
     {
       "epoch": 2.144725512176894,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2828,
+      "loss": 0.9684,
       "step": 132010
     },
     {
       "epoch": 2.1448879790742637,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2738,
+      "loss": 0.979,
       "step": 132020
     },
     {
       "epoch": 2.1450504459716333,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2944,
+      "loss": 0.9901,
       "step": 132030
     },
     {
       "epoch": 2.145212912869003,
-      "grad_norm": 6.53125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9391,
       "step": 132040
     },
     {
       "epoch": 2.1453753797663726,
-      "grad_norm": 2.140625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2468,
+      "loss": 0.9591,
       "step": 132050
     },
     {
       "epoch": 2.1455378466637423,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.257,
+      "loss": 0.9445,
       "step": 132060
     },
     {
       "epoch": 2.145700313561112,
-      "grad_norm": 3.328125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2462,
+      "loss": 0.9696,
       "step": 132070
     },
     {
       "epoch": 2.1458627804584816,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2522,
+      "loss": 0.9706,
       "step": 132080
     },
     {
       "epoch": 2.146025247355851,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2402,
+      "loss": 0.9884,
       "step": 132090
     },
     {
       "epoch": 2.146187714253221,
-      "grad_norm": 3.671875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2271,
+      "loss": 0.9705,
       "step": 132100
     },
     {
       "epoch": 2.1463501811505905,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 0.9602,
       "step": 132110
     },
     {
       "epoch": 2.14651264804796,
-      "grad_norm": 4.34375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.258,
+      "loss": 0.9685,
       "step": 132120
     },
     {
       "epoch": 2.14667511494533,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2322,
+      "loss": 0.9459,
       "step": 132130
     },
     {
       "epoch": 2.1468375818426995,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2614,
+      "loss": 0.9361,
       "step": 132140
     },
     {
       "epoch": 2.147000048740069,
-      "grad_norm": 3.578125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2591,
+      "loss": 0.9762,
       "step": 132150
     },
     {
       "epoch": 2.1471625156374388,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2737,
+      "loss": 0.9791,
       "step": 132160
     },
     {
       "epoch": 2.1473249825348084,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2718,
+      "loss": 0.9642,
       "step": 132170
     },
     {
       "epoch": 2.147487449432178,
-      "grad_norm": 3.546875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2456,
+      "loss": 0.9874,
       "step": 132180
     },
     {
       "epoch": 2.1476499163295477,
-      "grad_norm": 3.625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 0.9832,
       "step": 132190
     },
     {
       "epoch": 2.1478123832269174,
-      "grad_norm": 4.09375,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.237,
+      "loss": 0.9944,
       "step": 132200
     },
     {
       "epoch": 2.147974850124287,
-      "grad_norm": 3.09375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2636,
+      "loss": 0.996,
       "step": 132210
     },
     {
       "epoch": 2.1481373170216567,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 0.9528,
       "step": 132220
     },
     {
       "epoch": 2.1482997839190263,
-      "grad_norm": 3.640625,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2768,
+      "loss": 0.958,
       "step": 132230
     },
     {
       "epoch": 2.148462250816396,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9276,
       "step": 132240
     },
     {
       "epoch": 2.1486247177137656,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2757,
+      "loss": 0.9613,
       "step": 132250
     },
     {
       "epoch": 2.1487871846111357,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.28,
+      "loss": 0.9697,
       "step": 132260
     },
     {
       "epoch": 2.148949651508505,
-      "grad_norm": 3.53125,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9575,
       "step": 132270
     },
     {
       "epoch": 2.149112118405875,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9421,
       "step": 132280
     },
     {
       "epoch": 2.1492745853032447,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2458,
+      "loss": 1.0128,
       "step": 132290
     },
     {
       "epoch": 2.1494370522006143,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2788,
+      "loss": 0.9522,
       "step": 132300
     },
     {
       "epoch": 2.149599519097984,
-      "grad_norm": 4.625,
+      "grad_norm": 12.0,
       "learning_rate": 5e-05,
-      "loss": 0.2565,
+      "loss": 0.951,
       "step": 132310
     },
     {
       "epoch": 2.1497619859953536,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2504,
+      "loss": 0.9608,
       "step": 132320
     },
     {
       "epoch": 2.1499244528927233,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2492,
+      "loss": 0.9746,
       "step": 132330
     },
     {
       "epoch": 2.150086919790093,
-      "grad_norm": 3.375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.248,
+      "loss": 0.9422,
       "step": 132340
     },
     {
       "epoch": 2.1502493866874626,
-      "grad_norm": 3.390625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2281,
+      "loss": 0.9545,
       "step": 132350
     },
     {
       "epoch": 2.1504118535848322,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2636,
+      "loss": 0.9749,
       "step": 132360
     },
     {
       "epoch": 2.150574320482202,
-      "grad_norm": 3.15625,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.276,
+      "loss": 0.9894,
       "step": 132370
     },
     {
       "epoch": 2.1507367873795715,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2509,
+      "loss": 0.9581,
       "step": 132380
     },
     {
       "epoch": 2.150899254276941,
-      "grad_norm": 4.8125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9436,
       "step": 132390
     },
     {
       "epoch": 2.151061721174311,
-      "grad_norm": 3.234375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2642,
+      "loss": 0.9932,
       "step": 132400
     },
     {
       "epoch": 2.1512241880716805,
-      "grad_norm": 4.0,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2744,
+      "loss": 0.9568,
       "step": 132410
     },
     {
       "epoch": 2.15138665496905,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2542,
+      "loss": 0.981,
       "step": 132420
     },
     {
       "epoch": 2.15154912186642,
-      "grad_norm": 4.21875,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2359,
+      "loss": 0.9635,
       "step": 132430
     },
     {
       "epoch": 2.1517115887637894,
-      "grad_norm": 3.078125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2532,
+      "loss": 0.9714,
       "step": 132440
     },
     {
       "epoch": 2.151874055661159,
-      "grad_norm": 4.375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2382,
+      "loss": 0.9712,
       "step": 132450
     },
     {
       "epoch": 2.1520365225585287,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2599,
+      "loss": 0.9493,
       "step": 132460
     },
     {
       "epoch": 2.1521989894558984,
-      "grad_norm": 3.5,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2709,
+      "loss": 0.9774,
       "step": 132470
     },
     {
       "epoch": 2.152361456353268,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2818,
+      "loss": 0.9567,
       "step": 132480
     },
     {
       "epoch": 2.1525239232506377,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2644,
+      "loss": 0.9558,
       "step": 132490
     },
     {
       "epoch": 2.1526863901480073,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2675,
+      "loss": 0.9333,
       "step": 132500
     },
     {
       "epoch": 2.152848857045377,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9734,
       "step": 132510
     },
     {
       "epoch": 2.1530113239427466,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2506,
+      "loss": 0.9814,
       "step": 132520
     },
     {
       "epoch": 2.1531737908401163,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2507,
+      "loss": 0.971,
       "step": 132530
     },
     {
       "epoch": 2.153336257737486,
-      "grad_norm": 3.34375,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2678,
+      "loss": 0.9838,
       "step": 132540
     },
     {
       "epoch": 2.1534987246348556,
-      "grad_norm": 3.9375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2489,
+      "loss": 0.9083,
       "step": 132550
     },
     {
       "epoch": 2.1536611915322252,
-      "grad_norm": 4.125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2297,
+      "loss": 0.967,
       "step": 132560
     },
     {
       "epoch": 2.153823658429595,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9418,
       "step": 132570
     },
     {
       "epoch": 2.1539861253269645,
-      "grad_norm": 4.5,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2758,
+      "loss": 0.9624,
       "step": 132580
     },
     {
       "epoch": 2.154148592224334,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.251,
+      "loss": 0.9874,
       "step": 132590
     },
     {
       "epoch": 2.154311059121704,
-      "grad_norm": 3.78125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2569,
+      "loss": 0.9768,
       "step": 132600
     },
     {
       "epoch": 2.1544735260190735,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2643,
+      "loss": 0.9645,
       "step": 132610
     },
     {
       "epoch": 2.154635992916443,
-      "grad_norm": 3.625,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2606,
+      "loss": 0.9544,
       "step": 132620
     },
     {
       "epoch": 2.154798459813813,
-      "grad_norm": 3.765625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.237,
+      "loss": 0.9817,
       "step": 132630
     },
     {
       "epoch": 2.1549609267111824,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2364,
+      "loss": 0.9782,
       "step": 132640
     },
     {
       "epoch": 2.155123393608552,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2404,
+      "loss": 0.9821,
       "step": 132650
     },
     {
       "epoch": 2.1552858605059217,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2598,
+      "loss": 0.9448,
       "step": 132660
     },
     {
       "epoch": 2.1554483274032914,
-      "grad_norm": 5.09375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.255,
+      "loss": 0.9747,
       "step": 132670
     },
     {
       "epoch": 2.155610794300661,
-      "grad_norm": 3.578125,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.2773,
+      "loss": 0.9487,
       "step": 132680
     },
     {
       "epoch": 2.1557732611980307,
-      "grad_norm": 2.609375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2478,
+      "loss": 0.9644,
       "step": 132690
     },
     {
       "epoch": 2.1559357280954003,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.257,
+      "loss": 0.9467,
       "step": 132700
     },
     {
       "epoch": 2.1560981949927704,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2532,
+      "loss": 0.9988,
       "step": 132710
     },
     {
       "epoch": 2.1562606618901397,
-      "grad_norm": 3.234375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2714,
+      "loss": 0.9761,
       "step": 132720
     },
     {
       "epoch": 2.1564231287875097,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2515,
+      "loss": 0.9448,
       "step": 132730
     },
     {
       "epoch": 2.1565855956848794,
-      "grad_norm": 3.75,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.25,
+      "loss": 0.9442,
       "step": 132740
     },
     {
       "epoch": 2.156748062582249,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9738,
       "step": 132750
     },
     {
       "epoch": 2.1569105294796187,
-      "grad_norm": 5.53125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2365,
+      "loss": 0.9502,
       "step": 132760
     },
     {
       "epoch": 2.1570729963769883,
-      "grad_norm": 4.75,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2673,
+      "loss": 0.9441,
       "step": 132770
     },
     {
       "epoch": 2.157235463274358,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2758,
+      "loss": 0.9269,
       "step": 132780
     },
     {
       "epoch": 2.1573979301717277,
-      "grad_norm": 3.625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2627,
+      "loss": 0.9859,
       "step": 132790
     },
     {
       "epoch": 2.1575603970690973,
-      "grad_norm": 4.90625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2742,
+      "loss": 0.9601,
       "step": 132800
     },
     {
       "epoch": 2.157722863966467,
-      "grad_norm": 3.5,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.245,
+      "loss": 0.9849,
       "step": 132810
     },
     {
       "epoch": 2.1578853308638366,
-      "grad_norm": 5.90625,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2542,
+      "loss": 0.9586,
       "step": 132820
     },
     {
       "epoch": 2.1580477977612063,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 1.005,
       "step": 132830
     },
     {
       "epoch": 2.158210264658576,
-      "grad_norm": 4.84375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2817,
+      "loss": 0.9252,
       "step": 132840
     },
     {
       "epoch": 2.1583727315559456,
-      "grad_norm": 2.828125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2543,
+      "loss": 0.9897,
       "step": 132850
     },
     {
       "epoch": 2.158535198453315,
-      "grad_norm": 5.34375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2679,
+      "loss": 0.9892,
       "step": 132860
     },
     {
       "epoch": 2.158697665350685,
-      "grad_norm": 3.125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2517,
+      "loss": 1.0064,
       "step": 132870
     },
     {
       "epoch": 2.1588601322480545,
-      "grad_norm": 2.671875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9694,
       "step": 132880
     },
     {
       "epoch": 2.159022599145424,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2742,
+      "loss": 0.9414,
       "step": 132890
     },
     {
       "epoch": 2.159185066042794,
-      "grad_norm": 6.375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2399,
+      "loss": 0.9755,
       "step": 132900
     },
     {
       "epoch": 2.1593475329401635,
-      "grad_norm": 4.71875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9617,
       "step": 132910
     },
     {
       "epoch": 2.159509999837533,
-      "grad_norm": 4.53125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2548,
+      "loss": 0.941,
       "step": 132920
     },
     {
       "epoch": 2.1596724667349028,
-      "grad_norm": 4.375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2817,
+      "loss": 0.9827,
       "step": 132930
     },
     {
       "epoch": 2.1598349336322724,
-      "grad_norm": 3.75,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.9792,
       "step": 132940
     },
     {
       "epoch": 2.159997400529642,
-      "grad_norm": 4.8125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.943,
       "step": 132950
     },
     {
       "epoch": 2.1601598674270117,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 0.9637,
       "step": 132960
     },
     {
       "epoch": 2.1603223343243814,
-      "grad_norm": 4.40625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2536,
+      "loss": 0.9871,
       "step": 132970
     },
     {
       "epoch": 2.160484801221751,
-      "grad_norm": 4.40625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2882,
+      "loss": 0.9994,
       "step": 132980
     },
     {
       "epoch": 2.1606472681191207,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.252,
+      "loss": 0.9492,
       "step": 132990
     },
     {
       "epoch": 2.1608097350164903,
-      "grad_norm": 3.984375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2626,
+      "loss": 0.9831,
       "step": 133000
     },
     {
       "epoch": 2.16097220191386,
-      "grad_norm": 4.25,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2601,
+      "loss": 0.9585,
       "step": 133010
     },
     {
       "epoch": 2.1611346688112296,
-      "grad_norm": 4.125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.9562,
       "step": 133020
     },
     {
       "epoch": 2.1612971357085993,
-      "grad_norm": 4.75,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2869,
+      "loss": 0.9488,
       "step": 133030
     },
     {
       "epoch": 2.161459602605969,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2542,
+      "loss": 0.9489,
       "step": 133040
     },
     {
       "epoch": 2.1616220695033386,
-      "grad_norm": 3.75,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9977,
       "step": 133050
     },
     {
       "epoch": 2.1617845364007082,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2595,
+      "loss": 0.9784,
       "step": 133060
     },
     {
       "epoch": 2.161947003298078,
-      "grad_norm": 3.3125,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2337,
+      "loss": 0.996,
       "step": 133070
     },
     {
       "epoch": 2.1621094701954475,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2284,
+      "loss": 0.9572,
       "step": 133080
     },
     {
       "epoch": 2.162271937092817,
-      "grad_norm": 3.75,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9388,
       "step": 133090
     },
     {
       "epoch": 2.162434403990187,
-      "grad_norm": 2.96875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9544,
       "step": 133100
     },
     {
       "epoch": 2.1625968708875565,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2559,
+      "loss": 0.9652,
       "step": 133110
     },
     {
       "epoch": 2.162759337784926,
-      "grad_norm": 3.328125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2477,
+      "loss": 0.9633,
       "step": 133120
     },
     {
       "epoch": 2.162921804682296,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 0.9349,
       "step": 133130
     },
     {
       "epoch": 2.163084271579666,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.271,
+      "loss": 0.9559,
       "step": 133140
     },
     {
       "epoch": 2.163246738477035,
-      "grad_norm": 4.25,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2444,
+      "loss": 0.9845,
       "step": 133150
     },
     {
       "epoch": 2.163409205374405,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2529,
+      "loss": 1.0221,
       "step": 133160
     },
     {
       "epoch": 2.163571672271775,
-      "grad_norm": 3.78125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.24,
+      "loss": 0.963,
       "step": 133170
     },
     {
       "epoch": 2.1637341391691445,
-      "grad_norm": 3.984375,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2601,
+      "loss": 0.95,
       "step": 133180
     },
     {
       "epoch": 2.163896606066514,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.986,
       "step": 133190
     },
     {
       "epoch": 2.164059072963884,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2463,
+      "loss": 0.9752,
       "step": 133200
     },
     {
       "epoch": 2.1642215398612534,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2574,
+      "loss": 0.954,
       "step": 133210
     },
     {
       "epoch": 2.164384006758623,
-      "grad_norm": 3.328125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2444,
+      "loss": 0.9696,
       "step": 133220
     },
     {
       "epoch": 2.1645464736559927,
-      "grad_norm": 3.8125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2385,
+      "loss": 0.9737,
       "step": 133230
     },
     {
       "epoch": 2.1647089405533624,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9694,
       "step": 133240
     },
     {
       "epoch": 2.164871407450732,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9568,
       "step": 133250
     },
     {
       "epoch": 2.1650338743481017,
-      "grad_norm": 5.15625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2398,
+      "loss": 0.9877,
       "step": 133260
     },
     {
       "epoch": 2.1651963412454713,
-      "grad_norm": 3.578125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2811,
+      "loss": 0.9768,
       "step": 133270
     },
     {
       "epoch": 2.165358808142841,
-      "grad_norm": 4.5,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2824,
+      "loss": 0.9693,
       "step": 133280
     },
     {
       "epoch": 2.1655212750402106,
-      "grad_norm": 3.671875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2882,
+      "loss": 0.9966,
       "step": 133290
     },
     {
       "epoch": 2.1656837419375803,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9265,
       "step": 133300
     },
     {
       "epoch": 2.16584620883495,
-      "grad_norm": 4.65625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 1.018,
       "step": 133310
     },
     {
       "epoch": 2.1660086757323196,
-      "grad_norm": 3.875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2528,
+      "loss": 0.9562,
       "step": 133320
     },
     {
       "epoch": 2.1661711426296892,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2603,
+      "loss": 0.9581,
       "step": 133330
     },
     {
       "epoch": 2.166333609527059,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2675,
+      "loss": 0.9889,
       "step": 133340
     },
     {
       "epoch": 2.1664960764244285,
-      "grad_norm": 3.015625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2682,
+      "loss": 0.9552,
       "step": 133350
     },
     {
       "epoch": 2.166658543321798,
-      "grad_norm": 3.0625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9751,
       "step": 133360
     },
     {
       "epoch": 2.166821010219168,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9844,
       "step": 133370
     },
     {
       "epoch": 2.1669834771165375,
-      "grad_norm": 3.546875,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2457,
+      "loss": 0.9682,
       "step": 133380
     },
     {
       "epoch": 2.167145944013907,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.9777,
       "step": 133390
     },
     {
       "epoch": 2.167308410911277,
-      "grad_norm": 2.546875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2712,
+      "loss": 0.9571,
       "step": 133400
     },
     {
       "epoch": 2.1674708778086464,
-      "grad_norm": 3.171875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9366,
       "step": 133410
     },
     {
       "epoch": 2.167633344706016,
-      "grad_norm": 3.171875,
+      "grad_norm": 12.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2375,
+      "loss": 0.9672,
       "step": 133420
     },
     {
       "epoch": 2.1677958116033857,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2487,
+      "loss": 0.993,
       "step": 133430
     },
     {
       "epoch": 2.1679582785007554,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2796,
+      "loss": 0.9575,
       "step": 133440
     },
     {
       "epoch": 2.168120745398125,
-      "grad_norm": 4.125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9505,
       "step": 133450
     },
     {
       "epoch": 2.1682832122954947,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.25,
+      "loss": 1.0117,
       "step": 133460
     },
     {
       "epoch": 2.1684456791928644,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2529,
+      "loss": 0.9609,
       "step": 133470
     },
     {
       "epoch": 2.168608146090234,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2404,
+      "loss": 0.9955,
       "step": 133480
     },
     {
       "epoch": 2.1687706129876037,
-      "grad_norm": 3.546875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.244,
+      "loss": 0.9769,
       "step": 133490
     },
     {
       "epoch": 2.1689330798849733,
-      "grad_norm": 4.5,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2477,
+      "loss": 0.9411,
       "step": 133500
     },
     {
       "epoch": 2.169095546782343,
-      "grad_norm": 2.4375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.249,
+      "loss": 0.953,
       "step": 133510
     },
     {
       "epoch": 2.1692580136797126,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2461,
+      "loss": 0.9778,
       "step": 133520
     },
     {
       "epoch": 2.1694204805770823,
-      "grad_norm": 3.25,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2577,
+      "loss": 0.9867,
       "step": 133530
     },
     {
       "epoch": 2.169582947474452,
-      "grad_norm": 2.84375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9767,
       "step": 133540
     },
     {
       "epoch": 2.1697454143718216,
-      "grad_norm": 3.328125,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9804,
       "step": 133550
     },
     {
       "epoch": 2.169907881269191,
-      "grad_norm": 4.0,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.9939,
       "step": 133560
     },
     {
       "epoch": 2.170070348166561,
-      "grad_norm": 3.421875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2701,
+      "loss": 0.9679,
       "step": 133570
     },
     {
       "epoch": 2.1702328150639305,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2787,
+      "loss": 0.9873,
       "step": 133580
     },
     {
       "epoch": 2.1703952819613006,
-      "grad_norm": 5.125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2459,
+      "loss": 0.9873,
       "step": 133590
     },
     {
       "epoch": 2.17055774885867,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2546,
+      "loss": 0.9924,
       "step": 133600
     },
     {
       "epoch": 2.17072021575604,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2794,
+      "loss": 0.9447,
       "step": 133610
     },
     {
       "epoch": 2.1708826826534096,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 0.9371,
       "step": 133620
     },
     {
       "epoch": 2.171045149550779,
-      "grad_norm": 3.3125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2396,
+      "loss": 0.9601,
       "step": 133630
     },
     {
       "epoch": 2.171207616448149,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2396,
+      "loss": 0.968,
       "step": 133640
     },
     {
       "epoch": 2.1713700833455185,
-      "grad_norm": 4.84375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2378,
+      "loss": 0.9451,
       "step": 133650
     },
     {
       "epoch": 2.171532550242888,
-      "grad_norm": 4.34375,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.9887,
       "step": 133660
     },
     {
       "epoch": 2.171695017140258,
-      "grad_norm": 4.40625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2539,
+      "loss": 1.0097,
       "step": 133670
     },
     {
       "epoch": 2.1718574840376275,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2638,
+      "loss": 0.965,
       "step": 133680
     },
     {
       "epoch": 2.172019950934997,
-      "grad_norm": 4.71875,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2703,
+      "loss": 0.9747,
       "step": 133690
     },
     {
       "epoch": 2.1721824178323668,
-      "grad_norm": 3.625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2755,
+      "loss": 0.9799,
       "step": 133700
     },
     {
       "epoch": 2.1723448847297364,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.276,
+      "loss": 0.9732,
       "step": 133710
     },
     {
       "epoch": 2.172507351627106,
-      "grad_norm": 3.78125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2704,
+      "loss": 0.9926,
       "step": 133720
     },
     {
       "epoch": 2.1726698185244757,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9781,
       "step": 133730
     },
     {
       "epoch": 2.1728322854218454,
-      "grad_norm": 4.125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.965,
       "step": 133740
     },
     {
       "epoch": 2.172994752319215,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2493,
+      "loss": 0.9613,
       "step": 133750
     },
     {
       "epoch": 2.1731572192165847,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2528,
+      "loss": 1.0078,
       "step": 133760
     },
     {
       "epoch": 2.1733196861139543,
-      "grad_norm": 4.21875,
+      "grad_norm": 7.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 1.0053,
       "step": 133770
     },
     {
       "epoch": 2.173482153011324,
-      "grad_norm": 5.09375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2699,
+      "loss": 0.9571,
       "step": 133780
     },
     {
       "epoch": 2.1736446199086936,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9645,
       "step": 133790
     },
     {
       "epoch": 2.1738070868060633,
-      "grad_norm": 3.34375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2587,
+      "loss": 0.969,
       "step": 133800
     },
     {
       "epoch": 2.173969553703433,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9729,
       "step": 133810
     },
     {
       "epoch": 2.1741320206008026,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.242,
+      "loss": 0.9456,
       "step": 133820
     },
     {
       "epoch": 2.1742944874981722,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2626,
+      "loss": 0.9585,
       "step": 133830
     },
     {
       "epoch": 2.174456954395542,
-      "grad_norm": 3.484375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9859,
       "step": 133840
     },
     {
       "epoch": 2.1746194212929115,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9855,
       "step": 133850
     },
     {
       "epoch": 2.174781888190281,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2523,
+      "loss": 0.9771,
       "step": 133860
     },
     {
       "epoch": 2.174944355087651,
-      "grad_norm": 3.734375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 0.9513,
       "step": 133870
     },
     {
       "epoch": 2.1751068219850205,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2699,
+      "loss": 0.9962,
       "step": 133880
     },
     {
       "epoch": 2.17526928888239,
-      "grad_norm": 4.71875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2426,
+      "loss": 0.9851,
       "step": 133890
     },
     {
       "epoch": 2.17543175577976,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2611,
+      "loss": 0.9579,
       "step": 133900
     },
     {
       "epoch": 2.1755942226771294,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2708,
+      "loss": 0.9588,
       "step": 133910
     },
     {
       "epoch": 2.175756689574499,
-      "grad_norm": 3.046875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9697,
       "step": 133920
     },
     {
       "epoch": 2.1759191564718687,
-      "grad_norm": 4.03125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2417,
+      "loss": 0.9607,
       "step": 133930
     },
     {
       "epoch": 2.1760816233692384,
-      "grad_norm": 3.671875,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2443,
+      "loss": 0.976,
       "step": 133940
     },
     {
       "epoch": 2.176244090266608,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9588,
       "step": 133950
     },
     {
       "epoch": 2.1764065571639777,
-      "grad_norm": 3.546875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2526,
+      "loss": 0.9711,
       "step": 133960
     },
     {
       "epoch": 2.1765690240613473,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 1.0151,
       "step": 133970
     },
     {
       "epoch": 2.176731490958717,
-      "grad_norm": 3.265625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2676,
+      "loss": 0.9978,
       "step": 133980
     },
     {
       "epoch": 2.1768939578560866,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2516,
+      "loss": 0.935,
       "step": 133990
     },
     {
       "epoch": 2.1770564247534563,
-      "grad_norm": 3.1875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2231,
+      "loss": 0.9829,
       "step": 134000
     },
     {
       "epoch": 2.177218891650826,
-      "grad_norm": 3.453125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2302,
+      "loss": 0.989,
       "step": 134010
     },
     {
       "epoch": 2.177381358548196,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.227,
+      "loss": 0.9854,
       "step": 134020
     },
     {
       "epoch": 2.1775438254455652,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2169,
+      "loss": 1.0103,
       "step": 134030
     },
     {
       "epoch": 2.1777062923429353,
-      "grad_norm": 3.890625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.2258,
+      "loss": 0.9737,
       "step": 134040
     },
     {
       "epoch": 2.177868759240305,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2259,
+      "loss": 1.0034,
       "step": 134050
     },
     {
       "epoch": 2.1780312261376746,
-      "grad_norm": 4.0,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2578,
+      "loss": 0.9477,
       "step": 134060
     },
     {
       "epoch": 2.1781936930350443,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2403,
+      "loss": 0.9565,
       "step": 134070
     },
     {
       "epoch": 2.178356159932414,
-      "grad_norm": 4.25,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2614,
+      "loss": 0.9445,
       "step": 134080
     },
     {
       "epoch": 2.1785186268297836,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9564,
       "step": 134090
     },
     {
       "epoch": 2.1786810937271532,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2676,
+      "loss": 0.9723,
       "step": 134100
     },
     {
       "epoch": 2.178843560624523,
-      "grad_norm": 3.109375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2696,
+      "loss": 1.0047,
       "step": 134110
     },
     {
       "epoch": 2.1790060275218925,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2639,
+      "loss": 0.9696,
       "step": 134120
     },
     {
       "epoch": 2.179168494419262,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2728,
+      "loss": 0.9102,
       "step": 134130
     },
     {
       "epoch": 2.179330961316632,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2721,
+      "loss": 0.9106,
       "step": 134140
     },
     {
       "epoch": 2.1794934282140015,
-      "grad_norm": 5.1875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.281,
+      "loss": 0.9846,
       "step": 134150
     },
     {
       "epoch": 2.179655895111371,
-      "grad_norm": 3.671875,
+      "grad_norm": 13.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2732,
+      "loss": 0.9927,
       "step": 134160
     },
     {
       "epoch": 2.179818362008741,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.259,
+      "loss": 0.96,
       "step": 134170
     },
     {
       "epoch": 2.1799808289061104,
-      "grad_norm": 3.671875,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2481,
+      "loss": 0.9517,
       "step": 134180
     },
     {
       "epoch": 2.18014329580348,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2714,
+      "loss": 0.994,
       "step": 134190
     },
     {
       "epoch": 2.1803057627008497,
-      "grad_norm": 3.5,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.274,
+      "loss": 0.9786,
       "step": 134200
     },
     {
       "epoch": 2.1804682295982194,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.2637,
+      "loss": 0.9467,
       "step": 134210
     },
     {
       "epoch": 2.180630696495589,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.255,
+      "loss": 0.9967,
       "step": 134220
     },
     {
       "epoch": 2.1807931633929587,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9816,
       "step": 134230
     },
     {
       "epoch": 2.1809556302903284,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2627,
+      "loss": 0.9802,
       "step": 134240
     },
     {
       "epoch": 2.181118097187698,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2771,
+      "loss": 0.9553,
       "step": 134250
     },
     {
       "epoch": 2.1812805640850677,
-      "grad_norm": 2.5625,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2553,
+      "loss": 1.0221,
       "step": 134260
     },
     {
       "epoch": 2.1814430309824373,
-      "grad_norm": 3.375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9913,
       "step": 134270
     },
     {
       "epoch": 2.181605497879807,
-      "grad_norm": 3.1875,
+      "grad_norm": 12.125,
       "learning_rate": 5e-05,
-      "loss": 0.2675,
+      "loss": 0.964,
       "step": 134280
     },
     {
       "epoch": 2.1817679647771766,
-      "grad_norm": 3.625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2565,
+      "loss": 1.0085,
       "step": 134290
     },
     {
       "epoch": 2.1819304316745463,
-      "grad_norm": 4.0625,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2428,
+      "loss": 0.9271,
       "step": 134300
     },
     {
       "epoch": 2.182092898571916,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2322,
+      "loss": 0.98,
       "step": 134310
     },
     {
       "epoch": 2.1822553654692856,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2279,
+      "loss": 0.9551,
       "step": 134320
     },
     {
       "epoch": 2.182417832366655,
-      "grad_norm": 4.375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2655,
+      "loss": 0.9536,
       "step": 134330
     },
     {
       "epoch": 2.182580299264025,
-      "grad_norm": 4.03125,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2572,
+      "loss": 0.959,
       "step": 134340
     },
     {
       "epoch": 2.1827427661613945,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.957,
       "step": 134350
     },
     {
       "epoch": 2.182905233058764,
-      "grad_norm": 2.84375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2693,
+      "loss": 0.97,
       "step": 134360
     },
     {
       "epoch": 2.183067699956134,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2622,
+      "loss": 0.9795,
       "step": 134370
     },
     {
       "epoch": 2.1832301668535035,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2422,
+      "loss": 0.9657,
       "step": 134380
     },
     {
       "epoch": 2.183392633750873,
-      "grad_norm": 4.09375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9552,
       "step": 134390
     },
     {
       "epoch": 2.1835551006482428,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2553,
+      "loss": 0.9459,
       "step": 134400
     },
     {
       "epoch": 2.1837175675456124,
-      "grad_norm": 3.8125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9552,
       "step": 134410
     },
     {
       "epoch": 2.183880034442982,
-      "grad_norm": 3.6875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.9381,
       "step": 134420
     },
     {
       "epoch": 2.1840425013403517,
-      "grad_norm": 4.0,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 0.9647,
       "step": 134430
     },
     {
       "epoch": 2.1842049682377214,
-      "grad_norm": 5.03125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2437,
+      "loss": 0.9629,
       "step": 134440
     },
     {
       "epoch": 2.184367435135091,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2709,
+      "loss": 0.9681,
       "step": 134450
     },
     {
       "epoch": 2.1845299020324607,
-      "grad_norm": 3.921875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2664,
+      "loss": 0.951,
       "step": 134460
     },
     {
       "epoch": 2.1846923689298308,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 0.9687,
       "step": 134470
     },
     {
       "epoch": 2.1848548358272004,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2547,
+      "loss": 0.9256,
       "step": 134480
     },
     {
       "epoch": 2.18501730272457,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2771,
+      "loss": 0.9951,
       "step": 134490
     },
     {
       "epoch": 2.1851797696219397,
-      "grad_norm": 4.8125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2694,
+      "loss": 0.9549,
       "step": 134500
     },
     {
       "epoch": 2.1853422365193094,
-      "grad_norm": 4.125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2887,
+      "loss": 0.9625,
       "step": 134510
     },
     {
       "epoch": 2.185504703416679,
-      "grad_norm": 3.828125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2708,
+      "loss": 0.9597,
       "step": 134520
     },
     {
       "epoch": 2.1856671703140487,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 0.9574,
       "step": 134530
     },
     {
       "epoch": 2.1858296372114183,
-      "grad_norm": 4.25,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2811,
+      "loss": 1.005,
       "step": 134540
     },
     {
       "epoch": 2.185992104108788,
-      "grad_norm": 4.03125,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2623,
+      "loss": 1.0018,
       "step": 134550
     },
     {
       "epoch": 2.1861545710061576,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9756,
       "step": 134560
     },
     {
       "epoch": 2.1863170379035273,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.979,
       "step": 134570
     },
     {
       "epoch": 2.186479504800897,
-      "grad_norm": 5.71875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2629,
+      "loss": 0.9155,
       "step": 134580
     },
     {
       "epoch": 2.1866419716982666,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 0.9938,
       "step": 134590
     },
     {
       "epoch": 2.1868044385956362,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2764,
+      "loss": 0.9818,
       "step": 134600
     },
     {
       "epoch": 2.186966905493006,
-      "grad_norm": 3.5625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.3095,
+      "loss": 0.9888,
       "step": 134610
     },
     {
       "epoch": 2.1871293723903755,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2877,
+      "loss": 0.9763,
       "step": 134620
     },
     {
       "epoch": 2.187291839287745,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 0.9723,
       "step": 134630
     },
     {
       "epoch": 2.187454306185115,
-      "grad_norm": 3.359375,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2807,
+      "loss": 0.9487,
       "step": 134640
     },
     {
       "epoch": 2.1876167730824845,
-      "grad_norm": 3.546875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2851,
+      "loss": 0.9813,
       "step": 134650
     },
     {
       "epoch": 2.187779239979854,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2562,
+      "loss": 0.9631,
       "step": 134660
     },
     {
       "epoch": 2.187941706877224,
-      "grad_norm": 3.359375,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2467,
+      "loss": 0.9907,
       "step": 134670
     },
     {
       "epoch": 2.1881041737745934,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2606,
+      "loss": 1.0028,
       "step": 134680
     },
     {
       "epoch": 2.188266640671963,
-      "grad_norm": 3.8125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2583,
+      "loss": 0.9933,
       "step": 134690
     },
     {
       "epoch": 2.1884291075693327,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2535,
+      "loss": 1.0153,
       "step": 134700
     },
     {
       "epoch": 2.1885915744667024,
-      "grad_norm": 5.71875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.253,
+      "loss": 0.98,
       "step": 134710
     },
     {
       "epoch": 2.188754041364072,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2789,
+      "loss": 0.9426,
       "step": 134720
     },
     {
       "epoch": 2.1889165082614417,
-      "grad_norm": 4.125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2752,
+      "loss": 0.9884,
       "step": 134730
     },
     {
       "epoch": 2.1890789751588113,
-      "grad_norm": 3.703125,
+      "grad_norm": 7.625,
       "learning_rate": 5e-05,
-      "loss": 0.2737,
+      "loss": 0.9506,
       "step": 134740
     },
     {
       "epoch": 2.189241442056181,
-      "grad_norm": 3.96875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.273,
+      "loss": 0.9738,
       "step": 134750
     },
     {
       "epoch": 2.1894039089535506,
-      "grad_norm": 3.28125,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2466,
+      "loss": 0.9939,
       "step": 134760
     },
     {
       "epoch": 2.1895663758509203,
-      "grad_norm": 5.28125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2547,
+      "loss": 0.9789,
       "step": 134770
     },
     {
       "epoch": 2.18972884274829,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2384,
+      "loss": 0.9743,
       "step": 134780
     },
     {
       "epoch": 2.1898913096456596,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2432,
+      "loss": 0.9622,
       "step": 134790
     },
     {
       "epoch": 2.1900537765430292,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2212,
+      "loss": 1.019,
       "step": 134800
     },
     {
       "epoch": 2.190216243440399,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2447,
+      "loss": 0.9751,
       "step": 134810
     },
     {
       "epoch": 2.1903787103377685,
-      "grad_norm": 3.96875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.26,
+      "loss": 0.9944,
       "step": 134820
     },
     {
       "epoch": 2.190541177235138,
-      "grad_norm": 4.65625,
+      "grad_norm": 12.25,
       "learning_rate": 5e-05,
-      "loss": 0.2667,
+      "loss": 0.9603,
       "step": 134830
     },
     {
       "epoch": 2.190703644132508,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2605,
+      "loss": 0.9886,
       "step": 134840
     },
     {
       "epoch": 2.1908661110298775,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2433,
+      "loss": 0.954,
       "step": 134850
     },
     {
       "epoch": 2.191028577927247,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9989,
       "step": 134860
     },
     {
       "epoch": 2.191191044824617,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2601,
+      "loss": 0.9615,
       "step": 134870
     },
     {
       "epoch": 2.1913535117219864,
-      "grad_norm": 3.734375,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2575,
+      "loss": 0.9717,
       "step": 134880
     },
     {
       "epoch": 2.191515978619356,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 0.9929,
       "step": 134890
     },
     {
       "epoch": 2.191678445516726,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.276,
+      "loss": 0.9707,
       "step": 134900
     },
     {
       "epoch": 2.1918409124140954,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.272,
+      "loss": 0.9867,
       "step": 134910
     },
     {
       "epoch": 2.1920033793114655,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2588,
+      "loss": 0.9845,
       "step": 134920
     },
     {
       "epoch": 2.192165846208835,
-      "grad_norm": 3.140625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2537,
+      "loss": 0.9686,
       "step": 134930
     },
     {
       "epoch": 2.192328313106205,
-      "grad_norm": 5.28125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2664,
+      "loss": 0.9605,
       "step": 134940
     },
     {
       "epoch": 2.1924907800035744,
-      "grad_norm": 2.953125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2674,
+      "loss": 0.9642,
       "step": 134950
     },
     {
       "epoch": 2.192653246900944,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2829,
+      "loss": 0.9572,
       "step": 134960
     },
     {
       "epoch": 2.1928157137983137,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2753,
+      "loss": 0.9286,
       "step": 134970
     },
     {
       "epoch": 2.1929781806956834,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2725,
+      "loss": 0.9709,
       "step": 134980
     },
     {
       "epoch": 2.193140647593053,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2367,
+      "loss": 0.9696,
       "step": 134990
     },
     {
       "epoch": 2.1933031144904227,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2468,
+      "loss": 0.995,
       "step": 135000
     },
     {
       "epoch": 2.1934655813877924,
-      "grad_norm": 3.578125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.254,
+      "loss": 0.9717,
       "step": 135010
     },
     {
       "epoch": 2.193628048285162,
-      "grad_norm": 3.859375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2673,
+      "loss": 0.9961,
       "step": 135020
     },
     {
       "epoch": 2.1937905151825317,
-      "grad_norm": 2.9375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2759,
+      "loss": 0.9827,
       "step": 135030
     },
     {
       "epoch": 2.1939529820799013,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2677,
+      "loss": 0.9776,
       "step": 135040
     },
     {
       "epoch": 2.194115448977271,
-      "grad_norm": 4.625,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 0.962,
       "step": 135050
     },
     {
       "epoch": 2.1942779158746406,
-      "grad_norm": 4.75,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2479,
+      "loss": 0.9838,
       "step": 135060
     },
     {
       "epoch": 2.1944403827720103,
-      "grad_norm": 3.828125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2594,
+      "loss": 1.0074,
       "step": 135070
     },
     {
       "epoch": 2.19460284966938,
-      "grad_norm": 4.25,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2588,
+      "loss": 0.9735,
       "step": 135080
     },
     {
       "epoch": 2.1947653165667496,
-      "grad_norm": 4.34375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2887,
+      "loss": 0.9732,
       "step": 135090
     },
     {
       "epoch": 2.194927783464119,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2671,
+      "loss": 1.0093,
       "step": 135100
     },
     {
       "epoch": 2.195090250361489,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2747,
+      "loss": 0.9672,
       "step": 135110
     },
     {
       "epoch": 2.1952527172588585,
-      "grad_norm": 3.59375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2814,
+      "loss": 0.941,
       "step": 135120
     },
     {
       "epoch": 2.195415184156228,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.29,
+      "loss": 0.97,
       "step": 135130
     },
     {
       "epoch": 2.195577651053598,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2755,
+      "loss": 0.9726,
       "step": 135140
     },
     {
       "epoch": 2.1957401179509675,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.278,
+      "loss": 0.9857,
       "step": 135150
     },
     {
       "epoch": 2.195902584848337,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9809,
       "step": 135160
     },
     {
       "epoch": 2.1960650517457068,
-      "grad_norm": 3.0,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 1.0052,
       "step": 135170
     },
     {
       "epoch": 2.1962275186430764,
-      "grad_norm": 3.484375,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2907,
+      "loss": 0.9788,
       "step": 135180
     },
     {
       "epoch": 2.196389985540446,
-      "grad_norm": 4.46875,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.272,
+      "loss": 0.9988,
       "step": 135190
     },
     {
       "epoch": 2.1965524524378157,
-      "grad_norm": 2.921875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2976,
+      "loss": 0.9614,
       "step": 135200
     },
     {
       "epoch": 2.1967149193351854,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2664,
+      "loss": 0.9136,
       "step": 135210
     },
     {
       "epoch": 2.196877386232555,
-      "grad_norm": 4.625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2623,
+      "loss": 0.9933,
       "step": 135220
     },
     {
       "epoch": 2.1970398531299247,
-      "grad_norm": 3.40625,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2332,
+      "loss": 0.9629,
       "step": 135230
     },
     {
       "epoch": 2.1972023200272943,
-      "grad_norm": 3.15625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2448,
+      "loss": 0.9623,
       "step": 135240
     },
     {
       "epoch": 2.197364786924664,
-      "grad_norm": 5.53125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2342,
+      "loss": 0.967,
       "step": 135250
     },
     {
       "epoch": 2.1975272538220336,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2253,
+      "loss": 0.9633,
       "step": 135260
     },
     {
       "epoch": 2.1976897207194033,
-      "grad_norm": 4.5,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 0.9458,
       "step": 135270
     },
     {
       "epoch": 2.197852187616773,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2693,
+      "loss": 0.9859,
       "step": 135280
     },
     {
       "epoch": 2.1980146545141426,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2505,
+      "loss": 0.967,
       "step": 135290
     },
     {
       "epoch": 2.1981771214115122,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2605,
+      "loss": 0.9649,
       "step": 135300
     },
     {
       "epoch": 2.198339588308882,
-      "grad_norm": 4.5,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.97,
       "step": 135310
     },
     {
       "epoch": 2.1985020552062515,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9784,
       "step": 135320
     },
     {
       "epoch": 2.1986645221036216,
-      "grad_norm": 5.53125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2842,
+      "loss": 0.9542,
       "step": 135330
     },
     {
       "epoch": 2.198826989000991,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2542,
+      "loss": 0.9762,
       "step": 135340
     },
     {
       "epoch": 2.198989455898361,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2642,
+      "loss": 0.9552,
       "step": 135350
     },
     {
       "epoch": 2.1991519227957306,
-      "grad_norm": 4.125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2859,
+      "loss": 0.9342,
       "step": 135360
     },
     {
       "epoch": 2.1993143896931002,
-      "grad_norm": 3.375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2829,
+      "loss": 0.9857,
       "step": 135370
     },
     {
       "epoch": 2.19947685659047,
-      "grad_norm": 4.5,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.962,
       "step": 135380
     },
     {
       "epoch": 2.1996393234878395,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2795,
+      "loss": 1.018,
       "step": 135390
     },
     {
       "epoch": 2.199801790385209,
-      "grad_norm": 3.984375,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 1.0124,
       "step": 135400
     },
     {
       "epoch": 2.199964257282579,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2664,
+      "loss": 0.9665,
       "step": 135410
     },
     {
       "epoch": 2.2001267241799485,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2449,
+      "loss": 0.9734,
       "step": 135420
     },
     {
       "epoch": 2.200289191077318,
-      "grad_norm": 3.9375,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2593,
+      "loss": 0.9433,
       "step": 135430
     },
     {
       "epoch": 2.200451657974688,
-      "grad_norm": 2.953125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9599,
       "step": 135440
     },
     {
       "epoch": 2.2006141248720574,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.26,
+      "loss": 0.9371,
       "step": 135450
     },
     {
       "epoch": 2.200776591769427,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2562,
+      "loss": 0.9741,
       "step": 135460
     },
     {
       "epoch": 2.2009390586667967,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2659,
+      "loss": 0.9637,
       "step": 135470
     },
     {
       "epoch": 2.2011015255641664,
-      "grad_norm": 4.5,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2516,
+      "loss": 0.9619,
       "step": 135480
     },
     {
       "epoch": 2.201263992461536,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9904,
       "step": 135490
     },
     {
       "epoch": 2.2014264593589057,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2713,
+      "loss": 0.9883,
       "step": 135500
     },
     {
       "epoch": 2.2015889262562753,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9737,
       "step": 135510
     },
     {
       "epoch": 2.201751393153645,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9953,
       "step": 135520
     },
     {
       "epoch": 2.2019138600510146,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2801,
+      "loss": 0.979,
       "step": 135530
     },
     {
       "epoch": 2.2020763269483843,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2721,
+      "loss": 0.9797,
       "step": 135540
     },
     {
       "epoch": 2.202238793845754,
-      "grad_norm": 4.5,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2572,
+      "loss": 0.9251,
       "step": 135550
     },
     {
       "epoch": 2.2024012607431236,
-      "grad_norm": 5.0625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.259,
+      "loss": 0.9844,
       "step": 135560
     },
     {
       "epoch": 2.2025637276404932,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9942,
       "step": 135570
     },
     {
       "epoch": 2.202726194537863,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2484,
+      "loss": 1.0165,
       "step": 135580
     },
     {
       "epoch": 2.2028886614352325,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9458,
       "step": 135590
     },
     {
       "epoch": 2.203051128332602,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2462,
+      "loss": 0.9807,
       "step": 135600
     },
     {
       "epoch": 2.203213595229972,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9785,
       "step": 135610
     },
     {
       "epoch": 2.2033760621273415,
-      "grad_norm": 4.0,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2496,
+      "loss": 0.9552,
       "step": 135620
     },
     {
       "epoch": 2.203538529024711,
-      "grad_norm": 4.125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2575,
+      "loss": 0.9562,
       "step": 135630
     },
     {
       "epoch": 2.203700995922081,
-      "grad_norm": 4.5,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.257,
+      "loss": 0.9647,
       "step": 135640
     },
     {
       "epoch": 2.2038634628194504,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2491,
+      "loss": 0.9194,
       "step": 135650
     },
     {
       "epoch": 2.20402592971682,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2401,
+      "loss": 0.9658,
       "step": 135660
     },
     {
       "epoch": 2.2041883966141897,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2721,
+      "loss": 0.9869,
       "step": 135670
     },
     {
       "epoch": 2.2043508635115594,
-      "grad_norm": 3.390625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.9641,
       "step": 135680
     },
     {
       "epoch": 2.204513330408929,
-      "grad_norm": 3.140625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2717,
+      "loss": 0.9856,
       "step": 135690
     },
     {
       "epoch": 2.2046757973062987,
-      "grad_norm": 3.90625,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.279,
+      "loss": 0.9575,
       "step": 135700
     },
     {
       "epoch": 2.2048382642036684,
-      "grad_norm": 4.25,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2645,
+      "loss": 0.9992,
       "step": 135710
     },
     {
       "epoch": 2.205000731101038,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2779,
+      "loss": 0.9463,
       "step": 135720
     },
     {
       "epoch": 2.2051631979984077,
-      "grad_norm": 3.40625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2531,
+      "loss": 0.9551,
       "step": 135730
     },
     {
       "epoch": 2.2053256648957773,
-      "grad_norm": 4.90625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2484,
+      "loss": 0.9704,
       "step": 135740
     },
     {
       "epoch": 2.205488131793147,
-      "grad_norm": 4.40625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2436,
+      "loss": 0.9669,
       "step": 135750
     },
     {
       "epoch": 2.2056505986905166,
-      "grad_norm": 3.484375,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.2709,
+      "loss": 0.9797,
       "step": 135760
     },
     {
       "epoch": 2.2058130655878863,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.25,
+      "loss": 0.9644,
       "step": 135770
     },
     {
       "epoch": 2.2059755324852564,
-      "grad_norm": 3.859375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2813,
+      "loss": 0.9755,
       "step": 135780
     },
     {
       "epoch": 2.2061379993826256,
-      "grad_norm": 3.25,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2723,
+      "loss": 0.9866,
       "step": 135790
     },
     {
       "epoch": 2.2063004662799957,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2887,
+      "loss": 0.9378,
       "step": 135800
     },
     {
       "epoch": 2.2064629331773653,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2664,
+      "loss": 0.9785,
       "step": 135810
     },
     {
       "epoch": 2.206625400074735,
-      "grad_norm": 3.578125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2729,
+      "loss": 0.9783,
       "step": 135820
     },
     {
       "epoch": 2.2067878669721046,
-      "grad_norm": 4.53125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2625,
+      "loss": 0.9572,
       "step": 135830
     },
     {
       "epoch": 2.2069503338694743,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2951,
+      "loss": 0.9657,
       "step": 135840
     },
     {
       "epoch": 2.207112800766844,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2806,
+      "loss": 0.9655,
       "step": 135850
     },
     {
       "epoch": 2.2072752676642136,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2694,
+      "loss": 0.9702,
       "step": 135860
     },
     {
       "epoch": 2.207437734561583,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2364,
+      "loss": 0.9632,
       "step": 135870
     },
     {
       "epoch": 2.207600201458953,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2512,
+      "loss": 0.9616,
       "step": 135880
     },
     {
       "epoch": 2.2077626683563225,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2323,
+      "loss": 0.9456,
       "step": 135890
     },
     {
       "epoch": 2.207925135253692,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2626,
+      "loss": 0.9981,
       "step": 135900
     },
     {
       "epoch": 2.208087602151062,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 0.9848,
       "step": 135910
     },
     {
       "epoch": 2.2082500690484315,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.9581,
       "step": 135920
     },
     {
       "epoch": 2.208412535945801,
-      "grad_norm": 3.03125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2631,
+      "loss": 0.9768,
       "step": 135930
     },
     {
       "epoch": 2.2085750028431708,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2582,
+      "loss": 0.9338,
       "step": 135940
     },
     {
       "epoch": 2.2087374697405404,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2678,
+      "loss": 0.9668,
       "step": 135950
     },
     {
       "epoch": 2.20889993663791,
-      "grad_norm": 3.375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2691,
+      "loss": 1.0032,
       "step": 135960
     },
     {
       "epoch": 2.2090624035352797,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2576,
+      "loss": 0.9769,
       "step": 135970
     },
     {
       "epoch": 2.2092248704326494,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.2495,
+      "loss": 1.0184,
       "step": 135980
     },
     {
       "epoch": 2.209387337330019,
-      "grad_norm": 4.5,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2558,
+      "loss": 0.9622,
       "step": 135990
     },
     {
       "epoch": 2.2095498042273887,
-      "grad_norm": 4.9375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2611,
+      "loss": 0.9799,
       "step": 136000
     },
     {
       "epoch": 2.2097122711247583,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2386,
+      "loss": 0.9599,
       "step": 136010
     },
     {
       "epoch": 2.209874738022128,
-      "grad_norm": 3.75,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.246,
+      "loss": 0.9569,
       "step": 136020
     },
     {
       "epoch": 2.2100372049194976,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.262,
+      "loss": 0.9492,
       "step": 136030
     },
     {
       "epoch": 2.2101996718168673,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 1.0121,
       "step": 136040
     },
     {
       "epoch": 2.210362138714237,
-      "grad_norm": 4.125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2657,
+      "loss": 0.9669,
       "step": 136050
     },
     {
       "epoch": 2.2105246056116066,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2513,
+      "loss": 0.971,
       "step": 136060
     },
     {
       "epoch": 2.2106870725089762,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9847,
       "step": 136070
     },
     {
       "epoch": 2.210849539406346,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.9247,
       "step": 136080
     },
     {
       "epoch": 2.2110120063037155,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2361,
+      "loss": 0.9655,
       "step": 136090
     },
     {
       "epoch": 2.211174473201085,
-      "grad_norm": 3.546875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2558,
+      "loss": 0.9239,
       "step": 136100
     },
     {
       "epoch": 2.211336940098455,
-      "grad_norm": 3.25,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2344,
+      "loss": 0.9498,
       "step": 136110
     },
     {
       "epoch": 2.2114994069958245,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2484,
+      "loss": 0.9577,
       "step": 136120
     },
     {
       "epoch": 2.211661873893194,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2284,
+      "loss": 0.9843,
       "step": 136130
     },
     {
       "epoch": 2.211824340790564,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2422,
+      "loss": 1.001,
       "step": 136140
     },
     {
       "epoch": 2.2119868076879334,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2518,
+      "loss": 0.9798,
       "step": 136150
     },
     {
       "epoch": 2.212149274585303,
-      "grad_norm": 3.09375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.227,
+      "loss": 0.9882,
       "step": 136160
     },
     {
       "epoch": 2.2123117414826727,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2502,
+      "loss": 0.9644,
       "step": 136170
     },
     {
       "epoch": 2.2124742083800424,
-      "grad_norm": 5.71875,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2763,
+      "loss": 0.9624,
       "step": 136180
     },
     {
       "epoch": 2.212636675277412,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2484,
+      "loss": 0.9493,
       "step": 136190
     },
     {
       "epoch": 2.2127991421747817,
-      "grad_norm": 3.125,
+      "grad_norm": 12.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2357,
+      "loss": 0.9646,
       "step": 136200
     },
     {
       "epoch": 2.212961609072152,
-      "grad_norm": 4.03125,
+      "grad_norm": 12.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2359,
+      "loss": 0.9982,
       "step": 136210
     },
     {
       "epoch": 2.213124075969521,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2528,
+      "loss": 1.0076,
       "step": 136220
     },
     {
       "epoch": 2.213286542866891,
-      "grad_norm": 2.984375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2513,
+      "loss": 0.9575,
       "step": 136230
     },
     {
       "epoch": 2.2134490097642607,
-      "grad_norm": 3.078125,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.9593,
       "step": 136240
     },
     {
       "epoch": 2.2136114766616304,
-      "grad_norm": 3.59375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2741,
+      "loss": 0.9902,
       "step": 136250
     },
     {
       "epoch": 2.213773943559,
-      "grad_norm": 3.28125,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.271,
+      "loss": 0.9498,
       "step": 136260
     },
     {
       "epoch": 2.2139364104563697,
-      "grad_norm": 5.03125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.98,
       "step": 136270
     },
     {
       "epoch": 2.2140988773537393,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2812,
+      "loss": 0.9917,
       "step": 136280
     },
     {
       "epoch": 2.214261344251109,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2828,
+      "loss": 0.9902,
       "step": 136290
     },
     {
       "epoch": 2.2144238111484786,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2758,
+      "loss": 0.9663,
       "step": 136300
     },
     {
       "epoch": 2.2145862780458483,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.9795,
       "step": 136310
     },
     {
       "epoch": 2.214748744943218,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2678,
+      "loss": 0.9984,
       "step": 136320
     },
     {
       "epoch": 2.2149112118405876,
-      "grad_norm": 2.65625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.26,
+      "loss": 1.0229,
       "step": 136330
     },
     {
       "epoch": 2.2150736787379572,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2599,
+      "loss": 0.9519,
       "step": 136340
     },
     {
       "epoch": 2.215236145635327,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2825,
+      "loss": 1.0106,
       "step": 136350
     },
     {
       "epoch": 2.2153986125326965,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9383,
       "step": 136360
     },
     {
       "epoch": 2.215561079430066,
-      "grad_norm": 3.671875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2574,
+      "loss": 0.9665,
       "step": 136370
     },
     {
       "epoch": 2.215723546327436,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2478,
+      "loss": 1.0084,
       "step": 136380
     },
     {
       "epoch": 2.2158860132248055,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2654,
+      "loss": 1.0054,
       "step": 136390
     },
     {
       "epoch": 2.216048480122175,
-      "grad_norm": 2.875,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2862,
+      "loss": 0.9745,
       "step": 136400
     },
     {
       "epoch": 2.216210947019545,
-      "grad_norm": 3.5625,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.2711,
+      "loss": 1.0005,
       "step": 136410
     },
     {
       "epoch": 2.2163734139169144,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2526,
+      "loss": 0.9681,
       "step": 136420
     },
     {
       "epoch": 2.216535880814284,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2555,
+      "loss": 0.9694,
       "step": 136430
     },
     {
       "epoch": 2.2166983477116537,
-      "grad_norm": 3.59375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2747,
+      "loss": 0.9932,
       "step": 136440
     },
     {
       "epoch": 2.2168608146090234,
-      "grad_norm": 5.125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2931,
+      "loss": 0.9623,
       "step": 136450
     },
     {
       "epoch": 2.217023281506393,
-      "grad_norm": 3.625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 0.9668,
       "step": 136460
     },
     {
       "epoch": 2.2171857484037627,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2664,
+      "loss": 0.9943,
       "step": 136470
     },
     {
       "epoch": 2.2173482153011324,
-      "grad_norm": 5.25,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2693,
+      "loss": 1.0254,
       "step": 136480
     },
     {
       "epoch": 2.217510682198502,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.266,
+      "loss": 0.966,
       "step": 136490
     },
     {
       "epoch": 2.2176731490958717,
-      "grad_norm": 4.90625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2781,
+      "loss": 0.9368,
       "step": 136500
     },
     {
       "epoch": 2.2178356159932413,
-      "grad_norm": 4.375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2446,
+      "loss": 1.0251,
       "step": 136510
     },
     {
       "epoch": 2.217998082890611,
-      "grad_norm": 5.53125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2787,
+      "loss": 0.9902,
       "step": 136520
     },
     {
       "epoch": 2.2181605497879806,
-      "grad_norm": 2.8125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9965,
       "step": 136530
     },
     {
       "epoch": 2.2183230166853503,
-      "grad_norm": 4.78125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2684,
+      "loss": 0.9884,
       "step": 136540
     },
     {
       "epoch": 2.21848548358272,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2669,
+      "loss": 0.9951,
       "step": 136550
     },
     {
       "epoch": 2.2186479504800896,
-      "grad_norm": 5.90625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2492,
+      "loss": 0.9596,
       "step": 136560
     },
     {
       "epoch": 2.218810417377459,
-      "grad_norm": 5.125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 0.9825,
       "step": 136570
     },
     {
       "epoch": 2.218972884274829,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2485,
+      "loss": 1.0002,
       "step": 136580
     },
     {
       "epoch": 2.2191353511721985,
-      "grad_norm": 3.25,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2432,
+      "loss": 0.9755,
       "step": 136590
     },
     {
       "epoch": 2.219297818069568,
-      "grad_norm": 4.3125,
+      "grad_norm": 12.0,
       "learning_rate": 5e-05,
-      "loss": 0.2499,
+      "loss": 0.9781,
       "step": 136600
     },
     {
       "epoch": 2.219460284966938,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 0.958,
       "step": 136610
     },
     {
       "epoch": 2.2196227518643075,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2702,
+      "loss": 0.9586,
       "step": 136620
     },
     {
       "epoch": 2.219785218761677,
-      "grad_norm": 5.25,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2642,
+      "loss": 0.9721,
       "step": 136630
     },
     {
       "epoch": 2.2199476856590468,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9867,
       "step": 136640
     },
     {
       "epoch": 2.2201101525564164,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2498,
+      "loss": 0.9948,
       "step": 136650
     },
     {
       "epoch": 2.2202726194537865,
-      "grad_norm": 3.125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2738,
+      "loss": 0.9702,
       "step": 136660
     },
     {
       "epoch": 2.2204350863511557,
-      "grad_norm": 3.625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2614,
+      "loss": 0.9289,
       "step": 136670
     },
     {
       "epoch": 2.220597553248526,
-      "grad_norm": 5.0,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2328,
+      "loss": 1.0213,
       "step": 136680
     },
     {
       "epoch": 2.2207600201458955,
-      "grad_norm": 3.09375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2267,
+      "loss": 0.9762,
       "step": 136690
     },
     {
       "epoch": 2.220922487043265,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2937,
+      "loss": 0.9728,
       "step": 136700
     },
     {
       "epoch": 2.2210849539406348,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2665,
+      "loss": 0.9656,
       "step": 136710
     },
     {
       "epoch": 2.2212474208380044,
-      "grad_norm": 4.375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2821,
+      "loss": 0.9974,
       "step": 136720
     },
     {
       "epoch": 2.221409887735374,
-      "grad_norm": 3.390625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.9642,
       "step": 136730
     },
     {
       "epoch": 2.2215723546327437,
-      "grad_norm": 3.390625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2932,
+      "loss": 0.9626,
       "step": 136740
     },
     {
       "epoch": 2.2217348215301134,
-      "grad_norm": 4.375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2863,
+      "loss": 0.97,
       "step": 136750
     },
     {
       "epoch": 2.221897288427483,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2804,
+      "loss": 0.9961,
       "step": 136760
     },
     {
       "epoch": 2.2220597553248527,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2771,
+      "loss": 0.9662,
       "step": 136770
     },
     {
       "epoch": 2.2222222222222223,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2541,
+      "loss": 0.9876,
       "step": 136780
     },
     {
       "epoch": 2.222384689119592,
-      "grad_norm": 5.75,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.975,
       "step": 136790
     },
     {
       "epoch": 2.2225471560169616,
-      "grad_norm": 5.03125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2535,
+      "loss": 0.9464,
       "step": 136800
     },
     {
       "epoch": 2.2227096229143313,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2211,
+      "loss": 1.0248,
       "step": 136810
     },
     {
       "epoch": 2.222872089811701,
-      "grad_norm": 3.734375,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2444,
+      "loss": 0.9729,
       "step": 136820
     },
     {
       "epoch": 2.2230345567090706,
-      "grad_norm": 4.0,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2384,
+      "loss": 0.9514,
       "step": 136830
     },
     {
       "epoch": 2.2231970236064402,
-      "grad_norm": 4.71875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2743,
+      "loss": 1.0065,
       "step": 136840
     },
     {
       "epoch": 2.22335949050381,
-      "grad_norm": 4.125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2762,
+      "loss": 0.9805,
       "step": 136850
     },
     {
       "epoch": 2.2235219574011795,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2645,
+      "loss": 0.9922,
       "step": 136860
     },
     {
       "epoch": 2.223684424298549,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9958,
       "step": 136870
     },
     {
       "epoch": 2.223846891195919,
-      "grad_norm": 4.0,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2501,
+      "loss": 0.9869,
       "step": 136880
     },
     {
       "epoch": 2.2240093580932885,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2716,
+      "loss": 0.9946,
       "step": 136890
     },
     {
       "epoch": 2.224171824990658,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2493,
+      "loss": 0.9478,
       "step": 136900
     },
     {
       "epoch": 2.224334291888028,
-      "grad_norm": 4.25,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2486,
+      "loss": 1.0017,
       "step": 136910
     },
     {
       "epoch": 2.2244967587853974,
-      "grad_norm": 3.859375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.262,
+      "loss": 0.9553,
       "step": 136920
     },
     {
       "epoch": 2.224659225682767,
-      "grad_norm": 3.5,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 1.0252,
       "step": 136930
     },
     {
       "epoch": 2.2248216925801367,
-      "grad_norm": 4.78125,
+      "grad_norm": 12.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 0.999,
       "step": 136940
     },
     {
       "epoch": 2.2249841594775064,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 1.0201,
       "step": 136950
     },
     {
       "epoch": 2.225146626374876,
-      "grad_norm": 3.21875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2666,
+      "loss": 0.985,
       "step": 136960
     },
     {
       "epoch": 2.2253090932722457,
-      "grad_norm": 3.703125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2875,
+      "loss": 1.0018,
       "step": 136970
     },
     {
       "epoch": 2.2254715601696153,
-      "grad_norm": 3.515625,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2682,
+      "loss": 0.9416,
       "step": 136980
     },
     {
       "epoch": 2.225634027066985,
-      "grad_norm": 3.984375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2885,
+      "loss": 1.0072,
       "step": 136990
     },
     {
       "epoch": 2.2257964939643546,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2859,
+      "loss": 0.9195,
       "step": 137000
     },
     {
       "epoch": 2.2259589608617243,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2784,
+      "loss": 0.9969,
       "step": 137010
     },
     {
       "epoch": 2.226121427759094,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2762,
+      "loss": 0.9618,
       "step": 137020
     },
     {
       "epoch": 2.2262838946564636,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9787,
       "step": 137030
     },
     {
       "epoch": 2.2264463615538332,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2719,
+      "loss": 0.9519,
       "step": 137040
     },
     {
       "epoch": 2.226608828451203,
-      "grad_norm": 5.0625,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.988,
       "step": 137050
     },
     {
       "epoch": 2.2267712953485725,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2516,
+      "loss": 0.9843,
       "step": 137060
     },
     {
       "epoch": 2.226933762245942,
-      "grad_norm": 4.34375,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.9756,
       "step": 137070
     },
     {
       "epoch": 2.227096229143312,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.9948,
       "step": 137080
     },
     {
       "epoch": 2.227258696040682,
-      "grad_norm": 3.734375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2398,
+      "loss": 0.9607,
       "step": 137090
     },
     {
       "epoch": 2.227421162938051,
-      "grad_norm": 3.78125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2533,
+      "loss": 0.9428,
       "step": 137100
     },
     {
       "epoch": 2.2275836298354212,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2629,
+      "loss": 0.9615,
       "step": 137110
     },
     {
       "epoch": 2.227746096732791,
-      "grad_norm": 3.671875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2811,
+      "loss": 0.96,
       "step": 137120
     },
     {
       "epoch": 2.2279085636301605,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2624,
+      "loss": 0.977,
       "step": 137130
     },
     {
       "epoch": 2.22807103052753,
-      "grad_norm": 3.5625,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2619,
+      "loss": 0.9474,
       "step": 137140
     },
     {
       "epoch": 2.2282334974249,
-      "grad_norm": 3.875,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2563,
+      "loss": 0.9811,
       "step": 137150
     },
     {
       "epoch": 2.2283959643222695,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 1.0214,
       "step": 137160
     },
     {
       "epoch": 2.228558431219639,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9757,
       "step": 137170
     },
     {
       "epoch": 2.228720898117009,
-      "grad_norm": 4.78125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2824,
+      "loss": 0.9649,
       "step": 137180
     },
     {
       "epoch": 2.2288833650143784,
-      "grad_norm": 3.421875,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2922,
+      "loss": 0.9855,
       "step": 137190
     },
     {
       "epoch": 2.229045831911748,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2742,
+      "loss": 0.9839,
       "step": 137200
     },
     {
       "epoch": 2.2292082988091177,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2718,
+      "loss": 0.9615,
       "step": 137210
     },
     {
       "epoch": 2.2293707657064874,
-      "grad_norm": 3.09375,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2426,
+      "loss": 0.9857,
       "step": 137220
     },
     {
       "epoch": 2.229533232603857,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2377,
+      "loss": 0.9945,
       "step": 137230
     },
     {
       "epoch": 2.2296956995012267,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2533,
+      "loss": 0.9612,
       "step": 137240
     },
     {
       "epoch": 2.2298581663985964,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2753,
+      "loss": 0.9723,
       "step": 137250
     },
     {
       "epoch": 2.230020633295966,
-      "grad_norm": 2.828125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2911,
+      "loss": 0.9372,
       "step": 137260
     },
     {
       "epoch": 2.2301831001933357,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2611,
+      "loss": 0.9756,
       "step": 137270
     },
     {
       "epoch": 2.2303455670907053,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2847,
+      "loss": 0.9545,
       "step": 137280
     },
     {
       "epoch": 2.230508033988075,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2941,
+      "loss": 0.9552,
       "step": 137290
     },
     {
       "epoch": 2.2306705008854446,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2662,
+      "loss": 0.9652,
       "step": 137300
     },
     {
       "epoch": 2.2308329677828143,
-      "grad_norm": 2.96875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 0.9655,
       "step": 137310
     },
     {
       "epoch": 2.230995434680184,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2826,
+      "loss": 0.947,
       "step": 137320
     },
     {
       "epoch": 2.2311579015775536,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2472,
+      "loss": 0.9679,
       "step": 137330
     },
     {
       "epoch": 2.231320368474923,
-      "grad_norm": 4.53125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2337,
+      "loss": 0.9612,
       "step": 137340
     },
     {
       "epoch": 2.231482835372293,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2402,
+      "loss": 0.9555,
       "step": 137350
     },
     {
       "epoch": 2.2316453022696625,
-      "grad_norm": 4.4375,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2439,
+      "loss": 0.9552,
       "step": 137360
     },
     {
       "epoch": 2.231807769167032,
-      "grad_norm": 3.890625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.256,
+      "loss": 0.9584,
       "step": 137370
     },
     {
       "epoch": 2.231970236064402,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2622,
+      "loss": 0.9972,
       "step": 137380
     },
     {
       "epoch": 2.2321327029617715,
-      "grad_norm": 5.1875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2677,
+      "loss": 0.9896,
       "step": 137390
     },
     {
       "epoch": 2.232295169859141,
-      "grad_norm": 3.25,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2598,
+      "loss": 1.0018,
       "step": 137400
     },
     {
       "epoch": 2.2324576367565108,
-      "grad_norm": 4.0,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2527,
+      "loss": 0.968,
       "step": 137410
     },
     {
       "epoch": 2.2326201036538804,
-      "grad_norm": 3.875,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2467,
+      "loss": 0.9708,
       "step": 137420
     },
     {
       "epoch": 2.23278257055125,
-      "grad_norm": 5.40625,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9437,
       "step": 137430
     },
     {
       "epoch": 2.2329450374486197,
-      "grad_norm": 4.21875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2715,
+      "loss": 0.9861,
       "step": 137440
     },
     {
       "epoch": 2.2331075043459894,
-      "grad_norm": 3.5,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2485,
+      "loss": 0.9637,
       "step": 137450
     },
     {
       "epoch": 2.233269971243359,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2489,
+      "loss": 0.9568,
       "step": 137460
     },
     {
       "epoch": 2.2334324381407287,
-      "grad_norm": 5.65625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9653,
       "step": 137470
     },
     {
       "epoch": 2.2335949050380983,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2682,
+      "loss": 0.9696,
       "step": 137480
     },
     {
       "epoch": 2.233757371935468,
-      "grad_norm": 2.90625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2343,
+      "loss": 0.9813,
       "step": 137490
     },
     {
       "epoch": 2.2339198388328376,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2492,
+      "loss": 0.9891,
       "step": 137500
     },
     {
       "epoch": 2.2340823057302073,
-      "grad_norm": 5.78125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 0.9827,
       "step": 137510
     },
     {
       "epoch": 2.234244772627577,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2702,
+      "loss": 0.9852,
       "step": 137520
     },
     {
       "epoch": 2.2344072395249466,
-      "grad_norm": 4.65625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.263,
+      "loss": 0.961,
       "step": 137530
     },
     {
       "epoch": 2.2345697064223167,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2681,
+      "loss": 0.9809,
       "step": 137540
     },
     {
       "epoch": 2.2347321733196863,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2584,
+      "loss": 0.9702,
       "step": 137550
     },
     {
       "epoch": 2.234894640217056,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2641,
+      "loss": 0.986,
       "step": 137560
     },
     {
       "epoch": 2.2350571071144256,
-      "grad_norm": 3.015625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2652,
+      "loss": 0.9788,
       "step": 137570
     },
     {
       "epoch": 2.2352195740117953,
-      "grad_norm": 3.015625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2368,
+      "loss": 0.9684,
       "step": 137580
     },
     {
       "epoch": 2.235382040909165,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2723,
+      "loss": 0.9798,
       "step": 137590
     },
     {
       "epoch": 2.2355445078065346,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2695,
+      "loss": 0.967,
       "step": 137600
     },
     {
       "epoch": 2.2357069747039042,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2452,
+      "loss": 0.9947,
       "step": 137610
     },
     {
       "epoch": 2.235869441601274,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2724,
+      "loss": 0.983,
       "step": 137620
     },
     {
       "epoch": 2.2360319084986435,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2641,
+      "loss": 0.9802,
       "step": 137630
     },
     {
       "epoch": 2.236194375396013,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2701,
+      "loss": 0.9713,
       "step": 137640
     },
     {
       "epoch": 2.236356842293383,
-      "grad_norm": 4.3125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2393,
+      "loss": 0.9902,
       "step": 137650
     },
     {
       "epoch": 2.2365193091907525,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2403,
+      "loss": 0.9991,
       "step": 137660
     },
     {
       "epoch": 2.236681776088122,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9945,
       "step": 137670
     },
     {
       "epoch": 2.236844242985492,
-      "grad_norm": 3.421875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2418,
+      "loss": 0.9608,
       "step": 137680
     },
     {
       "epoch": 2.2370067098828614,
-      "grad_norm": 4.8125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2569,
+      "loss": 0.9841,
       "step": 137690
     },
     {
       "epoch": 2.237169176780231,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2564,
+      "loss": 1.0132,
       "step": 137700
     },
     {
       "epoch": 2.2373316436776007,
-      "grad_norm": 4.125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2448,
+      "loss": 0.9957,
       "step": 137710
     },
     {
       "epoch": 2.2374941105749704,
-      "grad_norm": 6.03125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2693,
+      "loss": 0.9611,
       "step": 137720
     },
     {
       "epoch": 2.23765657747234,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 0.9917,
       "step": 137730
     },
     {
       "epoch": 2.2378190443697097,
-      "grad_norm": 4.53125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2788,
+      "loss": 0.9687,
       "step": 137740
     },
     {
       "epoch": 2.2379815112670793,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2476,
+      "loss": 0.9752,
       "step": 137750
     },
     {
       "epoch": 2.238143978164449,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2483,
+      "loss": 0.9588,
       "step": 137760
     },
     {
       "epoch": 2.2383064450618186,
-      "grad_norm": 3.6875,
+      "grad_norm": 18.875,
       "learning_rate": 5e-05,
-      "loss": 0.2569,
+      "loss": 0.9622,
       "step": 137770
     },
     {
       "epoch": 2.2384689119591883,
-      "grad_norm": 3.484375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2496,
+      "loss": 0.9878,
       "step": 137780
     },
     {
       "epoch": 2.238631378856558,
-      "grad_norm": 3.25,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2413,
+      "loss": 1.0043,
       "step": 137790
     },
     {
       "epoch": 2.2387938457539276,
-      "grad_norm": 4.0,
+      "grad_norm": 8.0,
       "learning_rate": 5e-05,
-      "loss": 0.2403,
+      "loss": 0.958,
       "step": 137800
     },
     {
       "epoch": 2.2389563126512972,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2453,
+      "loss": 1.0025,
       "step": 137810
     },
     {
       "epoch": 2.239118779548667,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.221,
+      "loss": 0.948,
       "step": 137820
     },
     {
       "epoch": 2.2392812464460365,
-      "grad_norm": 4.53125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2301,
+      "loss": 0.9716,
       "step": 137830
     },
     {
       "epoch": 2.239443713343406,
-      "grad_norm": 2.78125,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2505,
+      "loss": 0.9863,
       "step": 137840
     },
     {
       "epoch": 2.239606180240776,
-      "grad_norm": 2.921875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2497,
+      "loss": 0.9957,
       "step": 137850
     },
     {
       "epoch": 2.2397686471381455,
-      "grad_norm": 3.5,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2481,
+      "loss": 1.0294,
       "step": 137860
     },
     {
       "epoch": 2.239931114035515,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2481,
+      "loss": 1.0201,
       "step": 137870
     },
     {
       "epoch": 2.240093580932885,
-      "grad_norm": 4.8125,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2561,
+      "loss": 0.9822,
       "step": 137880
     },
     {
       "epoch": 2.2402560478302544,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2794,
+      "loss": 0.9757,
       "step": 137890
     },
     {
       "epoch": 2.240418514727624,
-      "grad_norm": 3.109375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2656,
+      "loss": 0.9707,
       "step": 137900
     },
     {
       "epoch": 2.2405809816249938,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2444,
+      "loss": 0.9806,
       "step": 137910
     },
     {
       "epoch": 2.2407434485223634,
-      "grad_norm": 3.953125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2569,
+      "loss": 0.9809,
       "step": 137920
     },
     {
       "epoch": 2.240905915419733,
-      "grad_norm": 5.0,
+      "grad_norm": 12.8125,
       "learning_rate": 5e-05,
-      "loss": 0.25,
+      "loss": 0.9821,
       "step": 137930
     },
     {
       "epoch": 2.2410683823171027,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2734,
+      "loss": 0.9669,
       "step": 137940
     },
     {
       "epoch": 2.2412308492144724,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.25,
+      "loss": 0.9653,
       "step": 137950
     },
     {
       "epoch": 2.241393316111842,
-      "grad_norm": 5.3125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2493,
+      "loss": 0.956,
       "step": 137960
     },
     {
       "epoch": 2.241555783009212,
-      "grad_norm": 4.03125,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2562,
+      "loss": 0.959,
       "step": 137970
     },
     {
       "epoch": 2.2417182499065813,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2911,
+      "loss": 0.9831,
       "step": 137980
     },
     {
       "epoch": 2.2418807168039514,
-      "grad_norm": 3.375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2604,
+      "loss": 0.95,
       "step": 137990
     },
     {
       "epoch": 2.242043183701321,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.269,
+      "loss": 1.0075,
       "step": 138000
     },
     {
       "epoch": 2.2422056505986907,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.274,
+      "loss": 0.9694,
       "step": 138010
     },
     {
       "epoch": 2.2423681174960604,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2475,
+      "loss": 0.9903,
       "step": 138020
     },
     {
       "epoch": 2.24253058439343,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2787,
+      "loss": 1.0022,
       "step": 138030
     },
     {
       "epoch": 2.2426930512907997,
-      "grad_norm": 4.125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 0.9801,
       "step": 138040
     },
     {
       "epoch": 2.2428555181881693,
-      "grad_norm": 3.375,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2605,
+      "loss": 0.9909,
       "step": 138050
     },
     {
       "epoch": 2.243017985085539,
-      "grad_norm": 4.625,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2434,
+      "loss": 0.9743,
       "step": 138060
     },
     {
       "epoch": 2.2431804519829086,
-      "grad_norm": 4.40625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2342,
+      "loss": 0.9675,
       "step": 138070
     },
     {
       "epoch": 2.2433429188802783,
-      "grad_norm": 4.375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2576,
+      "loss": 0.9714,
       "step": 138080
     },
     {
       "epoch": 2.243505385777648,
-      "grad_norm": 3.09375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2548,
+      "loss": 0.9603,
       "step": 138090
     },
     {
       "epoch": 2.2436678526750176,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2478,
+      "loss": 1.0016,
       "step": 138100
     },
     {
       "epoch": 2.243830319572387,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2542,
+      "loss": 0.9559,
       "step": 138110
     },
     {
       "epoch": 2.243992786469757,
-      "grad_norm": 4.0,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.238,
+      "loss": 0.9793,
       "step": 138120
     },
     {
       "epoch": 2.2441552533671265,
-      "grad_norm": 3.921875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2536,
+      "loss": 1.0044,
       "step": 138130
     },
     {
       "epoch": 2.244317720264496,
-      "grad_norm": 4.5,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2566,
+      "loss": 0.9928,
       "step": 138140
     },
     {
       "epoch": 2.244480187161866,
-      "grad_norm": 3.390625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2555,
+      "loss": 0.9763,
       "step": 138150
     },
     {
       "epoch": 2.2446426540592355,
-      "grad_norm": 3.5,
+      "grad_norm": 7.96875,
       "learning_rate": 5e-05,
-      "loss": 0.2744,
+      "loss": 0.9933,
       "step": 138160
     },
     {
       "epoch": 2.244805120956605,
-      "grad_norm": 3.453125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2399,
+      "loss": 0.9517,
       "step": 138170
     },
     {
       "epoch": 2.2449675878539748,
-      "grad_norm": 3.828125,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2765,
+      "loss": 1.0194,
       "step": 138180
     },
     {
       "epoch": 2.2451300547513444,
-      "grad_norm": 4.34375,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2738,
+      "loss": 1.0111,
       "step": 138190
     },
     {
       "epoch": 2.245292521648714,
-      "grad_norm": 3.609375,
+      "grad_norm": 7.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2909,
+      "loss": 1.0131,
       "step": 138200
     },
     {
       "epoch": 2.2454549885460837,
-      "grad_norm": 4.84375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.284,
+      "loss": 0.9913,
       "step": 138210
     },
     {
       "epoch": 2.2456174554434534,
-      "grad_norm": 4.46875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2594,
+      "loss": 0.9618,
       "step": 138220
     },
     {
       "epoch": 2.245779922340823,
-      "grad_norm": 4.84375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2718,
+      "loss": 1.0065,
       "step": 138230
     },
     {
       "epoch": 2.2459423892381927,
-      "grad_norm": 3.171875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.248,
+      "loss": 0.9782,
       "step": 138240
     },
     {
       "epoch": 2.2461048561355623,
-      "grad_norm": 4.03125,
+      "grad_norm": 12.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2599,
+      "loss": 0.9787,
       "step": 138250
     },
     {
       "epoch": 2.246267323032932,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.255,
+      "loss": 0.9647,
       "step": 138260
     },
     {
       "epoch": 2.2464297899303016,
-      "grad_norm": 3.625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2473,
+      "loss": 0.9589,
       "step": 138270
     },
     {
       "epoch": 2.2465922568276713,
-      "grad_norm": 4.375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2696,
+      "loss": 0.9743,
       "step": 138280
     },
     {
       "epoch": 2.246754723725041,
-      "grad_norm": 3.125,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 0.9566,
       "step": 138290
     },
     {
       "epoch": 2.2469171906224106,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.281,
+      "loss": 1.0041,
       "step": 138300
     },
     {
       "epoch": 2.2470796575197802,
-      "grad_norm": 3.453125,
+      "grad_norm": 13.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2501,
+      "loss": 0.993,
       "step": 138310
     },
     {
       "epoch": 2.24724212441715,
-      "grad_norm": 4.375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.9628,
       "step": 138320
     },
     {
       "epoch": 2.2474045913145195,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2379,
+      "loss": 0.9942,
       "step": 138330
     },
     {
       "epoch": 2.247567058211889,
-      "grad_norm": 4.90625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2371,
+      "loss": 0.9925,
       "step": 138340
     },
     {
       "epoch": 2.247729525109259,
-      "grad_norm": 6.03125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2635,
+      "loss": 0.9511,
       "step": 138350
     },
     {
       "epoch": 2.2478919920066285,
-      "grad_norm": 3.265625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 0.9751,
       "step": 138360
     },
     {
       "epoch": 2.248054458903998,
-      "grad_norm": 3.953125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2583,
+      "loss": 0.9846,
       "step": 138370
     },
     {
       "epoch": 2.248216925801368,
-      "grad_norm": 4.96875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2539,
+      "loss": 0.9775,
       "step": 138380
     },
     {
       "epoch": 2.2483793926987374,
-      "grad_norm": 3.078125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9748,
       "step": 138390
     },
     {
       "epoch": 2.2485418595961075,
-      "grad_norm": 3.953125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.263,
+      "loss": 0.992,
       "step": 138400
     },
     {
       "epoch": 2.2487043264934767,
-      "grad_norm": 3.671875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2802,
+      "loss": 0.9554,
       "step": 138410
     },
     {
       "epoch": 2.248866793390847,
-      "grad_norm": 3.9375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2857,
+      "loss": 0.9849,
       "step": 138420
     },
     {
       "epoch": 2.2490292602882165,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2599,
+      "loss": 0.9214,
       "step": 138430
     },
     {
       "epoch": 2.249191727185586,
-      "grad_norm": 4.90625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2721,
+      "loss": 0.9401,
       "step": 138440
     },
     {
       "epoch": 2.249354194082956,
-      "grad_norm": 3.296875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.26,
+      "loss": 0.9697,
       "step": 138450
     },
     {
       "epoch": 2.2495166609803254,
-      "grad_norm": 3.703125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2763,
+      "loss": 0.9693,
       "step": 138460
     },
     {
       "epoch": 2.249679127877695,
-      "grad_norm": 3.3125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.256,
+      "loss": 0.9723,
       "step": 138470
     },
     {
       "epoch": 2.2498415947750647,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2375,
+      "loss": 0.9966,
       "step": 138480
     },
     {
       "epoch": 2.2500040616724344,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2288,
+      "loss": 0.9803,
       "step": 138490
     },
     {
       "epoch": 2.250166528569804,
-      "grad_norm": 3.9375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2493,
+      "loss": 0.9783,
       "step": 138500
     },
     {
       "epoch": 2.2503289954671737,
-      "grad_norm": 4.875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2576,
+      "loss": 0.9472,
       "step": 138510
     },
     {
       "epoch": 2.2504914623645433,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2674,
+      "loss": 0.9705,
       "step": 138520
     },
     {
       "epoch": 2.250653929261913,
-      "grad_norm": 4.0625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2426,
+      "loss": 1.0006,
       "step": 138530
     },
     {
       "epoch": 2.2508163961592826,
-      "grad_norm": 3.8125,
+      "grad_norm": 12.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2547,
+      "loss": 0.9747,
       "step": 138540
     },
     {
       "epoch": 2.2509788630566523,
-      "grad_norm": 4.625,
+      "grad_norm": 11.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2496,
+      "loss": 0.9786,
       "step": 138550
     },
     {
       "epoch": 2.251141329954022,
-      "grad_norm": 4.5,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2186,
+      "loss": 0.9942,
       "step": 138560
     },
     {
       "epoch": 2.2513037968513916,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2531,
+      "loss": 0.9709,
       "step": 138570
     },
     {
       "epoch": 2.2514662637487612,
-      "grad_norm": 4.25,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2401,
+      "loss": 0.9812,
       "step": 138580
     },
     {
       "epoch": 2.251628730646131,
-      "grad_norm": 4.65625,
+      "grad_norm": 11.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2282,
+      "loss": 0.9507,
       "step": 138590
     },
     {
       "epoch": 2.2517911975435005,
-      "grad_norm": 2.859375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2248,
+      "loss": 0.9564,
       "step": 138600
     },
     {
       "epoch": 2.25195366444087,
-      "grad_norm": 3.4375,
+      "grad_norm": 13.375,
       "learning_rate": 5e-05,
-      "loss": 0.2307,
+      "loss": 0.9784,
       "step": 138610
     },
     {
       "epoch": 2.25211613133824,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2855,
+      "loss": 0.9783,
       "step": 138620
     },
     {
       "epoch": 2.2522785982356095,
-      "grad_norm": 3.421875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2475,
+      "loss": 1.004,
       "step": 138630
     },
     {
       "epoch": 2.252441065132979,
-      "grad_norm": 3.625,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2526,
+      "loss": 0.9932,
       "step": 138640
     },
     {
       "epoch": 2.252603532030349,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.259,
+      "loss": 1.0391,
       "step": 138650
     },
     {
       "epoch": 2.2527659989277184,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2575,
+      "loss": 1.0035,
       "step": 138660
     },
     {
       "epoch": 2.252928465825088,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2619,
+      "loss": 0.9794,
       "step": 138670
     },
     {
       "epoch": 2.2530909327224578,
-      "grad_norm": 3.359375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2369,
+      "loss": 0.9705,
       "step": 138680
     },
     {
       "epoch": 2.2532533996198274,
-      "grad_norm": 3.28125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.258,
+      "loss": 0.9731,
       "step": 138690
     },
     {
       "epoch": 2.253415866517197,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2587,
+      "loss": 0.9643,
       "step": 138700
     },
     {
       "epoch": 2.2535783334145667,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2636,
+      "loss": 0.9792,
       "step": 138710
     },
     {
       "epoch": 2.2537408003119364,
-      "grad_norm": 3.953125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2731,
+      "loss": 0.9465,
       "step": 138720
     },
     {
       "epoch": 2.253903267209306,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2441,
+      "loss": 0.9991,
       "step": 138730
     },
     {
       "epoch": 2.2540657341066757,
-      "grad_norm": 5.3125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2537,
+      "loss": 0.9993,
       "step": 138740
     },
     {
       "epoch": 2.2542282010040453,
-      "grad_norm": 4.53125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2767,
+      "loss": 0.9723,
       "step": 138750
     },
     {
       "epoch": 2.254390667901415,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2631,
+      "loss": 0.9816,
       "step": 138760
     },
     {
       "epoch": 2.2545531347987846,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.269,
+      "loss": 0.9784,
       "step": 138770
     },
     {
       "epoch": 2.2547156016961543,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2798,
+      "loss": 0.9702,
       "step": 138780
     },
     {
       "epoch": 2.254878068593524,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2785,
+      "loss": 0.9917,
       "step": 138790
     },
     {
       "epoch": 2.2550405354908936,
-      "grad_norm": 4.46875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2597,
+      "loss": 0.9596,
       "step": 138800
     },
     {
       "epoch": 2.255203002388263,
-      "grad_norm": 3.734375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2425,
+      "loss": 0.9571,
       "step": 138810
     },
     {
       "epoch": 2.255365469285633,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2737,
+      "loss": 0.9386,
       "step": 138820
     },
     {
       "epoch": 2.255527936183003,
-      "grad_norm": 3.765625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9512,
       "step": 138830
     },
     {
       "epoch": 2.255690403080372,
-      "grad_norm": 3.421875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2786,
+      "loss": 0.9957,
       "step": 138840
     },
     {
       "epoch": 2.2558528699777423,
-      "grad_norm": 4.5,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 1.0357,
       "step": 138850
     },
     {
       "epoch": 2.2560153368751115,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2666,
+      "loss": 0.9793,
       "step": 138860
     },
     {
       "epoch": 2.2561778037724816,
-      "grad_norm": 4.25,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 0.9814,
       "step": 138870
     },
     {
       "epoch": 2.256340270669851,
-      "grad_norm": 3.1875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2684,
+      "loss": 0.9689,
       "step": 138880
     },
     {
       "epoch": 2.256502737567221,
-      "grad_norm": 3.234375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2874,
+      "loss": 0.9642,
       "step": 138890
     },
     {
       "epoch": 2.2566652044645905,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.3015,
+      "loss": 1.0189,
       "step": 138900
     },
     {
       "epoch": 2.25682767136196,
-      "grad_norm": 3.953125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2718,
+      "loss": 0.9803,
       "step": 138910
     },
     {
       "epoch": 2.25699013825933,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 1.005,
       "step": 138920
     },
     {
       "epoch": 2.2571526051566995,
-      "grad_norm": 4.9375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2589,
+      "loss": 0.967,
       "step": 138930
     },
     {
       "epoch": 2.257315072054069,
-      "grad_norm": 2.953125,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2652,
+      "loss": 0.9593,
       "step": 138940
     },
     {
       "epoch": 2.2574775389514388,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2759,
+      "loss": 0.9861,
       "step": 138950
     },
     {
       "epoch": 2.2576400058488084,
-      "grad_norm": 5.0625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2625,
+      "loss": 0.9759,
       "step": 138960
     },
     {
       "epoch": 2.257802472746178,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2759,
+      "loss": 0.987,
       "step": 138970
     },
     {
       "epoch": 2.2579649396435477,
-      "grad_norm": 3.46875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2822,
+      "loss": 0.9915,
       "step": 138980
     },
     {
       "epoch": 2.2581274065409174,
-      "grad_norm": 3.71875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2928,
+      "loss": 0.96,
       "step": 138990
     },
     {
       "epoch": 2.258289873438287,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9723,
       "step": 139000
     },
     {
       "epoch": 2.2584523403356567,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2544,
+      "loss": 0.976,
       "step": 139010
     },
     {
       "epoch": 2.2586148072330263,
-      "grad_norm": 3.578125,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2717,
+      "loss": 1.0036,
       "step": 139020
     },
     {
       "epoch": 2.258777274130396,
-      "grad_norm": 3.953125,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 1.0088,
       "step": 139030
     },
     {
       "epoch": 2.2589397410277656,
-      "grad_norm": 4.125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2819,
+      "loss": 1.0161,
       "step": 139040
     },
     {
       "epoch": 2.2591022079251353,
-      "grad_norm": 3.375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2615,
+      "loss": 1.0011,
       "step": 139050
     },
     {
       "epoch": 2.259264674822505,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2754,
+      "loss": 0.9677,
       "step": 139060
     },
     {
       "epoch": 2.2594271417198746,
-      "grad_norm": 5.40625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2809,
+      "loss": 0.9991,
       "step": 139070
     },
     {
       "epoch": 2.2595896086172442,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2894,
+      "loss": 0.9795,
       "step": 139080
     },
     {
       "epoch": 2.259752075514614,
-      "grad_norm": 3.265625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2677,
+      "loss": 1.0123,
       "step": 139090
     },
     {
       "epoch": 2.2599145424119835,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2458,
+      "loss": 0.9904,
       "step": 139100
     },
     {
       "epoch": 2.260077009309353,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2551,
+      "loss": 0.967,
       "step": 139110
     },
     {
       "epoch": 2.260239476206723,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9713,
       "step": 139120
     },
     {
       "epoch": 2.2604019431040925,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2445,
+      "loss": 1.0048,
       "step": 139130
     },
     {
       "epoch": 2.260564410001462,
-      "grad_norm": 2.875,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2386,
+      "loss": 0.9658,
       "step": 139140
     },
     {
       "epoch": 2.260726876898832,
-      "grad_norm": 4.78125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2577,
+      "loss": 0.9749,
       "step": 139150
     },
     {
       "epoch": 2.2608893437962014,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2457,
+      "loss": 0.9743,
       "step": 139160
     },
     {
       "epoch": 2.261051810693571,
-      "grad_norm": 3.234375,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 1.0086,
       "step": 139170
     },
     {
       "epoch": 2.2612142775909407,
-      "grad_norm": 3.5,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2625,
+      "loss": 0.9528,
       "step": 139180
     },
     {
       "epoch": 2.2613767444883104,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2436,
+      "loss": 0.9881,
       "step": 139190
     },
     {
       "epoch": 2.26153921138568,
-      "grad_norm": 3.6875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2534,
+      "loss": 0.9671,
       "step": 139200
     },
     {
       "epoch": 2.2617016782830497,
-      "grad_norm": 4.65625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9909,
       "step": 139210
     },
     {
       "epoch": 2.2618641451804193,
-      "grad_norm": 4.78125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 0.9882,
       "step": 139220
     },
     {
       "epoch": 2.262026612077789,
-      "grad_norm": 4.5,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2553,
+      "loss": 0.9848,
       "step": 139230
     },
     {
       "epoch": 2.2621890789751586,
-      "grad_norm": 2.984375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 0.9909,
       "step": 139240
     },
     {
       "epoch": 2.2623515458725283,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2474,
+      "loss": 0.9632,
       "step": 139250
     },
     {
       "epoch": 2.262514012769898,
-      "grad_norm": 4.34375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2627,
+      "loss": 0.9823,
       "step": 139260
     },
     {
       "epoch": 2.2626764796672676,
-      "grad_norm": 4.0,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2772,
+      "loss": 1.006,
       "step": 139270
     },
     {
       "epoch": 2.2628389465646377,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.9676,
       "step": 139280
     },
     {
       "epoch": 2.263001413462007,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2637,
+      "loss": 0.9661,
       "step": 139290
     },
     {
       "epoch": 2.263163880359377,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2572,
+      "loss": 0.9818,
       "step": 139300
     },
     {
       "epoch": 2.263326347256746,
-      "grad_norm": 4.25,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2821,
+      "loss": 0.9897,
       "step": 139310
     },
     {
       "epoch": 2.2634888141541163,
-      "grad_norm": 3.875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.978,
       "step": 139320
     },
     {
       "epoch": 2.263651281051486,
-      "grad_norm": 4.78125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9861,
       "step": 139330
     },
     {
       "epoch": 2.2638137479488556,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2554,
+      "loss": 0.9727,
       "step": 139340
     },
     {
       "epoch": 2.2639762148462252,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2469,
+      "loss": 0.9818,
       "step": 139350
     },
     {
       "epoch": 2.264138681743595,
-      "grad_norm": 3.984375,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2601,
+      "loss": 0.962,
       "step": 139360
     },
     {
       "epoch": 2.2643011486409645,
-      "grad_norm": 3.4375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 0.9649,
       "step": 139370
     },
     {
       "epoch": 2.264463615538334,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 1.0103,
       "step": 139380
     },
     {
       "epoch": 2.264626082435704,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2784,
+      "loss": 0.975,
       "step": 139390
     },
     {
       "epoch": 2.2647885493330735,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2909,
+      "loss": 0.9539,
       "step": 139400
     },
     {
       "epoch": 2.264951016230443,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2701,
+      "loss": 1.0031,
       "step": 139410
     },
     {
       "epoch": 2.265113483127813,
-      "grad_norm": 4.375,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2696,
+      "loss": 0.9712,
       "step": 139420
     },
     {
       "epoch": 2.2652759500251824,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2793,
+      "loss": 1.0044,
       "step": 139430
     },
     {
       "epoch": 2.265438416922552,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2665,
+      "loss": 0.9942,
       "step": 139440
     },
     {
       "epoch": 2.2656008838199218,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2719,
+      "loss": 0.9792,
       "step": 139450
     },
     {
       "epoch": 2.2657633507172914,
-      "grad_norm": 4.34375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9973,
       "step": 139460
     },
     {
       "epoch": 2.265925817614661,
-      "grad_norm": 4.625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2493,
+      "loss": 0.9937,
       "step": 139470
     },
     {
       "epoch": 2.2660882845120307,
-      "grad_norm": 3.53125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2283,
+      "loss": 0.9618,
       "step": 139480
     },
     {
       "epoch": 2.2662507514094004,
-      "grad_norm": 4.34375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2291,
+      "loss": 0.9821,
       "step": 139490
     },
     {
       "epoch": 2.26641321830677,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2458,
+      "loss": 0.9906,
       "step": 139500
     },
     {
       "epoch": 2.2665756852041397,
-      "grad_norm": 3.875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2645,
+      "loss": 0.9709,
       "step": 139510
     },
     {
       "epoch": 2.2667381521015093,
-      "grad_norm": 3.65625,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2712,
+      "loss": 1.0106,
       "step": 139520
     },
     {
       "epoch": 2.266900618998879,
-      "grad_norm": 3.90625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2842,
+      "loss": 0.9901,
       "step": 139530
     },
     {
       "epoch": 2.2670630858962486,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2671,
+      "loss": 0.9987,
       "step": 139540
     },
     {
       "epoch": 2.2672255527936183,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2645,
+      "loss": 0.9561,
       "step": 139550
     },
     {
       "epoch": 2.267388019690988,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9529,
       "step": 139560
     },
     {
       "epoch": 2.2675504865883576,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2845,
+      "loss": 0.9902,
       "step": 139570
     },
     {
       "epoch": 2.267712953485727,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3021,
+      "loss": 0.967,
       "step": 139580
     },
     {
       "epoch": 2.267875420383097,
-      "grad_norm": 3.078125,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 0.9914,
       "step": 139590
     },
     {
       "epoch": 2.2680378872804665,
-      "grad_norm": 2.890625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2483,
+      "loss": 0.9536,
       "step": 139600
     },
     {
       "epoch": 2.268200354177836,
-      "grad_norm": 4.375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 0.9645,
       "step": 139610
     },
     {
       "epoch": 2.268362821075206,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.235,
+      "loss": 0.9934,
       "step": 139620
     },
     {
       "epoch": 2.2685252879725755,
-      "grad_norm": 4.875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2686,
+      "loss": 0.9861,
       "step": 139630
     },
     {
       "epoch": 2.268687754869945,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2297,
+      "loss": 1.0222,
       "step": 139640
     },
     {
       "epoch": 2.2688502217673148,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.279,
+      "loss": 0.9446,
       "step": 139650
     },
     {
       "epoch": 2.2690126886646844,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2359,
+      "loss": 0.9884,
       "step": 139660
     },
     {
       "epoch": 2.269175155562054,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.267,
+      "loss": 0.9933,
       "step": 139670
     },
     {
       "epoch": 2.2693376224594237,
-      "grad_norm": 3.875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2492,
+      "loss": 0.9665,
       "step": 139680
     },
     {
       "epoch": 2.2695000893567934,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2657,
+      "loss": 0.9671,
       "step": 139690
     },
     {
       "epoch": 2.269662556254163,
-      "grad_norm": 2.921875,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2512,
+      "loss": 0.9693,
       "step": 139700
     },
     {
       "epoch": 2.269825023151533,
-      "grad_norm": 3.96875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2532,
+      "loss": 0.9611,
       "step": 139710
     },
     {
       "epoch": 2.2699874900489023,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2476,
+      "loss": 0.9654,
       "step": 139720
     },
     {
       "epoch": 2.2701499569462724,
-      "grad_norm": 4.4375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2397,
+      "loss": 0.9747,
       "step": 139730
     },
     {
       "epoch": 2.2703124238436416,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2432,
+      "loss": 0.9858,
       "step": 139740
     },
     {
       "epoch": 2.2704748907410117,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2219,
+      "loss": 1.0076,
       "step": 139750
     },
     {
       "epoch": 2.2706373576383814,
-      "grad_norm": 2.953125,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2394,
+      "loss": 0.968,
       "step": 139760
     },
     {
       "epoch": 2.270799824535751,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.249,
+      "loss": 0.9812,
       "step": 139770
     },
     {
       "epoch": 2.2709622914331207,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9828,
       "step": 139780
     },
     {
       "epoch": 2.2711247583304903,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2841,
+      "loss": 0.989,
       "step": 139790
     },
     {
       "epoch": 2.27128722522786,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2595,
+      "loss": 0.9918,
       "step": 139800
     },
     {
       "epoch": 2.2714496921252296,
-      "grad_norm": 4.875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2365,
+      "loss": 1.0054,
       "step": 139810
     },
     {
       "epoch": 2.2716121590225993,
-      "grad_norm": 5.25,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.241,
+      "loss": 0.989,
       "step": 139820
     },
     {
       "epoch": 2.271774625919969,
-      "grad_norm": 3.625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2512,
+      "loss": 1.002,
       "step": 139830
     },
     {
       "epoch": 2.2719370928173386,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2434,
+      "loss": 1.0103,
       "step": 139840
     },
     {
       "epoch": 2.2720995597147082,
-      "grad_norm": 4.0,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2414,
+      "loss": 0.9613,
       "step": 139850
     },
     {
       "epoch": 2.272262026612078,
-      "grad_norm": 5.15625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.267,
+      "loss": 0.9818,
       "step": 139860
     },
     {
       "epoch": 2.2724244935094475,
-      "grad_norm": 5.0,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2437,
+      "loss": 0.9795,
       "step": 139870
     },
     {
       "epoch": 2.272586960406817,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2503,
+      "loss": 0.9771,
       "step": 139880
     },
     {
       "epoch": 2.272749427304187,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2424,
+      "loss": 1.0372,
       "step": 139890
     },
     {
       "epoch": 2.2729118942015565,
-      "grad_norm": 3.75,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2518,
+      "loss": 1.0002,
       "step": 139900
     },
     {
       "epoch": 2.273074361098926,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 1.0153,
       "step": 139910
     },
     {
       "epoch": 2.273236827996296,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2564,
+      "loss": 0.964,
       "step": 139920
     },
     {
       "epoch": 2.2733992948936654,
-      "grad_norm": 3.390625,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 0.9982,
       "step": 139930
     },
     {
       "epoch": 2.273561761791035,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.262,
+      "loss": 0.9521,
       "step": 139940
     },
     {
       "epoch": 2.2737242286884047,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2705,
+      "loss": 1.031,
       "step": 139950
     },
     {
       "epoch": 2.2738866955857744,
-      "grad_norm": 3.390625,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2662,
+      "loss": 0.9895,
       "step": 139960
     },
     {
       "epoch": 2.274049162483144,
-      "grad_norm": 4.71875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2674,
+      "loss": 0.9663,
       "step": 139970
     },
     {
       "epoch": 2.2742116293805137,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2813,
+      "loss": 1.0117,
       "step": 139980
     },
     {
       "epoch": 2.2743740962778833,
-      "grad_norm": 5.1875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9729,
       "step": 139990
     },
     {
       "epoch": 2.274536563175253,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2837,
+      "loss": 0.9534,
       "step": 140000
     },
     {
       "epoch": 2.2746990300726226,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2723,
+      "loss": 1.0134,
       "step": 140010
     },
     {
       "epoch": 2.2748614969699923,
-      "grad_norm": 3.8125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.996,
       "step": 140020
     },
     {
       "epoch": 2.275023963867362,
-      "grad_norm": 3.625,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2702,
+      "loss": 1.0174,
       "step": 140030
     },
     {
       "epoch": 2.2751864307647316,
-      "grad_norm": 3.734375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2603,
+      "loss": 0.966,
       "step": 140040
     },
     {
       "epoch": 2.2753488976621012,
-      "grad_norm": 3.703125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9624,
       "step": 140050
     },
     {
       "epoch": 2.275511364559471,
-      "grad_norm": 6.34375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2938,
+      "loss": 1.0039,
       "step": 140060
     },
     {
       "epoch": 2.2756738314568405,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9837,
       "step": 140070
     },
     {
       "epoch": 2.27583629835421,
-      "grad_norm": 3.609375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2526,
+      "loss": 0.9914,
       "step": 140080
     },
     {
       "epoch": 2.27599876525158,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2611,
+      "loss": 1.0161,
       "step": 140090
     },
     {
       "epoch": 2.2761612321489495,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2692,
+      "loss": 0.9833,
       "step": 140100
     },
     {
       "epoch": 2.276323699046319,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 0.946,
       "step": 140110
     },
     {
       "epoch": 2.276486165943689,
-      "grad_norm": 4.125,
+      "grad_norm": 11.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2673,
+      "loss": 1.0027,
       "step": 140120
     },
     {
       "epoch": 2.2766486328410585,
-      "grad_norm": 4.90625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2675,
+      "loss": 0.9881,
       "step": 140130
     },
     {
       "epoch": 2.276811099738428,
-      "grad_norm": 3.5625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.285,
+      "loss": 1.0048,
       "step": 140140
     },
     {
       "epoch": 2.2769735666357978,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2685,
+      "loss": 0.9739,
       "step": 140150
     },
     {
       "epoch": 2.277136033533168,
-      "grad_norm": 4.625,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2752,
+      "loss": 0.9897,
       "step": 140160
     },
     {
       "epoch": 2.277298500430537,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2633,
+      "loss": 0.9978,
       "step": 140170
     },
     {
       "epoch": 2.277460967327907,
-      "grad_norm": 5.75,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2709,
+      "loss": 0.9848,
       "step": 140180
     },
     {
       "epoch": 2.2776234342252764,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.281,
+      "loss": 0.9707,
       "step": 140190
     },
     {
       "epoch": 2.2777859011226465,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2853,
+      "loss": 0.9751,
       "step": 140200
     },
     {
       "epoch": 2.277948368020016,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.3119,
+      "loss": 0.9765,
       "step": 140210
     },
     {
       "epoch": 2.2781108349173858,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2951,
+      "loss": 0.9938,
       "step": 140220
     },
     {
       "epoch": 2.2782733018147554,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2879,
+      "loss": 1.0128,
       "step": 140230
     },
     {
       "epoch": 2.278435768712125,
-      "grad_norm": 4.5,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2732,
+      "loss": 0.9877,
       "step": 140240
     },
     {
       "epoch": 2.2785982356094947,
-      "grad_norm": 4.65625,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2902,
+      "loss": 0.9774,
       "step": 140250
     },
     {
       "epoch": 2.2787607025068644,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2746,
+      "loss": 0.9679,
       "step": 140260
     },
     {
       "epoch": 2.278923169404234,
-      "grad_norm": 3.203125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2919,
+      "loss": 0.982,
       "step": 140270
     },
     {
       "epoch": 2.2790856363016037,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2813,
+      "loss": 0.9633,
       "step": 140280
     },
     {
       "epoch": 2.2792481031989733,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2669,
+      "loss": 0.989,
       "step": 140290
     },
     {
       "epoch": 2.279410570096343,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2587,
+      "loss": 1.0197,
       "step": 140300
     },
     {
       "epoch": 2.2795730369937126,
-      "grad_norm": 4.6875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2531,
+      "loss": 1.0186,
       "step": 140310
     },
     {
       "epoch": 2.2797355038910823,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9945,
       "step": 140320
     },
     {
       "epoch": 2.279897970788452,
-      "grad_norm": 3.953125,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2348,
+      "loss": 0.9875,
       "step": 140330
     },
     {
       "epoch": 2.2800604376858216,
-      "grad_norm": 4.71875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9791,
       "step": 140340
     },
     {
       "epoch": 2.280222904583191,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2395,
+      "loss": 0.9555,
       "step": 140350
     },
     {
       "epoch": 2.280385371480561,
-      "grad_norm": 3.265625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2515,
+      "loss": 0.9456,
       "step": 140360
     },
     {
       "epoch": 2.2805478383779305,
-      "grad_norm": 4.71875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 1.012,
       "step": 140370
     },
     {
       "epoch": 2.2807103052753,
-      "grad_norm": 3.296875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2541,
+      "loss": 0.9601,
       "step": 140380
     },
     {
       "epoch": 2.28087277217267,
-      "grad_norm": 4.75,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2682,
+      "loss": 1.0405,
       "step": 140390
     },
     {
       "epoch": 2.2810352390700395,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2516,
+      "loss": 0.9526,
       "step": 140400
     },
     {
       "epoch": 2.281197705967409,
-      "grad_norm": 5.46875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 1.0043,
       "step": 140410
     },
     {
       "epoch": 2.2813601728647788,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2487,
+      "loss": 1.0045,
       "step": 140420
     },
     {
       "epoch": 2.2815226397621484,
-      "grad_norm": 3.921875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 0.9741,
       "step": 140430
     },
     {
       "epoch": 2.281685106659518,
-      "grad_norm": 3.375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2494,
+      "loss": 0.9814,
       "step": 140440
     },
     {
       "epoch": 2.2818475735568877,
-      "grad_norm": 3.75,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2689,
+      "loss": 1.0026,
       "step": 140450
     },
     {
       "epoch": 2.2820100404542574,
-      "grad_norm": 4.84375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.989,
       "step": 140460
     },
     {
       "epoch": 2.282172507351627,
-      "grad_norm": 2.734375,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2612,
+      "loss": 0.9857,
       "step": 140470
     },
     {
       "epoch": 2.2823349742489967,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2634,
+      "loss": 0.9712,
       "step": 140480
     },
     {
       "epoch": 2.2824974411463663,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2601,
+      "loss": 0.9619,
       "step": 140490
     },
     {
       "epoch": 2.282659908043736,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 1.0164,
       "step": 140500
     },
     {
       "epoch": 2.2828223749411056,
-      "grad_norm": 3.34375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 1.0072,
       "step": 140510
     },
     {
       "epoch": 2.2829848418384753,
-      "grad_norm": 4.96875,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2667,
+      "loss": 0.9575,
       "step": 140520
     },
     {
       "epoch": 2.283147308735845,
-      "grad_norm": 5.6875,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2489,
+      "loss": 1.0081,
       "step": 140530
     },
     {
       "epoch": 2.2833097756332146,
-      "grad_norm": 3.5,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2575,
+      "loss": 1.0042,
       "step": 140540
     },
     {
       "epoch": 2.2834722425305842,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2592,
+      "loss": 0.9642,
       "step": 140550
     },
     {
       "epoch": 2.283634709427954,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2475,
+      "loss": 1.0018,
       "step": 140560
     },
     {
       "epoch": 2.2837971763253235,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2916,
+      "loss": 0.9945,
       "step": 140570
     },
     {
       "epoch": 2.283959643222693,
-      "grad_norm": 3.40625,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2638,
+      "loss": 0.9753,
       "step": 140580
     },
     {
       "epoch": 2.2841221101200633,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2515,
+      "loss": 0.9868,
       "step": 140590
     },
     {
       "epoch": 2.2842845770174325,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2521,
+      "loss": 0.9718,
       "step": 140600
     },
     {
       "epoch": 2.2844470439148026,
-      "grad_norm": 3.625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2702,
+      "loss": 0.9902,
       "step": 140610
     },
     {
       "epoch": 2.284609510812172,
-      "grad_norm": 2.96875,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2544,
+      "loss": 0.9733,
       "step": 140620
     },
     {
       "epoch": 2.284771977709542,
-      "grad_norm": 3.859375,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 0.9774,
       "step": 140630
     },
     {
       "epoch": 2.2849344446069115,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 1.0068,
       "step": 140640
     },
     {
       "epoch": 2.285096911504281,
-      "grad_norm": 3.890625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2812,
+      "loss": 0.9919,
       "step": 140650
     },
     {
       "epoch": 2.285259378401651,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9534,
       "step": 140660
     },
     {
       "epoch": 2.2854218452990205,
-      "grad_norm": 3.34375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2388,
+      "loss": 1.0182,
       "step": 140670
     },
     {
       "epoch": 2.28558431219639,
-      "grad_norm": 4.90625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2645,
+      "loss": 0.9861,
       "step": 140680
     },
     {
       "epoch": 2.28574677909376,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2681,
+      "loss": 1.0023,
       "step": 140690
     },
     {
       "epoch": 2.2859092459911294,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9734,
       "step": 140700
     },
     {
       "epoch": 2.286071712888499,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2588,
+      "loss": 0.9816,
       "step": 140710
     },
     {
       "epoch": 2.2862341797858687,
-      "grad_norm": 4.65625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9974,
       "step": 140720
     },
     {
       "epoch": 2.2863966466832384,
-      "grad_norm": 4.46875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2732,
+      "loss": 0.993,
       "step": 140730
     },
     {
       "epoch": 2.286559113580608,
-      "grad_norm": 3.421875,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2814,
+      "loss": 0.994,
       "step": 140740
     },
     {
       "epoch": 2.2867215804779777,
-      "grad_norm": 3.296875,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 1.0097,
       "step": 140750
     },
     {
       "epoch": 2.2868840473753473,
-      "grad_norm": 4.53125,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9971,
       "step": 140760
     },
     {
       "epoch": 2.287046514272717,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 0.9888,
       "step": 140770
     },
     {
       "epoch": 2.2872089811700866,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.9978,
       "step": 140780
     },
     {
       "epoch": 2.2873714480674563,
-      "grad_norm": 5.0625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2616,
+      "loss": 0.9875,
       "step": 140790
     },
     {
       "epoch": 2.287533914964826,
-      "grad_norm": 4.25,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 0.999,
       "step": 140800
     },
     {
       "epoch": 2.2876963818621956,
-      "grad_norm": 4.34375,
+      "grad_norm": 11.5,
       "learning_rate": 5e-05,
-      "loss": 0.2221,
+      "loss": 0.9956,
       "step": 140810
     },
     {
       "epoch": 2.2878588487595652,
-      "grad_norm": 5.5625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2319,
+      "loss": 1.0161,
       "step": 140820
     },
     {
       "epoch": 2.288021315656935,
-      "grad_norm": 5.84375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2724,
+      "loss": 0.9471,
       "step": 140830
     },
     {
       "epoch": 2.2881837825543045,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2457,
+      "loss": 1.0021,
       "step": 140840
     },
     {
       "epoch": 2.288346249451674,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.2578,
+      "loss": 0.9937,
       "step": 140850
     },
     {
       "epoch": 2.288508716349044,
-      "grad_norm": 3.625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2674,
+      "loss": 0.9642,
       "step": 140860
     },
     {
       "epoch": 2.2886711832464135,
-      "grad_norm": 3.9375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2873,
+      "loss": 0.979,
       "step": 140870
     },
     {
       "epoch": 2.288833650143783,
-      "grad_norm": 3.640625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2508,
+      "loss": 0.9991,
       "step": 140880
     },
     {
       "epoch": 2.288996117041153,
-      "grad_norm": 5.375,
+      "grad_norm": 11.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2672,
+      "loss": 0.9914,
       "step": 140890
     },
     {
       "epoch": 2.2891585839385225,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2591,
+      "loss": 1.0069,
       "step": 140900
     },
     {
       "epoch": 2.289321050835892,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2658,
+      "loss": 0.9982,
       "step": 140910
     },
     {
       "epoch": 2.2894835177332618,
-      "grad_norm": 3.75,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2555,
+      "loss": 0.9851,
       "step": 140920
     },
     {
       "epoch": 2.2896459846306314,
-      "grad_norm": 5.4375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.289,
+      "loss": 0.9936,
       "step": 140930
     },
     {
       "epoch": 2.289808451528001,
-      "grad_norm": 4.5,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2563,
+      "loss": 0.9584,
       "step": 140940
     },
     {
       "epoch": 2.2899709184253707,
-      "grad_norm": 4.71875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2781,
+      "loss": 0.9691,
       "step": 140950
     },
     {
       "epoch": 2.2901333853227404,
-      "grad_norm": 3.1875,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2604,
+      "loss": 0.9951,
       "step": 140960
     },
     {
       "epoch": 2.29029585222011,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2554,
+      "loss": 1.0058,
       "step": 140970
     },
     {
       "epoch": 2.2904583191174797,
-      "grad_norm": 4.96875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2636,
+      "loss": 1.0009,
       "step": 140980
     },
     {
       "epoch": 2.2906207860148493,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2703,
+      "loss": 0.978,
       "step": 140990
     },
     {
       "epoch": 2.290783252912219,
-      "grad_norm": 3.90625,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2688,
+      "loss": 0.992,
       "step": 141000
     },
     {
       "epoch": 2.2909457198095886,
-      "grad_norm": 3.5625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2719,
+      "loss": 1.0028,
       "step": 141010
     },
     {
       "epoch": 2.2911081867069587,
-      "grad_norm": 3.03125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 1.009,
       "step": 141020
     },
     {
       "epoch": 2.291270653604328,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2667,
+      "loss": 0.9809,
       "step": 141030
     },
     {
       "epoch": 2.291433120501698,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2744,
+      "loss": 0.9933,
       "step": 141040
     },
     {
       "epoch": 2.291595587399067,
-      "grad_norm": 3.375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2782,
+      "loss": 1.0453,
       "step": 141050
     },
     {
       "epoch": 2.2917580542964373,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 0.9976,
       "step": 141060
     },
     {
       "epoch": 2.2919205211938065,
-      "grad_norm": 3.953125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2467,
+      "loss": 0.9707,
       "step": 141070
     },
     {
       "epoch": 2.2920829880911766,
-      "grad_norm": 3.625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.9693,
       "step": 141080
     },
     {
       "epoch": 2.2922454549885463,
-      "grad_norm": 3.40625,
+      "grad_norm": 13.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2648,
+      "loss": 0.9489,
       "step": 141090
     },
     {
       "epoch": 2.292407921885916,
-      "grad_norm": 3.9375,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.275,
+      "loss": 0.9623,
       "step": 141100
     },
     {
       "epoch": 2.2925703887832856,
-      "grad_norm": 5.15625,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2789,
+      "loss": 0.9536,
       "step": 141110
     },
     {
       "epoch": 2.292732855680655,
-      "grad_norm": 4.78125,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2871,
+      "loss": 0.9897,
       "step": 141120
     },
     {
       "epoch": 2.292895322578025,
-      "grad_norm": 3.28125,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2723,
+      "loss": 0.9792,
       "step": 141130
     },
     {
       "epoch": 2.2930577894753945,
-      "grad_norm": 3.46875,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2696,
+      "loss": 0.9732,
       "step": 141140
     },
     {
       "epoch": 2.293220256372764,
-      "grad_norm": 3.6875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2678,
+      "loss": 1.0259,
       "step": 141150
     },
     {
       "epoch": 2.293382723270134,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2554,
+      "loss": 1.005,
       "step": 141160
     },
     {
       "epoch": 2.2935451901675035,
-      "grad_norm": 3.203125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.271,
+      "loss": 0.9931,
       "step": 141170
     },
     {
       "epoch": 2.293707657064873,
-      "grad_norm": 3.78125,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2498,
+      "loss": 0.9878,
       "step": 141180
     },
     {
       "epoch": 2.2938701239622428,
-      "grad_norm": 3.765625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.277,
+      "loss": 0.953,
       "step": 141190
     },
     {
       "epoch": 2.2940325908596124,
-      "grad_norm": 4.0,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2555,
+      "loss": 1.0121,
       "step": 141200
     },
     {
       "epoch": 2.294195057756982,
-      "grad_norm": 3.875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9857,
       "step": 141210
     },
     {
       "epoch": 2.2943575246543517,
-      "grad_norm": 3.125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2427,
+      "loss": 0.9567,
       "step": 141220
     },
     {
       "epoch": 2.2945199915517214,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.2442,
+      "loss": 0.9596,
       "step": 141230
     },
     {
       "epoch": 2.294682458449091,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2573,
+      "loss": 0.9902,
       "step": 141240
     },
     {
       "epoch": 2.2948449253464607,
-      "grad_norm": 3.359375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2391,
+      "loss": 0.9824,
       "step": 141250
     },
     {
       "epoch": 2.2950073922438303,
-      "grad_norm": 4.0625,
+      "grad_norm": 7.65625,
       "learning_rate": 5e-05,
-      "loss": 0.2507,
+      "loss": 0.9768,
       "step": 141260
     },
     {
       "epoch": 2.2951698591412,
-      "grad_norm": 4.65625,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2379,
+      "loss": 0.9752,
       "step": 141270
     },
     {
       "epoch": 2.2953323260385696,
-      "grad_norm": 4.28125,
+      "grad_norm": 11.0,
       "learning_rate": 5e-05,
-      "loss": 0.2683,
+      "loss": 1.0227,
       "step": 141280
     },
     {
       "epoch": 2.2954947929359393,
-      "grad_norm": 3.8125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2939,
+      "loss": 0.9638,
       "step": 141290
     },
     {
       "epoch": 2.295657259833309,
-      "grad_norm": 3.734375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2645,
+      "loss": 1.0012,
       "step": 141300
     },
     {
       "epoch": 2.2958197267306786,
-      "grad_norm": 5.28125,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.282,
+      "loss": 0.9902,
       "step": 141310
     },
     {
       "epoch": 2.2959821936280482,
-      "grad_norm": 5.0625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2657,
+      "loss": 0.9679,
       "step": 141320
     },
     {
       "epoch": 2.296144660525418,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2729,
+      "loss": 0.9965,
       "step": 141330
     },
     {
       "epoch": 2.2963071274227875,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9776,
       "step": 141340
     },
     {
       "epoch": 2.296469594320157,
-      "grad_norm": 5.0625,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2673,
+      "loss": 1.0024,
       "step": 141350
     },
     {
       "epoch": 2.296632061217527,
-      "grad_norm": 4.5625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2517,
+      "loss": 0.9688,
       "step": 141360
     },
     {
       "epoch": 2.2967945281148965,
-      "grad_norm": 3.875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2585,
+      "loss": 0.9489,
       "step": 141370
     },
     {
       "epoch": 2.296956995012266,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2462,
+      "loss": 0.9974,
       "step": 141380
     },
     {
       "epoch": 2.297119461909636,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2625,
+      "loss": 0.9966,
       "step": 141390
     },
     {
       "epoch": 2.2972819288070054,
-      "grad_norm": 4.21875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2617,
+      "loss": 0.9993,
       "step": 141400
     },
     {
       "epoch": 2.297444395704375,
-      "grad_norm": 4.21875,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2599,
+      "loss": 0.975,
       "step": 141410
     },
     {
       "epoch": 2.2976068626017447,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2443,
+      "loss": 0.9632,
       "step": 141420
     },
     {
       "epoch": 2.2977693294991144,
-      "grad_norm": 5.40625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2576,
+      "loss": 0.943,
       "step": 141430
     },
     {
       "epoch": 2.297931796396484,
-      "grad_norm": 4.6875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2693,
+      "loss": 0.9748,
       "step": 141440
     },
     {
       "epoch": 2.2980942632938537,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2655,
+      "loss": 0.9827,
       "step": 141450
     },
     {
       "epoch": 2.2982567301912233,
-      "grad_norm": 4.375,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2549,
+      "loss": 0.9813,
       "step": 141460
     },
     {
       "epoch": 2.2984191970885934,
-      "grad_norm": 4.65625,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2735,
+      "loss": 0.9899,
       "step": 141470
     },
     {
       "epoch": 2.2985816639859626,
-      "grad_norm": 3.484375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2723,
+      "loss": 0.955,
       "step": 141480
     },
     {
       "epoch": 2.2987441308833327,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2802,
+      "loss": 0.9664,
       "step": 141490
     },
     {
       "epoch": 2.298906597780702,
-      "grad_norm": 3.71875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 1.0124,
       "step": 141500
     },
     {
       "epoch": 2.299069064678072,
-      "grad_norm": 4.5625,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.256,
+      "loss": 0.9882,
       "step": 141510
     },
     {
       "epoch": 2.2992315315754417,
-      "grad_norm": 4.8125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2794,
+      "loss": 0.9603,
       "step": 141520
     },
     {
       "epoch": 2.2993939984728113,
-      "grad_norm": 3.8125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.253,
+      "loss": 0.9791,
       "step": 141530
     },
     {
       "epoch": 2.299556465370181,
-      "grad_norm": 3.796875,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.9869,
       "step": 141540
     },
     {
       "epoch": 2.2997189322675506,
-      "grad_norm": 3.546875,
+      "grad_norm": 11.625,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9868,
       "step": 141550
     },
     {
       "epoch": 2.2998813991649203,
-      "grad_norm": 3.1875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2683,
+      "loss": 0.9663,
       "step": 141560
     },
     {
       "epoch": 2.30004386606229,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 1.0082,
       "step": 141570
     },
     {
       "epoch": 2.3002063329596596,
-      "grad_norm": 3.953125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.281,
+      "loss": 0.9931,
       "step": 141580
     },
     {
       "epoch": 2.3003687998570292,
-      "grad_norm": 3.703125,
+      "grad_norm": 8.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2701,
+      "loss": 1.0075,
       "step": 141590
     },
     {
       "epoch": 2.300531266754399,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2517,
+      "loss": 1.0015,
       "step": 141600
     },
     {
       "epoch": 2.3006937336517685,
-      "grad_norm": 5.0625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.255,
+      "loss": 0.9803,
       "step": 141610
     },
     {
       "epoch": 2.300856200549138,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2696,
+      "loss": 0.9291,
       "step": 141620
     },
     {
       "epoch": 2.301018667446508,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.972,
       "step": 141630
     },
     {
       "epoch": 2.3011811343438775,
-      "grad_norm": 5.03125,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2967,
+      "loss": 0.9698,
       "step": 141640
     },
     {
       "epoch": 2.301343601241247,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2684,
+      "loss": 0.9934,
       "step": 141650
     },
     {
       "epoch": 2.301506068138617,
-      "grad_norm": 4.96875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.247,
+      "loss": 1.008,
       "step": 141660
     },
     {
       "epoch": 2.3016685350359865,
-      "grad_norm": 4.0,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2696,
+      "loss": 0.9841,
       "step": 141670
     },
     {
       "epoch": 2.301831001933356,
-      "grad_norm": 4.375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2595,
+      "loss": 1.0373,
       "step": 141680
     },
     {
       "epoch": 2.3019934688307258,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.237,
+      "loss": 0.9969,
       "step": 141690
     },
     {
       "epoch": 2.3021559357280954,
-      "grad_norm": 3.109375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2566,
+      "loss": 0.9664,
       "step": 141700
     },
     {
       "epoch": 2.302318402625465,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2499,
+      "loss": 0.9494,
       "step": 141710
     },
     {
       "epoch": 2.3024808695228347,
-      "grad_norm": 4.5,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2338,
+      "loss": 0.9575,
       "step": 141720
     },
     {
       "epoch": 2.3026433364202044,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.254,
+      "loss": 0.9879,
       "step": 141730
     },
     {
       "epoch": 2.302805803317574,
-      "grad_norm": 5.21875,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2634,
+      "loss": 0.991,
       "step": 141740
     },
     {
       "epoch": 2.3029682702149437,
-      "grad_norm": 4.1875,
+      "grad_norm": 8.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2745,
+      "loss": 0.9891,
       "step": 141750
     },
     {
       "epoch": 2.3031307371123133,
-      "grad_norm": 2.765625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 1.0124,
       "step": 141760
     },
     {
       "epoch": 2.303293204009683,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2594,
+      "loss": 0.9705,
       "step": 141770
     },
     {
       "epoch": 2.3034556709070526,
-      "grad_norm": 3.921875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2751,
+      "loss": 0.9854,
       "step": 141780
     },
     {
       "epoch": 2.3036181378044223,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2679,
+      "loss": 0.9816,
       "step": 141790
     },
     {
       "epoch": 2.303780604701792,
-      "grad_norm": 4.125,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2986,
+      "loss": 0.9797,
       "step": 141800
     },
     {
       "epoch": 2.3039430715991616,
-      "grad_norm": 4.03125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2814,
+      "loss": 0.9692,
       "step": 141810
     },
     {
       "epoch": 2.304105538496531,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2571,
+      "loss": 0.9966,
       "step": 141820
     },
     {
       "epoch": 2.304268005393901,
-      "grad_norm": 3.8125,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2548,
+      "loss": 0.9988,
       "step": 141830
     },
     {
       "epoch": 2.3044304722912705,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2747,
+      "loss": 0.9952,
       "step": 141840
     },
     {
       "epoch": 2.30459293918864,
-      "grad_norm": 4.375,
+      "grad_norm": 11.75,
       "learning_rate": 5e-05,
-      "loss": 0.27,
+      "loss": 0.9814,
       "step": 141850
     },
     {
       "epoch": 2.30475540608601,
-      "grad_norm": 3.96875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.3033,
+      "loss": 1.0064,
       "step": 141860
     },
     {
       "epoch": 2.3049178729833795,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2862,
+      "loss": 0.9669,
       "step": 141870
     },
     {
       "epoch": 2.305080339880749,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.75,
       "learning_rate": 5e-05,
-      "loss": 0.2976,
+      "loss": 0.9744,
       "step": 141880
     },
     {
       "epoch": 2.3052428067781188,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.875,
       "learning_rate": 5e-05,
-      "loss": 0.3063,
+      "loss": 0.9987,
       "step": 141890
     },
     {
       "epoch": 2.305405273675489,
-      "grad_norm": 3.71875,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 1.0178,
       "step": 141900
     },
     {
       "epoch": 2.305567740572858,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.3038,
+      "loss": 0.9609,
       "step": 141910
     },
     {
       "epoch": 2.305730207470228,
-      "grad_norm": 4.4375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2722,
+      "loss": 0.9993,
       "step": 141920
     },
     {
       "epoch": 2.3058926743675974,
-      "grad_norm": 3.25,
+      "grad_norm": 11.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2727,
+      "loss": 0.9703,
       "step": 141930
     },
     {
       "epoch": 2.3060551412649675,
-      "grad_norm": 3.890625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2535,
+      "loss": 1.0188,
       "step": 141940
     },
     {
       "epoch": 2.3062176081623367,
-      "grad_norm": 3.828125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.231,
+      "loss": 1.0007,
       "step": 141950
     },
     {
       "epoch": 2.3063800750597068,
-      "grad_norm": 3.0625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2231,
+      "loss": 1.0001,
       "step": 141960
     },
     {
       "epoch": 2.3065425419570764,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2546,
+      "loss": 0.9918,
       "step": 141970
     },
     {
       "epoch": 2.306705008854446,
-      "grad_norm": 4.25,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2452,
+      "loss": 0.9627,
       "step": 141980
     },
     {
       "epoch": 2.3068674757518157,
-      "grad_norm": 3.546875,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 1.0145,
       "step": 141990
     },
     {
       "epoch": 2.3070299426491854,
-      "grad_norm": 2.96875,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2527,
+      "loss": 0.9713,
       "step": 142000
     },
     {
       "epoch": 2.307192409546555,
-      "grad_norm": 3.484375,
+      "grad_norm": 12.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2425,
+      "loss": 1.0018,
       "step": 142010
     },
     {
       "epoch": 2.3073548764439247,
-      "grad_norm": 4.53125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2791,
+      "loss": 0.9819,
       "step": 142020
     },
     {
       "epoch": 2.3075173433412943,
-      "grad_norm": 3.390625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2556,
+      "loss": 0.9723,
       "step": 142030
     },
     {
       "epoch": 2.307679810238664,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2569,
+      "loss": 0.9789,
       "step": 142040
     },
     {
       "epoch": 2.3078422771360336,
-      "grad_norm": 4.75,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2886,
+      "loss": 0.9912,
       "step": 142050
     },
     {
       "epoch": 2.3080047440334033,
-      "grad_norm": 3.234375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2739,
+      "loss": 0.9616,
       "step": 142060
     },
     {
       "epoch": 2.308167210930773,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2797,
+      "loss": 0.9987,
       "step": 142070
     },
     {
       "epoch": 2.3083296778281426,
-      "grad_norm": 3.859375,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2881,
+      "loss": 0.9756,
       "step": 142080
     },
     {
       "epoch": 2.3084921447255122,
-      "grad_norm": 2.703125,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2538,
+      "loss": 0.9565,
       "step": 142090
     },
     {
       "epoch": 2.308654611622882,
-      "grad_norm": 3.4375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2419,
+      "loss": 0.9518,
       "step": 142100
     },
     {
       "epoch": 2.3088170785202515,
-      "grad_norm": 4.53125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2543,
+      "loss": 1.0086,
       "step": 142110
     },
     {
       "epoch": 2.308979545417621,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2742,
+      "loss": 1.0129,
       "step": 142120
     },
     {
       "epoch": 2.309142012314991,
-      "grad_norm": 3.671875,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2547,
+      "loss": 0.9737,
       "step": 142130
     },
     {
       "epoch": 2.3093044792123605,
-      "grad_norm": 5.25,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2474,
+      "loss": 0.9872,
       "step": 142140
     },
     {
       "epoch": 2.30946694610973,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2448,
+      "loss": 1.0469,
       "step": 142150
     },
     {
       "epoch": 2.3096294130071,
-      "grad_norm": 4.8125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2509,
+      "loss": 0.9696,
       "step": 142160
     },
     {
       "epoch": 2.3097918799044694,
-      "grad_norm": 4.25,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2638,
+      "loss": 0.9665,
       "step": 142170
     },
     {
       "epoch": 2.309954346801839,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2949,
+      "loss": 0.9594,
       "step": 142180
     },
     {
       "epoch": 2.3101168136992087,
-      "grad_norm": 4.5625,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.279,
+      "loss": 0.9984,
       "step": 142190
     },
     {
       "epoch": 2.3102792805965784,
-      "grad_norm": 3.84375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2794,
+      "loss": 0.9811,
       "step": 142200
     },
     {
       "epoch": 2.310441747493948,
-      "grad_norm": 3.3125,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2722,
+      "loss": 1.0161,
       "step": 142210
     },
     {
       "epoch": 2.3106042143913177,
-      "grad_norm": 3.5625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2626,
+      "loss": 1.0253,
       "step": 142220
     },
     {
       "epoch": 2.3107666812886873,
-      "grad_norm": 4.65625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2656,
+      "loss": 0.9925,
       "step": 142230
     },
     {
       "epoch": 2.310929148186057,
-      "grad_norm": 5.5,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2472,
+      "loss": 0.9948,
       "step": 142240
     },
     {
       "epoch": 2.3110916150834266,
-      "grad_norm": 4.5,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2857,
+      "loss": 1.0219,
       "step": 142250
     },
     {
       "epoch": 2.3112540819807963,
-      "grad_norm": 5.25,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2746,
+      "loss": 1.0576,
       "step": 142260
     },
     {
       "epoch": 2.311416548878166,
-      "grad_norm": 3.125,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2777,
+      "loss": 0.957,
       "step": 142270
     },
     {
       "epoch": 2.3115790157755356,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.264,
+      "loss": 0.9875,
       "step": 142280
     },
     {
       "epoch": 2.3117414826729052,
-      "grad_norm": 5.125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2826,
+      "loss": 0.9606,
       "step": 142290
     },
     {
       "epoch": 2.311903949570275,
-      "grad_norm": 4.3125,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2776,
+      "loss": 0.9955,
       "step": 142300
     },
     {
       "epoch": 2.3120664164676445,
-      "grad_norm": 3.921875,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.3054,
+      "loss": 0.9796,
       "step": 142310
     },
     {
       "epoch": 2.312228883365014,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2851,
+      "loss": 0.9955,
       "step": 142320
     },
     {
       "epoch": 2.312391350262384,
-      "grad_norm": 4.28125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2794,
+      "loss": 1.0171,
       "step": 142330
     },
     {
       "epoch": 2.3125538171597535,
-      "grad_norm": 3.453125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2932,
+      "loss": 0.9982,
       "step": 142340
     },
     {
       "epoch": 2.3127162840571236,
-      "grad_norm": 3.578125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2703,
+      "loss": 0.9876,
       "step": 142350
     },
     {
       "epoch": 2.312878750954493,
-      "grad_norm": 4.53125,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2618,
+      "loss": 0.988,
       "step": 142360
     },
     {
       "epoch": 2.313041217851863,
-      "grad_norm": 3.6875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9977,
       "step": 142370
     },
     {
       "epoch": 2.313203684749232,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2491,
+      "loss": 1.0183,
       "step": 142380
     },
     {
       "epoch": 2.313366151646602,
-      "grad_norm": 5.5,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2652,
+      "loss": 1.0087,
       "step": 142390
     },
     {
       "epoch": 2.313528618543972,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2754,
+      "loss": 0.9824,
       "step": 142400
     },
     {
       "epoch": 2.3136910854413415,
-      "grad_norm": 3.171875,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2697,
+      "loss": 0.9546,
       "step": 142410
     },
     {
       "epoch": 2.313853552338711,
-      "grad_norm": 4.125,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2698,
+      "loss": 1.058,
       "step": 142420
     },
     {
       "epoch": 2.314016019236081,
-      "grad_norm": 5.0,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.271,
+      "loss": 0.9971,
       "step": 142430
     },
     {
       "epoch": 2.3141784861334505,
-      "grad_norm": 3.9375,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2602,
+      "loss": 0.9714,
       "step": 142440
     },
     {
       "epoch": 2.31434095303082,
-      "grad_norm": 5.875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2611,
+      "loss": 0.9911,
       "step": 142450
     },
     {
       "epoch": 2.3145034199281898,
-      "grad_norm": 4.5,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2545,
+      "loss": 0.9722,
       "step": 142460
     },
     {
       "epoch": 2.3146658868255594,
-      "grad_norm": 3.5625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.257,
+      "loss": 0.9871,
       "step": 142470
     },
     {
       "epoch": 2.314828353722929,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2508,
+      "loss": 0.9868,
       "step": 142480
     },
     {
       "epoch": 2.3149908206202987,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2749,
+      "loss": 0.9623,
       "step": 142490
     },
     {
       "epoch": 2.3151532875176684,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2678,
+      "loss": 1.0085,
       "step": 142500
     },
     {
       "epoch": 2.315315754415038,
-      "grad_norm": 2.796875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2764,
+      "loss": 0.9772,
       "step": 142510
     },
     {
       "epoch": 2.3154782213124077,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2829,
+      "loss": 0.9974,
       "step": 142520
     },
     {
       "epoch": 2.3156406882097773,
-      "grad_norm": 3.734375,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2703,
+      "loss": 0.9888,
       "step": 142530
     },
     {
       "epoch": 2.315803155107147,
-      "grad_norm": 5.09375,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2807,
+      "loss": 0.9536,
       "step": 142540
     },
     {
       "epoch": 2.3159656220045166,
-      "grad_norm": 3.96875,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 0.9937,
       "step": 142550
     },
     {
       "epoch": 2.3161280889018863,
-      "grad_norm": 2.734375,
+      "grad_norm": 12.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2707,
+      "loss": 1.0208,
       "step": 142560
     },
     {
       "epoch": 2.316290555799256,
-      "grad_norm": 3.515625,
+      "grad_norm": 8.625,
       "learning_rate": 5e-05,
-      "loss": 0.2706,
+      "loss": 0.9648,
       "step": 142570
     },
     {
       "epoch": 2.3164530226966256,
-      "grad_norm": 3.6875,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.281,
+      "loss": 0.9983,
       "step": 142580
     },
     {
       "epoch": 2.316615489593995,
-      "grad_norm": 4.65625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2571,
+      "loss": 0.9711,
       "step": 142590
     },
     {
       "epoch": 2.316777956491365,
-      "grad_norm": 4.65625,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.2808,
+      "loss": 1.0166,
       "step": 142600
     },
     {
       "epoch": 2.3169404233887345,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2779,
+      "loss": 0.9755,
       "step": 142610
     },
     {
       "epoch": 2.317102890286104,
-      "grad_norm": 5.0625,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2983,
+      "loss": 0.9788,
       "step": 142620
     },
     {
       "epoch": 2.317265357183474,
-      "grad_norm": 4.28125,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2815,
+      "loss": 0.9706,
       "step": 142630
     },
     {
       "epoch": 2.3174278240808435,
-      "grad_norm": 3.59375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2653,
+      "loss": 0.9955,
       "step": 142640
     },
     {
       "epoch": 2.317590290978213,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2662,
+      "loss": 0.9698,
       "step": 142650
     },
     {
       "epoch": 2.3177527578755828,
-      "grad_norm": 5.03125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.246,
+      "loss": 1.0209,
       "step": 142660
     },
     {
       "epoch": 2.3179152247729524,
-      "grad_norm": 3.875,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2798,
+      "loss": 1.0103,
       "step": 142670
     },
     {
       "epoch": 2.318077691670322,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2752,
+      "loss": 0.9764,
       "step": 142680
     },
     {
       "epoch": 2.3182401585676917,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2605,
+      "loss": 0.9711,
       "step": 142690
     },
     {
       "epoch": 2.3184026254650614,
-      "grad_norm": 4.84375,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9979,
       "step": 142700
     },
     {
       "epoch": 2.318565092362431,
-      "grad_norm": 4.0,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2574,
+      "loss": 1.0068,
       "step": 142710
     },
     {
       "epoch": 2.3187275592598007,
-      "grad_norm": 3.984375,
+      "grad_norm": 11.375,
       "learning_rate": 5e-05,
-      "loss": 0.2856,
+      "loss": 0.992,
       "step": 142720
     },
     {
       "epoch": 2.3188900261571703,
-      "grad_norm": 3.328125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2791,
+      "loss": 0.9952,
       "step": 142730
     },
     {
       "epoch": 2.31905249305454,
-      "grad_norm": 4.0,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2804,
+      "loss": 0.9906,
       "step": 142740
     },
     {
       "epoch": 2.3192149599519096,
-      "grad_norm": 4.625,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2713,
+      "loss": 1.0103,
       "step": 142750
     },
     {
       "epoch": 2.3193774268492793,
-      "grad_norm": 4.25,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2792,
+      "loss": 0.9874,
       "step": 142760
     },
     {
       "epoch": 2.319539893746649,
-      "grad_norm": 4.03125,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2898,
+      "loss": 0.9841,
       "step": 142770
     },
     {
       "epoch": 2.319702360644019,
-      "grad_norm": 3.453125,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2728,
+      "loss": 0.9618,
       "step": 142780
     },
     {
       "epoch": 2.3198648275413882,
-      "grad_norm": 3.046875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2724,
+      "loss": 0.9905,
       "step": 142790
     },
     {
       "epoch": 2.3200272944387583,
-      "grad_norm": 4.5,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2783,
+      "loss": 0.9863,
       "step": 142800
     },
     {
       "epoch": 2.3201897613361275,
-      "grad_norm": 4.3125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2762,
+      "loss": 0.9802,
       "step": 142810
     },
     {
       "epoch": 2.3203522282334976,
-      "grad_norm": 3.28125,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2787,
+      "loss": 0.9645,
       "step": 142820
     },
     {
       "epoch": 2.3205146951308673,
-      "grad_norm": 3.296875,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2651,
+      "loss": 1.0011,
       "step": 142830
     },
     {
       "epoch": 2.320677162028237,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2726,
+      "loss": 1.0124,
       "step": 142840
     },
     {
       "epoch": 2.3208396289256066,
-      "grad_norm": 3.6875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2804,
+      "loss": 1.0008,
       "step": 142850
     },
     {
       "epoch": 2.3210020958229762,
-      "grad_norm": 4.21875,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2578,
+      "loss": 0.9789,
       "step": 142860
     },
     {
       "epoch": 2.321164562720346,
-      "grad_norm": 5.40625,
+      "grad_norm": 8.125,
       "learning_rate": 5e-05,
-      "loss": 0.2663,
+      "loss": 0.995,
       "step": 142870
     },
     {
       "epoch": 2.3213270296177155,
-      "grad_norm": 4.375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2506,
+      "loss": 0.9838,
       "step": 142880
     },
     {
       "epoch": 2.321489496515085,
-      "grad_norm": 4.5,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2434,
+      "loss": 0.9706,
       "step": 142890
     },
     {
       "epoch": 2.321651963412455,
-      "grad_norm": 4.375,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2631,
+      "loss": 0.9887,
       "step": 142900
     },
     {
       "epoch": 2.3218144303098245,
-      "grad_norm": 3.609375,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 1.0005,
       "step": 142910
     },
     {
       "epoch": 2.321976897207194,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2613,
+      "loss": 0.9777,
       "step": 142920
     },
     {
       "epoch": 2.322139364104564,
-      "grad_norm": 3.984375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2596,
+      "loss": 0.9592,
       "step": 142930
     },
     {
       "epoch": 2.3223018310019334,
-      "grad_norm": 5.3125,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2798,
+      "loss": 1.0101,
       "step": 142940
     },
     {
       "epoch": 2.322464297899303,
-      "grad_norm": 4.40625,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2713,
+      "loss": 0.9855,
       "step": 142950
     },
     {
       "epoch": 2.3226267647966727,
-      "grad_norm": 3.25,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2646,
+      "loss": 0.9603,
       "step": 142960
     },
     {
       "epoch": 2.3227892316940424,
-      "grad_norm": 4.15625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2433,
+      "loss": 1.0213,
       "step": 142970
     },
     {
       "epoch": 2.322951698591412,
-      "grad_norm": 4.25,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2755,
+      "loss": 1.0087,
       "step": 142980
     },
     {
       "epoch": 2.3231141654887817,
-      "grad_norm": 5.5625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2668,
+      "loss": 0.9624,
       "step": 142990
     },
     {
       "epoch": 2.3232766323861513,
-      "grad_norm": 4.59375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2818,
+      "loss": 0.9497,
       "step": 143000
     },
     {
       "epoch": 2.323439099283521,
-      "grad_norm": 4.125,
+      "grad_norm": 8.375,
       "learning_rate": 5e-05,
-      "loss": 0.2608,
+      "loss": 0.9879,
       "step": 143010
     },
     {
       "epoch": 2.3236015661808906,
-      "grad_norm": 3.421875,
+      "grad_norm": 10.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2623,
+      "loss": 1.0193,
       "step": 143020
     },
     {
       "epoch": 2.3237640330782603,
-      "grad_norm": 4.75,
+      "grad_norm": 7.9375,
       "learning_rate": 5e-05,
-      "loss": 0.3046,
+      "loss": 1.031,
       "step": 143030
     },
     {
       "epoch": 2.32392649997563,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.256,
+      "loss": 0.9863,
       "step": 143040
     },
     {
       "epoch": 2.3240889668729996,
-      "grad_norm": 2.984375,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2763,
+      "loss": 1.0069,
       "step": 143050
     },
     {
       "epoch": 2.3242514337703692,
-      "grad_norm": 3.453125,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2741,
+      "loss": 1.0206,
       "step": 143060
     },
     {
       "epoch": 2.324413900667739,
-      "grad_norm": 4.75,
+      "grad_norm": 12.0,
       "learning_rate": 5e-05,
-      "loss": 0.2912,
+      "loss": 0.9762,
       "step": 143070
     },
     {
       "epoch": 2.3245763675651085,
-      "grad_norm": 3.765625,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2649,
+      "loss": 0.993,
       "step": 143080
     },
     {
       "epoch": 2.324738834462478,
-      "grad_norm": 3.640625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2578,
+      "loss": 0.9917,
       "step": 143090
     },
     {
       "epoch": 2.324901301359848,
-      "grad_norm": 4.71875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2821,
+      "loss": 0.99,
       "step": 143100
     },
     {
       "epoch": 2.3250637682572175,
-      "grad_norm": 4.34375,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.268,
+      "loss": 0.9861,
       "step": 143110
     },
     {
       "epoch": 2.325226235154587,
-      "grad_norm": 3.65625,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2588,
+      "loss": 0.9755,
       "step": 143120
     },
     {
       "epoch": 2.325388702051957,
-      "grad_norm": 4.46875,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2474,
+      "loss": 0.9642,
       "step": 143130
     },
     {
       "epoch": 2.3255511689493265,
-      "grad_norm": 4.09375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2485,
+      "loss": 1.0342,
       "step": 143140
     },
     {
       "epoch": 2.325713635846696,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2465,
+      "loss": 1.0156,
       "step": 143150
     },
     {
       "epoch": 2.3258761027440658,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.9755,
       "step": 143160
     },
     {
       "epoch": 2.3260385696414354,
-      "grad_norm": 4.375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.288,
+      "loss": 0.9893,
       "step": 143170
     },
     {
       "epoch": 2.326201036538805,
-      "grad_norm": 3.609375,
+      "grad_norm": 10.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2597,
+      "loss": 0.9645,
       "step": 143180
     },
     {
       "epoch": 2.3263635034361747,
-      "grad_norm": 3.828125,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2392,
+      "loss": 0.9607,
       "step": 143190
     },
     {
       "epoch": 2.3265259703335444,
-      "grad_norm": 3.375,
+      "grad_norm": 8.75,
       "learning_rate": 5e-05,
-      "loss": 0.2427,
+      "loss": 0.952,
       "step": 143200
     },
     {
       "epoch": 2.326688437230914,
-      "grad_norm": 4.53125,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2756,
+      "loss": 1.0021,
       "step": 143210
     },
     {
       "epoch": 2.3268509041282837,
-      "grad_norm": 4.3125,
+      "grad_norm": 8.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2905,
+      "loss": 0.9727,
       "step": 143220
     },
     {
       "epoch": 2.3270133710256538,
-      "grad_norm": 4.375,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2524,
+      "loss": 0.9675,
       "step": 143230
     },
     {
       "epoch": 2.327175837923023,
-      "grad_norm": 5.28125,
+      "grad_norm": 11.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2774,
+      "loss": 1.0121,
       "step": 143240
     },
     {
       "epoch": 2.327338304820393,
-      "grad_norm": 3.15625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2628,
+      "loss": 0.9917,
       "step": 143250
     },
     {
       "epoch": 2.3275007717177623,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.266,
+      "loss": 1.0198,
       "step": 143260
     },
     {
       "epoch": 2.3276632386151324,
-      "grad_norm": 3.515625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9694,
       "step": 143270
     },
     {
       "epoch": 2.327825705512502,
-      "grad_norm": 4.21875,
+      "grad_norm": 9.75,
       "learning_rate": 5e-05,
-      "loss": 0.2568,
+      "loss": 0.9926,
       "step": 143280
     },
     {
       "epoch": 2.3279881724098717,
-      "grad_norm": 5.0,
+      "grad_norm": 8.25,
       "learning_rate": 5e-05,
-      "loss": 0.2621,
+      "loss": 0.9852,
       "step": 143290
     },
     {
       "epoch": 2.3281506393072413,
-      "grad_norm": 4.15625,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2488,
+      "loss": 1.0239,
       "step": 143300
     },
     {
       "epoch": 2.328313106204611,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2182,
+      "loss": 0.9845,
       "step": 143310
     },
     {
       "epoch": 2.3284755731019806,
-      "grad_norm": 5.125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2672,
+      "loss": 0.9889,
       "step": 143320
     },
     {
       "epoch": 2.3286380399993503,
-      "grad_norm": 5.15625,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2523,
+      "loss": 1.0098,
       "step": 143330
     },
     {
       "epoch": 2.32880050689672,
-      "grad_norm": 3.046875,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.248,
+      "loss": 0.9901,
       "step": 143340
     },
     {
       "epoch": 2.3289629737940896,
-      "grad_norm": 5.21875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2587,
+      "loss": 1.0034,
       "step": 143350
     },
     {
       "epoch": 2.329125440691459,
-      "grad_norm": 5.15625,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2411,
+      "loss": 0.9927,
       "step": 143360
     },
     {
       "epoch": 2.329287907588829,
-      "grad_norm": 4.75,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2667,
+      "loss": 0.9846,
       "step": 143370
     },
     {
       "epoch": 2.3294503744861985,
-      "grad_norm": 4.0,
+      "grad_norm": 9.0,
       "learning_rate": 5e-05,
-      "loss": 0.2541,
+      "loss": 1.0202,
       "step": 143380
     },
     {
       "epoch": 2.329612841383568,
-      "grad_norm": 3.09375,
+      "grad_norm": 10.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2789,
+      "loss": 1.0159,
       "step": 143390
     },
     {
       "epoch": 2.329775308280938,
-      "grad_norm": 3.328125,
+      "grad_norm": 12.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2796,
+      "loss": 0.992,
       "step": 143400
     },
     {
       "epoch": 2.3299377751783075,
-      "grad_norm": 3.453125,
+      "grad_norm": 9.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2916,
+      "loss": 1.0077,
       "step": 143410
     },
     {
       "epoch": 2.330100242075677,
-      "grad_norm": 3.734375,
+      "grad_norm": 11.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2781,
+      "loss": 0.9933,
       "step": 143420
     },
     {
       "epoch": 2.3302627089730468,
-      "grad_norm": 4.5,
+      "grad_norm": 9.125,
       "learning_rate": 5e-05,
-      "loss": 0.2762,
+      "loss": 0.9847,
       "step": 143430
     },
     {
       "epoch": 2.3304251758704164,
-      "grad_norm": 3.90625,
+      "grad_norm": 7.875,
       "learning_rate": 5e-05,
-      "loss": 0.2764,
+      "loss": 0.9906,
       "step": 143440
     },
     {
       "epoch": 2.330587642767786,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.254,
+      "loss": 0.9535,
       "step": 143450
     },
     {
       "epoch": 2.3307501096651557,
-      "grad_norm": 3.46875,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2519,
+      "loss": 0.9894,
       "step": 143460
     },
     {
       "epoch": 2.3309125765625254,
-      "grad_norm": 4.375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.254,
+      "loss": 1.0133,
       "step": 143470
     },
     {
       "epoch": 2.331075043459895,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2394,
+      "loss": 0.9964,
       "step": 143480
     },
     {
       "epoch": 2.3312375103572647,
-      "grad_norm": 4.5,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2662,
+      "loss": 1.0298,
       "step": 143490
     },
     {
       "epoch": 2.3313999772546343,
-      "grad_norm": 4.34375,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2781,
+      "loss": 1.0009,
       "step": 143500
     },
     {
       "epoch": 2.331562444152004,
-      "grad_norm": 4.28125,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2579,
+      "loss": 0.9795,
       "step": 143510
     },
     {
       "epoch": 2.3317249110493736,
-      "grad_norm": 4.40625,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2842,
+      "loss": 0.9988,
       "step": 143520
     },
     {
       "epoch": 2.3318873779467433,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.25,
       "learning_rate": 5e-05,
-      "loss": 0.2637,
+      "loss": 0.9831,
       "step": 143530
     },
     {
       "epoch": 2.332049844844113,
-      "grad_norm": 4.0625,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.249,
+      "loss": 0.9721,
       "step": 143540
     },
     {
       "epoch": 2.3322123117414826,
-      "grad_norm": 3.484375,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2607,
+      "loss": 0.9814,
       "step": 143550
     },
     {
       "epoch": 2.3323747786388522,
-      "grad_norm": 4.1875,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2816,
+      "loss": 0.9956,
       "step": 143560
     },
     {
       "epoch": 2.332537245536222,
-      "grad_norm": 3.875,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2898,
+      "loss": 0.9808,
       "step": 143570
     },
     {
       "epoch": 2.3326997124335915,
-      "grad_norm": 3.703125,
+      "grad_norm": 9.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2831,
+      "loss": 0.9962,
       "step": 143580
     },
     {
       "epoch": 2.332862179330961,
-      "grad_norm": 4.75,
+      "grad_norm": 10.375,
       "learning_rate": 5e-05,
-      "loss": 0.2724,
+      "loss": 0.9581,
       "step": 143590
     },
     {
       "epoch": 2.333024646228331,
-      "grad_norm": 3.578125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.9672,
       "step": 143600
     },
     {
       "epoch": 2.3331871131257005,
-      "grad_norm": 4.4375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2756,
+      "loss": 0.9978,
       "step": 143610
     },
     {
       "epoch": 2.33334958002307,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2544,
+      "loss": 0.9883,
       "step": 143620
     },
     {
       "epoch": 2.33351204692044,
-      "grad_norm": 4.78125,
+      "grad_norm": 11.125,
       "learning_rate": 5e-05,
-      "loss": 0.2345,
+      "loss": 1.0177,
       "step": 143630
     },
     {
       "epoch": 2.3336745138178094,
-      "grad_norm": 4.09375,
+      "grad_norm": 8.8125,
       "learning_rate": 5e-05,
-      "loss": 0.2447,
+      "loss": 1.0032,
       "step": 143640
     },
     {
       "epoch": 2.333836980715179,
-      "grad_norm": 4.0,
+      "grad_norm": 8.5,
       "learning_rate": 5e-05,
-      "loss": 0.274,
+      "loss": 0.9625,
       "step": 143650
     },
     {
       "epoch": 2.333999447612549,
-      "grad_norm": 3.5625,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.289,
+      "loss": 0.9875,
       "step": 143660
     },
     {
       "epoch": 2.3341619145099184,
-      "grad_norm": 3.734375,
+      "grad_norm": 9.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2589,
+      "loss": 1.0138,
       "step": 143670
     },
     {
       "epoch": 2.3343243814072885,
-      "grad_norm": 4.625,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2801,
+      "loss": 0.9993,
       "step": 143680
     },
     {
       "epoch": 2.3344868483046577,
-      "grad_norm": 3.203125,
+      "grad_norm": 8.875,
       "learning_rate": 5e-05,
-      "loss": 0.2793,
+      "loss": 1.0037,
       "step": 143690
     },
     {
       "epoch": 2.334649315202028,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2756,
+      "loss": 0.961,
       "step": 143700
     },
     {
       "epoch": 2.3348117820993974,
-      "grad_norm": 3.609375,
+      "grad_norm": 8.5625,
       "learning_rate": 5e-05,
-      "loss": 0.2583,
+      "loss": 1.0087,
       "step": 143710
     },
     {
       "epoch": 2.334974248996767,
-      "grad_norm": 4.5,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2799,
+      "loss": 0.9828,
       "step": 143720
     },
     {
       "epoch": 2.3351367158941367,
-      "grad_norm": 3.875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2796,
+      "loss": 0.978,
       "step": 143730
     },
     {
       "epoch": 2.3352991827915064,
-      "grad_norm": 3.421875,
+      "grad_norm": 10.6875,
       "learning_rate": 5e-05,
-      "loss": 0.2852,
+      "loss": 1.0078,
       "step": 143740
     },
     {
       "epoch": 2.335461649688876,
-      "grad_norm": 3.875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.3069,
+      "loss": 1.0351,
       "step": 143750
     },
     {
       "epoch": 2.3356241165862457,
-      "grad_norm": 3.78125,
+      "grad_norm": 9.5,
       "learning_rate": 5e-05,
-      "loss": 0.2656,
+      "loss": 1.0122,
       "step": 143760
     },
     {
       "epoch": 2.3357865834836153,
-      "grad_norm": 3.625,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2614,
+      "loss": 0.9892,
       "step": 143770
     },
     {
       "epoch": 2.335949050380985,
-      "grad_norm": 4.09375,
+      "grad_norm": 10.625,
       "learning_rate": 5e-05,
-      "loss": 0.2609,
+      "loss": 0.993,
       "step": 143780
     },
     {
       "epoch": 2.3361115172783546,
-      "grad_norm": 4.84375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2769,
+      "loss": 0.9969,
       "step": 143790
     },
     {
       "epoch": 2.3362739841757243,
-      "grad_norm": 4.78125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2904,
+      "loss": 0.9634,
       "step": 143800
     },
     {
       "epoch": 2.336436451073094,
-      "grad_norm": 3.84375,
+      "grad_norm": 9.875,
       "learning_rate": 5e-05,
-      "loss": 0.2812,
+      "loss": 0.9738,
       "step": 143810
     },
     {
       "epoch": 2.3365989179704636,
-      "grad_norm": 4.40625,
+      "grad_norm": 11.875,
       "learning_rate": 5e-05,
-      "loss": 0.2513,
+      "loss": 0.9936,
       "step": 143820
     },
     {
       "epoch": 2.3367613848678332,
-      "grad_norm": 2.875,
+      "grad_norm": 10.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2761,
+      "loss": 0.9781,
       "step": 143830
     },
     {
       "epoch": 2.336923851765203,
-      "grad_norm": 3.796875,
+      "grad_norm": 9.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2662,
+      "loss": 0.965,
       "step": 143840
     },
     {
       "epoch": 2.3370863186625725,
-      "grad_norm": 3.0625,
+      "grad_norm": 10.4375,
       "learning_rate": 5e-05,
-      "loss": 0.2687,
+      "loss": 0.97,
       "step": 143850
     },
     {
       "epoch": 2.337248785559942,
-      "grad_norm": 4.15625,
+      "grad_norm": 9.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2712,
+      "loss": 0.9988,
       "step": 143860
     },
     {
       "epoch": 2.337411252457312,
-      "grad_norm": 3.28125,
+      "grad_norm": 9.1875,
       "learning_rate": 5e-05,
-      "loss": 0.2498,
+      "loss": 1.0007,
       "step": 143870
     },
     {
       "epoch": 2.3375737193546815,
-      "grad_norm": 3.875,
+      "grad_norm": 9.375,
       "learning_rate": 5e-05,
-      "loss": 0.2547,
+      "loss": 1.0265,
       "step": 143880
     },
     {
       "epoch": 2.337736186252051,
-      "grad_norm": 3.515625,
+      "grad_norm": 10.0,
       "learning_rate": 5e-05,
-      "loss": 0.2626,
+      "loss": 0.9877,
       "step": 143890
     },
     {
       "epoch": 2.337898653149421,
-      "grad_norm": 4.25,
+      "grad_norm": 11.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2514,
+      "loss": 1.0134,
       "step": 143900
     },
     {
       "epoch": 2.3380611200467905,
-      "grad_norm": 4.59375,
+      "grad_norm": 11.25,
       "learning_rate": 5e-05,
-      "loss": 0.2855,
+      "loss": 1.0292,
       "step": 143910
     },
     {
       "epoch": 2.33822358694416,
-      "grad_norm": 4.34375,
+      "grad_norm": 10.125,
       "learning_rate": 5e-05,
-      "loss": 0.2661,
+      "loss": 0.9853,
       "step": 143920
     },
     {
       "epoch": 2.3383860538415298,
-      "grad_norm": 3.78125,
+      "grad_norm": 8.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2498,
+      "loss": 0.99,
       "step": 143930
     },
     {
       "epoch": 2.3385485207388994,
-      "grad_norm": 4.875,
+      "grad_norm": 9.0625,
       "learning_rate": 5e-05,
-      "loss": 0.2581,
+      "loss": 1.0027,
       "step": 143940
     },
     {
       "epoch": 2.338710987636269,
-      "grad_norm": 3.796875,
+      "grad_norm": 10.3125,
       "learning_rate": 5e-05,
-      "loss": 0.2493,
+      "loss": 0.9766,
       "step": 143950
     },
     {
       "epoch": 2.3388734545336387,
-      "grad_norm": 3.65625,
+      "grad_norm": 10.5,
       "learning_rate": 5e-05,
-      "loss": 0.2458,
+      "loss": 1.0029,
       "step": 143960
     },
     {
       "epoch": 2.3390359214310084,
-      "grad_norm": 3.359375,
+      "grad_norm": 9.625,
       "learning_rate": 5e-05,
-      "loss": 0.2422,
+      "loss": 0.9588,
       "step": 143970
     },
     {
       "epoch": 2.339198388328378,
-      "grad_norm": 4.1875,
+      "grad_norm": 10.25,
       "learning_rate": 5e-05,
-      "loss": 0.2708,
+      "loss": 1.0014,
       "step": 143980
     },
     {
       "epoch": 2.3393608552257477,
-      "grad_norm": 3.53125,
+      "grad_norm": 9.4375,
       "learning_rate": 5e-05,
-      "loss": 0.273,
+      "loss": 1.0253,
       "step": 143990
     },
     {
       "epoch": 2.3395233221231173,
-      "grad_norm": 3.84375,
+      "grad_norm": 8.9375,
       "learning_rate": 5e-05,
-      "loss": 0.2572,
+      "loss": 0.9446,
       "step": 144000
-    },
-    {
-      "epoch": 2.339685789020487,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 144010
-    },
-    {
-      "epoch": 2.3398482559178566,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 144020
-    },
-    {
-      "epoch": 2.3400107228152263,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 144030
-    },
-    {
-      "epoch": 2.340173189712596,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 144040
-    },
-    {
-      "epoch": 2.3403356566099656,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2545,
-      "step": 144050
-    },
-    {
-      "epoch": 2.340498123507335,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 144060
-    },
-    {
-      "epoch": 2.340660590404705,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2688,
-      "step": 144070
-    },
-    {
-      "epoch": 2.3408230573020745,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.29,
-      "step": 144080
-    },
-    {
-      "epoch": 2.3409855241994446,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.246,
-      "step": 144090
-    },
-    {
-      "epoch": 2.341147991096814,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 144100
-    },
-    {
-      "epoch": 2.341310457994184,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2393,
-      "step": 144110
-    },
-    {
-      "epoch": 2.341472924891553,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 144120
-    },
-    {
-      "epoch": 2.341635391788923,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2513,
-      "step": 144130
-    },
-    {
-      "epoch": 2.3417978586862924,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 144140
-    },
-    {
-      "epoch": 2.3419603255836625,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2492,
-      "step": 144150
-    },
-    {
-      "epoch": 2.342122792481032,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 144160
-    },
-    {
-      "epoch": 2.342285259378402,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 144170
-    },
-    {
-      "epoch": 2.3424477262757715,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2681,
-      "step": 144180
-    },
-    {
-      "epoch": 2.342610193173141,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 144190
-    },
-    {
-      "epoch": 2.3427726600705108,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 144200
-    },
-    {
-      "epoch": 2.3429351269678804,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 144210
-    },
-    {
-      "epoch": 2.34309759386525,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 144220
-    },
-    {
-      "epoch": 2.3432600607626197,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 144230
-    },
-    {
-      "epoch": 2.3434225276599894,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 144240
-    },
-    {
-      "epoch": 2.343584994557359,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2475,
-      "step": 144250
-    },
-    {
-      "epoch": 2.3437474614547287,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 144260
-    },
-    {
-      "epoch": 2.3439099283520983,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 144270
-    },
-    {
-      "epoch": 2.344072395249468,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 144280
-    },
-    {
-      "epoch": 2.3442348621468376,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 144290
-    },
-    {
-      "epoch": 2.3443973290442073,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2482,
-      "step": 144300
-    },
-    {
-      "epoch": 2.344559795941577,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 144310
-    },
-    {
-      "epoch": 2.3447222628389466,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 144320
-    },
-    {
-      "epoch": 2.3448847297363162,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2445,
-      "step": 144330
-    },
-    {
-      "epoch": 2.345047196633686,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2594,
-      "step": 144340
-    },
-    {
-      "epoch": 2.3452096635310555,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 144350
-    },
-    {
-      "epoch": 2.345372130428425,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 144360
-    },
-    {
-      "epoch": 2.345534597325795,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2657,
-      "step": 144370
-    },
-    {
-      "epoch": 2.3456970642231645,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2649,
-      "step": 144380
-    },
-    {
-      "epoch": 2.345859531120534,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2568,
-      "step": 144390
-    },
-    {
-      "epoch": 2.346021998017904,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 144400
-    },
-    {
-      "epoch": 2.3461844649152734,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2551,
-      "step": 144410
-    },
-    {
-      "epoch": 2.346346931812643,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2663,
-      "step": 144420
-    },
-    {
-      "epoch": 2.3465093987100127,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.261,
-      "step": 144430
-    },
-    {
-      "epoch": 2.3466718656073824,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 144440
-    },
-    {
-      "epoch": 2.346834332504752,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 144450
-    },
-    {
-      "epoch": 2.3469967994021217,
-      "grad_norm": 3.046875,
-      "learning_rate": 5e-05,
-      "loss": 0.2569,
-      "step": 144460
-    },
-    {
-      "epoch": 2.3471592662994913,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 144470
-    },
-    {
-      "epoch": 2.347321733196861,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 144480
-    },
-    {
-      "epoch": 2.3474842000942306,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 144490
-    },
-    {
-      "epoch": 2.3476466669916003,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 144500
-    },
-    {
-      "epoch": 2.34780913388897,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 144510
-    },
-    {
-      "epoch": 2.3479716007863396,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 144520
-    },
-    {
-      "epoch": 2.3481340676837092,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2542,
-      "step": 144530
-    },
-    {
-      "epoch": 2.3482965345810793,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 144540
-    },
-    {
-      "epoch": 2.3484590014784485,
-      "grad_norm": 2.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 144550
-    },
-    {
-      "epoch": 2.3486214683758186,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 144560
-    },
-    {
-      "epoch": 2.348783935273188,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2964,
-      "step": 144570
-    },
-    {
-      "epoch": 2.348946402170558,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 144580
-    },
-    {
-      "epoch": 2.3491088690679276,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 144590
-    },
-    {
-      "epoch": 2.3492713359652972,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 144600
-    },
-    {
-      "epoch": 2.349433802862667,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 144610
-    },
-    {
-      "epoch": 2.3495962697600365,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2455,
-      "step": 144620
-    },
-    {
-      "epoch": 2.349758736657406,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2431,
-      "step": 144630
-    },
-    {
-      "epoch": 2.349921203554776,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2521,
-      "step": 144640
-    },
-    {
-      "epoch": 2.3500836704521455,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 144650
-    },
-    {
-      "epoch": 2.350246137349515,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 144660
-    },
-    {
-      "epoch": 2.350408604246885,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2562,
-      "step": 144670
-    },
-    {
-      "epoch": 2.3505710711442545,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 144680
-    },
-    {
-      "epoch": 2.350733538041624,
-      "grad_norm": 2.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2466,
-      "step": 144690
-    },
-    {
-      "epoch": 2.3508960049389938,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2496,
-      "step": 144700
-    },
-    {
-      "epoch": 2.3510584718363634,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 144710
-    },
-    {
-      "epoch": 2.351220938733733,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 144720
-    },
-    {
-      "epoch": 2.3513834056311027,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2571,
-      "step": 144730
-    },
-    {
-      "epoch": 2.3515458725284724,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2468,
-      "step": 144740
-    },
-    {
-      "epoch": 2.351708339425842,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2575,
-      "step": 144750
-    },
-    {
-      "epoch": 2.3518708063232117,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 144760
-    },
-    {
-      "epoch": 2.3520332732205813,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2522,
-      "step": 144770
-    },
-    {
-      "epoch": 2.352195740117951,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2624,
-      "step": 144780
-    },
-    {
-      "epoch": 2.3523582070153206,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 144790
-    },
-    {
-      "epoch": 2.3525206739126903,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.25,
-      "step": 144800
-    },
-    {
-      "epoch": 2.35268314081006,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2707,
-      "step": 144810
-    },
-    {
-      "epoch": 2.3528456077074296,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 144820
-    },
-    {
-      "epoch": 2.353008074604799,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2552,
-      "step": 144830
-    },
-    {
-      "epoch": 2.353170541502169,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 144840
-    },
-    {
-      "epoch": 2.3533330083995385,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 144850
-    },
-    {
-      "epoch": 2.353495475296908,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 144860
-    },
-    {
-      "epoch": 2.353657942194278,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2665,
-      "step": 144870
-    },
-    {
-      "epoch": 2.3538204090916475,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.252,
-      "step": 144880
-    },
-    {
-      "epoch": 2.353982875989017,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2793,
-      "step": 144890
-    },
-    {
-      "epoch": 2.3541453428863868,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 144900
-    },
-    {
-      "epoch": 2.3543078097837564,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.24,
-      "step": 144910
-    },
-    {
-      "epoch": 2.354470276681126,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 144920
-    },
-    {
-      "epoch": 2.3546327435784957,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 144930
-    },
-    {
-      "epoch": 2.3547952104758654,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2654,
-      "step": 144940
-    },
-    {
-      "epoch": 2.354957677373235,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2635,
-      "step": 144950
-    },
-    {
-      "epoch": 2.3551201442706047,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 144960
-    },
-    {
-      "epoch": 2.3552826111679748,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 144970
-    },
-    {
-      "epoch": 2.355445078065344,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 144980
-    },
-    {
-      "epoch": 2.355607544962714,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2681,
-      "step": 144990
-    },
-    {
-      "epoch": 2.3557700118600833,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2523,
-      "step": 145000
-    },
-    {
-      "epoch": 2.3559324787574534,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2449,
-      "step": 145010
-    },
-    {
-      "epoch": 2.3560949456548226,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 145020
-    },
-    {
-      "epoch": 2.3562574125521927,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2534,
-      "step": 145030
-    },
-    {
-      "epoch": 2.3564198794495623,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 145040
-    },
-    {
-      "epoch": 2.356582346346932,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 145050
-    },
-    {
-      "epoch": 2.3567448132443016,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 145060
-    },
-    {
-      "epoch": 2.3569072801416713,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 145070
-    },
-    {
-      "epoch": 2.357069747039041,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2634,
-      "step": 145080
-    },
-    {
-      "epoch": 2.3572322139364106,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 145090
-    },
-    {
-      "epoch": 2.3573946808337802,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 145100
-    },
-    {
-      "epoch": 2.35755714773115,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 145110
-    },
-    {
-      "epoch": 2.3577196146285195,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 145120
-    },
-    {
-      "epoch": 2.357882081525889,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2679,
-      "step": 145130
-    },
-    {
-      "epoch": 2.358044548423259,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2563,
-      "step": 145140
-    },
-    {
-      "epoch": 2.3582070153206285,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 145150
-    },
-    {
-      "epoch": 2.358369482217998,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2594,
-      "step": 145160
-    },
-    {
-      "epoch": 2.358531949115368,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 145170
-    },
-    {
-      "epoch": 2.3586944160127374,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2596,
-      "step": 145180
-    },
-    {
-      "epoch": 2.358856882910107,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 145190
-    },
-    {
-      "epoch": 2.3590193498074767,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 145200
-    },
-    {
-      "epoch": 2.3591818167048464,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2578,
-      "step": 145210
-    },
-    {
-      "epoch": 2.359344283602216,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 145220
-    },
-    {
-      "epoch": 2.3595067504995857,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 145230
-    },
-    {
-      "epoch": 2.3596692173969553,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2897,
-      "step": 145240
-    },
-    {
-      "epoch": 2.359831684294325,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.274,
-      "step": 145250
-    },
-    {
-      "epoch": 2.3599941511916946,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.261,
-      "step": 145260
-    },
-    {
-      "epoch": 2.3601566180890643,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 145270
-    },
-    {
-      "epoch": 2.360319084986434,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 145280
-    },
-    {
-      "epoch": 2.3604815518838036,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 145290
-    },
-    {
-      "epoch": 2.3606440187811732,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2508,
-      "step": 145300
-    },
-    {
-      "epoch": 2.360806485678543,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 145310
-    },
-    {
-      "epoch": 2.3609689525759125,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2558,
-      "step": 145320
-    },
-    {
-      "epoch": 2.361131419473282,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 145330
-    },
-    {
-      "epoch": 2.361293886370652,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2563,
-      "step": 145340
-    },
-    {
-      "epoch": 2.3614563532680215,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 145350
-    },
-    {
-      "epoch": 2.361618820165391,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 145360
-    },
-    {
-      "epoch": 2.361781287062761,
-      "grad_norm": 2.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2639,
-      "step": 145370
-    },
-    {
-      "epoch": 2.3619437539601305,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2702,
-      "step": 145380
-    },
-    {
-      "epoch": 2.3621062208575,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2585,
-      "step": 145390
-    },
-    {
-      "epoch": 2.3622686877548698,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 145400
-    },
-    {
-      "epoch": 2.3624311546522394,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2939,
-      "step": 145410
-    },
-    {
-      "epoch": 2.3625936215496095,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 145420
-    },
-    {
-      "epoch": 2.3627560884469787,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 145430
-    },
-    {
-      "epoch": 2.362918555344349,
-      "grad_norm": 2.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 145440
-    },
-    {
-      "epoch": 2.363081022241718,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2635,
-      "step": 145450
-    },
-    {
-      "epoch": 2.363243489139088,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 145460
-    },
-    {
-      "epoch": 2.3634059560364578,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 145470
-    },
-    {
-      "epoch": 2.3635684229338274,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 145480
-    },
-    {
-      "epoch": 2.363730889831197,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 145490
-    },
-    {
-      "epoch": 2.3638933567285667,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 145500
-    },
-    {
-      "epoch": 2.3640558236259364,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 145510
-    },
-    {
-      "epoch": 2.364218290523306,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 145520
-    },
-    {
-      "epoch": 2.3643807574206757,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 145530
-    },
-    {
-      "epoch": 2.3645432243180453,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2456,
-      "step": 145540
-    },
-    {
-      "epoch": 2.364705691215415,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2719,
-      "step": 145550
-    },
-    {
-      "epoch": 2.3648681581127846,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 145560
-    },
-    {
-      "epoch": 2.3650306250101543,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 145570
-    },
-    {
-      "epoch": 2.365193091907524,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 145580
-    },
-    {
-      "epoch": 2.3653555588048936,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2551,
-      "step": 145590
-    },
-    {
-      "epoch": 2.365518025702263,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 145600
-    },
-    {
-      "epoch": 2.365680492599633,
-      "grad_norm": 3.046875,
-      "learning_rate": 5e-05,
-      "loss": 0.2606,
-      "step": 145610
-    },
-    {
-      "epoch": 2.3658429594970025,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 145620
-    },
-    {
-      "epoch": 2.366005426394372,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 145630
-    },
-    {
-      "epoch": 2.366167893291742,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 145640
-    },
-    {
-      "epoch": 2.3663303601891115,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 145650
-    },
-    {
-      "epoch": 2.366492827086481,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 145660
-    },
-    {
-      "epoch": 2.3666552939838508,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2654,
-      "step": 145670
-    },
-    {
-      "epoch": 2.3668177608812204,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 145680
-    },
-    {
-      "epoch": 2.36698022777859,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2451,
-      "step": 145690
-    },
-    {
-      "epoch": 2.3671426946759597,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2401,
-      "step": 145700
-    },
-    {
-      "epoch": 2.3673051615733294,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2605,
-      "step": 145710
-    },
-    {
-      "epoch": 2.367467628470699,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2591,
-      "step": 145720
-    },
-    {
-      "epoch": 2.3676300953680687,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 145730
-    },
-    {
-      "epoch": 2.3677925622654383,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 145740
-    },
-    {
-      "epoch": 2.367955029162808,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 145750
-    },
-    {
-      "epoch": 2.3681174960601776,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 145760
-    },
-    {
-      "epoch": 2.3682799629575473,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 145770
-    },
-    {
-      "epoch": 2.368442429854917,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2422,
-      "step": 145780
-    },
-    {
-      "epoch": 2.3686048967522866,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2541,
-      "step": 145790
-    },
-    {
-      "epoch": 2.3687673636496562,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2619,
-      "step": 145800
-    },
-    {
-      "epoch": 2.368929830547026,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.257,
-      "step": 145810
-    },
-    {
-      "epoch": 2.3690922974443955,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2531,
-      "step": 145820
-    },
-    {
-      "epoch": 2.369254764341765,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2455,
-      "step": 145830
-    },
-    {
-      "epoch": 2.369417231239135,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2527,
-      "step": 145840
-    },
-    {
-      "epoch": 2.369579698136505,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2499,
-      "step": 145850
-    },
-    {
-      "epoch": 2.369742165033874,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 145860
-    },
-    {
-      "epoch": 2.3699046319312442,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2445,
-      "step": 145870
-    },
-    {
-      "epoch": 2.3700670988286134,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2562,
-      "step": 145880
-    },
-    {
-      "epoch": 2.3702295657259835,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2495,
-      "step": 145890
-    },
-    {
-      "epoch": 2.370392032623353,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2379,
-      "step": 145900
-    },
-    {
-      "epoch": 2.370554499520723,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2426,
-      "step": 145910
-    },
-    {
-      "epoch": 2.3707169664180925,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2579,
-      "step": 145920
-    },
-    {
-      "epoch": 2.370879433315462,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2494,
-      "step": 145930
-    },
-    {
-      "epoch": 2.371041900212832,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2617,
-      "step": 145940
-    },
-    {
-      "epoch": 2.3712043671102014,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2735,
-      "step": 145950
-    },
-    {
-      "epoch": 2.371366834007571,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 145960
-    },
-    {
-      "epoch": 2.3715293009049407,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 145970
-    },
-    {
-      "epoch": 2.3716917678023104,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2611,
-      "step": 145980
-    },
-    {
-      "epoch": 2.37185423469968,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2548,
-      "step": 145990
-    },
-    {
-      "epoch": 2.3720167015970497,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 146000
-    },
-    {
-      "epoch": 2.3721791684944193,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 146010
-    },
-    {
-      "epoch": 2.372341635391789,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2628,
-      "step": 146020
-    },
-    {
-      "epoch": 2.3725041022891586,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 146030
-    },
-    {
-      "epoch": 2.3726665691865283,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 146040
-    },
-    {
-      "epoch": 2.372829036083898,
-      "grad_norm": 2.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 146050
-    },
-    {
-      "epoch": 2.3729915029812676,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 146060
-    },
-    {
-      "epoch": 2.3731539698786372,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 146070
-    },
-    {
-      "epoch": 2.373316436776007,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 146080
-    },
-    {
-      "epoch": 2.3734789036733765,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2608,
-      "step": 146090
-    },
-    {
-      "epoch": 2.373641370570746,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2595,
-      "step": 146100
-    },
-    {
-      "epoch": 2.373803837468116,
-      "grad_norm": 3.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2639,
-      "step": 146110
-    },
-    {
-      "epoch": 2.3739663043654855,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 146120
-    },
-    {
-      "epoch": 2.374128771262855,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2535,
-      "step": 146130
-    },
-    {
-      "epoch": 2.374291238160225,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 146140
-    },
-    {
-      "epoch": 2.3744537050575945,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2681,
-      "step": 146150
-    },
-    {
-      "epoch": 2.374616171954964,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2545,
-      "step": 146160
-    },
-    {
-      "epoch": 2.3747786388523338,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 146170
-    },
-    {
-      "epoch": 2.3749411057497034,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 146180
-    },
-    {
-      "epoch": 2.375103572647073,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 146190
-    },
-    {
-      "epoch": 2.3752660395444427,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 146200
-    },
-    {
-      "epoch": 2.3754285064418124,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 146210
-    },
-    {
-      "epoch": 2.375590973339182,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2511,
-      "step": 146220
-    },
-    {
-      "epoch": 2.3757534402365517,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 146230
-    },
-    {
-      "epoch": 2.3759159071339213,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2676,
-      "step": 146240
-    },
-    {
-      "epoch": 2.376078374031291,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2409,
-      "step": 146250
-    },
-    {
-      "epoch": 2.3762408409286606,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2583,
-      "step": 146260
-    },
-    {
-      "epoch": 2.3764033078260303,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2519,
-      "step": 146270
-    },
-    {
-      "epoch": 2.3765657747234,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2423,
-      "step": 146280
-    },
-    {
-      "epoch": 2.3767282416207696,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2697,
-      "step": 146290
-    },
-    {
-      "epoch": 2.3768907085181397,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 146300
-    },
-    {
-      "epoch": 2.377053175415509,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 146310
-    },
-    {
-      "epoch": 2.377215642312879,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 146320
-    },
-    {
-      "epoch": 2.377378109210248,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 146330
-    },
-    {
-      "epoch": 2.3775405761076183,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 146340
-    },
-    {
-      "epoch": 2.377703043004988,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2553,
-      "step": 146350
-    },
-    {
-      "epoch": 2.3778655099023576,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 146360
-    },
-    {
-      "epoch": 2.378027976799727,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 146370
-    },
-    {
-      "epoch": 2.378190443697097,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 146380
-    },
-    {
-      "epoch": 2.3783529105944665,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 146390
-    },
-    {
-      "epoch": 2.378515377491836,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2574,
-      "step": 146400
-    },
-    {
-      "epoch": 2.378677844389206,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 146410
-    },
-    {
-      "epoch": 2.3788403112865755,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 146420
-    },
-    {
-      "epoch": 2.379002778183945,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.276,
-      "step": 146430
-    },
-    {
-      "epoch": 2.3791652450813148,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2765,
-      "step": 146440
-    },
-    {
-      "epoch": 2.3793277119786844,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 146450
-    },
-    {
-      "epoch": 2.379490178876054,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 146460
-    },
-    {
-      "epoch": 2.3796526457734237,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 146470
-    },
-    {
-      "epoch": 2.3798151126707934,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 146480
-    },
-    {
-      "epoch": 2.379977579568163,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 146490
-    },
-    {
-      "epoch": 2.3801400464655327,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 146500
-    },
-    {
-      "epoch": 2.3803025133629023,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 146510
-    },
-    {
-      "epoch": 2.380464980260272,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2697,
-      "step": 146520
-    },
-    {
-      "epoch": 2.3806274471576416,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2625,
-      "step": 146530
-    },
-    {
-      "epoch": 2.3807899140550113,
-      "grad_norm": 2.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 146540
-    },
-    {
-      "epoch": 2.380952380952381,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2531,
-      "step": 146550
-    },
-    {
-      "epoch": 2.3811148478497506,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2502,
-      "step": 146560
-    },
-    {
-      "epoch": 2.3812773147471202,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2612,
-      "step": 146570
-    },
-    {
-      "epoch": 2.38143978164449,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2506,
-      "step": 146580
-    },
-    {
-      "epoch": 2.3816022485418595,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2545,
-      "step": 146590
-    },
-    {
-      "epoch": 2.381764715439229,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 146600
-    },
-    {
-      "epoch": 2.381927182336599,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 146610
-    },
-    {
-      "epoch": 2.3820896492339685,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 146620
-    },
-    {
-      "epoch": 2.382252116131338,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 146630
-    },
-    {
-      "epoch": 2.382414583028708,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2648,
-      "step": 146640
-    },
-    {
-      "epoch": 2.3825770499260774,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2852,
-      "step": 146650
-    },
-    {
-      "epoch": 2.382739516823447,
-      "grad_norm": 2.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 146660
-    },
-    {
-      "epoch": 2.3829019837208167,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.255,
-      "step": 146670
-    },
-    {
-      "epoch": 2.3830644506181864,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2493,
-      "step": 146680
-    },
-    {
-      "epoch": 2.383226917515556,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2605,
-      "step": 146690
-    },
-    {
-      "epoch": 2.3833893844129257,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2442,
-      "step": 146700
-    },
-    {
-      "epoch": 2.3835518513102953,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2713,
-      "step": 146710
-    },
-    {
-      "epoch": 2.383714318207665,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 146720
-    },
-    {
-      "epoch": 2.383876785105035,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 146730
-    },
-    {
-      "epoch": 2.3840392520024043,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2627,
-      "step": 146740
-    },
-    {
-      "epoch": 2.3842017188997744,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 146750
-    },
-    {
-      "epoch": 2.3843641857971436,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 146760
-    },
-    {
-      "epoch": 2.3845266526945137,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 146770
-    },
-    {
-      "epoch": 2.3846891195918833,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2592,
-      "step": 146780
-    },
-    {
-      "epoch": 2.384851586489253,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2657,
-      "step": 146790
-    },
-    {
-      "epoch": 2.3850140533866226,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 146800
-    },
-    {
-      "epoch": 2.3851765202839923,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 146810
-    },
-    {
-      "epoch": 2.385338987181362,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 146820
-    },
-    {
-      "epoch": 2.3855014540787316,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3026,
-      "step": 146830
-    },
-    {
-      "epoch": 2.3856639209761012,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 146840
-    },
-    {
-      "epoch": 2.385826387873471,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2565,
-      "step": 146850
-    },
-    {
-      "epoch": 2.3859888547708406,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 146860
-    },
-    {
-      "epoch": 2.38615132166821,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2955,
-      "step": 146870
-    },
-    {
-      "epoch": 2.38631378856558,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3073,
-      "step": 146880
-    },
-    {
-      "epoch": 2.3864762554629495,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 146890
-    },
-    {
-      "epoch": 2.386638722360319,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 146900
-    },
-    {
-      "epoch": 2.386801189257689,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 146910
-    },
-    {
-      "epoch": 2.3869636561550585,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 146920
-    },
-    {
-      "epoch": 2.387126123052428,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 146930
-    },
-    {
-      "epoch": 2.3872885899497978,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 146940
-    },
-    {
-      "epoch": 2.3874510568471674,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2595,
-      "step": 146950
-    },
-    {
-      "epoch": 2.387613523744537,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 146960
-    },
-    {
-      "epoch": 2.3877759906419067,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 146970
-    },
-    {
-      "epoch": 2.3879384575392764,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 146980
-    },
-    {
-      "epoch": 2.388100924436646,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 146990
-    },
-    {
-      "epoch": 2.3882633913340157,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2478,
-      "step": 147000
-    },
-    {
-      "epoch": 2.3884258582313853,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2617,
-      "step": 147010
-    },
-    {
-      "epoch": 2.388588325128755,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 147020
-    },
-    {
-      "epoch": 2.3887507920261246,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2654,
-      "step": 147030
-    },
-    {
-      "epoch": 2.3889132589234943,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2535,
-      "step": 147040
-    },
-    {
-      "epoch": 2.389075725820864,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 147050
-    },
-    {
-      "epoch": 2.3892381927182336,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 147060
-    },
-    {
-      "epoch": 2.389400659615603,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2587,
-      "step": 147070
-    },
-    {
-      "epoch": 2.389563126512973,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 147080
-    },
-    {
-      "epoch": 2.3897255934103425,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 147090
-    },
-    {
-      "epoch": 2.389888060307712,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2569,
-      "step": 147100
-    },
-    {
-      "epoch": 2.390050527205082,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2486,
-      "step": 147110
-    },
-    {
-      "epoch": 2.3902129941024515,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2412,
-      "step": 147120
-    },
-    {
-      "epoch": 2.390375460999821,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2593,
-      "step": 147130
-    },
-    {
-      "epoch": 2.3905379278971908,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2968,
-      "step": 147140
-    },
-    {
-      "epoch": 2.3907003947945604,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 147150
-    },
-    {
-      "epoch": 2.3908628616919305,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 147160
-    },
-    {
-      "epoch": 2.3910253285892997,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 147170
-    },
-    {
-      "epoch": 2.39118779548667,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 147180
-    },
-    {
-      "epoch": 2.391350262384039,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2991,
-      "step": 147190
-    },
-    {
-      "epoch": 2.391512729281409,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 147200
-    },
-    {
-      "epoch": 2.3916751961787783,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2765,
-      "step": 147210
-    },
-    {
-      "epoch": 2.3918376630761484,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2617,
-      "step": 147220
-    },
-    {
-      "epoch": 2.392000129973518,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2468,
-      "step": 147230
-    },
-    {
-      "epoch": 2.3921625968708877,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 147240
-    },
-    {
-      "epoch": 2.3923250637682574,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2538,
-      "step": 147250
-    },
-    {
-      "epoch": 2.392487530665627,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2567,
-      "step": 147260
-    },
-    {
-      "epoch": 2.3926499975629967,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 147270
-    },
-    {
-      "epoch": 2.3928124644603663,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 147280
-    },
-    {
-      "epoch": 2.392974931357736,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 147290
-    },
-    {
-      "epoch": 2.3931373982551056,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 147300
-    },
-    {
-      "epoch": 2.3932998651524753,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 147310
-    },
-    {
-      "epoch": 2.393462332049845,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 147320
-    },
-    {
-      "epoch": 2.3936247989472146,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2548,
-      "step": 147330
-    },
-    {
-      "epoch": 2.3937872658445842,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2603,
-      "step": 147340
-    },
-    {
-      "epoch": 2.393949732741954,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 147350
-    },
-    {
-      "epoch": 2.3941121996393235,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2568,
-      "step": 147360
-    },
-    {
-      "epoch": 2.394274666536693,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2426,
-      "step": 147370
-    },
-    {
-      "epoch": 2.394437133434063,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2611,
-      "step": 147380
-    },
-    {
-      "epoch": 2.3945996003314325,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2818,
-      "step": 147390
-    },
-    {
-      "epoch": 2.394762067228802,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 147400
-    },
-    {
-      "epoch": 2.394924534126172,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 147410
-    },
-    {
-      "epoch": 2.3950870010235414,
-      "grad_norm": 2.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2544,
-      "step": 147420
-    },
-    {
-      "epoch": 2.395249467920911,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 147430
-    },
-    {
-      "epoch": 2.3954119348182807,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 147440
-    },
-    {
-      "epoch": 2.3955744017156504,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 147450
-    },
-    {
-      "epoch": 2.39573686861302,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 147460
-    },
-    {
-      "epoch": 2.3958993355103897,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 147470
-    },
-    {
-      "epoch": 2.3960618024077593,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 147480
-    },
-    {
-      "epoch": 2.396224269305129,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 147490
-    },
-    {
-      "epoch": 2.3963867362024986,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 147500
-    },
-    {
-      "epoch": 2.3965492030998683,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 147510
-    },
-    {
-      "epoch": 2.396711669997238,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 147520
-    },
-    {
-      "epoch": 2.3968741368946076,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 147530
-    },
-    {
-      "epoch": 2.3970366037919772,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 147540
-    },
-    {
-      "epoch": 2.397199070689347,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 147550
-    },
-    {
-      "epoch": 2.3973615375867166,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 147560
-    },
-    {
-      "epoch": 2.397524004484086,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 147570
-    },
-    {
-      "epoch": 2.397686471381456,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 147580
-    },
-    {
-      "epoch": 2.3978489382788255,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.3039,
-      "step": 147590
-    },
-    {
-      "epoch": 2.398011405176195,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2585,
-      "step": 147600
-    },
-    {
-      "epoch": 2.3981738720735652,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 147610
-    },
-    {
-      "epoch": 2.3983363389709345,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 147620
-    },
-    {
-      "epoch": 2.3984988058683046,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 147630
-    },
-    {
-      "epoch": 2.3986612727656738,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.274,
-      "step": 147640
-    },
-    {
-      "epoch": 2.398823739663044,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2542,
-      "step": 147650
-    },
-    {
-      "epoch": 2.3989862065604135,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2501,
-      "step": 147660
-    },
-    {
-      "epoch": 2.399148673457783,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 147670
-    },
-    {
-      "epoch": 2.399311140355153,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 147680
-    },
-    {
-      "epoch": 2.3994736072525225,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.297,
-      "step": 147690
-    },
-    {
-      "epoch": 2.399636074149892,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2969,
-      "step": 147700
-    },
-    {
-      "epoch": 2.3997985410472618,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 147710
-    },
-    {
-      "epoch": 2.3999610079446314,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 147720
-    },
-    {
-      "epoch": 2.400123474842001,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 147730
-    },
-    {
-      "epoch": 2.4002859417393707,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 147740
-    },
-    {
-      "epoch": 2.4004484086367404,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 147750
-    },
-    {
-      "epoch": 2.40061087553411,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 147760
-    },
-    {
-      "epoch": 2.4007733424314797,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 147770
-    },
-    {
-      "epoch": 2.4009358093288493,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.3024,
-      "step": 147780
-    },
-    {
-      "epoch": 2.401098276226219,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 147790
-    },
-    {
-      "epoch": 2.4012607431235886,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2931,
-      "step": 147800
-    },
-    {
-      "epoch": 2.4014232100209583,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 147810
-    },
-    {
-      "epoch": 2.401585676918328,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2709,
-      "step": 147820
-    },
-    {
-      "epoch": 2.4017481438156976,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 147830
-    },
-    {
-      "epoch": 2.401910610713067,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 147840
-    },
-    {
-      "epoch": 2.402073077610437,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 147850
-    },
-    {
-      "epoch": 2.4022355445078065,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2603,
-      "step": 147860
-    },
-    {
-      "epoch": 2.402398011405176,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 147870
-    },
-    {
-      "epoch": 2.402560478302546,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 147880
-    },
-    {
-      "epoch": 2.4027229451999155,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 147890
-    },
-    {
-      "epoch": 2.402885412097285,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3034,
-      "step": 147900
-    },
-    {
-      "epoch": 2.4030478789946548,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.264,
-      "step": 147910
-    },
-    {
-      "epoch": 2.4032103458920244,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 147920
-    },
-    {
-      "epoch": 2.403372812789394,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2559,
-      "step": 147930
-    },
-    {
-      "epoch": 2.4035352796867637,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 147940
-    },
-    {
-      "epoch": 2.4036977465841334,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2543,
-      "step": 147950
-    },
-    {
-      "epoch": 2.403860213481503,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 147960
-    },
-    {
-      "epoch": 2.4040226803788727,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 147970
-    },
-    {
-      "epoch": 2.4041851472762423,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 147980
-    },
-    {
-      "epoch": 2.404347614173612,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 147990
-    },
-    {
-      "epoch": 2.4045100810709816,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2713,
-      "step": 148000
-    },
-    {
-      "epoch": 2.4046725479683513,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2587,
-      "step": 148010
-    },
-    {
-      "epoch": 2.404835014865721,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2931,
-      "step": 148020
-    },
-    {
-      "epoch": 2.4049974817630906,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 148030
-    },
-    {
-      "epoch": 2.4051599486604607,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 148040
-    },
-    {
-      "epoch": 2.40532241555783,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2461,
-      "step": 148050
-    },
-    {
-      "epoch": 2.4054848824552,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 148060
-    },
-    {
-      "epoch": 2.405647349352569,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 148070
-    },
-    {
-      "epoch": 2.4058098162499393,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 148080
-    },
-    {
-      "epoch": 2.4059722831473085,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 148090
-    },
-    {
-      "epoch": 2.4061347500446786,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 148100
-    },
-    {
-      "epoch": 2.4062972169420482,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.264,
-      "step": 148110
-    },
-    {
-      "epoch": 2.406459683839418,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2539,
-      "step": 148120
-    },
-    {
-      "epoch": 2.4066221507367875,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 148130
-    },
-    {
-      "epoch": 2.406784617634157,
-      "grad_norm": 9.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 148140
-    },
-    {
-      "epoch": 2.406947084531527,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 148150
-    },
-    {
-      "epoch": 2.4071095514288965,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2536,
-      "step": 148160
-    },
-    {
-      "epoch": 2.407272018326266,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2445,
-      "step": 148170
-    },
-    {
-      "epoch": 2.407434485223636,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2631,
-      "step": 148180
-    },
-    {
-      "epoch": 2.4075969521210054,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 148190
-    },
-    {
-      "epoch": 2.407759419018375,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2599,
-      "step": 148200
-    },
-    {
-      "epoch": 2.4079218859157447,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 148210
-    },
-    {
-      "epoch": 2.4080843528131144,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 148220
-    },
-    {
-      "epoch": 2.408246819710484,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 148230
-    },
-    {
-      "epoch": 2.4084092866078537,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2541,
-      "step": 148240
-    },
-    {
-      "epoch": 2.4085717535052233,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2612,
-      "step": 148250
-    },
-    {
-      "epoch": 2.408734220402593,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 148260
-    },
-    {
-      "epoch": 2.4088966872999626,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2634,
-      "step": 148270
-    },
-    {
-      "epoch": 2.4090591541973323,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2582,
-      "step": 148280
-    },
-    {
-      "epoch": 2.409221621094702,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 148290
-    },
-    {
-      "epoch": 2.4093840879920716,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 148300
-    },
-    {
-      "epoch": 2.4095465548894412,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 148310
-    },
-    {
-      "epoch": 2.409709021786811,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2962,
-      "step": 148320
-    },
-    {
-      "epoch": 2.4098714886841806,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 148330
-    },
-    {
-      "epoch": 2.41003395558155,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2654,
-      "step": 148340
-    },
-    {
-      "epoch": 2.41019642247892,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2543,
-      "step": 148350
-    },
-    {
-      "epoch": 2.4103588893762895,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2447,
-      "step": 148360
-    },
-    {
-      "epoch": 2.410521356273659,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 148370
-    },
-    {
-      "epoch": 2.410683823171029,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2548,
-      "step": 148380
-    },
-    {
-      "epoch": 2.4108462900683985,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2556,
-      "step": 148390
-    },
-    {
-      "epoch": 2.411008756965768,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2719,
-      "step": 148400
-    },
-    {
-      "epoch": 2.4111712238631378,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2549,
-      "step": 148410
-    },
-    {
-      "epoch": 2.4113336907605074,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2612,
-      "step": 148420
-    },
-    {
-      "epoch": 2.411496157657877,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2813,
-      "step": 148430
-    },
-    {
-      "epoch": 2.4116586245552467,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 148440
-    },
-    {
-      "epoch": 2.4118210914526164,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2514,
-      "step": 148450
-    },
-    {
-      "epoch": 2.411983558349986,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2455,
-      "step": 148460
-    },
-    {
-      "epoch": 2.4121460252473557,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.263,
-      "step": 148470
-    },
-    {
-      "epoch": 2.4123084921447253,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 148480
-    },
-    {
-      "epoch": 2.4124709590420954,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2422,
-      "step": 148490
-    },
-    {
-      "epoch": 2.4126334259394646,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2504,
-      "step": 148500
-    },
-    {
-      "epoch": 2.4127958928368347,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.241,
-      "step": 148510
-    },
-    {
-      "epoch": 2.412958359734204,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 148520
-    },
-    {
-      "epoch": 2.413120826631574,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2675,
-      "step": 148530
-    },
-    {
-      "epoch": 2.4132832935289437,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2599,
-      "step": 148540
-    },
-    {
-      "epoch": 2.4134457604263133,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2506,
-      "step": 148550
-    },
-    {
-      "epoch": 2.413608227323683,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.229,
-      "step": 148560
-    },
-    {
-      "epoch": 2.4137706942210526,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2476,
-      "step": 148570
-    },
-    {
-      "epoch": 2.4139331611184223,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2575,
-      "step": 148580
-    },
-    {
-      "epoch": 2.414095628015792,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 148590
-    },
-    {
-      "epoch": 2.4142580949131616,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.251,
-      "step": 148600
-    },
-    {
-      "epoch": 2.414420561810531,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.258,
-      "step": 148610
-    },
-    {
-      "epoch": 2.414583028707901,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 148620
-    },
-    {
-      "epoch": 2.4147454956052705,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2709,
-      "step": 148630
-    },
-    {
-      "epoch": 2.41490796250264,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 148640
-    },
-    {
-      "epoch": 2.41507042940001,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2675,
-      "step": 148650
-    },
-    {
-      "epoch": 2.4152328962973795,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2549,
-      "step": 148660
-    },
-    {
-      "epoch": 2.415395363194749,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 148670
-    },
-    {
-      "epoch": 2.4155578300921188,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 148680
-    },
-    {
-      "epoch": 2.4157202969894884,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2987,
-      "step": 148690
-    },
-    {
-      "epoch": 2.415882763886858,
-      "grad_norm": 2.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 148700
-    },
-    {
-      "epoch": 2.4160452307842277,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 148710
-    },
-    {
-      "epoch": 2.4162076976815974,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 148720
-    },
-    {
-      "epoch": 2.416370164578967,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2649,
-      "step": 148730
-    },
-    {
-      "epoch": 2.4165326314763367,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2676,
-      "step": 148740
-    },
-    {
-      "epoch": 2.4166950983737063,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2736,
-      "step": 148750
-    },
-    {
-      "epoch": 2.416857565271076,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 148760
-    },
-    {
-      "epoch": 2.4170200321684456,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 148770
-    },
-    {
-      "epoch": 2.4171824990658153,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 148780
-    },
-    {
-      "epoch": 2.417344965963185,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 148790
-    },
-    {
-      "epoch": 2.4175074328605546,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 148800
-    },
-    {
-      "epoch": 2.4176698997579242,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2518,
-      "step": 148810
-    },
-    {
-      "epoch": 2.417832366655294,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2649,
-      "step": 148820
-    },
-    {
-      "epoch": 2.4179948335526635,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2669,
-      "step": 148830
-    },
-    {
-      "epoch": 2.418157300450033,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 148840
-    },
-    {
-      "epoch": 2.418319767347403,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 148850
-    },
-    {
-      "epoch": 2.4184822342447725,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2467,
-      "step": 148860
-    },
-    {
-      "epoch": 2.418644701142142,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2507,
-      "step": 148870
-    },
-    {
-      "epoch": 2.418807168039512,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2513,
-      "step": 148880
-    },
-    {
-      "epoch": 2.4189696349368814,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2721,
-      "step": 148890
-    },
-    {
-      "epoch": 2.419132101834251,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 148900
-    },
-    {
-      "epoch": 2.4192945687316207,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 148910
-    },
-    {
-      "epoch": 2.419457035628991,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 148920
-    },
-    {
-      "epoch": 2.41961950252636,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 148930
-    },
-    {
-      "epoch": 2.41978196942373,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 148940
-    },
-    {
-      "epoch": 2.4199444363210993,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2713,
-      "step": 148950
-    },
-    {
-      "epoch": 2.4201069032184694,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 148960
-    },
-    {
-      "epoch": 2.420269370115839,
-      "grad_norm": 2.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2669,
-      "step": 148970
-    },
-    {
-      "epoch": 2.4204318370132087,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 148980
-    },
-    {
-      "epoch": 2.4205943039105784,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2592,
-      "step": 148990
-    },
-    {
-      "epoch": 2.420756770807948,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.276,
-      "step": 149000
-    },
-    {
-      "epoch": 2.4209192377053177,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2423,
-      "step": 149010
-    },
-    {
-      "epoch": 2.4210817046026873,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2574,
-      "step": 149020
-    },
-    {
-      "epoch": 2.421244171500057,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.277,
-      "step": 149030
-    },
-    {
-      "epoch": 2.4214066383974266,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2589,
-      "step": 149040
-    },
-    {
-      "epoch": 2.4215691052947963,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2685,
-      "step": 149050
-    },
-    {
-      "epoch": 2.421731572192166,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2395,
-      "step": 149060
-    },
-    {
-      "epoch": 2.4218940390895356,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2488,
-      "step": 149070
-    },
-    {
-      "epoch": 2.4220565059869053,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2528,
-      "step": 149080
-    },
-    {
-      "epoch": 2.422218972884275,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2548,
-      "step": 149090
-    },
-    {
-      "epoch": 2.4223814397816446,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 149100
-    },
-    {
-      "epoch": 2.422543906679014,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 149110
-    },
-    {
-      "epoch": 2.422706373576384,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2569,
-      "step": 149120
-    },
-    {
-      "epoch": 2.4228688404737535,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 149130
-    },
-    {
-      "epoch": 2.423031307371123,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 149140
-    },
-    {
-      "epoch": 2.423193774268493,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 149150
-    },
-    {
-      "epoch": 2.4233562411658625,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 149160
-    },
-    {
-      "epoch": 2.423518708063232,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 149170
-    },
-    {
-      "epoch": 2.4236811749606018,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2466,
-      "step": 149180
-    },
-    {
-      "epoch": 2.4238436418579714,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2654,
-      "step": 149190
-    },
-    {
-      "epoch": 2.424006108755341,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 149200
-    },
-    {
-      "epoch": 2.4241685756527107,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 149210
-    },
-    {
-      "epoch": 2.4243310425500804,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.246,
-      "step": 149220
-    },
-    {
-      "epoch": 2.42449350944745,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2342,
-      "step": 149230
-    },
-    {
-      "epoch": 2.4246559763448197,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2813,
-      "step": 149240
-    },
-    {
-      "epoch": 2.4248184432421893,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 149250
-    },
-    {
-      "epoch": 2.424980910139559,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 149260
-    },
-    {
-      "epoch": 2.4251433770369286,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 149270
-    },
-    {
-      "epoch": 2.4253058439342983,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3089,
-      "step": 149280
-    },
-    {
-      "epoch": 2.425468310831668,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 149290
-    },
-    {
-      "epoch": 2.4256307777290376,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 149300
-    },
-    {
-      "epoch": 2.425793244626407,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 149310
-    },
-    {
-      "epoch": 2.425955711523777,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 149320
-    },
-    {
-      "epoch": 2.4261181784211465,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 149330
-    },
-    {
-      "epoch": 2.426280645318516,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 149340
-    },
-    {
-      "epoch": 2.426443112215886,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 149350
-    },
-    {
-      "epoch": 2.4266055791132555,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 149360
-    },
-    {
-      "epoch": 2.4267680460106256,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 149370
-    },
-    {
-      "epoch": 2.4269305129079948,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 149380
-    },
-    {
-      "epoch": 2.427092979805365,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 149390
-    },
-    {
-      "epoch": 2.427255446702734,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 149400
-    },
-    {
-      "epoch": 2.427417913600104,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2496,
-      "step": 149410
-    },
-    {
-      "epoch": 2.427580380497474,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 149420
-    },
-    {
-      "epoch": 2.4277428473948435,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 149430
-    },
-    {
-      "epoch": 2.427905314292213,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 149440
-    },
-    {
-      "epoch": 2.4280677811895828,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 149450
-    },
-    {
-      "epoch": 2.4282302480869524,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 149460
-    },
-    {
-      "epoch": 2.428392714984322,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 149470
-    },
-    {
-      "epoch": 2.4285551818816917,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 149480
-    },
-    {
-      "epoch": 2.4287176487790614,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2516,
-      "step": 149490
-    },
-    {
-      "epoch": 2.428880115676431,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 149500
-    },
-    {
-      "epoch": 2.4290425825738007,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 149510
-    },
-    {
-      "epoch": 2.4292050494711703,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2459,
-      "step": 149520
-    },
-    {
-      "epoch": 2.42936751636854,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2518,
-      "step": 149530
-    },
-    {
-      "epoch": 2.4295299832659096,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2594,
-      "step": 149540
-    },
-    {
-      "epoch": 2.4296924501632793,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.243,
-      "step": 149550
-    },
-    {
-      "epoch": 2.429854917060649,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2484,
-      "step": 149560
-    },
-    {
-      "epoch": 2.4300173839580186,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 149570
-    },
-    {
-      "epoch": 2.4301798508553882,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.274,
-      "step": 149580
-    },
-    {
-      "epoch": 2.430342317752758,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2538,
-      "step": 149590
-    },
-    {
-      "epoch": 2.4305047846501275,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2625,
-      "step": 149600
-    },
-    {
-      "epoch": 2.430667251547497,
-      "grad_norm": 3.078125,
-      "learning_rate": 5e-05,
-      "loss": 0.2581,
-      "step": 149610
-    },
-    {
-      "epoch": 2.430829718444867,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 149620
-    },
-    {
-      "epoch": 2.4309921853422365,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2685,
-      "step": 149630
-    },
-    {
-      "epoch": 2.431154652239606,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 149640
-    },
-    {
-      "epoch": 2.431317119136976,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 149650
-    },
-    {
-      "epoch": 2.4314795860343454,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 149660
-    },
-    {
-      "epoch": 2.431642052931715,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 149670
-    },
-    {
-      "epoch": 2.4318045198290847,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.3129,
-      "step": 149680
-    },
-    {
-      "epoch": 2.4319669867264544,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 149690
-    },
-    {
-      "epoch": 2.432129453623824,
-      "grad_norm": 2.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 149700
-    },
-    {
-      "epoch": 2.4322919205211937,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 149710
-    },
-    {
-      "epoch": 2.4324543874185633,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 149720
-    },
-    {
-      "epoch": 2.432616854315933,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 149730
-    },
-    {
-      "epoch": 2.4327793212133026,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2713,
-      "step": 149740
-    },
-    {
-      "epoch": 2.4329417881106723,
-      "grad_norm": 2.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 149750
-    },
-    {
-      "epoch": 2.433104255008042,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 149760
-    },
-    {
-      "epoch": 2.4332667219054116,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 149770
-    },
-    {
-      "epoch": 2.4334291888027813,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 149780
-    },
-    {
-      "epoch": 2.433591655700151,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.3055,
-      "step": 149790
-    },
-    {
-      "epoch": 2.433754122597521,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 149800
-    },
-    {
-      "epoch": 2.43391658949489,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 149810
-    },
-    {
-      "epoch": 2.4340790563922603,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2919,
-      "step": 149820
-    },
-    {
-      "epoch": 2.4342415232896295,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2721,
-      "step": 149830
-    },
-    {
-      "epoch": 2.4344039901869996,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 149840
-    },
-    {
-      "epoch": 2.4345664570843693,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 149850
-    },
-    {
-      "epoch": 2.434728923981739,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2502,
-      "step": 149860
-    },
-    {
-      "epoch": 2.4348913908791086,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2459,
-      "step": 149870
-    },
-    {
-      "epoch": 2.435053857776478,
-      "grad_norm": 2.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2414,
-      "step": 149880
-    },
-    {
-      "epoch": 2.435216324673848,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 149890
-    },
-    {
-      "epoch": 2.4353787915712175,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2465,
-      "step": 149900
-    },
-    {
-      "epoch": 2.435541258468587,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 149910
-    },
-    {
-      "epoch": 2.435703725365957,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 149920
-    },
-    {
-      "epoch": 2.4358661922633265,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 149930
-    },
-    {
-      "epoch": 2.436028659160696,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 149940
-    },
-    {
-      "epoch": 2.4361911260580658,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 149950
-    },
-    {
-      "epoch": 2.4363535929554354,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2702,
-      "step": 149960
-    },
-    {
-      "epoch": 2.436516059852805,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 149970
-    },
-    {
-      "epoch": 2.4366785267501747,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2651,
-      "step": 149980
-    },
-    {
-      "epoch": 2.4368409936475444,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2407,
-      "step": 149990
-    },
-    {
-      "epoch": 2.437003460544914,
-      "grad_norm": 2.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 150000
-    },
-    {
-      "epoch": 2.4371659274422837,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2608,
-      "step": 150010
-    },
-    {
-      "epoch": 2.4373283943396533,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 150020
-    },
-    {
-      "epoch": 2.437490861237023,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 150030
-    },
-    {
-      "epoch": 2.4376533281343926,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 150040
-    },
-    {
-      "epoch": 2.4378157950317623,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 150050
-    },
-    {
-      "epoch": 2.437978261929132,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2576,
-      "step": 150060
-    },
-    {
-      "epoch": 2.4381407288265016,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 150070
-    },
-    {
-      "epoch": 2.438303195723871,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 150080
-    },
-    {
-      "epoch": 2.438465662621241,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2596,
-      "step": 150090
-    },
-    {
-      "epoch": 2.4386281295186105,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 150100
-    },
-    {
-      "epoch": 2.43879059641598,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2561,
-      "step": 150110
-    },
-    {
-      "epoch": 2.43895306331335,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3091,
-      "step": 150120
-    },
-    {
-      "epoch": 2.4391155302107195,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 150130
-    },
-    {
-      "epoch": 2.439277997108089,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 150140
-    },
-    {
-      "epoch": 2.4394404640054588,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 150150
-    },
-    {
-      "epoch": 2.4396029309028284,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2562,
-      "step": 150160
-    },
-    {
-      "epoch": 2.439765397800198,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2719,
-      "step": 150170
-    },
-    {
-      "epoch": 2.4399278646975677,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 150180
-    },
-    {
-      "epoch": 2.4400903315949374,
-      "grad_norm": 2.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 150190
-    },
-    {
-      "epoch": 2.440252798492307,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2985,
-      "step": 150200
-    },
-    {
-      "epoch": 2.4404152653896767,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 150210
-    },
-    {
-      "epoch": 2.4405777322870463,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 150220
-    },
-    {
-      "epoch": 2.440740199184416,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 150230
-    },
-    {
-      "epoch": 2.4409026660817856,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 150240
-    },
-    {
-      "epoch": 2.4410651329791557,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.3113,
-      "step": 150250
-    },
-    {
-      "epoch": 2.441227599876525,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 150260
-    },
-    {
-      "epoch": 2.441390066773895,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2645,
-      "step": 150270
-    },
-    {
-      "epoch": 2.4415525336712642,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 150280
-    },
-    {
-      "epoch": 2.4417150005686343,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 150290
-    },
-    {
-      "epoch": 2.441877467466004,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2709,
-      "step": 150300
-    },
-    {
-      "epoch": 2.4420399343633736,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 150310
-    },
-    {
-      "epoch": 2.4422024012607433,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 150320
-    },
-    {
-      "epoch": 2.442364868158113,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 150330
-    },
-    {
-      "epoch": 2.4425273350554826,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 150340
-    },
-    {
-      "epoch": 2.4426898019528522,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 150350
-    },
-    {
-      "epoch": 2.442852268850222,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 150360
-    },
-    {
-      "epoch": 2.4430147357475915,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2931,
-      "step": 150370
-    },
-    {
-      "epoch": 2.443177202644961,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2613,
-      "step": 150380
-    },
-    {
-      "epoch": 2.443339669542331,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 150390
-    },
-    {
-      "epoch": 2.4435021364397005,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 150400
-    },
-    {
-      "epoch": 2.44366460333707,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2537,
-      "step": 150410
-    },
-    {
-      "epoch": 2.44382707023444,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2616,
-      "step": 150420
-    },
-    {
-      "epoch": 2.4439895371318094,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2545,
-      "step": 150430
-    },
-    {
-      "epoch": 2.444152004029179,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.246,
-      "step": 150440
-    },
-    {
-      "epoch": 2.4443144709265487,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 150450
-    },
-    {
-      "epoch": 2.4444769378239184,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2606,
-      "step": 150460
-    },
-    {
-      "epoch": 2.444639404721288,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 150470
-    },
-    {
-      "epoch": 2.4448018716186577,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3156,
-      "step": 150480
-    },
-    {
-      "epoch": 2.4449643385160273,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 150490
-    },
-    {
-      "epoch": 2.445126805413397,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.277,
-      "step": 150500
-    },
-    {
-      "epoch": 2.4452892723107666,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2583,
-      "step": 150510
-    },
-    {
-      "epoch": 2.4454517392081363,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 150520
-    },
-    {
-      "epoch": 2.445614206105506,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2549,
-      "step": 150530
-    },
-    {
-      "epoch": 2.4457766730028756,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 150540
-    },
-    {
-      "epoch": 2.4459391399002453,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 150550
-    },
-    {
-      "epoch": 2.446101606797615,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 150560
-    },
-    {
-      "epoch": 2.4462640736949846,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 150570
-    },
-    {
-      "epoch": 2.446426540592354,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 150580
-    },
-    {
-      "epoch": 2.446589007489724,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 150590
-    },
-    {
-      "epoch": 2.4467514743870935,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 150600
-    },
-    {
-      "epoch": 2.446913941284463,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 150610
-    },
-    {
-      "epoch": 2.447076408181833,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2663,
-      "step": 150620
-    },
-    {
-      "epoch": 2.4472388750792025,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2957,
-      "step": 150630
-    },
-    {
-      "epoch": 2.447401341976572,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 150640
-    },
-    {
-      "epoch": 2.4475638088739418,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 150650
-    },
-    {
-      "epoch": 2.4477262757713114,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3014,
-      "step": 150660
-    },
-    {
-      "epoch": 2.447888742668681,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 150670
-    },
-    {
-      "epoch": 2.448051209566051,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 150680
-    },
-    {
-      "epoch": 2.4482136764634204,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 150690
-    },
-    {
-      "epoch": 2.4483761433607905,
-      "grad_norm": 2.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 150700
-    },
-    {
-      "epoch": 2.4485386102581597,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2933,
-      "step": 150710
-    },
-    {
-      "epoch": 2.4487010771555298,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 150720
-    },
-    {
-      "epoch": 2.4488635440528994,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 150730
-    },
-    {
-      "epoch": 2.449026010950269,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2705,
-      "step": 150740
-    },
-    {
-      "epoch": 2.4491884778476387,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2573,
-      "step": 150750
-    },
-    {
-      "epoch": 2.4493509447450084,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 150760
-    },
-    {
-      "epoch": 2.449513411642378,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 150770
-    },
-    {
-      "epoch": 2.4496758785397477,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 150780
-    },
-    {
-      "epoch": 2.4498383454371173,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 150790
-    },
-    {
-      "epoch": 2.450000812334487,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2622,
-      "step": 150800
-    },
-    {
-      "epoch": 2.4501632792318566,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 150810
-    },
-    {
-      "epoch": 2.4503257461292263,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 150820
-    },
-    {
-      "epoch": 2.450488213026596,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2291,
-      "step": 150830
-    },
-    {
-      "epoch": 2.4506506799239656,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 150840
-    },
-    {
-      "epoch": 2.450813146821335,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2613,
-      "step": 150850
-    },
-    {
-      "epoch": 2.450975613718705,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2481,
-      "step": 150860
-    },
-    {
-      "epoch": 2.4511380806160745,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2559,
-      "step": 150870
-    },
-    {
-      "epoch": 2.451300547513444,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2353,
-      "step": 150880
-    },
-    {
-      "epoch": 2.451463014410814,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2476,
-      "step": 150890
-    },
-    {
-      "epoch": 2.4516254813081835,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2295,
-      "step": 150900
-    },
-    {
-      "epoch": 2.451787948205553,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 150910
-    },
-    {
-      "epoch": 2.4519504151029228,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2628,
-      "step": 150920
-    },
-    {
-      "epoch": 2.4521128820002924,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2517,
-      "step": 150930
-    },
-    {
-      "epoch": 2.452275348897662,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 150940
-    },
-    {
-      "epoch": 2.4524378157950317,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 150950
-    },
-    {
-      "epoch": 2.4526002826924014,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 150960
-    },
-    {
-      "epoch": 2.452762749589771,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2449,
-      "step": 150970
-    },
-    {
-      "epoch": 2.4529252164871407,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2508,
-      "step": 150980
-    },
-    {
-      "epoch": 2.4530876833845103,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2551,
-      "step": 150990
-    },
-    {
-      "epoch": 2.45325015028188,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2553,
-      "step": 151000
-    },
-    {
-      "epoch": 2.4534126171792496,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 151010
-    },
-    {
-      "epoch": 2.4535750840766193,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 151020
-    },
-    {
-      "epoch": 2.453737550973989,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 151030
-    },
-    {
-      "epoch": 2.4539000178713586,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 151040
-    },
-    {
-      "epoch": 2.4540624847687282,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2357,
-      "step": 151050
-    },
-    {
-      "epoch": 2.454224951666098,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 151060
-    },
-    {
-      "epoch": 2.4543874185634675,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 151070
-    },
-    {
-      "epoch": 2.454549885460837,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 151080
-    },
-    {
-      "epoch": 2.454712352358207,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 151090
-    },
-    {
-      "epoch": 2.4548748192555765,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 151100
-    },
-    {
-      "epoch": 2.4550372861529466,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 151110
-    },
-    {
-      "epoch": 2.455199753050316,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2384,
-      "step": 151120
-    },
-    {
-      "epoch": 2.455362219947686,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2634,
-      "step": 151130
-    },
-    {
-      "epoch": 2.455524686845055,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 151140
-    },
-    {
-      "epoch": 2.455687153742425,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 151150
-    },
-    {
-      "epoch": 2.4558496206397944,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 151160
-    },
-    {
-      "epoch": 2.4560120875371645,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 151170
-    },
-    {
-      "epoch": 2.456174554434534,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 151180
-    },
-    {
-      "epoch": 2.456337021331904,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 151190
-    },
-    {
-      "epoch": 2.4564994882292734,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 151200
-    },
-    {
-      "epoch": 2.456661955126643,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 151210
-    },
-    {
-      "epoch": 2.4568244220240127,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 151220
-    },
-    {
-      "epoch": 2.4569868889213824,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 151230
-    },
-    {
-      "epoch": 2.457149355818752,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 151240
-    },
-    {
-      "epoch": 2.4573118227161217,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 151250
-    },
-    {
-      "epoch": 2.4574742896134913,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 151260
-    },
-    {
-      "epoch": 2.457636756510861,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2577,
-      "step": 151270
-    },
-    {
-      "epoch": 2.4577992234082306,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2534,
-      "step": 151280
-    },
-    {
-      "epoch": 2.4579616903056003,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2557,
-      "step": 151290
-    },
-    {
-      "epoch": 2.45812415720297,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 151300
-    },
-    {
-      "epoch": 2.4582866241003396,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 151310
-    },
-    {
-      "epoch": 2.4584490909977093,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2688,
-      "step": 151320
-    },
-    {
-      "epoch": 2.458611557895079,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 151330
-    },
-    {
-      "epoch": 2.4587740247924486,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2416,
-      "step": 151340
-    },
-    {
-      "epoch": 2.458936491689818,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2709,
-      "step": 151350
-    },
-    {
-      "epoch": 2.459098958587188,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.252,
-      "step": 151360
-    },
-    {
-      "epoch": 2.4592614254845575,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 151370
-    },
-    {
-      "epoch": 2.459423892381927,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.251,
-      "step": 151380
-    },
-    {
-      "epoch": 2.459586359279297,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2634,
-      "step": 151390
-    },
-    {
-      "epoch": 2.4597488261766665,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 151400
-    },
-    {
-      "epoch": 2.459911293074036,
-      "grad_norm": 3.078125,
-      "learning_rate": 5e-05,
-      "loss": 0.2931,
-      "step": 151410
-    },
-    {
-      "epoch": 2.4600737599714058,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2793,
-      "step": 151420
-    },
-    {
-      "epoch": 2.4602362268687754,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 151430
-    },
-    {
-      "epoch": 2.460398693766145,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 151440
-    },
-    {
-      "epoch": 2.4605611606635147,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 151450
-    },
-    {
-      "epoch": 2.4607236275608844,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 151460
-    },
-    {
-      "epoch": 2.460886094458254,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 151470
-    },
-    {
-      "epoch": 2.4610485613556237,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 151480
-    },
-    {
-      "epoch": 2.4612110282529933,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 151490
-    },
-    {
-      "epoch": 2.461373495150363,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 151500
-    },
-    {
-      "epoch": 2.4615359620477326,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 151510
-    },
-    {
-      "epoch": 2.4616984289451023,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 151520
-    },
-    {
-      "epoch": 2.461860895842472,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 151530
-    },
-    {
-      "epoch": 2.4620233627398416,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 151540
-    },
-    {
-      "epoch": 2.462185829637211,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 151550
-    },
-    {
-      "epoch": 2.4623482965345813,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 151560
-    },
-    {
-      "epoch": 2.4625107634319505,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 151570
-    },
-    {
-      "epoch": 2.4626732303293206,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 151580
-    },
-    {
-      "epoch": 2.46283569722669,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.26,
-      "step": 151590
-    },
-    {
-      "epoch": 2.46299816412406,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 151600
-    },
-    {
-      "epoch": 2.4631606310214296,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 151610
-    },
-    {
-      "epoch": 2.463323097918799,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2575,
-      "step": 151620
-    },
-    {
-      "epoch": 2.463485564816169,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 151630
-    },
-    {
-      "epoch": 2.4636480317135385,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 151640
-    },
-    {
-      "epoch": 2.463810498610908,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2589,
-      "step": 151650
-    },
-    {
-      "epoch": 2.463972965508278,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 151660
-    },
-    {
-      "epoch": 2.4641354324056475,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2539,
-      "step": 151670
-    },
-    {
-      "epoch": 2.464297899303017,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 151680
-    },
-    {
-      "epoch": 2.4644603662003868,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2484,
-      "step": 151690
-    },
-    {
-      "epoch": 2.4646228330977564,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 151700
-    },
-    {
-      "epoch": 2.464785299995126,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 151710
-    },
-    {
-      "epoch": 2.4649477668924957,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2635,
-      "step": 151720
-    },
-    {
-      "epoch": 2.4651102337898654,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2991,
-      "step": 151730
-    },
-    {
-      "epoch": 2.465272700687235,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 151740
-    },
-    {
-      "epoch": 2.4654351675846047,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2547,
-      "step": 151750
-    },
-    {
-      "epoch": 2.4655976344819743,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2669,
-      "step": 151760
-    },
-    {
-      "epoch": 2.465760101379344,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.259,
-      "step": 151770
-    },
-    {
-      "epoch": 2.4659225682767136,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 151780
-    },
-    {
-      "epoch": 2.4660850351740833,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 151790
-    },
-    {
-      "epoch": 2.466247502071453,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 151800
-    },
-    {
-      "epoch": 2.4664099689688226,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2592,
-      "step": 151810
-    },
-    {
-      "epoch": 2.4665724358661922,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2776,
-      "step": 151820
-    },
-    {
-      "epoch": 2.466734902763562,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 151830
-    },
-    {
-      "epoch": 2.4668973696609315,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 151840
-    },
-    {
-      "epoch": 2.467059836558301,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.277,
-      "step": 151850
-    },
-    {
-      "epoch": 2.467222303455671,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 151860
-    },
-    {
-      "epoch": 2.4673847703530405,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 151870
-    },
-    {
-      "epoch": 2.46754723725041,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 151880
-    },
-    {
-      "epoch": 2.46770970414778,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2905,
-      "step": 151890
-    },
-    {
-      "epoch": 2.4678721710451494,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 151900
-    },
-    {
-      "epoch": 2.468034637942519,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 151910
-    },
-    {
-      "epoch": 2.4681971048398887,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 151920
-    },
-    {
-      "epoch": 2.4683595717372584,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 151930
-    },
-    {
-      "epoch": 2.468522038634628,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2539,
-      "step": 151940
-    },
-    {
-      "epoch": 2.4686845055319977,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 151950
-    },
-    {
-      "epoch": 2.4688469724293673,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 151960
-    },
-    {
-      "epoch": 2.469009439326737,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2653,
-      "step": 151970
-    },
-    {
-      "epoch": 2.4691719062241066,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2616,
-      "step": 151980
-    },
-    {
-      "epoch": 2.4693343731214767,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 151990
-    },
-    {
-      "epoch": 2.469496840018846,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2916,
-      "step": 152000
-    },
-    {
-      "epoch": 2.469659306916216,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 152010
-    },
-    {
-      "epoch": 2.4698217738135853,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 152020
-    },
-    {
-      "epoch": 2.4699842407109553,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 152030
-    },
-    {
-      "epoch": 2.4701467076083246,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 152040
-    },
-    {
-      "epoch": 2.4703091745056946,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 152050
-    },
-    {
-      "epoch": 2.4704716414030643,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 152060
-    },
-    {
-      "epoch": 2.470634108300434,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2593,
-      "step": 152070
-    },
-    {
-      "epoch": 2.4707965751978036,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 152080
-    },
-    {
-      "epoch": 2.4709590420951733,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.274,
-      "step": 152090
-    },
-    {
-      "epoch": 2.471121508992543,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 152100
-    },
-    {
-      "epoch": 2.4712839758899126,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 152110
-    },
-    {
-      "epoch": 2.471446442787282,
-      "grad_norm": 2.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2567,
-      "step": 152120
-    },
-    {
-      "epoch": 2.471608909684652,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 152130
-    },
-    {
-      "epoch": 2.4717713765820215,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3005,
-      "step": 152140
-    },
-    {
-      "epoch": 2.471933843479391,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 152150
-    },
-    {
-      "epoch": 2.472096310376761,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2531,
-      "step": 152160
-    },
-    {
-      "epoch": 2.4722587772741305,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2449,
-      "step": 152170
-    },
-    {
-      "epoch": 2.4724212441715,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 152180
-    },
-    {
-      "epoch": 2.4725837110688698,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2645,
-      "step": 152190
-    },
-    {
-      "epoch": 2.4727461779662394,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2474,
-      "step": 152200
-    },
-    {
-      "epoch": 2.472908644863609,
-      "grad_norm": 3.078125,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 152210
-    },
-    {
-      "epoch": 2.4730711117609787,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2482,
-      "step": 152220
-    },
-    {
-      "epoch": 2.4732335786583484,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 152230
-    },
-    {
-      "epoch": 2.473396045555718,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 152240
-    },
-    {
-      "epoch": 2.4735585124530877,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2535,
-      "step": 152250
-    },
-    {
-      "epoch": 2.4737209793504573,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 152260
-    },
-    {
-      "epoch": 2.473883446247827,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 152270
-    },
-    {
-      "epoch": 2.4740459131451966,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 152280
-    },
-    {
-      "epoch": 2.4742083800425663,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 152290
-    },
-    {
-      "epoch": 2.474370846939936,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2695,
-      "step": 152300
-    },
-    {
-      "epoch": 2.4745333138373056,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 152310
-    },
-    {
-      "epoch": 2.474695780734675,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 152320
-    },
-    {
-      "epoch": 2.474858247632045,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 152330
-    },
-    {
-      "epoch": 2.4750207145294145,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2685,
-      "step": 152340
-    },
-    {
-      "epoch": 2.475183181426784,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 152350
-    },
-    {
-      "epoch": 2.475345648324154,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 152360
-    },
-    {
-      "epoch": 2.4755081152215235,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 152370
-    },
-    {
-      "epoch": 2.475670582118893,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 152380
-    },
-    {
-      "epoch": 2.4758330490162628,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2583,
-      "step": 152390
-    },
-    {
-      "epoch": 2.4759955159136324,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 152400
-    },
-    {
-      "epoch": 2.476157982811002,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 152410
-    },
-    {
-      "epoch": 2.4763204497083717,
-      "grad_norm": 3.015625,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 152420
-    },
-    {
-      "epoch": 2.4764829166057414,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 152430
-    },
-    {
-      "epoch": 2.4766453835031115,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2474,
-      "step": 152440
-    },
-    {
-      "epoch": 2.4768078504004807,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 152450
-    },
-    {
-      "epoch": 2.4769703172978508,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2565,
-      "step": 152460
-    },
-    {
-      "epoch": 2.47713278419522,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2588,
-      "step": 152470
-    },
-    {
-      "epoch": 2.47729525109259,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2568,
-      "step": 152480
-    },
-    {
-      "epoch": 2.4774577179899597,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2718,
-      "step": 152490
-    },
-    {
-      "epoch": 2.4776201848873294,
-      "grad_norm": 2.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 152500
-    },
-    {
-      "epoch": 2.477782651784699,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 152510
-    },
-    {
-      "epoch": 2.4779451186820687,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 152520
-    },
-    {
-      "epoch": 2.4781075855794383,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 152530
-    },
-    {
-      "epoch": 2.478270052476808,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 152540
-    },
-    {
-      "epoch": 2.4784325193741776,
-      "grad_norm": 2.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 152550
-    },
-    {
-      "epoch": 2.4785949862715473,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 152560
-    },
-    {
-      "epoch": 2.478757453168917,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 152570
-    },
-    {
-      "epoch": 2.4789199200662866,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 152580
-    },
-    {
-      "epoch": 2.4790823869636562,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2624,
-      "step": 152590
-    },
-    {
-      "epoch": 2.479244853861026,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2517,
-      "step": 152600
-    },
-    {
-      "epoch": 2.4794073207583955,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.247,
-      "step": 152610
-    },
-    {
-      "epoch": 2.479569787655765,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.226,
-      "step": 152620
-    },
-    {
-      "epoch": 2.479732254553135,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2509,
-      "step": 152630
-    },
-    {
-      "epoch": 2.4798947214505045,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 152640
-    },
-    {
-      "epoch": 2.480057188347874,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2489,
-      "step": 152650
-    },
-    {
-      "epoch": 2.480219655245244,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 152660
-    },
-    {
-      "epoch": 2.4803821221426134,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2735,
-      "step": 152670
-    },
-    {
-      "epoch": 2.480544589039983,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2613,
-      "step": 152680
-    },
-    {
-      "epoch": 2.4807070559373527,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 152690
-    },
-    {
-      "epoch": 2.4808695228347224,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 152700
-    },
-    {
-      "epoch": 2.481031989732092,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 152710
-    },
-    {
-      "epoch": 2.4811944566294617,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2653,
-      "step": 152720
-    },
-    {
-      "epoch": 2.4813569235268313,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 152730
-    },
-    {
-      "epoch": 2.481519390424201,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 152740
-    },
-    {
-      "epoch": 2.4816818573215707,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 152750
-    },
-    {
-      "epoch": 2.4818443242189403,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 152760
-    },
-    {
-      "epoch": 2.48200679111631,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 152770
-    },
-    {
-      "epoch": 2.4821692580136796,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 152780
-    },
-    {
-      "epoch": 2.4823317249110493,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 152790
-    },
-    {
-      "epoch": 2.482494191808419,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 152800
-    },
-    {
-      "epoch": 2.4826566587057886,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 152810
-    },
-    {
-      "epoch": 2.482819125603158,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 152820
-    },
-    {
-      "epoch": 2.482981592500528,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 152830
-    },
-    {
-      "epoch": 2.4831440593978975,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2974,
-      "step": 152840
-    },
-    {
-      "epoch": 2.483306526295267,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 152850
-    },
-    {
-      "epoch": 2.483468993192637,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 152860
-    },
-    {
-      "epoch": 2.483631460090007,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 152870
-    },
-    {
-      "epoch": 2.483793926987376,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 152880
-    },
-    {
-      "epoch": 2.483956393884746,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 152890
-    },
-    {
-      "epoch": 2.4841188607821154,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 152900
-    },
-    {
-      "epoch": 2.4842813276794855,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 152910
-    },
-    {
-      "epoch": 2.484443794576855,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2652,
-      "step": 152920
-    },
-    {
-      "epoch": 2.484606261474225,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 152930
-    },
-    {
-      "epoch": 2.4847687283715945,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 152940
-    },
-    {
-      "epoch": 2.484931195268964,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 152950
-    },
-    {
-      "epoch": 2.4850936621663338,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 152960
-    },
-    {
-      "epoch": 2.4852561290637034,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 152970
-    },
-    {
-      "epoch": 2.485418595961073,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 152980
-    },
-    {
-      "epoch": 2.4855810628584427,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 152990
-    },
-    {
-      "epoch": 2.4857435297558124,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 153000
-    },
-    {
-      "epoch": 2.485905996653182,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2988,
-      "step": 153010
-    },
-    {
-      "epoch": 2.4860684635505517,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2964,
-      "step": 153020
-    },
-    {
-      "epoch": 2.4862309304479213,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 153030
-    },
-    {
-      "epoch": 2.486393397345291,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 153040
-    },
-    {
-      "epoch": 2.4865558642426606,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3005,
-      "step": 153050
-    },
-    {
-      "epoch": 2.4867183311400303,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 153060
-    },
-    {
-      "epoch": 2.4868807980374,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 153070
-    },
-    {
-      "epoch": 2.4870432649347696,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2902,
-      "step": 153080
-    },
-    {
-      "epoch": 2.487205731832139,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 153090
-    },
-    {
-      "epoch": 2.487368198729509,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 153100
-    },
-    {
-      "epoch": 2.4875306656268785,
-      "grad_norm": 2.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2647,
-      "step": 153110
-    },
-    {
-      "epoch": 2.487693132524248,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2919,
-      "step": 153120
-    },
-    {
-      "epoch": 2.487855599421618,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 153130
-    },
-    {
-      "epoch": 2.4880180663189875,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 153140
-    },
-    {
-      "epoch": 2.488180533216357,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3058,
-      "step": 153150
-    },
-    {
-      "epoch": 2.4883430001137268,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 153160
-    },
-    {
-      "epoch": 2.4885054670110964,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 153170
-    },
-    {
-      "epoch": 2.488667933908466,
-      "grad_norm": 2.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 153180
-    },
-    {
-      "epoch": 2.4888304008058357,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 153190
-    },
-    {
-      "epoch": 2.4889928677032054,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 153200
-    },
-    {
-      "epoch": 2.489155334600575,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2534,
-      "step": 153210
-    },
-    {
-      "epoch": 2.4893178014979447,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.3083,
-      "step": 153220
-    },
-    {
-      "epoch": 2.4894802683953143,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 153230
-    },
-    {
-      "epoch": 2.489642735292684,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 153240
-    },
-    {
-      "epoch": 2.4898052021900536,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2698,
-      "step": 153250
-    },
-    {
-      "epoch": 2.4899676690874233,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 153260
-    },
-    {
-      "epoch": 2.490130135984793,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 153270
-    },
-    {
-      "epoch": 2.4902926028821626,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 153280
-    },
-    {
-      "epoch": 2.4904550697795322,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 153290
-    },
-    {
-      "epoch": 2.490617536676902,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 153300
-    },
-    {
-      "epoch": 2.4907800035742715,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 153310
-    },
-    {
-      "epoch": 2.4909424704716416,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2916,
-      "step": 153320
-    },
-    {
-      "epoch": 2.491104937369011,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 153330
-    },
-    {
-      "epoch": 2.491267404266381,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2621,
-      "step": 153340
-    },
-    {
-      "epoch": 2.49142987116375,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 153350
-    },
-    {
-      "epoch": 2.4915923380611202,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 153360
-    },
-    {
-      "epoch": 2.49175480495849,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 153370
-    },
-    {
-      "epoch": 2.4919172718558595,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2939,
-      "step": 153380
-    },
-    {
-      "epoch": 2.492079738753229,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 153390
-    },
-    {
-      "epoch": 2.492242205650599,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 153400
-    },
-    {
-      "epoch": 2.4924046725479685,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 153410
-    },
-    {
-      "epoch": 2.492567139445338,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 153420
-    },
-    {
-      "epoch": 2.492729606342708,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3017,
-      "step": 153430
-    },
-    {
-      "epoch": 2.4928920732400774,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 153440
-    },
-    {
-      "epoch": 2.493054540137447,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2718,
-      "step": 153450
-    },
-    {
-      "epoch": 2.4932170070348167,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 153460
-    },
-    {
-      "epoch": 2.4933794739321864,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 153470
-    },
-    {
-      "epoch": 2.493541940829556,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 153480
-    },
-    {
-      "epoch": 2.4937044077269257,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2377,
-      "step": 153490
-    },
-    {
-      "epoch": 2.4938668746242953,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 153500
-    },
-    {
-      "epoch": 2.494029341521665,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 153510
-    },
-    {
-      "epoch": 2.4941918084190347,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 153520
-    },
-    {
-      "epoch": 2.4943542753164043,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 153530
-    },
-    {
-      "epoch": 2.494516742213774,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.3018,
-      "step": 153540
-    },
-    {
-      "epoch": 2.4946792091111436,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 153550
-    },
-    {
-      "epoch": 2.4948416760085133,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 153560
-    },
-    {
-      "epoch": 2.495004142905883,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3021,
-      "step": 153570
-    },
-    {
-      "epoch": 2.4951666098032526,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 153580
-    },
-    {
-      "epoch": 2.495329076700622,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 153590
-    },
-    {
-      "epoch": 2.495491543597992,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 153600
-    },
-    {
-      "epoch": 2.4956540104953615,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 153610
-    },
-    {
-      "epoch": 2.495816477392731,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 153620
-    },
-    {
-      "epoch": 2.495978944290101,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 153630
-    },
-    {
-      "epoch": 2.4961414111874705,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 153640
-    },
-    {
-      "epoch": 2.49630387808484,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 153650
-    },
-    {
-      "epoch": 2.4964663449822098,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2529,
-      "step": 153660
-    },
-    {
-      "epoch": 2.4966288118795794,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2585,
-      "step": 153670
-    },
-    {
-      "epoch": 2.496791278776949,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 153680
-    },
-    {
-      "epoch": 2.4969537456743187,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 153690
-    },
-    {
-      "epoch": 2.4971162125716884,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 153700
-    },
-    {
-      "epoch": 2.497278679469058,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 153710
-    },
-    {
-      "epoch": 2.4974411463664277,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2503,
-      "step": 153720
-    },
-    {
-      "epoch": 2.4976036132637973,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2593,
-      "step": 153730
-    },
-    {
-      "epoch": 2.497766080161167,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 153740
-    },
-    {
-      "epoch": 2.497928547058537,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 153750
-    },
-    {
-      "epoch": 2.4980910139559063,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2663,
-      "step": 153760
-    },
-    {
-      "epoch": 2.4982534808532764,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2613,
-      "step": 153770
-    },
-    {
-      "epoch": 2.4984159477506456,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 153780
-    },
-    {
-      "epoch": 2.4985784146480157,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 153790
-    },
-    {
-      "epoch": 2.4987408815453853,
-      "grad_norm": 3.046875,
-      "learning_rate": 5e-05,
-      "loss": 0.258,
-      "step": 153800
-    },
-    {
-      "epoch": 2.498903348442755,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 153810
-    },
-    {
-      "epoch": 2.4990658153401246,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 153820
-    },
-    {
-      "epoch": 2.4992282822374943,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 153830
-    },
-    {
-      "epoch": 2.499390749134864,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 153840
-    },
-    {
-      "epoch": 2.4995532160322336,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 153850
-    },
-    {
-      "epoch": 2.4997156829296032,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 153860
-    },
-    {
-      "epoch": 2.499878149826973,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 153870
-    },
-    {
-      "epoch": 2.5000406167243425,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 153880
-    },
-    {
-      "epoch": 2.500203083621712,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 153890
-    },
-    {
-      "epoch": 2.500365550519082,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2989,
-      "step": 153900
-    },
-    {
-      "epoch": 2.5005280174164515,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 153910
-    },
-    {
-      "epoch": 2.500690484313821,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.3091,
-      "step": 153920
-    },
-    {
-      "epoch": 2.5008529512111908,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 153930
-    },
-    {
-      "epoch": 2.5010154181085604,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2685,
-      "step": 153940
-    },
-    {
-      "epoch": 2.50117788500593,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2707,
-      "step": 153950
-    },
-    {
-      "epoch": 2.5013403519032997,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 153960
-    },
-    {
-      "epoch": 2.5015028188006694,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 153970
-    },
-    {
-      "epoch": 2.501665285698039,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2532,
-      "step": 153980
-    },
-    {
-      "epoch": 2.5018277525954087,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2688,
-      "step": 153990
-    },
-    {
-      "epoch": 2.5019902194927783,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2541,
-      "step": 154000
-    },
-    {
-      "epoch": 2.502152686390148,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 154010
-    },
-    {
-      "epoch": 2.5023151532875176,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3113,
-      "step": 154020
-    },
-    {
-      "epoch": 2.5024776201848873,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2955,
-      "step": 154030
-    },
-    {
-      "epoch": 2.502640087082257,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 154040
-    },
-    {
-      "epoch": 2.5028025539796266,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 154050
-    },
-    {
-      "epoch": 2.5029650208769962,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2705,
-      "step": 154060
-    },
-    {
-      "epoch": 2.503127487774366,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 154070
-    },
-    {
-      "epoch": 2.5032899546717355,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2702,
-      "step": 154080
-    },
-    {
-      "epoch": 2.503452421569105,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2638,
-      "step": 154090
-    },
-    {
-      "epoch": 2.503614888466475,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 154100
-    },
-    {
-      "epoch": 2.5037773553638445,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 154110
-    },
-    {
-      "epoch": 2.503939822261214,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2651,
-      "step": 154120
-    },
-    {
-      "epoch": 2.504102289158584,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 154130
-    },
-    {
-      "epoch": 2.5042647560559534,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2638,
-      "step": 154140
-    },
-    {
-      "epoch": 2.504427222953323,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 154150
-    },
-    {
-      "epoch": 2.504589689850693,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2418,
-      "step": 154160
-    },
-    {
-      "epoch": 2.5047521567480624,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2468,
-      "step": 154170
-    },
-    {
-      "epoch": 2.5049146236454325,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 154180
-    },
-    {
-      "epoch": 2.5050770905428017,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2588,
-      "step": 154190
-    },
-    {
-      "epoch": 2.505239557440172,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 154200
-    },
-    {
-      "epoch": 2.505402024337541,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2567,
-      "step": 154210
-    },
-    {
-      "epoch": 2.505564491234911,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 154220
-    },
-    {
-      "epoch": 2.5057269581322803,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 154230
-    },
-    {
-      "epoch": 2.5058894250296504,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 154240
-    },
-    {
-      "epoch": 2.5060518919270196,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 154250
-    },
-    {
-      "epoch": 2.5062143588243897,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 154260
-    },
-    {
-      "epoch": 2.5063768257217593,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 154270
-    },
-    {
-      "epoch": 2.506539292619129,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 154280
-    },
-    {
-      "epoch": 2.5067017595164987,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 154290
-    },
-    {
-      "epoch": 2.5068642264138683,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.297,
-      "step": 154300
-    },
-    {
-      "epoch": 2.507026693311238,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 154310
-    },
-    {
-      "epoch": 2.5071891602086076,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 154320
-    },
-    {
-      "epoch": 2.5073516271059773,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2635,
-      "step": 154330
-    },
-    {
-      "epoch": 2.507514094003347,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 154340
-    },
-    {
-      "epoch": 2.5076765609007166,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2516,
-      "step": 154350
-    },
-    {
-      "epoch": 2.507839027798086,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2681,
-      "step": 154360
-    },
-    {
-      "epoch": 2.508001494695456,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3062,
-      "step": 154370
-    },
-    {
-      "epoch": 2.5081639615928255,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 154380
-    },
-    {
-      "epoch": 2.508326428490195,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 154390
-    },
-    {
-      "epoch": 2.508488895387565,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 154400
-    },
-    {
-      "epoch": 2.5086513622849345,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2487,
-      "step": 154410
-    },
-    {
-      "epoch": 2.508813829182304,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2552,
-      "step": 154420
-    },
-    {
-      "epoch": 2.5089762960796738,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2442,
-      "step": 154430
-    },
-    {
-      "epoch": 2.5091387629770434,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2522,
-      "step": 154440
-    },
-    {
-      "epoch": 2.509301229874413,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 154450
-    },
-    {
-      "epoch": 2.5094636967717827,
-      "grad_norm": 2.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 154460
-    },
-    {
-      "epoch": 2.5096261636691524,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 154470
-    },
-    {
-      "epoch": 2.509788630566522,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 154480
-    },
-    {
-      "epoch": 2.5099510974638917,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2647,
-      "step": 154490
-    },
-    {
-      "epoch": 2.5101135643612613,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 154500
-    },
-    {
-      "epoch": 2.510276031258631,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2495,
-      "step": 154510
-    },
-    {
-      "epoch": 2.5104384981560006,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 154520
-    },
-    {
-      "epoch": 2.5106009650533703,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 154530
-    },
-    {
-      "epoch": 2.51076343195074,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 154540
-    },
-    {
-      "epoch": 2.5109258988481096,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 154550
-    },
-    {
-      "epoch": 2.5110883657454792,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 154560
-    },
-    {
-      "epoch": 2.511250832642849,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 154570
-    },
-    {
-      "epoch": 2.5114132995402185,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 154580
-    },
-    {
-      "epoch": 2.511575766437588,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 154590
-    },
-    {
-      "epoch": 2.511738233334958,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2484,
-      "step": 154600
-    },
-    {
-      "epoch": 2.511900700232328,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2641,
-      "step": 154610
-    },
-    {
-      "epoch": 2.512063167129697,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 154620
-    },
-    {
-      "epoch": 2.5122256340270672,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2685,
-      "step": 154630
-    },
-    {
-      "epoch": 2.5123881009244364,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 154640
-    },
-    {
-      "epoch": 2.5125505678218065,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 154650
-    },
-    {
-      "epoch": 2.5127130347191757,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 154660
-    },
-    {
-      "epoch": 2.512875501616546,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3001,
-      "step": 154670
-    },
-    {
-      "epoch": 2.513037968513915,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 154680
-    },
-    {
-      "epoch": 2.513200435411285,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 154690
-    },
-    {
-      "epoch": 2.5133629023086543,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2718,
-      "step": 154700
-    },
-    {
-      "epoch": 2.5135253692060244,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 154710
-    },
-    {
-      "epoch": 2.513687836103394,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 154720
-    },
-    {
-      "epoch": 2.5138503030007637,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2721,
-      "step": 154730
-    },
-    {
-      "epoch": 2.5140127698981334,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 154740
-    },
-    {
-      "epoch": 2.514175236795503,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 154750
-    },
-    {
-      "epoch": 2.5143377036928727,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 154760
-    },
-    {
-      "epoch": 2.5145001705902423,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 154770
-    },
-    {
-      "epoch": 2.514662637487612,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 154780
-    },
-    {
-      "epoch": 2.5148251043849816,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2712,
-      "step": 154790
-    },
-    {
-      "epoch": 2.5149875712823513,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2986,
-      "step": 154800
-    },
-    {
-      "epoch": 2.515150038179721,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 154810
-    },
-    {
-      "epoch": 2.5153125050770906,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 154820
-    },
-    {
-      "epoch": 2.5154749719744602,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 154830
-    },
-    {
-      "epoch": 2.51563743887183,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 154840
-    },
-    {
-      "epoch": 2.5157999057691995,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3083,
-      "step": 154850
-    },
-    {
-      "epoch": 2.515962372666569,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 154860
-    },
-    {
-      "epoch": 2.516124839563939,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 154870
-    },
-    {
-      "epoch": 2.5162873064613085,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 154880
-    },
-    {
-      "epoch": 2.516449773358678,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 154890
-    },
-    {
-      "epoch": 2.516612240256048,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 154900
-    },
-    {
-      "epoch": 2.5167747071534174,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2673,
-      "step": 154910
-    },
-    {
-      "epoch": 2.516937174050787,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 154920
-    },
-    {
-      "epoch": 2.5170996409481567,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 154930
-    },
-    {
-      "epoch": 2.5172621078455264,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 154940
-    },
-    {
-      "epoch": 2.517424574742896,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2969,
-      "step": 154950
-    },
-    {
-      "epoch": 2.5175870416402657,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 154960
-    },
-    {
-      "epoch": 2.5177495085376354,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 154970
-    },
-    {
-      "epoch": 2.517911975435005,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.301,
-      "step": 154980
-    },
-    {
-      "epoch": 2.5180744423323747,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 154990
-    },
-    {
-      "epoch": 2.5182369092297443,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 155000
-    },
-    {
-      "epoch": 2.518399376127114,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 155010
-    },
-    {
-      "epoch": 2.5185618430244836,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 155020
-    },
-    {
-      "epoch": 2.5187243099218533,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.264,
-      "step": 155030
-    },
-    {
-      "epoch": 2.5188867768192233,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2698,
-      "step": 155040
-    },
-    {
-      "epoch": 2.5190492437165926,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 155050
-    },
-    {
-      "epoch": 2.5192117106139627,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2432,
-      "step": 155060
-    },
-    {
-      "epoch": 2.519374177511332,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2547,
-      "step": 155070
-    },
-    {
-      "epoch": 2.519536644408702,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 155080
-    },
-    {
-      "epoch": 2.519699111306071,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2686,
-      "step": 155090
-    },
-    {
-      "epoch": 2.5198615782034413,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 155100
-    },
-    {
-      "epoch": 2.5200240451008105,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2625,
-      "step": 155110
-    },
-    {
-      "epoch": 2.5201865119981806,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 155120
-    },
-    {
-      "epoch": 2.5203489788955498,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2537,
-      "step": 155130
-    },
-    {
-      "epoch": 2.52051144579292,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 155140
-    },
-    {
-      "epoch": 2.5206739126902895,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2527,
-      "step": 155150
-    },
-    {
-      "epoch": 2.520836379587659,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2517,
-      "step": 155160
-    },
-    {
-      "epoch": 2.520998846485029,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 155170
-    },
-    {
-      "epoch": 2.5211613133823985,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2608,
-      "step": 155180
-    },
-    {
-      "epoch": 2.521323780279768,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 155190
-    },
-    {
-      "epoch": 2.5214862471771378,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 155200
-    },
-    {
-      "epoch": 2.5216487140745074,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 155210
-    },
-    {
-      "epoch": 2.521811180971877,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 155220
-    },
-    {
-      "epoch": 2.5219736478692467,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 155230
-    },
-    {
-      "epoch": 2.5221361147666164,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 155240
-    },
-    {
-      "epoch": 2.522298581663986,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2617,
-      "step": 155250
-    },
-    {
-      "epoch": 2.5224610485613557,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 155260
-    },
-    {
-      "epoch": 2.5226235154587253,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 155270
-    },
-    {
-      "epoch": 2.522785982356095,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 155280
-    },
-    {
-      "epoch": 2.5229484492534646,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2624,
-      "step": 155290
-    },
-    {
-      "epoch": 2.5231109161508343,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2558,
-      "step": 155300
-    },
-    {
-      "epoch": 2.523273383048204,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 155310
-    },
-    {
-      "epoch": 2.5234358499455736,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 155320
-    },
-    {
-      "epoch": 2.5235983168429432,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 155330
-    },
-    {
-      "epoch": 2.523760783740313,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2516,
-      "step": 155340
-    },
-    {
-      "epoch": 2.5239232506376825,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2538,
-      "step": 155350
-    },
-    {
-      "epoch": 2.524085717535052,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2449,
-      "step": 155360
-    },
-    {
-      "epoch": 2.524248184432422,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2495,
-      "step": 155370
-    },
-    {
-      "epoch": 2.5244106513297915,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 155380
-    },
-    {
-      "epoch": 2.524573118227161,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 155390
-    },
-    {
-      "epoch": 2.524735585124531,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 155400
-    },
-    {
-      "epoch": 2.5248980520219004,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 155410
-    },
-    {
-      "epoch": 2.52506051891927,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 155420
-    },
-    {
-      "epoch": 2.5252229858166397,
-      "grad_norm": 2.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2568,
-      "step": 155430
-    },
-    {
-      "epoch": 2.5253854527140094,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 155440
-    },
-    {
-      "epoch": 2.525547919611379,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 155450
-    },
-    {
-      "epoch": 2.5257103865087487,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.2541,
-      "step": 155460
-    },
-    {
-      "epoch": 2.525872853406119,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 155470
-    },
-    {
-      "epoch": 2.526035320303488,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 155480
-    },
-    {
-      "epoch": 2.526197787200858,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 155490
-    },
-    {
-      "epoch": 2.5263602540982273,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2971,
-      "step": 155500
-    },
-    {
-      "epoch": 2.5265227209955974,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 155510
-    },
-    {
-      "epoch": 2.5266851878929666,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 155520
-    },
-    {
-      "epoch": 2.5268476547903367,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 155530
-    },
-    {
-      "epoch": 2.527010121687706,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3101,
-      "step": 155540
-    },
-    {
-      "epoch": 2.527172588585076,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 155550
-    },
-    {
-      "epoch": 2.527335055482445,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 155560
-    },
-    {
-      "epoch": 2.5274975223798153,
-      "grad_norm": 2.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 155570
-    },
-    {
-      "epoch": 2.527659989277185,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 155580
-    },
-    {
-      "epoch": 2.5278224561745546,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2526,
-      "step": 155590
-    },
-    {
-      "epoch": 2.5279849230719242,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2905,
-      "step": 155600
-    },
-    {
-      "epoch": 2.528147389969294,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 155610
-    },
-    {
-      "epoch": 2.5283098568666635,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 155620
-    },
-    {
-      "epoch": 2.528472323764033,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 155630
-    },
-    {
-      "epoch": 2.528634790661403,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 155640
-    },
-    {
-      "epoch": 2.5287972575587725,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 155650
-    },
-    {
-      "epoch": 2.528959724456142,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 155660
-    },
-    {
-      "epoch": 2.529122191353512,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 155670
-    },
-    {
-      "epoch": 2.5292846582508814,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 155680
-    },
-    {
-      "epoch": 2.529447125148251,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 155690
-    },
-    {
-      "epoch": 2.5296095920456207,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 155700
-    },
-    {
-      "epoch": 2.5297720589429904,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.2576,
-      "step": 155710
-    },
-    {
-      "epoch": 2.52993452584036,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2456,
-      "step": 155720
-    },
-    {
-      "epoch": 2.5300969927377297,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 155730
-    },
-    {
-      "epoch": 2.5302594596350994,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 155740
-    },
-    {
-      "epoch": 2.530421926532469,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2765,
-      "step": 155750
-    },
-    {
-      "epoch": 2.5305843934298387,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2584,
-      "step": 155760
-    },
-    {
-      "epoch": 2.5307468603272083,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2515,
-      "step": 155770
-    },
-    {
-      "epoch": 2.530909327224578,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 155780
-    },
-    {
-      "epoch": 2.5310717941219476,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 155790
-    },
-    {
-      "epoch": 2.5312342610193173,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2712,
-      "step": 155800
-    },
-    {
-      "epoch": 2.531396727916687,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2721,
-      "step": 155810
-    },
-    {
-      "epoch": 2.5315591948140566,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 155820
-    },
-    {
-      "epoch": 2.531721661711426,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 155830
-    },
-    {
-      "epoch": 2.531884128608796,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 155840
-    },
-    {
-      "epoch": 2.5320465955061655,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 155850
-    },
-    {
-      "epoch": 2.532209062403535,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 155860
-    },
-    {
-      "epoch": 2.532371529300905,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 155870
-    },
-    {
-      "epoch": 2.5325339961982745,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2818,
-      "step": 155880
-    },
-    {
-      "epoch": 2.532696463095644,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 155890
-    },
-    {
-      "epoch": 2.5328589299930138,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 155900
-    },
-    {
-      "epoch": 2.5330213968903834,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2468,
-      "step": 155910
-    },
-    {
-      "epoch": 2.5331838637877535,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 155920
-    },
-    {
-      "epoch": 2.5333463306851227,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2673,
-      "step": 155930
-    },
-    {
-      "epoch": 2.533508797582493,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 155940
-    },
-    {
-      "epoch": 2.533671264479862,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 155950
-    },
-    {
-      "epoch": 2.533833731377232,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2671,
-      "step": 155960
-    },
-    {
-      "epoch": 2.5339961982746013,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 155970
-    },
-    {
-      "epoch": 2.5341586651719714,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 155980
-    },
-    {
-      "epoch": 2.5343211320693406,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 155990
-    },
-    {
-      "epoch": 2.5344835989667107,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2544,
-      "step": 156000
-    },
-    {
-      "epoch": 2.53464606586408,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2552,
-      "step": 156010
-    },
-    {
-      "epoch": 2.53480853276145,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2648,
-      "step": 156020
-    },
-    {
-      "epoch": 2.5349709996588197,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2423,
-      "step": 156030
-    },
-    {
-      "epoch": 2.5351334665561893,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2562,
-      "step": 156040
-    },
-    {
-      "epoch": 2.535295933453559,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 156050
-    },
-    {
-      "epoch": 2.5354584003509286,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 156060
-    },
-    {
-      "epoch": 2.5356208672482983,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 156070
-    },
-    {
-      "epoch": 2.535783334145668,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3115,
-      "step": 156080
-    },
-    {
-      "epoch": 2.5359458010430376,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 156090
-    },
-    {
-      "epoch": 2.5361082679404072,
-      "grad_norm": 2.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2577,
-      "step": 156100
-    },
-    {
-      "epoch": 2.536270734837777,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2964,
-      "step": 156110
-    },
-    {
-      "epoch": 2.5364332017351465,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 156120
-    },
-    {
-      "epoch": 2.536595668632516,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 156130
-    },
-    {
-      "epoch": 2.536758135529886,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 156140
-    },
-    {
-      "epoch": 2.5369206024272555,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 156150
-    },
-    {
-      "epoch": 2.537083069324625,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 156160
-    },
-    {
-      "epoch": 2.537245536221995,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2497,
-      "step": 156170
-    },
-    {
-      "epoch": 2.5374080031193644,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 156180
-    },
-    {
-      "epoch": 2.537570470016734,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 156190
-    },
-    {
-      "epoch": 2.5377329369141037,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 156200
-    },
-    {
-      "epoch": 2.5378954038114734,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2518,
-      "step": 156210
-    },
-    {
-      "epoch": 2.538057870708843,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2606,
-      "step": 156220
-    },
-    {
-      "epoch": 2.5382203376062127,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2627,
-      "step": 156230
-    },
-    {
-      "epoch": 2.5383828045035823,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2591,
-      "step": 156240
-    },
-    {
-      "epoch": 2.538545271400952,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 156250
-    },
-    {
-      "epoch": 2.5387077382983216,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 156260
-    },
-    {
-      "epoch": 2.5388702051956913,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 156270
-    },
-    {
-      "epoch": 2.539032672093061,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 156280
-    },
-    {
-      "epoch": 2.5391951389904306,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 156290
-    },
-    {
-      "epoch": 2.5393576058878002,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2939,
-      "step": 156300
-    },
-    {
-      "epoch": 2.53952007278517,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 156310
-    },
-    {
-      "epoch": 2.5396825396825395,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 156320
-    },
-    {
-      "epoch": 2.539845006579909,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 156330
-    },
-    {
-      "epoch": 2.540007473477279,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2956,
-      "step": 156340
-    },
-    {
-      "epoch": 2.540169940374649,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 156350
-    },
-    {
-      "epoch": 2.540332407272018,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 156360
-    },
-    {
-      "epoch": 2.5404948741693882,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2431,
-      "step": 156370
-    },
-    {
-      "epoch": 2.5406573410667574,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2475,
-      "step": 156380
-    },
-    {
-      "epoch": 2.5408198079641275,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 156390
-    },
-    {
-      "epoch": 2.5409822748614967,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2561,
-      "step": 156400
-    },
-    {
-      "epoch": 2.541144741758867,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 156410
-    },
-    {
-      "epoch": 2.541307208656236,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2934,
-      "step": 156420
-    },
-    {
-      "epoch": 2.541469675553606,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 156430
-    },
-    {
-      "epoch": 2.5416321424509754,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 156440
-    },
-    {
-      "epoch": 2.5417946093483454,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 156450
-    },
-    {
-      "epoch": 2.541957076245715,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 156460
-    },
-    {
-      "epoch": 2.5421195431430847,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 156470
-    },
-    {
-      "epoch": 2.5422820100404544,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 156480
-    },
-    {
-      "epoch": 2.542444476937824,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 156490
-    },
-    {
-      "epoch": 2.5426069438351937,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 156500
-    },
-    {
-      "epoch": 2.5427694107325634,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2515,
-      "step": 156510
-    },
-    {
-      "epoch": 2.542931877629933,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 156520
-    },
-    {
-      "epoch": 2.5430943445273027,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 156530
-    },
-    {
-      "epoch": 2.5432568114246723,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 156540
-    },
-    {
-      "epoch": 2.543419278322042,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 156550
-    },
-    {
-      "epoch": 2.5435817452194116,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 156560
-    },
-    {
-      "epoch": 2.5437442121167813,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 156570
-    },
-    {
-      "epoch": 2.543906679014151,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 156580
-    },
-    {
-      "epoch": 2.5440691459115206,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2657,
-      "step": 156590
-    },
-    {
-      "epoch": 2.54423161280889,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 156600
-    },
-    {
-      "epoch": 2.54439407970626,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2702,
-      "step": 156610
-    },
-    {
-      "epoch": 2.5445565466036295,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 156620
-    },
-    {
-      "epoch": 2.544719013500999,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2477,
-      "step": 156630
-    },
-    {
-      "epoch": 2.544881480398369,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 156640
-    },
-    {
-      "epoch": 2.5450439472957385,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 156650
-    },
-    {
-      "epoch": 2.545206414193108,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 156660
-    },
-    {
-      "epoch": 2.5453688810904778,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 156670
-    },
-    {
-      "epoch": 2.5455313479878474,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2781,
-      "step": 156680
-    },
-    {
-      "epoch": 2.545693814885217,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 156690
-    },
-    {
-      "epoch": 2.5458562817825867,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 156700
-    },
-    {
-      "epoch": 2.5460187486799564,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 156710
-    },
-    {
-      "epoch": 2.546181215577326,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3148,
-      "step": 156720
-    },
-    {
-      "epoch": 2.5463436824746957,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 156730
-    },
-    {
-      "epoch": 2.5465061493720653,
-      "grad_norm": 3.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 156740
-    },
-    {
-      "epoch": 2.546668616269435,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 156750
-    },
-    {
-      "epoch": 2.5468310831668046,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 156760
-    },
-    {
-      "epoch": 2.5469935500641743,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 156770
-    },
-    {
-      "epoch": 2.547156016961544,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.3027,
-      "step": 156780
-    },
-    {
-      "epoch": 2.5473184838589136,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2442,
-      "step": 156790
-    },
-    {
-      "epoch": 2.5474809507562837,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 156800
-    },
-    {
-      "epoch": 2.547643417653653,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2595,
-      "step": 156810
-    },
-    {
-      "epoch": 2.547805884551023,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 156820
-    },
-    {
-      "epoch": 2.547968351448392,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 156830
-    },
-    {
-      "epoch": 2.5481308183457623,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 156840
-    },
-    {
-      "epoch": 2.5482932852431315,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.291,
-      "step": 156850
-    },
-    {
-      "epoch": 2.5484557521405016,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 156860
-    },
-    {
-      "epoch": 2.548618219037871,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2608,
-      "step": 156870
-    },
-    {
-      "epoch": 2.548780685935241,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 156880
-    },
-    {
-      "epoch": 2.54894315283261,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 156890
-    },
-    {
-      "epoch": 2.54910561972998,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 156900
-    },
-    {
-      "epoch": 2.54926808662735,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 156910
-    },
-    {
-      "epoch": 2.5494305535247195,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 156920
-    },
-    {
-      "epoch": 2.549593020422089,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 156930
-    },
-    {
-      "epoch": 2.549755487319459,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3004,
-      "step": 156940
-    },
-    {
-      "epoch": 2.5499179542168284,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 156950
-    },
-    {
-      "epoch": 2.550080421114198,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 156960
-    },
-    {
-      "epoch": 2.5502428880115677,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2622,
-      "step": 156970
-    },
-    {
-      "epoch": 2.5504053549089374,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2455,
-      "step": 156980
-    },
-    {
-      "epoch": 2.550567821806307,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 156990
-    },
-    {
-      "epoch": 2.5507302887036767,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 157000
-    },
-    {
-      "epoch": 2.5508927556010463,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 157010
-    },
-    {
-      "epoch": 2.551055222498416,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 157020
-    },
-    {
-      "epoch": 2.5512176893957856,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 157030
-    },
-    {
-      "epoch": 2.5513801562931553,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 157040
-    },
-    {
-      "epoch": 2.551542623190525,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 157050
-    },
-    {
-      "epoch": 2.5517050900878946,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 157060
-    },
-    {
-      "epoch": 2.5518675569852642,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 157070
-    },
-    {
-      "epoch": 2.552030023882634,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2955,
-      "step": 157080
-    },
-    {
-      "epoch": 2.5521924907800035,
-      "grad_norm": 2.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 157090
-    },
-    {
-      "epoch": 2.552354957677373,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 157100
-    },
-    {
-      "epoch": 2.552517424574743,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 157110
-    },
-    {
-      "epoch": 2.5526798914721125,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 157120
-    },
-    {
-      "epoch": 2.552842358369482,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2962,
-      "step": 157130
-    },
-    {
-      "epoch": 2.553004825266852,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3063,
-      "step": 157140
-    },
-    {
-      "epoch": 2.5531672921642214,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 157150
-    },
-    {
-      "epoch": 2.553329759061591,
-      "grad_norm": 2.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 157160
-    },
-    {
-      "epoch": 2.5534922259589607,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2982,
-      "step": 157170
-    },
-    {
-      "epoch": 2.5536546928563304,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 157180
-    },
-    {
-      "epoch": 2.5538171597537,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 157190
-    },
-    {
-      "epoch": 2.5539796266510697,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 157200
-    },
-    {
-      "epoch": 2.5541420935484394,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 157210
-    },
-    {
-      "epoch": 2.554304560445809,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2514,
-      "step": 157220
-    },
-    {
-      "epoch": 2.554467027343179,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 157230
-    },
-    {
-      "epoch": 2.5546294942405483,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 157240
-    },
-    {
-      "epoch": 2.5547919611379184,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 157250
-    },
-    {
-      "epoch": 2.5549544280352876,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2583,
-      "step": 157260
-    },
-    {
-      "epoch": 2.5551168949326577,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2702,
-      "step": 157270
-    },
-    {
-      "epoch": 2.555279361830027,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 157280
-    },
-    {
-      "epoch": 2.555441828727397,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 157290
-    },
-    {
-      "epoch": 2.555604295624766,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 157300
-    },
-    {
-      "epoch": 2.5557667625221363,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.29,
-      "step": 157310
-    },
-    {
-      "epoch": 2.5559292294195055,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 157320
-    },
-    {
-      "epoch": 2.5560916963168756,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 157330
-    },
-    {
-      "epoch": 2.5562541632142453,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 157340
-    },
-    {
-      "epoch": 2.556416630111615,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 157350
-    },
-    {
-      "epoch": 2.5565790970089846,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 157360
-    },
-    {
-      "epoch": 2.556741563906354,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 157370
-    },
-    {
-      "epoch": 2.556904030803724,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 157380
-    },
-    {
-      "epoch": 2.5570664977010935,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 157390
-    },
-    {
-      "epoch": 2.557228964598463,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 157400
-    },
-    {
-      "epoch": 2.557391431495833,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 157410
-    },
-    {
-      "epoch": 2.5575538983932025,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 157420
-    },
-    {
-      "epoch": 2.557716365290572,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2615,
-      "step": 157430
-    },
-    {
-      "epoch": 2.5578788321879418,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2538,
-      "step": 157440
-    },
-    {
-      "epoch": 2.5580412990853114,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 157450
-    },
-    {
-      "epoch": 2.558203765982681,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2863,
-      "step": 157460
-    },
-    {
-      "epoch": 2.5583662328800507,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2677,
-      "step": 157470
-    },
-    {
-      "epoch": 2.5585286997774204,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 157480
-    },
-    {
-      "epoch": 2.55869116667479,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 157490
-    },
-    {
-      "epoch": 2.5588536335721597,
-      "grad_norm": 2.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2627,
-      "step": 157500
-    },
-    {
-      "epoch": 2.5590161004695293,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 157510
-    },
-    {
-      "epoch": 2.559178567366899,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 157520
-    },
-    {
-      "epoch": 2.5593410342642686,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 157530
-    },
-    {
-      "epoch": 2.5595035011616383,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2652,
-      "step": 157540
-    },
-    {
-      "epoch": 2.559665968059008,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 157550
-    },
-    {
-      "epoch": 2.5598284349563776,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 157560
-    },
-    {
-      "epoch": 2.5599909018537472,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 157570
-    },
-    {
-      "epoch": 2.560153368751117,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 157580
-    },
-    {
-      "epoch": 2.5603158356484865,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 157590
-    },
-    {
-      "epoch": 2.560478302545856,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 157600
-    },
-    {
-      "epoch": 2.560640769443226,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 157610
-    },
-    {
-      "epoch": 2.5608032363405955,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 157620
-    },
-    {
-      "epoch": 2.560965703237965,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2639,
-      "step": 157630
-    },
-    {
-      "epoch": 2.561128170135335,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 157640
-    },
-    {
-      "epoch": 2.5612906370327044,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 157650
-    },
-    {
-      "epoch": 2.561453103930074,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 157660
-    },
-    {
-      "epoch": 2.5616155708274437,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2605,
-      "step": 157670
-    },
-    {
-      "epoch": 2.561778037724814,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 157680
-    },
-    {
-      "epoch": 2.561940504622183,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 157690
-    },
-    {
-      "epoch": 2.562102971519553,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 157700
-    },
-    {
-      "epoch": 2.5622654384169223,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 157710
-    },
-    {
-      "epoch": 2.5624279053142924,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 157720
-    },
-    {
-      "epoch": 2.5625903722116616,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 157730
-    },
-    {
-      "epoch": 2.5627528391090317,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 157740
-    },
-    {
-      "epoch": 2.562915306006401,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3125,
-      "step": 157750
-    },
-    {
-      "epoch": 2.563077772903771,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 157760
-    },
-    {
-      "epoch": 2.5632402398011402,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 157770
-    },
-    {
-      "epoch": 2.5634027066985103,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 157780
-    },
-    {
-      "epoch": 2.56356517359588,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 157790
-    },
-    {
-      "epoch": 2.5637276404932496,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 157800
-    },
-    {
-      "epoch": 2.5638901073906193,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 157810
-    },
-    {
-      "epoch": 2.564052574287989,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 157820
-    },
-    {
-      "epoch": 2.5642150411853586,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 157830
-    },
-    {
-      "epoch": 2.5643775080827282,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 157840
-    },
-    {
-      "epoch": 2.564539974980098,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2987,
-      "step": 157850
-    },
-    {
-      "epoch": 2.5647024418774675,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 157860
-    },
-    {
-      "epoch": 2.564864908774837,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 157870
-    },
-    {
-      "epoch": 2.565027375672207,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2574,
-      "step": 157880
-    },
-    {
-      "epoch": 2.5651898425695765,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 157890
-    },
-    {
-      "epoch": 2.565352309466946,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 157900
-    },
-    {
-      "epoch": 2.565514776364316,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 157910
-    },
-    {
-      "epoch": 2.5656772432616854,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2686,
-      "step": 157920
-    },
-    {
-      "epoch": 2.565839710159055,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 157930
-    },
-    {
-      "epoch": 2.5660021770564247,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 157940
-    },
-    {
-      "epoch": 2.5661646439537944,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 157950
-    },
-    {
-      "epoch": 2.566327110851164,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 157960
-    },
-    {
-      "epoch": 2.5664895777485337,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 157970
-    },
-    {
-      "epoch": 2.5666520446459034,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2645,
-      "step": 157980
-    },
-    {
-      "epoch": 2.566814511543273,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 157990
-    },
-    {
-      "epoch": 2.5669769784406427,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2537,
-      "step": 158000
-    },
-    {
-      "epoch": 2.5671394453380123,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 158010
-    },
-    {
-      "epoch": 2.567301912235382,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3007,
-      "step": 158020
-    },
-    {
-      "epoch": 2.5674643791327516,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 158030
-    },
-    {
-      "epoch": 2.5676268460301213,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 158040
-    },
-    {
-      "epoch": 2.567789312927491,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 158050
-    },
-    {
-      "epoch": 2.5679517798248606,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 158060
-    },
-    {
-      "epoch": 2.56811424672223,
-      "grad_norm": 2.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2628,
-      "step": 158070
-    },
-    {
-      "epoch": 2.5682767136196,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 158080
-    },
-    {
-      "epoch": 2.5684391805169695,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 158090
-    },
-    {
-      "epoch": 2.568601647414339,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 158100
-    },
-    {
-      "epoch": 2.5687641143117093,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 158110
-    },
-    {
-      "epoch": 2.5689265812090785,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2569,
-      "step": 158120
-    },
-    {
-      "epoch": 2.5690890481064486,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 158130
-    },
-    {
-      "epoch": 2.5692515150038178,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.264,
-      "step": 158140
-    },
-    {
-      "epoch": 2.569413981901188,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 158150
-    },
-    {
-      "epoch": 2.569576448798557,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 158160
-    },
-    {
-      "epoch": 2.569738915695927,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 158170
-    },
-    {
-      "epoch": 2.5699013825932964,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 158180
-    },
-    {
-      "epoch": 2.5700638494906665,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 158190
-    },
-    {
-      "epoch": 2.5702263163880357,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 158200
-    },
-    {
-      "epoch": 2.5703887832854058,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2641,
-      "step": 158210
-    },
-    {
-      "epoch": 2.5705512501827754,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 158220
-    },
-    {
-      "epoch": 2.570713717080145,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.249,
-      "step": 158230
-    },
-    {
-      "epoch": 2.5708761839775147,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2418,
-      "step": 158240
-    },
-    {
-      "epoch": 2.5710386508748844,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2521,
-      "step": 158250
-    },
-    {
-      "epoch": 2.571201117772254,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 158260
-    },
-    {
-      "epoch": 2.5713635846696237,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 158270
-    },
-    {
-      "epoch": 2.5715260515669933,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.258,
-      "step": 158280
-    },
-    {
-      "epoch": 2.571688518464363,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2813,
-      "step": 158290
-    },
-    {
-      "epoch": 2.5718509853617326,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 158300
-    },
-    {
-      "epoch": 2.5720134522591023,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 158310
-    },
-    {
-      "epoch": 2.572175919156472,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2676,
-      "step": 158320
-    },
-    {
-      "epoch": 2.5723383860538416,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 158330
-    },
-    {
-      "epoch": 2.5725008529512112,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 158340
-    },
-    {
-      "epoch": 2.572663319848581,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 158350
-    },
-    {
-      "epoch": 2.5728257867459505,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 158360
-    },
-    {
-      "epoch": 2.57298825364332,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3014,
-      "step": 158370
-    },
-    {
-      "epoch": 2.57315072054069,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2647,
-      "step": 158380
-    },
-    {
-      "epoch": 2.5733131874380595,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 158390
-    },
-    {
-      "epoch": 2.573475654335429,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 158400
-    },
-    {
-      "epoch": 2.573638121232799,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 158410
-    },
-    {
-      "epoch": 2.5738005881301684,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2909,
-      "step": 158420
-    },
-    {
-      "epoch": 2.573963055027538,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 158430
-    },
-    {
-      "epoch": 2.5741255219249077,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 158440
-    },
-    {
-      "epoch": 2.5742879888222774,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 158450
-    },
-    {
-      "epoch": 2.574450455719647,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 158460
-    },
-    {
-      "epoch": 2.5746129226170167,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 158470
-    },
-    {
-      "epoch": 2.5747753895143863,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 158480
-    },
-    {
-      "epoch": 2.574937856411756,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2589,
-      "step": 158490
-    },
-    {
-      "epoch": 2.5751003233091256,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 158500
-    },
-    {
-      "epoch": 2.5752627902064953,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 158510
-    },
-    {
-      "epoch": 2.575425257103865,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3067,
-      "step": 158520
-    },
-    {
-      "epoch": 2.5755877240012346,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.289,
-      "step": 158530
-    },
-    {
-      "epoch": 2.5757501908986047,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2712,
-      "step": 158540
-    },
-    {
-      "epoch": 2.575912657795974,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 158550
-    },
-    {
-      "epoch": 2.576075124693344,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 158560
-    },
-    {
-      "epoch": 2.576237591590713,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2612,
-      "step": 158570
-    },
-    {
-      "epoch": 2.5764000584880833,
-      "grad_norm": 15.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 158580
-    },
-    {
-      "epoch": 2.5765625253854525,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2588,
-      "step": 158590
-    },
-    {
-      "epoch": 2.5767249922828226,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3141,
-      "step": 158600
-    },
-    {
-      "epoch": 2.576887459180192,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 158610
-    },
-    {
-      "epoch": 2.577049926077562,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2688,
-      "step": 158620
-    },
-    {
-      "epoch": 2.577212392974931,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3004,
-      "step": 158630
-    },
-    {
-      "epoch": 2.577374859872301,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 158640
-    },
-    {
-      "epoch": 2.577537326769671,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2473,
-      "step": 158650
-    },
-    {
-      "epoch": 2.5776997936670405,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2558,
-      "step": 158660
-    },
-    {
-      "epoch": 2.57786226056441,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2446,
-      "step": 158670
-    },
-    {
-      "epoch": 2.57802472746178,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 158680
-    },
-    {
-      "epoch": 2.5781871943591494,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 158690
-    },
-    {
-      "epoch": 2.578349661256519,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 158700
-    },
-    {
-      "epoch": 2.5785121281538887,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 158710
-    },
-    {
-      "epoch": 2.5786745950512584,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 158720
-    },
-    {
-      "epoch": 2.578837061948628,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 158730
-    },
-    {
-      "epoch": 2.5789995288459977,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 158740
-    },
-    {
-      "epoch": 2.5791619957433674,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2621,
-      "step": 158750
-    },
-    {
-      "epoch": 2.579324462640737,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3015,
-      "step": 158760
-    },
-    {
-      "epoch": 2.5794869295381067,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 158770
-    },
-    {
-      "epoch": 2.5796493964354763,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 158780
-    },
-    {
-      "epoch": 2.579811863332846,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 158790
-    },
-    {
-      "epoch": 2.5799743302302156,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 158800
-    },
-    {
-      "epoch": 2.5801367971275853,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 158810
-    },
-    {
-      "epoch": 2.580299264024955,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 158820
-    },
-    {
-      "epoch": 2.5804617309223246,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2909,
-      "step": 158830
-    },
-    {
-      "epoch": 2.580624197819694,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3037,
-      "step": 158840
-    },
-    {
-      "epoch": 2.580786664717064,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.297,
-      "step": 158850
-    },
-    {
-      "epoch": 2.5809491316144335,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 158860
-    },
-    {
-      "epoch": 2.581111598511803,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 158870
-    },
-    {
-      "epoch": 2.581274065409173,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.305,
-      "step": 158880
-    },
-    {
-      "epoch": 2.5814365323065425,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 158890
-    },
-    {
-      "epoch": 2.581598999203912,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 158900
-    },
-    {
-      "epoch": 2.5817614661012818,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 158910
-    },
-    {
-      "epoch": 2.5819239329986514,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 158920
-    },
-    {
-      "epoch": 2.582086399896021,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 158930
-    },
-    {
-      "epoch": 2.5822488667933907,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 158940
-    },
-    {
-      "epoch": 2.5824113336907604,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 158950
-    },
-    {
-      "epoch": 2.58257380058813,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 158960
-    },
-    {
-      "epoch": 2.5827362674854997,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 158970
-    },
-    {
-      "epoch": 2.5828987343828693,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 158980
-    },
-    {
-      "epoch": 2.5830612012802394,
-      "grad_norm": 2.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 158990
-    },
-    {
-      "epoch": 2.5832236681776086,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3117,
-      "step": 159000
-    },
-    {
-      "epoch": 2.5833861350749787,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2971,
-      "step": 159010
-    },
-    {
-      "epoch": 2.583548601972348,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.29,
-      "step": 159020
-    },
-    {
-      "epoch": 2.583711068869718,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 159030
-    },
-    {
-      "epoch": 2.5838735357670872,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2607,
-      "step": 159040
-    },
-    {
-      "epoch": 2.5840360026644573,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 159050
-    },
-    {
-      "epoch": 2.5841984695618265,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2615,
-      "step": 159060
-    },
-    {
-      "epoch": 2.5843609364591966,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 159070
-    },
-    {
-      "epoch": 2.584523403356566,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 159080
-    },
-    {
-      "epoch": 2.584685870253936,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 159090
-    },
-    {
-      "epoch": 2.5848483371513056,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 159100
-    },
-    {
-      "epoch": 2.5850108040486752,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 159110
-    },
-    {
-      "epoch": 2.585173270946045,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 159120
-    },
-    {
-      "epoch": 2.5853357378434145,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 159130
-    },
-    {
-      "epoch": 2.585498204740784,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 159140
-    },
-    {
-      "epoch": 2.585660671638154,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.304,
-      "step": 159150
-    },
-    {
-      "epoch": 2.5858231385355235,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 159160
-    },
-    {
-      "epoch": 2.585985605432893,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 159170
-    },
-    {
-      "epoch": 2.586148072330263,
-      "grad_norm": 2.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 159180
-    },
-    {
-      "epoch": 2.5863105392276324,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 159190
-    },
-    {
-      "epoch": 2.586473006125002,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 159200
-    },
-    {
-      "epoch": 2.5866354730223717,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 159210
-    },
-    {
-      "epoch": 2.5867979399197414,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 159220
-    },
-    {
-      "epoch": 2.586960406817111,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 159230
-    },
-    {
-      "epoch": 2.5871228737144807,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 159240
-    },
-    {
-      "epoch": 2.5872853406118503,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 159250
-    },
-    {
-      "epoch": 2.58744780750922,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 159260
-    },
-    {
-      "epoch": 2.5876102744065896,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 159270
-    },
-    {
-      "epoch": 2.5877727413039593,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2702,
-      "step": 159280
-    },
-    {
-      "epoch": 2.587935208201329,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 159290
-    },
-    {
-      "epoch": 2.5880976750986986,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 159300
-    },
-    {
-      "epoch": 2.5882601419960682,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 159310
-    },
-    {
-      "epoch": 2.588422608893438,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2813,
-      "step": 159320
-    },
-    {
-      "epoch": 2.5885850757908075,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 159330
-    },
-    {
-      "epoch": 2.588747542688177,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2686,
-      "step": 159340
-    },
-    {
-      "epoch": 2.588910009585547,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.261,
-      "step": 159350
-    },
-    {
-      "epoch": 2.5890724764829165,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 159360
-    },
-    {
-      "epoch": 2.589234943380286,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2695,
-      "step": 159370
-    },
-    {
-      "epoch": 2.589397410277656,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 159380
-    },
-    {
-      "epoch": 2.5895598771750254,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 159390
-    },
-    {
-      "epoch": 2.589722344072395,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 159400
-    },
-    {
-      "epoch": 2.5898848109697648,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.299,
-      "step": 159410
-    },
-    {
-      "epoch": 2.590047277867135,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 159420
-    },
-    {
-      "epoch": 2.590209744764504,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 159430
-    },
-    {
-      "epoch": 2.590372211661874,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 159440
-    },
-    {
-      "epoch": 2.5905346785592434,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 159450
-    },
-    {
-      "epoch": 2.5906971454566134,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 159460
-    },
-    {
-      "epoch": 2.5908596123539827,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 159470
-    },
-    {
-      "epoch": 2.5910220792513527,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 159480
-    },
-    {
-      "epoch": 2.591184546148722,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 159490
-    },
-    {
-      "epoch": 2.591347013046092,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 159500
-    },
-    {
-      "epoch": 2.5915094799434613,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 159510
-    },
-    {
-      "epoch": 2.5916719468408314,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 159520
-    },
-    {
-      "epoch": 2.591834413738201,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 159530
-    },
-    {
-      "epoch": 2.5919968806355707,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 159540
-    },
-    {
-      "epoch": 2.5921593475329403,
-      "grad_norm": 2.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.26,
-      "step": 159550
-    },
-    {
-      "epoch": 2.59232181443031,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 159560
-    },
-    {
-      "epoch": 2.5924842813276796,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2793,
-      "step": 159570
-    },
-    {
-      "epoch": 2.5926467482250493,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 159580
-    },
-    {
-      "epoch": 2.592809215122419,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 159590
-    },
-    {
-      "epoch": 2.5929716820197886,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 159600
-    },
-    {
-      "epoch": 2.593134148917158,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 159610
-    },
-    {
-      "epoch": 2.593296615814528,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 159620
-    },
-    {
-      "epoch": 2.5934590827118975,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 159630
-    },
-    {
-      "epoch": 2.593621549609267,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 159640
-    },
-    {
-      "epoch": 2.593784016506637,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 159650
-    },
-    {
-      "epoch": 2.5939464834040065,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 159660
-    },
-    {
-      "epoch": 2.594108950301376,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 159670
-    },
-    {
-      "epoch": 2.5942714171987458,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2916,
-      "step": 159680
-    },
-    {
-      "epoch": 2.5944338840961154,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 159690
-    },
-    {
-      "epoch": 2.594596350993485,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2547,
-      "step": 159700
-    },
-    {
-      "epoch": 2.5947588178908547,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 159710
-    },
-    {
-      "epoch": 2.5949212847882244,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 159720
-    },
-    {
-      "epoch": 2.595083751685594,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 159730
-    },
-    {
-      "epoch": 2.5952462185829637,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2669,
-      "step": 159740
-    },
-    {
-      "epoch": 2.5954086854803333,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2558,
-      "step": 159750
-    },
-    {
-      "epoch": 2.595571152377703,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 159760
-    },
-    {
-      "epoch": 2.5957336192750726,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 159770
-    },
-    {
-      "epoch": 2.5958960861724423,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 159780
-    },
-    {
-      "epoch": 2.596058553069812,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2959,
-      "step": 159790
-    },
-    {
-      "epoch": 2.5962210199671816,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 159800
-    },
-    {
-      "epoch": 2.5963834868645512,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 159810
-    },
-    {
-      "epoch": 2.596545953761921,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 159820
-    },
-    {
-      "epoch": 2.5967084206592905,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.257,
-      "step": 159830
-    },
-    {
-      "epoch": 2.59687088755666,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2506,
-      "step": 159840
-    },
-    {
-      "epoch": 2.59703335445403,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2378,
-      "step": 159850
-    },
-    {
-      "epoch": 2.5971958213513995,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2514,
-      "step": 159860
-    },
-    {
-      "epoch": 2.5973582882487696,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 159870
-    },
-    {
-      "epoch": 2.597520755146139,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 159880
-    },
-    {
-      "epoch": 2.597683222043509,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 159890
-    },
-    {
-      "epoch": 2.597845688940878,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 159900
-    },
-    {
-      "epoch": 2.598008155838248,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 159910
-    },
-    {
-      "epoch": 2.5981706227356174,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.3004,
-      "step": 159920
-    },
-    {
-      "epoch": 2.5983330896329875,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.299,
-      "step": 159930
-    },
-    {
-      "epoch": 2.5984955565303567,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2909,
-      "step": 159940
-    },
-    {
-      "epoch": 2.598658023427727,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 159950
-    },
-    {
-      "epoch": 2.598820490325096,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 159960
-    },
-    {
-      "epoch": 2.598982957222466,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 159970
-    },
-    {
-      "epoch": 2.5991454241198357,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 159980
-    },
-    {
-      "epoch": 2.5993078910172054,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 159990
-    },
-    {
-      "epoch": 2.599470357914575,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 160000
-    },
-    {
-      "epoch": 2.5996328248119447,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 160010
-    },
-    {
-      "epoch": 2.5997952917093143,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 160020
-    },
-    {
-      "epoch": 2.599957758606684,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 160030
-    },
-    {
-      "epoch": 2.6001202255040536,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 160040
-    },
-    {
-      "epoch": 2.6002826924014233,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 160050
-    },
-    {
-      "epoch": 2.600445159298793,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 160060
-    },
-    {
-      "epoch": 2.6006076261961626,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 160070
-    },
-    {
-      "epoch": 2.6007700930935322,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2541,
-      "step": 160080
-    },
-    {
-      "epoch": 2.600932559990902,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 160090
-    },
-    {
-      "epoch": 2.6010950268882715,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2875,
-      "step": 160100
-    },
-    {
-      "epoch": 2.601257493785641,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2576,
-      "step": 160110
-    },
-    {
-      "epoch": 2.601419960683011,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2601,
-      "step": 160120
-    },
-    {
-      "epoch": 2.6015824275803805,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 160130
-    },
-    {
-      "epoch": 2.60174489447775,
-      "grad_norm": 2.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2587,
-      "step": 160140
-    },
-    {
-      "epoch": 2.60190736137512,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2601,
-      "step": 160150
-    },
-    {
-      "epoch": 2.6020698282724894,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.263,
-      "step": 160160
-    },
-    {
-      "epoch": 2.602232295169859,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2522,
-      "step": 160170
-    },
-    {
-      "epoch": 2.6023947620672288,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2563,
-      "step": 160180
-    },
-    {
-      "epoch": 2.6025572289645984,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2599,
-      "step": 160190
-    },
-    {
-      "epoch": 2.602719695861968,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 160200
-    },
-    {
-      "epoch": 2.6028821627593377,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3004,
-      "step": 160210
-    },
-    {
-      "epoch": 2.6030446296567074,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 160220
-    },
-    {
-      "epoch": 2.603207096554077,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 160230
-    },
-    {
-      "epoch": 2.6033695634514467,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.289,
-      "step": 160240
-    },
-    {
-      "epoch": 2.6035320303488163,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 160250
-    },
-    {
-      "epoch": 2.603694497246186,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3018,
-      "step": 160260
-    },
-    {
-      "epoch": 2.6038569641435556,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 160270
-    },
-    {
-      "epoch": 2.6040194310409253,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2378,
-      "step": 160280
-    },
-    {
-      "epoch": 2.604181897938295,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2533,
-      "step": 160290
-    },
-    {
-      "epoch": 2.604344364835665,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 160300
-    },
-    {
-      "epoch": 2.604506831733034,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 160310
-    },
-    {
-      "epoch": 2.6046692986304043,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 160320
-    },
-    {
-      "epoch": 2.6048317655277735,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 160330
-    },
-    {
-      "epoch": 2.6049942324251436,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 160340
-    },
-    {
-      "epoch": 2.605156699322513,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 160350
-    },
-    {
-      "epoch": 2.605319166219883,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 160360
-    },
-    {
-      "epoch": 2.605481633117252,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.3052,
-      "step": 160370
-    },
-    {
-      "epoch": 2.605644100014622,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 160380
-    },
-    {
-      "epoch": 2.6058065669119914,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3046,
-      "step": 160390
-    },
-    {
-      "epoch": 2.6059690338093615,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 160400
-    },
-    {
-      "epoch": 2.606131500706731,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2961,
-      "step": 160410
-    },
-    {
-      "epoch": 2.606293967604101,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 160420
-    },
-    {
-      "epoch": 2.6064564345014705,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 160430
-    },
-    {
-      "epoch": 2.60661890139884,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 160440
-    },
-    {
-      "epoch": 2.6067813682962098,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 160450
-    },
-    {
-      "epoch": 2.6069438351935794,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 160460
-    },
-    {
-      "epoch": 2.607106302090949,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.291,
-      "step": 160470
-    },
-    {
-      "epoch": 2.6072687689883187,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 160480
-    },
-    {
-      "epoch": 2.6074312358856884,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 160490
-    },
-    {
-      "epoch": 2.607593702783058,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 160500
-    },
-    {
-      "epoch": 2.6077561696804277,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 160510
-    },
-    {
-      "epoch": 2.6079186365777973,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2686,
-      "step": 160520
-    },
-    {
-      "epoch": 2.608081103475167,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 160530
-    },
-    {
-      "epoch": 2.6082435703725366,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 160540
-    },
-    {
-      "epoch": 2.6084060372699063,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2992,
-      "step": 160550
-    },
-    {
-      "epoch": 2.608568504167276,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 160560
-    },
-    {
-      "epoch": 2.6087309710646456,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 160570
-    },
-    {
-      "epoch": 2.6088934379620152,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 160580
-    },
-    {
-      "epoch": 2.609055904859385,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2551,
-      "step": 160590
-    },
-    {
-      "epoch": 2.6092183717567545,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3002,
-      "step": 160600
-    },
-    {
-      "epoch": 2.609380838654124,
-      "grad_norm": 2.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 160610
-    },
-    {
-      "epoch": 2.609543305551494,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 160620
-    },
-    {
-      "epoch": 2.6097057724488635,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2616,
-      "step": 160630
-    },
-    {
-      "epoch": 2.609868239346233,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 160640
-    },
-    {
-      "epoch": 2.610030706243603,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 160650
-    },
-    {
-      "epoch": 2.6101931731409724,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 160660
-    },
-    {
-      "epoch": 2.610355640038342,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 160670
-    },
-    {
-      "epoch": 2.6105181069357117,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 160680
-    },
-    {
-      "epoch": 2.6106805738330814,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 160690
-    },
-    {
-      "epoch": 2.610843040730451,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.294,
-      "step": 160700
-    },
-    {
-      "epoch": 2.6110055076278207,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 160710
-    },
-    {
-      "epoch": 2.6111679745251903,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 160720
-    },
-    {
-      "epoch": 2.61133044142256,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 160730
-    },
-    {
-      "epoch": 2.6114929083199296,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 160740
-    },
-    {
-      "epoch": 2.6116553752172997,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 160750
-    },
-    {
-      "epoch": 2.611817842114669,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 160760
-    },
-    {
-      "epoch": 2.611980309012039,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 160770
-    },
-    {
-      "epoch": 2.6121427759094082,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 160780
-    },
-    {
-      "epoch": 2.6123052428067783,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 160790
-    },
-    {
-      "epoch": 2.6124677097041475,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 160800
-    },
-    {
-      "epoch": 2.6126301766015176,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 160810
-    },
-    {
-      "epoch": 2.612792643498887,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 160820
-    },
-    {
-      "epoch": 2.612955110396257,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 160830
-    },
-    {
-      "epoch": 2.613117577293626,
-      "grad_norm": 2.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 160840
-    },
-    {
-      "epoch": 2.6132800441909962,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 160850
-    },
-    {
-      "epoch": 2.613442511088366,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3046,
-      "step": 160860
-    },
-    {
-      "epoch": 2.6136049779857355,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 160870
-    },
-    {
-      "epoch": 2.613767444883105,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 160880
-    },
-    {
-      "epoch": 2.613929911780475,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 160890
-    },
-    {
-      "epoch": 2.6140923786778445,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 160900
-    },
-    {
-      "epoch": 2.614254845575214,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 160910
-    },
-    {
-      "epoch": 2.614417312472584,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2886,
-      "step": 160920
-    },
-    {
-      "epoch": 2.6145797793699534,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 160930
-    },
-    {
-      "epoch": 2.614742246267323,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 160940
-    },
-    {
-      "epoch": 2.6149047131646928,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2606,
-      "step": 160950
-    },
-    {
-      "epoch": 2.6150671800620624,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 160960
-    },
-    {
-      "epoch": 2.615229646959432,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 160970
-    },
-    {
-      "epoch": 2.6153921138568017,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2986,
-      "step": 160980
-    },
-    {
-      "epoch": 2.6155545807541714,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.3078,
-      "step": 160990
-    },
-    {
-      "epoch": 2.615717047651541,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 161000
-    },
-    {
-      "epoch": 2.6158795145489107,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 161010
-    },
-    {
-      "epoch": 2.6160419814462803,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 161020
-    },
-    {
-      "epoch": 2.61620444834365,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2713,
-      "step": 161030
-    },
-    {
-      "epoch": 2.6163669152410196,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 161040
-    },
-    {
-      "epoch": 2.6165293821383893,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 161050
-    },
-    {
-      "epoch": 2.616691849035759,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 161060
-    },
-    {
-      "epoch": 2.6168543159331286,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 161070
-    },
-    {
-      "epoch": 2.617016782830498,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 161080
-    },
-    {
-      "epoch": 2.617179249727868,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 161090
-    },
-    {
-      "epoch": 2.6173417166252375,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 161100
-    },
-    {
-      "epoch": 2.617504183522607,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 161110
-    },
-    {
-      "epoch": 2.617666650419977,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 161120
-    },
-    {
-      "epoch": 2.6178291173173465,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 161130
-    },
-    {
-      "epoch": 2.617991584214716,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 161140
-    },
-    {
-      "epoch": 2.6181540511120858,
-      "grad_norm": 2.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2673,
-      "step": 161150
-    },
-    {
-      "epoch": 2.6183165180094554,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 161160
-    },
-    {
-      "epoch": 2.618478984906825,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 161170
-    },
-    {
-      "epoch": 2.618641451804195,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 161180
-    },
-    {
-      "epoch": 2.6188039187015644,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 161190
-    },
-    {
-      "epoch": 2.6189663855989345,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 161200
-    },
-    {
-      "epoch": 2.6191288524963037,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 161210
-    },
-    {
-      "epoch": 2.6192913193936738,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 161220
-    },
-    {
-      "epoch": 2.619453786291043,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 161230
-    },
-    {
-      "epoch": 2.619616253188413,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 161240
-    },
-    {
-      "epoch": 2.6197787200857823,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 161250
-    },
-    {
-      "epoch": 2.6199411869831524,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2483,
-      "step": 161260
-    },
-    {
-      "epoch": 2.6201036538805216,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 161270
-    },
-    {
-      "epoch": 2.6202661207778917,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2607,
-      "step": 161280
-    },
-    {
-      "epoch": 2.6204285876752613,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2818,
-      "step": 161290
-    },
-    {
-      "epoch": 2.620591054572631,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 161300
-    },
-    {
-      "epoch": 2.6207535214700006,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 161310
-    },
-    {
-      "epoch": 2.6209159883673703,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 161320
-    },
-    {
-      "epoch": 2.62107845526474,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2639,
-      "step": 161330
-    },
-    {
-      "epoch": 2.6212409221621096,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 161340
-    },
-    {
-      "epoch": 2.6214033890594792,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 161350
-    },
-    {
-      "epoch": 2.621565855956849,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2649,
-      "step": 161360
-    },
-    {
-      "epoch": 2.6217283228542185,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2478,
-      "step": 161370
-    },
-    {
-      "epoch": 2.621890789751588,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 161380
-    },
-    {
-      "epoch": 2.622053256648958,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 161390
-    },
-    {
-      "epoch": 2.6222157235463275,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 161400
-    },
-    {
-      "epoch": 2.622378190443697,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 161410
-    },
-    {
-      "epoch": 2.622540657341067,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 161420
-    },
-    {
-      "epoch": 2.6227031242384364,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 161430
-    },
-    {
-      "epoch": 2.622865591135806,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 161440
-    },
-    {
-      "epoch": 2.6230280580331757,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.3047,
-      "step": 161450
-    },
-    {
-      "epoch": 2.6231905249305454,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3016,
-      "step": 161460
-    },
-    {
-      "epoch": 2.623352991827915,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 161470
-    },
-    {
-      "epoch": 2.6235154587252847,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 161480
-    },
-    {
-      "epoch": 2.6236779256226543,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3075,
-      "step": 161490
-    },
-    {
-      "epoch": 2.623840392520024,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.3114,
-      "step": 161500
-    },
-    {
-      "epoch": 2.6240028594173936,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3106,
-      "step": 161510
-    },
-    {
-      "epoch": 2.6241653263147633,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2984,
-      "step": 161520
-    },
-    {
-      "epoch": 2.624327793212133,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2886,
-      "step": 161530
-    },
-    {
-      "epoch": 2.6244902601095026,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 161540
-    },
-    {
-      "epoch": 2.6246527270068722,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3223,
-      "step": 161550
-    },
-    {
-      "epoch": 2.624815193904242,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3067,
-      "step": 161560
-    },
-    {
-      "epoch": 2.6249776608016115,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 161570
-    },
-    {
-      "epoch": 2.625140127698981,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2943,
-      "step": 161580
-    },
-    {
-      "epoch": 2.625302594596351,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 161590
-    },
-    {
-      "epoch": 2.6254650614937205,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 161600
-    },
-    {
-      "epoch": 2.6256275283910906,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 161610
-    },
-    {
-      "epoch": 2.62578999528846,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2512,
-      "step": 161620
-    },
-    {
-      "epoch": 2.62595246218583,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 161630
-    },
-    {
-      "epoch": 2.626114929083199,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 161640
-    },
-    {
-      "epoch": 2.626277395980569,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2663,
-      "step": 161650
-    },
-    {
-      "epoch": 2.6264398628779384,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2676,
-      "step": 161660
-    },
-    {
-      "epoch": 2.6266023297753085,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 161670
-    },
-    {
-      "epoch": 2.6267647966726777,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 161680
-    },
-    {
-      "epoch": 2.626927263570048,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2487,
-      "step": 161690
-    },
-    {
-      "epoch": 2.627089730467417,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2875,
-      "step": 161700
-    },
-    {
-      "epoch": 2.627252197364787,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 161710
-    },
-    {
-      "epoch": 2.6274146642621568,
-      "grad_norm": 2.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 161720
-    },
-    {
-      "epoch": 2.6275771311595264,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 161730
-    },
-    {
-      "epoch": 2.627739598056896,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2961,
-      "step": 161740
-    },
-    {
-      "epoch": 2.6279020649542657,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 161750
-    },
-    {
-      "epoch": 2.6280645318516354,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 161760
-    },
-    {
-      "epoch": 2.628226998749005,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 161770
-    },
-    {
-      "epoch": 2.6283894656463747,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 161780
-    },
-    {
-      "epoch": 2.6285519325437443,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 161790
-    },
-    {
-      "epoch": 2.628714399441114,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2521,
-      "step": 161800
-    },
-    {
-      "epoch": 2.6288768663384836,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 161810
-    },
-    {
-      "epoch": 2.6290393332358533,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 161820
-    },
-    {
-      "epoch": 2.629201800133223,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2638,
-      "step": 161830
-    },
-    {
-      "epoch": 2.6293642670305926,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2641,
-      "step": 161840
-    },
-    {
-      "epoch": 2.629526733927962,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 161850
-    },
-    {
-      "epoch": 2.629689200825332,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 161860
-    },
-    {
-      "epoch": 2.6298516677227015,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 161870
-    },
-    {
-      "epoch": 2.630014134620071,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 161880
-    },
-    {
-      "epoch": 2.630176601517441,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 161890
-    },
-    {
-      "epoch": 2.6303390684148105,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 161900
-    },
-    {
-      "epoch": 2.63050153531218,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 161910
-    },
-    {
-      "epoch": 2.6306640022095498,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3041,
-      "step": 161920
-    },
-    {
-      "epoch": 2.6308264691069194,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 161930
-    },
-    {
-      "epoch": 2.630988936004289,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2986,
-      "step": 161940
-    },
-    {
-      "epoch": 2.6311514029016587,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2738,
-      "step": 161950
-    },
-    {
-      "epoch": 2.6313138697990284,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 161960
-    },
-    {
-      "epoch": 2.631476336696398,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 161970
-    },
-    {
-      "epoch": 2.6316388035937677,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 161980
-    },
-    {
-      "epoch": 2.6318012704911373,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 161990
-    },
-    {
-      "epoch": 2.631963737388507,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 162000
-    },
-    {
-      "epoch": 2.6321262042858766,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 162010
-    },
-    {
-      "epoch": 2.6322886711832463,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 162020
-    },
-    {
-      "epoch": 2.632451138080616,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2686,
-      "step": 162030
-    },
-    {
-      "epoch": 2.6326136049779856,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 162040
-    },
-    {
-      "epoch": 2.6327760718753552,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 162050
-    },
-    {
-      "epoch": 2.6329385387727253,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 162060
-    },
-    {
-      "epoch": 2.6331010056700945,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 162070
-    },
-    {
-      "epoch": 2.6332634725674646,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2575,
-      "step": 162080
-    },
-    {
-      "epoch": 2.633425939464834,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 162090
-    },
-    {
-      "epoch": 2.633588406362204,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 162100
-    },
-    {
-      "epoch": 2.633750873259573,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2933,
-      "step": 162110
-    },
-    {
-      "epoch": 2.6339133401569432,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 162120
-    },
-    {
-      "epoch": 2.6340758070543124,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 162130
-    },
-    {
-      "epoch": 2.6342382739516825,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 162140
-    },
-    {
-      "epoch": 2.6344007408490517,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2679,
-      "step": 162150
-    },
-    {
-      "epoch": 2.634563207746422,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 162160
-    },
-    {
-      "epoch": 2.6347256746437915,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 162170
-    },
-    {
-      "epoch": 2.634888141541161,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 162180
-    },
-    {
-      "epoch": 2.635050608438531,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 162190
-    },
-    {
-      "epoch": 2.6352130753359004,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 162200
-    },
-    {
-      "epoch": 2.63537554223327,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 162210
-    },
-    {
-      "epoch": 2.6355380091306397,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 162220
-    },
-    {
-      "epoch": 2.6357004760280094,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 162230
-    },
-    {
-      "epoch": 2.635862942925379,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 162240
-    },
-    {
-      "epoch": 2.6360254098227487,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 162250
-    },
-    {
-      "epoch": 2.6361878767201183,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.276,
-      "step": 162260
-    },
-    {
-      "epoch": 2.636350343617488,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 162270
-    },
-    {
-      "epoch": 2.6365128105148576,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2589,
-      "step": 162280
-    },
-    {
-      "epoch": 2.6366752774122273,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 162290
-    },
-    {
-      "epoch": 2.636837744309597,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 162300
-    },
-    {
-      "epoch": 2.6370002112069666,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 162310
-    },
-    {
-      "epoch": 2.6371626781043362,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 162320
-    },
-    {
-      "epoch": 2.637325145001706,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 162330
-    },
-    {
-      "epoch": 2.6374876118990755,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2934,
-      "step": 162340
-    },
-    {
-      "epoch": 2.637650078796445,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 162350
-    },
-    {
-      "epoch": 2.637812545693815,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 162360
-    },
-    {
-      "epoch": 2.6379750125911845,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2906,
-      "step": 162370
-    },
-    {
-      "epoch": 2.638137479488554,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 162380
-    },
-    {
-      "epoch": 2.638299946385924,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 162390
-    },
-    {
-      "epoch": 2.6384624132832935,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 162400
-    },
-    {
-      "epoch": 2.638624880180663,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2934,
-      "step": 162410
-    },
-    {
-      "epoch": 2.6387873470780328,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 162420
-    },
-    {
-      "epoch": 2.6389498139754024,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 162430
-    },
-    {
-      "epoch": 2.639112280872772,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 162440
-    },
-    {
-      "epoch": 2.6392747477701417,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2628,
-      "step": 162450
-    },
-    {
-      "epoch": 2.6394372146675114,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2863,
-      "step": 162460
-    },
-    {
-      "epoch": 2.639599681564881,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2504,
-      "step": 162470
-    },
-    {
-      "epoch": 2.6397621484622507,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 162480
-    },
-    {
-      "epoch": 2.6399246153596208,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 162490
-    },
-    {
-      "epoch": 2.64008708225699,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 162500
-    },
-    {
-      "epoch": 2.64024954915436,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 162510
-    },
-    {
-      "epoch": 2.6404120160517293,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2931,
-      "step": 162520
-    },
-    {
-      "epoch": 2.6405744829490994,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 162530
-    },
-    {
-      "epoch": 2.6407369498464686,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2971,
-      "step": 162540
-    },
-    {
-      "epoch": 2.6408994167438387,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 162550
-    },
-    {
-      "epoch": 2.641061883641208,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 162560
-    },
-    {
-      "epoch": 2.641224350538578,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 162570
-    },
-    {
-      "epoch": 2.641386817435947,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 162580
-    },
-    {
-      "epoch": 2.6415492843333173,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 162590
-    },
-    {
-      "epoch": 2.641711751230687,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 162600
-    },
-    {
-      "epoch": 2.6418742181280566,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 162610
-    },
-    {
-      "epoch": 2.642036685025426,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 162620
-    },
-    {
-      "epoch": 2.642199151922796,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 162630
-    },
-    {
-      "epoch": 2.6423616188201655,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 162640
-    },
-    {
-      "epoch": 2.642524085717535,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.3012,
-      "step": 162650
-    },
-    {
-      "epoch": 2.642686552614905,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 162660
-    },
-    {
-      "epoch": 2.6428490195122745,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 162670
-    },
-    {
-      "epoch": 2.643011486409644,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 162680
-    },
-    {
-      "epoch": 2.6431739533070138,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 162690
-    },
-    {
-      "epoch": 2.6433364202043834,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 162700
-    },
-    {
-      "epoch": 2.643498887101753,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 162710
-    },
-    {
-      "epoch": 2.6436613539991227,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2369,
-      "step": 162720
-    },
-    {
-      "epoch": 2.6438238208964924,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 162730
-    },
-    {
-      "epoch": 2.643986287793862,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2686,
-      "step": 162740
-    },
-    {
-      "epoch": 2.6441487546912317,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 162750
-    },
-    {
-      "epoch": 2.6443112215886013,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 162760
-    },
-    {
-      "epoch": 2.644473688485971,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2522,
-      "step": 162770
-    },
-    {
-      "epoch": 2.6446361553833406,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2482,
-      "step": 162780
-    },
-    {
-      "epoch": 2.6447986222807103,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3008,
-      "step": 162790
-    },
-    {
-      "epoch": 2.64496108917808,
-      "grad_norm": 2.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2673,
-      "step": 162800
-    },
-    {
-      "epoch": 2.6451235560754496,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 162810
-    },
-    {
-      "epoch": 2.6452860229728192,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2582,
-      "step": 162820
-    },
-    {
-      "epoch": 2.645448489870189,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2567,
-      "step": 162830
-    },
-    {
-      "epoch": 2.6456109567675585,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 162840
-    },
-    {
-      "epoch": 2.645773423664928,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3184,
-      "step": 162850
-    },
-    {
-      "epoch": 2.645935890562298,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2621,
-      "step": 162860
-    },
-    {
-      "epoch": 2.6460983574596675,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 162870
-    },
-    {
-      "epoch": 2.646260824357037,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.3061,
-      "step": 162880
-    },
-    {
-      "epoch": 2.646423291254407,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 162890
-    },
-    {
-      "epoch": 2.6465857581517764,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 162900
-    },
-    {
-      "epoch": 2.646748225049146,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 162910
-    },
-    {
-      "epoch": 2.6469106919465157,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 162920
-    },
-    {
-      "epoch": 2.6470731588438854,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 162930
-    },
-    {
-      "epoch": 2.6472356257412555,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2631,
-      "step": 162940
-    },
-    {
-      "epoch": 2.6473980926386247,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2677,
-      "step": 162950
-    },
-    {
-      "epoch": 2.647560559535995,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 162960
-    },
-    {
-      "epoch": 2.647723026433364,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2617,
-      "step": 162970
-    },
-    {
-      "epoch": 2.647885493330734,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 162980
-    },
-    {
-      "epoch": 2.6480479602281033,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 162990
-    },
-    {
-      "epoch": 2.6482104271254734,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2571,
-      "step": 163000
-    },
-    {
-      "epoch": 2.6483728940228426,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 163010
-    },
-    {
-      "epoch": 2.6485353609202127,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 163020
-    },
-    {
-      "epoch": 2.648697827817582,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 163030
-    },
-    {
-      "epoch": 2.648860294714952,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 163040
-    },
-    {
-      "epoch": 2.6490227616123216,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 163050
-    },
-    {
-      "epoch": 2.6491852285096913,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 163060
-    },
-    {
-      "epoch": 2.649347695407061,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 163070
-    },
-    {
-      "epoch": 2.6495101623044306,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 163080
-    },
-    {
-      "epoch": 2.6496726292018002,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2619,
-      "step": 163090
-    },
-    {
-      "epoch": 2.64983509609917,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 163100
-    },
-    {
-      "epoch": 2.6499975629965395,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 163110
-    },
-    {
-      "epoch": 2.650160029893909,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 163120
-    },
-    {
-      "epoch": 2.650322496791279,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 163130
-    },
-    {
-      "epoch": 2.6504849636886485,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 163140
-    },
-    {
-      "epoch": 2.650647430586018,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 163150
-    },
-    {
-      "epoch": 2.650809897483388,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2587,
-      "step": 163160
-    },
-    {
-      "epoch": 2.6509723643807575,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 163170
-    },
-    {
-      "epoch": 2.651134831278127,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3059,
-      "step": 163180
-    },
-    {
-      "epoch": 2.6512972981754968,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 163190
-    },
-    {
-      "epoch": 2.6514597650728664,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 163200
-    },
-    {
-      "epoch": 2.651622231970236,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 163210
-    },
-    {
-      "epoch": 2.6517846988676057,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 163220
-    },
-    {
-      "epoch": 2.6519471657649754,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2666,
-      "step": 163230
-    },
-    {
-      "epoch": 2.652109632662345,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 163240
-    },
-    {
-      "epoch": 2.6522720995597147,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 163250
-    },
-    {
-      "epoch": 2.6524345664570843,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 163260
-    },
-    {
-      "epoch": 2.652597033354454,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 163270
-    },
-    {
-      "epoch": 2.6527595002518236,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 163280
-    },
-    {
-      "epoch": 2.6529219671491933,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 163290
-    },
-    {
-      "epoch": 2.653084434046563,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 163300
-    },
-    {
-      "epoch": 2.6532469009439326,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 163310
-    },
-    {
-      "epoch": 2.653409367841302,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 163320
-    },
-    {
-      "epoch": 2.653571834738672,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2565,
-      "step": 163330
-    },
-    {
-      "epoch": 2.6537343016360415,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 163340
-    },
-    {
-      "epoch": 2.653896768533411,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 163350
-    },
-    {
-      "epoch": 2.654059235430781,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 163360
-    },
-    {
-      "epoch": 2.654221702328151,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2781,
-      "step": 163370
-    },
-    {
-      "epoch": 2.65438416922552,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 163380
-    },
-    {
-      "epoch": 2.65454663612289,
-      "grad_norm": 2.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2541,
-      "step": 163390
-    },
-    {
-      "epoch": 2.6547091030202594,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2713,
-      "step": 163400
-    },
-    {
-      "epoch": 2.6548715699176295,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 163410
-    },
-    {
-      "epoch": 2.6550340368149987,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 163420
-    },
-    {
-      "epoch": 2.655196503712369,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 163430
-    },
-    {
-      "epoch": 2.655358970609738,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 163440
-    },
-    {
-      "epoch": 2.655521437507108,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 163450
-    },
-    {
-      "epoch": 2.6556839044044773,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 163460
-    },
-    {
-      "epoch": 2.6558463713018474,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 163470
-    },
-    {
-      "epoch": 2.656008838199217,
-      "grad_norm": 2.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 163480
-    },
-    {
-      "epoch": 2.6561713050965867,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 163490
-    },
-    {
-      "epoch": 2.6563337719939564,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 163500
-    },
-    {
-      "epoch": 2.656496238891326,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.261,
-      "step": 163510
-    },
-    {
-      "epoch": 2.6566587057886957,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 163520
-    },
-    {
-      "epoch": 2.6568211726860653,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.26,
-      "step": 163530
-    },
-    {
-      "epoch": 2.656983639583435,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 163540
-    },
-    {
-      "epoch": 2.6571461064808046,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2654,
-      "step": 163550
-    },
-    {
-      "epoch": 2.6573085733781743,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 163560
-    },
-    {
-      "epoch": 2.657471040275544,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 163570
-    },
-    {
-      "epoch": 2.6576335071729136,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 163580
-    },
-    {
-      "epoch": 2.6577959740702832,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 163590
-    },
-    {
-      "epoch": 2.657958440967653,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 163600
-    },
-    {
-      "epoch": 2.6581209078650225,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3169,
-      "step": 163610
-    },
-    {
-      "epoch": 2.658283374762392,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2934,
-      "step": 163620
-    },
-    {
-      "epoch": 2.658445841659762,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 163630
-    },
-    {
-      "epoch": 2.6586083085571315,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2982,
-      "step": 163640
-    },
-    {
-      "epoch": 2.658770775454501,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 163650
-    },
-    {
-      "epoch": 2.658933242351871,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2653,
-      "step": 163660
-    },
-    {
-      "epoch": 2.6590957092492404,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 163670
-    },
-    {
-      "epoch": 2.65925817614661,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 163680
-    },
-    {
-      "epoch": 2.6594206430439797,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3099,
-      "step": 163690
-    },
-    {
-      "epoch": 2.6595831099413494,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 163700
-    },
-    {
-      "epoch": 2.659745576838719,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 163710
-    },
-    {
-      "epoch": 2.6599080437360887,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 163720
-    },
-    {
-      "epoch": 2.6600705106334583,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2886,
-      "step": 163730
-    },
-    {
-      "epoch": 2.660232977530828,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 163740
-    },
-    {
-      "epoch": 2.6603954444281976,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 163750
-    },
-    {
-      "epoch": 2.6605579113255673,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 163760
-    },
-    {
-      "epoch": 2.660720378222937,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2671,
-      "step": 163770
-    },
-    {
-      "epoch": 2.6608828451203066,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2813,
-      "step": 163780
-    },
-    {
-      "epoch": 2.6610453120176762,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2982,
-      "step": 163790
-    },
-    {
-      "epoch": 2.661207778915046,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 163800
-    },
-    {
-      "epoch": 2.6613702458124155,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 163810
-    },
-    {
-      "epoch": 2.6615327127097856,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2986,
-      "step": 163820
-    },
-    {
-      "epoch": 2.661695179607155,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2994,
-      "step": 163830
-    },
-    {
-      "epoch": 2.661857646504525,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 163840
-    },
-    {
-      "epoch": 2.662020113401894,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 163850
-    },
-    {
-      "epoch": 2.6621825802992642,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 163860
-    },
-    {
-      "epoch": 2.6623450471966335,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.29,
-      "step": 163870
-    },
-    {
-      "epoch": 2.6625075140940035,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 163880
-    },
-    {
-      "epoch": 2.6626699809913728,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 163890
-    },
-    {
-      "epoch": 2.662832447888743,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 163900
-    },
-    {
-      "epoch": 2.662994914786112,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2653,
-      "step": 163910
-    },
-    {
-      "epoch": 2.663157381683482,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 163920
-    },
-    {
-      "epoch": 2.663319848580852,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 163930
-    },
-    {
-      "epoch": 2.6634823154782215,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 163940
-    },
-    {
-      "epoch": 2.663644782375591,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 163950
-    },
-    {
-      "epoch": 2.6638072492729608,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 163960
-    },
-    {
-      "epoch": 2.6639697161703304,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2863,
-      "step": 163970
-    },
-    {
-      "epoch": 2.6641321830677,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 163980
-    },
-    {
-      "epoch": 2.6642946499650697,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.304,
-      "step": 163990
-    },
-    {
-      "epoch": 2.6644571168624394,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3002,
-      "step": 164000
-    },
-    {
-      "epoch": 2.664619583759809,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 164010
-    },
-    {
-      "epoch": 2.6647820506571787,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 164020
-    },
-    {
-      "epoch": 2.6649445175545483,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 164030
-    },
-    {
-      "epoch": 2.665106984451918,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 164040
-    },
-    {
-      "epoch": 2.6652694513492876,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2705,
-      "step": 164050
-    },
-    {
-      "epoch": 2.6654319182466573,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 164060
-    },
-    {
-      "epoch": 2.665594385144027,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2669,
-      "step": 164070
-    },
-    {
-      "epoch": 2.6657568520413966,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 164080
-    },
-    {
-      "epoch": 2.665919318938766,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 164090
-    },
-    {
-      "epoch": 2.666081785836136,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2657,
-      "step": 164100
-    },
-    {
-      "epoch": 2.6662442527335055,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 164110
-    },
-    {
-      "epoch": 2.666406719630875,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2594,
-      "step": 164120
-    },
-    {
-      "epoch": 2.666569186528245,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 164130
-    },
-    {
-      "epoch": 2.6667316534256145,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2688,
-      "step": 164140
-    },
-    {
-      "epoch": 2.666894120322984,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 164150
-    },
-    {
-      "epoch": 2.6670565872203538,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2482,
-      "step": 164160
-    },
-    {
-      "epoch": 2.6672190541177234,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 164170
-    },
-    {
-      "epoch": 2.667381521015093,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 164180
-    },
-    {
-      "epoch": 2.6675439879124627,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 164190
-    },
-    {
-      "epoch": 2.6677064548098324,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.258,
-      "step": 164200
-    },
-    {
-      "epoch": 2.667868921707202,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 164210
-    },
-    {
-      "epoch": 2.6680313886045717,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3127,
-      "step": 164220
-    },
-    {
-      "epoch": 2.6681938555019413,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 164230
-    },
-    {
-      "epoch": 2.668356322399311,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 164240
-    },
-    {
-      "epoch": 2.668518789296681,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 164250
-    },
-    {
-      "epoch": 2.6686812561940503,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3088,
-      "step": 164260
-    },
-    {
-      "epoch": 2.6688437230914204,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 164270
-    },
-    {
-      "epoch": 2.6690061899887896,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 164280
-    },
-    {
-      "epoch": 2.6691686568861597,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 164290
-    },
-    {
-      "epoch": 2.669331123783529,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 164300
-    },
-    {
-      "epoch": 2.669493590680899,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 164310
-    },
-    {
-      "epoch": 2.669656057578268,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 164320
-    },
-    {
-      "epoch": 2.6698185244756383,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 164330
-    },
-    {
-      "epoch": 2.6699809913730075,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 164340
-    },
-    {
-      "epoch": 2.6701434582703776,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 164350
-    },
-    {
-      "epoch": 2.6703059251677472,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 164360
-    },
-    {
-      "epoch": 2.670468392065117,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 164370
-    },
-    {
-      "epoch": 2.6706308589624865,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2604,
-      "step": 164380
-    },
-    {
-      "epoch": 2.670793325859856,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 164390
-    },
-    {
-      "epoch": 2.670955792757226,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2657,
-      "step": 164400
-    },
-    {
-      "epoch": 2.6711182596545955,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 164410
-    },
-    {
-      "epoch": 2.671280726551965,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 164420
-    },
-    {
-      "epoch": 2.671443193449335,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 164430
-    },
-    {
-      "epoch": 2.6716056603467044,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3018,
-      "step": 164440
-    },
-    {
-      "epoch": 2.671768127244074,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2619,
-      "step": 164450
-    },
-    {
-      "epoch": 2.6719305941414437,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2511,
-      "step": 164460
-    },
-    {
-      "epoch": 2.6720930610388134,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2596,
-      "step": 164470
-    },
-    {
-      "epoch": 2.672255527936183,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3094,
-      "step": 164480
-    },
-    {
-      "epoch": 2.6724179948335527,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2579,
-      "step": 164490
-    },
-    {
-      "epoch": 2.6725804617309223,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3,
-      "step": 164500
-    },
-    {
-      "epoch": 2.672742928628292,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 164510
-    },
-    {
-      "epoch": 2.6729053955256616,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 164520
-    },
-    {
-      "epoch": 2.6730678624230313,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 164530
-    },
-    {
-      "epoch": 2.673230329320401,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3055,
-      "step": 164540
-    },
-    {
-      "epoch": 2.6733927962177706,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2939,
-      "step": 164550
-    },
-    {
-      "epoch": 2.6735552631151402,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2956,
-      "step": 164560
-    },
-    {
-      "epoch": 2.67371773001251,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 164570
-    },
-    {
-      "epoch": 2.6738801969098795,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 164580
-    },
-    {
-      "epoch": 2.674042663807249,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 164590
-    },
-    {
-      "epoch": 2.674205130704619,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2996,
-      "step": 164600
-    },
-    {
-      "epoch": 2.6743675976019885,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 164610
-    },
-    {
-      "epoch": 2.674530064499358,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.302,
-      "step": 164620
-    },
-    {
-      "epoch": 2.674692531396728,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 164630
-    },
-    {
-      "epoch": 2.6748549982940975,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 164640
-    },
-    {
-      "epoch": 2.675017465191467,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2506,
-      "step": 164650
-    },
-    {
-      "epoch": 2.6751799320888368,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 164660
-    },
-    {
-      "epoch": 2.6753423989862064,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 164670
-    },
-    {
-      "epoch": 2.675504865883576,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2991,
-      "step": 164680
-    },
-    {
-      "epoch": 2.6756673327809457,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 164690
-    },
-    {
-      "epoch": 2.675829799678316,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 164700
-    },
-    {
-      "epoch": 2.675992266575685,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 164710
-    },
-    {
-      "epoch": 2.676154733473055,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 164720
-    },
-    {
-      "epoch": 2.6763172003704243,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 164730
-    },
-    {
-      "epoch": 2.6764796672677944,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 164740
-    },
-    {
-      "epoch": 2.6766421341651636,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 164750
-    },
-    {
-      "epoch": 2.6768046010625337,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 164760
-    },
-    {
-      "epoch": 2.676967067959903,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2566,
-      "step": 164770
-    },
-    {
-      "epoch": 2.677129534857273,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 164780
-    },
-    {
-      "epoch": 2.677292001754642,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.259,
-      "step": 164790
-    },
-    {
-      "epoch": 2.6774544686520123,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 164800
-    },
-    {
-      "epoch": 2.677616935549382,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2678,
-      "step": 164810
-    },
-    {
-      "epoch": 2.6777794024467516,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 164820
-    },
-    {
-      "epoch": 2.6779418693441213,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 164830
-    },
-    {
-      "epoch": 2.678104336241491,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3105,
-      "step": 164840
-    },
-    {
-      "epoch": 2.6782668031388606,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 164850
-    },
-    {
-      "epoch": 2.67842927003623,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3012,
-      "step": 164860
-    },
-    {
-      "epoch": 2.6785917369336,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2934,
-      "step": 164870
-    },
-    {
-      "epoch": 2.6787542038309695,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2965,
-      "step": 164880
-    },
-    {
-      "epoch": 2.678916670728339,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 164890
-    },
-    {
-      "epoch": 2.679079137625709,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2619,
-      "step": 164900
-    },
-    {
-      "epoch": 2.6792416045230785,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 164910
-    },
-    {
-      "epoch": 2.679404071420448,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 164920
-    },
-    {
-      "epoch": 2.6795665383178178,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 164930
-    },
-    {
-      "epoch": 2.6797290052151874,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 164940
-    },
-    {
-      "epoch": 2.679891472112557,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2574,
-      "step": 164950
-    },
-    {
-      "epoch": 2.6800539390099267,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 164960
-    },
-    {
-      "epoch": 2.6802164059072964,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 164970
-    },
-    {
-      "epoch": 2.680378872804666,
-      "grad_norm": 2.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2495,
-      "step": 164980
-    },
-    {
-      "epoch": 2.6805413397020357,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 164990
-    },
-    {
-      "epoch": 2.6807038065994053,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 165000
-    },
-    {
-      "epoch": 2.680866273496775,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 165010
-    },
-    {
-      "epoch": 2.6810287403941446,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 165020
-    },
-    {
-      "epoch": 2.6811912072915143,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3008,
-      "step": 165030
-    },
-    {
-      "epoch": 2.681353674188884,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 165040
-    },
-    {
-      "epoch": 2.6815161410862536,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 165050
-    },
-    {
-      "epoch": 2.6816786079836232,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 165060
-    },
-    {
-      "epoch": 2.681841074880993,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2968,
-      "step": 165070
-    },
-    {
-      "epoch": 2.6820035417783625,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 165080
-    },
-    {
-      "epoch": 2.682166008675732,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2956,
-      "step": 165090
-    },
-    {
-      "epoch": 2.682328475573102,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 165100
-    },
-    {
-      "epoch": 2.6824909424704715,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 165110
-    },
-    {
-      "epoch": 2.682653409367841,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 165120
-    },
-    {
-      "epoch": 2.6828158762652112,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 165130
-    },
-    {
-      "epoch": 2.6829783431625804,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 165140
-    },
-    {
-      "epoch": 2.6831408100599505,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 165150
-    },
-    {
-      "epoch": 2.6833032769573197,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 165160
-    },
-    {
-      "epoch": 2.68346574385469,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 165170
-    },
-    {
-      "epoch": 2.683628210752059,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 165180
-    },
-    {
-      "epoch": 2.683790677649429,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 165190
-    },
-    {
-      "epoch": 2.6839531445467983,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 165200
-    },
-    {
-      "epoch": 2.6841156114441684,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 165210
-    },
-    {
-      "epoch": 2.6842780783415376,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2595,
-      "step": 165220
-    },
-    {
-      "epoch": 2.6844405452389077,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 165230
-    },
-    {
-      "epoch": 2.6846030121362774,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 165240
-    },
-    {
-      "epoch": 2.684765479033647,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 165250
-    },
-    {
-      "epoch": 2.6849279459310167,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 165260
-    },
-    {
-      "epoch": 2.6850904128283863,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2955,
-      "step": 165270
-    },
-    {
-      "epoch": 2.685252879725756,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 165280
-    },
-    {
-      "epoch": 2.6854153466231256,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 165290
-    },
-    {
-      "epoch": 2.6855778135204953,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 165300
-    },
-    {
-      "epoch": 2.685740280417865,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 165310
-    },
-    {
-      "epoch": 2.6859027473152346,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 165320
-    },
-    {
-      "epoch": 2.6860652142126042,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2707,
-      "step": 165330
-    },
-    {
-      "epoch": 2.686227681109974,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2897,
-      "step": 165340
-    },
-    {
-      "epoch": 2.6863901480073435,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 165350
-    },
-    {
-      "epoch": 2.686552614904713,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3032,
-      "step": 165360
-    },
-    {
-      "epoch": 2.686715081802083,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2906,
-      "step": 165370
-    },
-    {
-      "epoch": 2.6868775486994525,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.3008,
-      "step": 165380
-    },
-    {
-      "epoch": 2.687040015596822,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 165390
-    },
-    {
-      "epoch": 2.687202482494192,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 165400
-    },
-    {
-      "epoch": 2.6873649493915615,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 165410
-    },
-    {
-      "epoch": 2.687527416288931,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 165420
-    },
-    {
-      "epoch": 2.6876898831863008,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 165430
-    },
-    {
-      "epoch": 2.6878523500836704,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2979,
-      "step": 165440
-    },
-    {
-      "epoch": 2.68801481698104,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2927,
-      "step": 165450
-    },
-    {
-      "epoch": 2.6881772838784097,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 165460
-    },
-    {
-      "epoch": 2.6883397507757794,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2679,
-      "step": 165470
-    },
-    {
-      "epoch": 2.688502217673149,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3205,
-      "step": 165480
-    },
-    {
-      "epoch": 2.6886646845705187,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 165490
-    },
-    {
-      "epoch": 2.6888271514678883,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2485,
-      "step": 165500
-    },
-    {
-      "epoch": 2.688989618365258,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 165510
-    },
-    {
-      "epoch": 2.6891520852626276,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2999,
-      "step": 165520
-    },
-    {
-      "epoch": 2.6893145521599973,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 165530
-    },
-    {
-      "epoch": 2.689477019057367,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 165540
-    },
-    {
-      "epoch": 2.6896394859547366,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 165550
-    },
-    {
-      "epoch": 2.6898019528521067,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2793,
-      "step": 165560
-    },
-    {
-      "epoch": 2.689964419749476,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 165570
-    },
-    {
-      "epoch": 2.690126886646846,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 165580
-    },
-    {
-      "epoch": 2.690289353544215,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3031,
-      "step": 165590
-    },
-    {
-      "epoch": 2.6904518204415853,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 165600
-    },
-    {
-      "epoch": 2.6906142873389545,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 165610
-    },
-    {
-      "epoch": 2.6907767542363246,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.312,
-      "step": 165620
-    },
-    {
-      "epoch": 2.6909392211336938,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 165630
-    },
-    {
-      "epoch": 2.691101688031064,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2968,
-      "step": 165640
-    },
-    {
-      "epoch": 2.691264154928433,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 165650
-    },
-    {
-      "epoch": 2.691426621825803,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2884,
-      "step": 165660
-    },
-    {
-      "epoch": 2.691589088723173,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 165670
-    },
-    {
-      "epoch": 2.6917515556205425,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 165680
-    },
-    {
-      "epoch": 2.691914022517912,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 165690
-    },
-    {
-      "epoch": 2.6920764894152818,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2599,
-      "step": 165700
-    },
-    {
-      "epoch": 2.6922389563126514,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 165710
-    },
-    {
-      "epoch": 2.692401423210021,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3041,
-      "step": 165720
-    },
-    {
-      "epoch": 2.6925638901073907,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2578,
-      "step": 165730
-    },
-    {
-      "epoch": 2.6927263570047604,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 165740
-    },
-    {
-      "epoch": 2.69288882390213,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 165750
-    },
-    {
-      "epoch": 2.6930512907994997,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 165760
-    },
-    {
-      "epoch": 2.6932137576968693,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3029,
-      "step": 165770
-    },
-    {
-      "epoch": 2.693376224594239,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2561,
-      "step": 165780
-    },
-    {
-      "epoch": 2.6935386914916086,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2475,
-      "step": 165790
-    },
-    {
-      "epoch": 2.6937011583889783,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2459,
-      "step": 165800
-    },
-    {
-      "epoch": 2.693863625286348,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2398,
-      "step": 165810
-    },
-    {
-      "epoch": 2.6940260921837176,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 165820
-    },
-    {
-      "epoch": 2.6941885590810872,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2633,
-      "step": 165830
-    },
-    {
-      "epoch": 2.694351025978457,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 165840
-    },
-    {
-      "epoch": 2.6945134928758265,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 165850
-    },
-    {
-      "epoch": 2.694675959773196,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 165860
-    },
-    {
-      "epoch": 2.694838426670566,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 165870
-    },
-    {
-      "epoch": 2.6950008935679355,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 165880
-    },
-    {
-      "epoch": 2.695163360465305,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.26,
-      "step": 165890
-    },
-    {
-      "epoch": 2.695325827362675,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3067,
-      "step": 165900
-    },
-    {
-      "epoch": 2.6954882942600444,
-      "grad_norm": 2.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 165910
-    },
-    {
-      "epoch": 2.695650761157414,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 165920
-    },
-    {
-      "epoch": 2.6958132280547837,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.299,
-      "step": 165930
-    },
-    {
-      "epoch": 2.6959756949521534,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2735,
-      "step": 165940
-    },
-    {
-      "epoch": 2.696138161849523,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 165950
-    },
-    {
-      "epoch": 2.6963006287468927,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 165960
-    },
-    {
-      "epoch": 2.6964630956442623,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 165970
-    },
-    {
-      "epoch": 2.696625562541632,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2577,
-      "step": 165980
-    },
-    {
-      "epoch": 2.6967880294390016,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 165990
-    },
-    {
-      "epoch": 2.6969504963363713,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 166000
-    },
-    {
-      "epoch": 2.6971129632337414,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.291,
-      "step": 166010
-    },
-    {
-      "epoch": 2.6972754301311106,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2921,
-      "step": 166020
-    },
-    {
-      "epoch": 2.6974378970284807,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 166030
-    },
-    {
-      "epoch": 2.69760036392585,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 166040
-    },
-    {
-      "epoch": 2.69776283082322,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2631,
-      "step": 166050
-    },
-    {
-      "epoch": 2.697925297720589,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2675,
-      "step": 166060
-    },
-    {
-      "epoch": 2.6980877646179593,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.29,
-      "step": 166070
-    },
-    {
-      "epoch": 2.6982502315153285,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 166080
-    },
-    {
-      "epoch": 2.6984126984126986,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 166090
-    },
-    {
-      "epoch": 2.698575165310068,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 166100
-    },
-    {
-      "epoch": 2.698737632207438,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 166110
-    },
-    {
-      "epoch": 2.6989000991048075,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 166120
-    },
-    {
-      "epoch": 2.699062566002177,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.307,
-      "step": 166130
-    },
-    {
-      "epoch": 2.699225032899547,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 166140
-    },
-    {
-      "epoch": 2.6993874997969165,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 166150
-    },
-    {
-      "epoch": 2.699549966694286,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 166160
-    },
-    {
-      "epoch": 2.699712433591656,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3029,
-      "step": 166170
-    },
-    {
-      "epoch": 2.6998749004890255,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 166180
-    },
-    {
-      "epoch": 2.700037367386395,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 166190
-    },
-    {
-      "epoch": 2.7001998342837648,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 166200
-    },
-    {
-      "epoch": 2.7003623011811344,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 166210
-    },
-    {
-      "epoch": 2.700524768078504,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 166220
-    },
-    {
-      "epoch": 2.7006872349758737,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 166230
-    },
-    {
-      "epoch": 2.7008497018732434,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2735,
-      "step": 166240
-    },
-    {
-      "epoch": 2.701012168770613,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 166250
-    },
-    {
-      "epoch": 2.7011746356679827,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 166260
-    },
-    {
-      "epoch": 2.7013371025653523,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2981,
-      "step": 166270
-    },
-    {
-      "epoch": 2.701499569462722,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 166280
-    },
-    {
-      "epoch": 2.7016620363600916,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3032,
-      "step": 166290
-    },
-    {
-      "epoch": 2.7018245032574613,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 166300
-    },
-    {
-      "epoch": 2.701986970154831,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 166310
-    },
-    {
-      "epoch": 2.7021494370522006,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2639,
-      "step": 166320
-    },
-    {
-      "epoch": 2.70231190394957,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 166330
-    },
-    {
-      "epoch": 2.70247437084694,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 166340
-    },
-    {
-      "epoch": 2.7026368377443095,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 166350
-    },
-    {
-      "epoch": 2.702799304641679,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3001,
-      "step": 166360
-    },
-    {
-      "epoch": 2.702961771539049,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 166370
-    },
-    {
-      "epoch": 2.7031242384364185,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2906,
-      "step": 166380
-    },
-    {
-      "epoch": 2.703286705333788,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 166390
-    },
-    {
-      "epoch": 2.7034491722311578,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 166400
-    },
-    {
-      "epoch": 2.7036116391285274,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 166410
-    },
-    {
-      "epoch": 2.703774106025897,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2657,
-      "step": 166420
-    },
-    {
-      "epoch": 2.7039365729232667,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2965,
-      "step": 166430
-    },
-    {
-      "epoch": 2.704099039820637,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2992,
-      "step": 166440
-    },
-    {
-      "epoch": 2.704261506718006,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 166450
-    },
-    {
-      "epoch": 2.704423973615376,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 166460
-    },
-    {
-      "epoch": 2.7045864405127453,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2581,
-      "step": 166470
-    },
-    {
-      "epoch": 2.7047489074101154,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 166480
-    },
-    {
-      "epoch": 2.7049113743074846,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.289,
-      "step": 166490
-    },
-    {
-      "epoch": 2.7050738412048547,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2728,
-      "step": 166500
-    },
-    {
-      "epoch": 2.705236308102224,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2577,
-      "step": 166510
-    },
-    {
-      "epoch": 2.705398774999594,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2581,
-      "step": 166520
-    },
-    {
-      "epoch": 2.7055612418969632,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 166530
-    },
-    {
-      "epoch": 2.7057237087943333,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 166540
-    },
-    {
-      "epoch": 2.705886175691703,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2979,
-      "step": 166550
-    },
-    {
-      "epoch": 2.7060486425890726,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2638,
-      "step": 166560
-    },
-    {
-      "epoch": 2.7062111094864423,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 166570
-    },
-    {
-      "epoch": 2.706373576383812,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.3117,
-      "step": 166580
-    },
-    {
-      "epoch": 2.7065360432811816,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 166590
-    },
-    {
-      "epoch": 2.7066985101785512,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 166600
-    },
-    {
-      "epoch": 2.706860977075921,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 166610
-    },
-    {
-      "epoch": 2.7070234439732905,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 166620
-    },
-    {
-      "epoch": 2.70718591087066,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 166630
-    },
-    {
-      "epoch": 2.70734837776803,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 166640
-    },
-    {
-      "epoch": 2.7075108446653995,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 166650
-    },
-    {
-      "epoch": 2.707673311562769,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.25,
-      "step": 166660
-    },
-    {
-      "epoch": 2.707835778460139,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 166670
-    },
-    {
-      "epoch": 2.7079982453575084,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 166680
-    },
-    {
-      "epoch": 2.708160712254878,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 166690
-    },
-    {
-      "epoch": 2.7083231791522477,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 166700
-    },
-    {
-      "epoch": 2.7084856460496174,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 166710
-    },
-    {
-      "epoch": 2.708648112946987,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 166720
-    },
-    {
-      "epoch": 2.7088105798443567,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 166730
-    },
-    {
-      "epoch": 2.7089730467417263,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 166740
-    },
-    {
-      "epoch": 2.709135513639096,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2578,
-      "step": 166750
-    },
-    {
-      "epoch": 2.7092979805364656,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 166760
-    },
-    {
-      "epoch": 2.7094604474338353,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2679,
-      "step": 166770
-    },
-    {
-      "epoch": 2.709622914331205,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 166780
-    },
-    {
-      "epoch": 2.7097853812285746,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 166790
-    },
-    {
-      "epoch": 2.7099478481259442,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 166800
-    },
-    {
-      "epoch": 2.710110315023314,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 166810
-    },
-    {
-      "epoch": 2.7102727819206835,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 166820
-    },
-    {
-      "epoch": 2.710435248818053,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 166830
-    },
-    {
-      "epoch": 2.710597715715423,
-      "grad_norm": 3.203125,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 166840
-    },
-    {
-      "epoch": 2.7107601826127925,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 166850
-    },
-    {
-      "epoch": 2.710922649510162,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 166860
-    },
-    {
-      "epoch": 2.711085116407532,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 166870
-    },
-    {
-      "epoch": 2.7112475833049015,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2974,
-      "step": 166880
-    },
-    {
-      "epoch": 2.7114100502022715,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 166890
-    },
-    {
-      "epoch": 2.7115725170996408,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 166900
-    },
-    {
-      "epoch": 2.711734983997011,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 166910
-    },
-    {
-      "epoch": 2.71189745089438,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 166920
-    },
-    {
-      "epoch": 2.71205991779175,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2531,
-      "step": 166930
-    },
-    {
-      "epoch": 2.7122223846891194,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2595,
-      "step": 166940
-    },
-    {
-      "epoch": 2.7123848515864895,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2875,
-      "step": 166950
-    },
-    {
-      "epoch": 2.7125473184838587,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 166960
-    },
-    {
-      "epoch": 2.7127097853812288,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2781,
-      "step": 166970
-    },
-    {
-      "epoch": 2.712872252278598,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 166980
-    },
-    {
-      "epoch": 2.713034719175968,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 166990
-    },
-    {
-      "epoch": 2.7131971860733377,
-      "grad_norm": 3.078125,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 167000
-    },
-    {
-      "epoch": 2.7133596529707074,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 167010
-    },
-    {
-      "epoch": 2.713522119868077,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 167020
-    },
-    {
-      "epoch": 2.7136845867654467,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.3081,
-      "step": 167030
-    },
-    {
-      "epoch": 2.7138470536628163,
-      "grad_norm": 3.015625,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 167040
-    },
-    {
-      "epoch": 2.714009520560186,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 167050
-    },
-    {
-      "epoch": 2.7141719874575556,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3087,
-      "step": 167060
-    },
-    {
-      "epoch": 2.7143344543549253,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3049,
-      "step": 167070
-    },
-    {
-      "epoch": 2.714496921252295,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2624,
-      "step": 167080
-    },
-    {
-      "epoch": 2.7146593881496646,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 167090
-    },
-    {
-      "epoch": 2.714821855047034,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 167100
-    },
-    {
-      "epoch": 2.714984321944404,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2985,
-      "step": 167110
-    },
-    {
-      "epoch": 2.7151467888417735,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 167120
-    },
-    {
-      "epoch": 2.715309255739143,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 167130
-    },
-    {
-      "epoch": 2.715471722636513,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3094,
-      "step": 167140
-    },
-    {
-      "epoch": 2.7156341895338825,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 167150
-    },
-    {
-      "epoch": 2.715796656431252,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2631,
-      "step": 167160
-    },
-    {
-      "epoch": 2.7159591233286218,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 167170
-    },
-    {
-      "epoch": 2.7161215902259914,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 167180
-    },
-    {
-      "epoch": 2.716284057123361,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2968,
-      "step": 167190
-    },
-    {
-      "epoch": 2.7164465240207307,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 167200
-    },
-    {
-      "epoch": 2.7166089909181004,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2627,
-      "step": 167210
-    },
-    {
-      "epoch": 2.71677145781547,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 167220
-    },
-    {
-      "epoch": 2.7169339247128397,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 167230
-    },
-    {
-      "epoch": 2.7170963916102093,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 167240
-    },
-    {
-      "epoch": 2.717258858507579,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 167250
-    },
-    {
-      "epoch": 2.7174213254049486,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 167260
-    },
-    {
-      "epoch": 2.7175837923023183,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.299,
-      "step": 167270
-    },
-    {
-      "epoch": 2.717746259199688,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 167280
-    },
-    {
-      "epoch": 2.7179087260970576,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 167290
-    },
-    {
-      "epoch": 2.7180711929944272,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 167300
-    },
-    {
-      "epoch": 2.718233659891797,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.3057,
-      "step": 167310
-    },
-    {
-      "epoch": 2.718396126789167,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 167320
-    },
-    {
-      "epoch": 2.718558593686536,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 167330
-    },
-    {
-      "epoch": 2.7187210605839063,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.312,
-      "step": 167340
-    },
-    {
-      "epoch": 2.7188835274812755,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 167350
-    },
-    {
-      "epoch": 2.7190459943786456,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 167360
-    },
-    {
-      "epoch": 2.719208461276015,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 167370
-    },
-    {
-      "epoch": 2.719370928173385,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 167380
-    },
-    {
-      "epoch": 2.719533395070754,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 167390
-    },
-    {
-      "epoch": 2.719695861968124,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 167400
-    },
-    {
-      "epoch": 2.7198583288654934,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 167410
-    },
-    {
-      "epoch": 2.7200207957628635,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 167420
-    },
-    {
-      "epoch": 2.720183262660233,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2638,
-      "step": 167430
-    },
-    {
-      "epoch": 2.720345729557603,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 167440
-    },
-    {
-      "epoch": 2.7205081964549724,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 167450
-    },
-    {
-      "epoch": 2.720670663352342,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 167460
-    },
-    {
-      "epoch": 2.7208331302497117,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 167470
-    },
-    {
-      "epoch": 2.7209955971470814,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 167480
-    },
-    {
-      "epoch": 2.721158064044451,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 167490
-    },
-    {
-      "epoch": 2.7213205309418207,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 167500
-    },
-    {
-      "epoch": 2.7214829978391903,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.306,
-      "step": 167510
-    },
-    {
-      "epoch": 2.72164546473656,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 167520
-    },
-    {
-      "epoch": 2.7218079316339296,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 167530
-    },
-    {
-      "epoch": 2.7219703985312993,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 167540
-    },
-    {
-      "epoch": 2.722132865428669,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 167550
-    },
-    {
-      "epoch": 2.7222953323260386,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 167560
-    },
-    {
-      "epoch": 2.7224577992234082,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2574,
-      "step": 167570
-    },
-    {
-      "epoch": 2.722620266120778,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 167580
-    },
-    {
-      "epoch": 2.7227827330181475,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 167590
-    },
-    {
-      "epoch": 2.722945199915517,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 167600
-    },
-    {
-      "epoch": 2.723107666812887,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2902,
-      "step": 167610
-    },
-    {
-      "epoch": 2.7232701337102565,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2934,
-      "step": 167620
-    },
-    {
-      "epoch": 2.723432600607626,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 167630
-    },
-    {
-      "epoch": 2.723595067504996,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 167640
-    },
-    {
-      "epoch": 2.7237575344023655,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 167650
-    },
-    {
-      "epoch": 2.723920001299735,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 167660
-    },
-    {
-      "epoch": 2.7240824681971048,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 167670
-    },
-    {
-      "epoch": 2.7242449350944744,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3139,
-      "step": 167680
-    },
-    {
-      "epoch": 2.724407401991844,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2981,
-      "step": 167690
-    },
-    {
-      "epoch": 2.7245698688892137,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 167700
-    },
-    {
-      "epoch": 2.7247323357865834,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 167710
-    },
-    {
-      "epoch": 2.724894802683953,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 167720
-    },
-    {
-      "epoch": 2.7250572695813227,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3034,
-      "step": 167730
-    },
-    {
-      "epoch": 2.7252197364786923,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 167740
-    },
-    {
-      "epoch": 2.725382203376062,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 167750
-    },
-    {
-      "epoch": 2.7255446702734316,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 167760
-    },
-    {
-      "epoch": 2.7257071371708017,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 167770
-    },
-    {
-      "epoch": 2.725869604068171,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 167780
-    },
-    {
-      "epoch": 2.726032070965541,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 167790
-    },
-    {
-      "epoch": 2.72619453786291,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 167800
-    },
-    {
-      "epoch": 2.7263570047602803,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2818,
-      "step": 167810
-    },
-    {
-      "epoch": 2.7265194716576495,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2738,
-      "step": 167820
-    },
-    {
-      "epoch": 2.7266819385550196,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2585,
-      "step": 167830
-    },
-    {
-      "epoch": 2.726844405452389,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2477,
-      "step": 167840
-    },
-    {
-      "epoch": 2.727006872349759,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3042,
-      "step": 167850
-    },
-    {
-      "epoch": 2.727169339247128,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.3122,
-      "step": 167860
-    },
-    {
-      "epoch": 2.727331806144498,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 167870
-    },
-    {
-      "epoch": 2.727494273041868,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 167880
-    },
-    {
-      "epoch": 2.7276567399392375,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 167890
-    },
-    {
-      "epoch": 2.727819206836607,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 167900
-    },
-    {
-      "epoch": 2.727981673733977,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 167910
-    },
-    {
-      "epoch": 2.7281441406313465,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 167920
-    },
-    {
-      "epoch": 2.728306607528716,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 167930
-    },
-    {
-      "epoch": 2.7284690744260858,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.3028,
-      "step": 167940
-    },
-    {
-      "epoch": 2.7286315413234554,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 167950
-    },
-    {
-      "epoch": 2.728794008220825,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 167960
-    },
-    {
-      "epoch": 2.7289564751181947,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 167970
-    },
-    {
-      "epoch": 2.7291189420155644,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 167980
-    },
-    {
-      "epoch": 2.729281408912934,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2987,
-      "step": 167990
-    },
-    {
-      "epoch": 2.7294438758103037,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.3055,
-      "step": 168000
-    },
-    {
-      "epoch": 2.7296063427076733,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 168010
-    },
-    {
-      "epoch": 2.729768809605043,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 168020
-    },
-    {
-      "epoch": 2.7299312765024126,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2526,
-      "step": 168030
-    },
-    {
-      "epoch": 2.7300937433997823,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2563,
-      "step": 168040
-    },
-    {
-      "epoch": 2.730256210297152,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 168050
-    },
-    {
-      "epoch": 2.7304186771945216,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 168060
-    },
-    {
-      "epoch": 2.7305811440918912,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3012,
-      "step": 168070
-    },
-    {
-      "epoch": 2.730743610989261,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3046,
-      "step": 168080
-    },
-    {
-      "epoch": 2.7309060778866305,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 168090
-    },
-    {
-      "epoch": 2.731068544784,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2895,
-      "step": 168100
-    },
-    {
-      "epoch": 2.73123101168137,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 168110
-    },
-    {
-      "epoch": 2.7313934785787395,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 168120
-    },
-    {
-      "epoch": 2.731555945476109,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3242,
-      "step": 168130
-    },
-    {
-      "epoch": 2.731718412373479,
-      "grad_norm": 2.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.3021,
-      "step": 168140
-    },
-    {
-      "epoch": 2.7318808792708484,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 168150
-    },
-    {
-      "epoch": 2.732043346168218,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3023,
-      "step": 168160
-    },
-    {
-      "epoch": 2.7322058130655877,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 168170
-    },
-    {
-      "epoch": 2.7323682799629574,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3116,
-      "step": 168180
-    },
-    {
-      "epoch": 2.732530746860327,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 168190
-    },
-    {
-      "epoch": 2.732693213757697,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 168200
-    },
-    {
-      "epoch": 2.7328556806550663,
-      "grad_norm": 2.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2594,
-      "step": 168210
-    },
-    {
-      "epoch": 2.7330181475524364,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2593,
-      "step": 168220
-    },
-    {
-      "epoch": 2.7331806144498056,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 168230
-    },
-    {
-      "epoch": 2.7333430813471757,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3163,
-      "step": 168240
-    },
-    {
-      "epoch": 2.733505548244545,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 168250
-    },
-    {
-      "epoch": 2.733668015141915,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 168260
-    },
-    {
-      "epoch": 2.7338304820392842,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 168270
-    },
-    {
-      "epoch": 2.7339929489366543,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.304,
-      "step": 168280
-    },
-    {
-      "epoch": 2.7341554158340236,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 168290
-    },
-    {
-      "epoch": 2.7343178827313936,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 168300
-    },
-    {
-      "epoch": 2.7344803496287633,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2625,
-      "step": 168310
-    },
-    {
-      "epoch": 2.734642816526133,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 168320
-    },
-    {
-      "epoch": 2.7348052834235026,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2852,
-      "step": 168330
-    },
-    {
-      "epoch": 2.7349677503208722,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 168340
-    },
-    {
-      "epoch": 2.735130217218242,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 168350
-    },
-    {
-      "epoch": 2.7352926841156116,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 168360
-    },
-    {
-      "epoch": 2.735455151012981,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2718,
-      "step": 168370
-    },
-    {
-      "epoch": 2.735617617910351,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2905,
-      "step": 168380
-    },
-    {
-      "epoch": 2.7357800848077205,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 168390
-    },
-    {
-      "epoch": 2.73594255170509,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 168400
-    },
-    {
-      "epoch": 2.73610501860246,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 168410
-    },
-    {
-      "epoch": 2.7362674854998295,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 168420
-    },
-    {
-      "epoch": 2.736429952397199,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 168430
-    },
-    {
-      "epoch": 2.7365924192945688,
-      "grad_norm": 2.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.3015,
-      "step": 168440
-    },
-    {
-      "epoch": 2.7367548861919384,
-      "grad_norm": 2.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 168450
-    },
-    {
-      "epoch": 2.736917353089308,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 168460
-    },
-    {
-      "epoch": 2.7370798199866777,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 168470
-    },
-    {
-      "epoch": 2.7372422868840474,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 168480
-    },
-    {
-      "epoch": 2.737404753781417,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 168490
-    },
-    {
-      "epoch": 2.7375672206787867,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 168500
-    },
-    {
-      "epoch": 2.7377296875761563,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 168510
-    },
-    {
-      "epoch": 2.737892154473526,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 168520
-    },
-    {
-      "epoch": 2.7380546213708956,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2519,
-      "step": 168530
-    },
-    {
-      "epoch": 2.7382170882682653,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 168540
-    },
-    {
-      "epoch": 2.738379555165635,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 168550
-    },
-    {
-      "epoch": 2.7385420220630046,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2565,
-      "step": 168560
-    },
-    {
-      "epoch": 2.738704488960374,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2507,
-      "step": 168570
-    },
-    {
-      "epoch": 2.738866955857744,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 168580
-    },
-    {
-      "epoch": 2.7390294227551135,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 168590
-    },
-    {
-      "epoch": 2.739191889652483,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 168600
-    },
-    {
-      "epoch": 2.739354356549853,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3092,
-      "step": 168610
-    },
-    {
-      "epoch": 2.7395168234472225,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.3146,
-      "step": 168620
-    },
-    {
-      "epoch": 2.7396792903445926,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.3023,
-      "step": 168630
-    },
-    {
-      "epoch": 2.7398417572419618,
-      "grad_norm": 2.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3013,
-      "step": 168640
-    },
-    {
-      "epoch": 2.740004224139332,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2909,
-      "step": 168650
-    },
-    {
-      "epoch": 2.740166691036701,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3031,
-      "step": 168660
-    },
-    {
-      "epoch": 2.740329157934071,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2962,
-      "step": 168670
-    },
-    {
-      "epoch": 2.7404916248314404,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.3041,
-      "step": 168680
-    },
-    {
-      "epoch": 2.7406540917288105,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 168690
-    },
-    {
-      "epoch": 2.7408165586261797,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 168700
-    },
-    {
-      "epoch": 2.7409790255235498,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2697,
-      "step": 168710
-    },
-    {
-      "epoch": 2.741141492420919,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3007,
-      "step": 168720
-    },
-    {
-      "epoch": 2.741303959318289,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 168730
-    },
-    {
-      "epoch": 2.7414664262156587,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2649,
-      "step": 168740
-    },
-    {
-      "epoch": 2.7416288931130284,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 168750
-    },
-    {
-      "epoch": 2.741791360010398,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 168760
-    },
-    {
-      "epoch": 2.7419538269077677,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 168770
-    },
-    {
-      "epoch": 2.7421162938051373,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3108,
-      "step": 168780
-    },
-    {
-      "epoch": 2.742278760702507,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 168790
-    },
-    {
-      "epoch": 2.7424412275998766,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 168800
-    },
-    {
-      "epoch": 2.7426036944972463,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.3007,
-      "step": 168810
-    },
-    {
-      "epoch": 2.742766161394616,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 168820
-    },
-    {
-      "epoch": 2.7429286282919856,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 168830
-    },
-    {
-      "epoch": 2.7430910951893552,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 168840
-    },
-    {
-      "epoch": 2.743253562086725,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 168850
-    },
-    {
-      "epoch": 2.7434160289840945,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 168860
-    },
-    {
-      "epoch": 2.743578495881464,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 168870
-    },
-    {
-      "epoch": 2.743740962778834,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 168880
-    },
-    {
-      "epoch": 2.7439034296762035,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2992,
-      "step": 168890
-    },
-    {
-      "epoch": 2.744065896573573,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 168900
-    },
-    {
-      "epoch": 2.744228363470943,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 168910
-    },
-    {
-      "epoch": 2.7443908303683124,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 168920
-    },
-    {
-      "epoch": 2.744553297265682,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.2314,
-      "step": 168930
-    },
-    {
-      "epoch": 2.7447157641630517,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 168940
-    },
-    {
-      "epoch": 2.7448782310604214,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 168950
-    },
-    {
-      "epoch": 2.745040697957791,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 168960
-    },
-    {
-      "epoch": 2.7452031648551607,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 168970
-    },
-    {
-      "epoch": 2.7453656317525303,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2456,
-      "step": 168980
-    },
-    {
-      "epoch": 2.7455280986499,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2616,
-      "step": 168990
-    },
-    {
-      "epoch": 2.7456905655472696,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 169000
-    },
-    {
-      "epoch": 2.7458530324446393,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 169010
-    },
-    {
-      "epoch": 2.746015499342009,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2718,
-      "step": 169020
-    },
-    {
-      "epoch": 2.7461779662393786,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.298,
-      "step": 169030
-    },
-    {
-      "epoch": 2.7463404331367482,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 169040
-    },
-    {
-      "epoch": 2.746502900034118,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 169050
-    },
-    {
-      "epoch": 2.7466653669314876,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 169060
-    },
-    {
-      "epoch": 2.746827833828857,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 169070
-    },
-    {
-      "epoch": 2.7469903007262273,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2919,
-      "step": 169080
-    },
-    {
-      "epoch": 2.7471527676235965,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 169090
-    },
-    {
-      "epoch": 2.7473152345209666,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.317,
-      "step": 169100
-    },
-    {
-      "epoch": 2.747477701418336,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 169110
-    },
-    {
-      "epoch": 2.747640168315706,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 169120
-    },
-    {
-      "epoch": 2.747802635213075,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 169130
-    },
-    {
-      "epoch": 2.747965102110445,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 169140
-    },
-    {
-      "epoch": 2.7481275690078144,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3017,
-      "step": 169150
-    },
-    {
-      "epoch": 2.7482900359051845,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 169160
-    },
-    {
-      "epoch": 2.7484525028025537,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 169170
-    },
-    {
-      "epoch": 2.748614969699924,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2886,
-      "step": 169180
-    },
-    {
-      "epoch": 2.7487774365972935,
-      "grad_norm": 2.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 169190
-    },
-    {
-      "epoch": 2.748939903494663,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 169200
-    },
-    {
-      "epoch": 2.7491023703920328,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.259,
-      "step": 169210
-    },
-    {
-      "epoch": 2.7492648372894024,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 169220
-    },
-    {
-      "epoch": 2.749427304186772,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 169230
-    },
-    {
-      "epoch": 2.7495897710841417,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2697,
-      "step": 169240
-    },
-    {
-      "epoch": 2.7497522379815114,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 169250
-    },
-    {
-      "epoch": 2.749914704878881,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 169260
-    },
-    {
-      "epoch": 2.7500771717762507,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2884,
-      "step": 169270
-    },
-    {
-      "epoch": 2.7502396386736203,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3058,
-      "step": 169280
-    },
-    {
-      "epoch": 2.75040210557099,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 169290
-    },
-    {
-      "epoch": 2.7505645724683596,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2927,
-      "step": 169300
-    },
-    {
-      "epoch": 2.7507270393657293,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 169310
-    },
-    {
-      "epoch": 2.750889506263099,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3075,
-      "step": 169320
-    },
-    {
-      "epoch": 2.7510519731604686,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 169330
-    },
-    {
-      "epoch": 2.751214440057838,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 169340
-    },
-    {
-      "epoch": 2.751376906955208,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.29,
-      "step": 169350
-    },
-    {
-      "epoch": 2.7515393738525775,
-      "grad_norm": 2.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3044,
-      "step": 169360
-    },
-    {
-      "epoch": 2.751701840749947,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.306,
-      "step": 169370
-    },
-    {
-      "epoch": 2.751864307647317,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 169380
-    },
-    {
-      "epoch": 2.7520267745446865,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 169390
-    },
-    {
-      "epoch": 2.752189241442056,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3053,
-      "step": 169400
-    },
-    {
-      "epoch": 2.7523517083394258,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 169410
-    },
-    {
-      "epoch": 2.7525141752367954,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 169420
-    },
-    {
-      "epoch": 2.752676642134165,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 169430
-    },
-    {
-      "epoch": 2.7528391090315347,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 169440
-    },
-    {
-      "epoch": 2.7530015759289044,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 169450
-    },
-    {
-      "epoch": 2.753164042826274,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2852,
-      "step": 169460
-    },
-    {
-      "epoch": 2.7533265097236437,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3155,
-      "step": 169470
-    },
-    {
-      "epoch": 2.7534889766210133,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3148,
-      "step": 169480
-    },
-    {
-      "epoch": 2.753651443518383,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 169490
-    },
-    {
-      "epoch": 2.7538139104157526,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 169500
-    },
-    {
-      "epoch": 2.7539763773131227,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3129,
-      "step": 169510
-    },
-    {
-      "epoch": 2.754138844210492,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 169520
-    },
-    {
-      "epoch": 2.754301311107862,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 169530
-    },
-    {
-      "epoch": 2.7544637780052312,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 169540
-    },
-    {
-      "epoch": 2.7546262449026013,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 169550
-    },
-    {
-      "epoch": 2.7547887117999705,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 169560
-    },
-    {
-      "epoch": 2.7549511786973406,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2676,
-      "step": 169570
-    },
-    {
-      "epoch": 2.75511364559471,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 169580
-    },
-    {
-      "epoch": 2.75527611249208,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 169590
-    },
-    {
-      "epoch": 2.755438579389449,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 169600
-    },
-    {
-      "epoch": 2.7556010462868192,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 169610
-    },
-    {
-      "epoch": 2.755763513184189,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 169620
-    },
-    {
-      "epoch": 2.7559259800815585,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 169630
-    },
-    {
-      "epoch": 2.756088446978928,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 169640
-    },
-    {
-      "epoch": 2.756250913876298,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 169650
-    },
-    {
-      "epoch": 2.7564133807736675,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 169660
-    },
-    {
-      "epoch": 2.756575847671037,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 169670
-    },
-    {
-      "epoch": 2.756738314568407,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 169680
-    },
-    {
-      "epoch": 2.7569007814657764,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 169690
-    },
-    {
-      "epoch": 2.757063248363146,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 169700
-    },
-    {
-      "epoch": 2.7572257152605157,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 169710
-    },
-    {
-      "epoch": 2.7573881821578854,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 169720
-    },
-    {
-      "epoch": 2.757550649055255,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3115,
-      "step": 169730
-    },
-    {
-      "epoch": 2.7577131159526247,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.294,
-      "step": 169740
-    },
-    {
-      "epoch": 2.7578755828499943,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 169750
-    },
-    {
-      "epoch": 2.758038049747364,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 169760
-    },
-    {
-      "epoch": 2.7582005166447336,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3087,
-      "step": 169770
-    },
-    {
-      "epoch": 2.7583629835421033,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 169780
-    },
-    {
-      "epoch": 2.758525450439473,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 169790
-    },
-    {
-      "epoch": 2.7586879173368426,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 169800
-    },
-    {
-      "epoch": 2.7588503842342122,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 169810
-    },
-    {
-      "epoch": 2.759012851131582,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2677,
-      "step": 169820
-    },
-    {
-      "epoch": 2.7591753180289516,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2999,
-      "step": 169830
-    },
-    {
-      "epoch": 2.759337784926321,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 169840
-    },
-    {
-      "epoch": 2.759500251823691,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.276,
-      "step": 169850
-    },
-    {
-      "epoch": 2.7596627187210605,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 169860
-    },
-    {
-      "epoch": 2.75982518561843,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 169870
-    },
-    {
-      "epoch": 2.7599876525158,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 169880
-    },
-    {
-      "epoch": 2.7601501194131695,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 169890
-    },
-    {
-      "epoch": 2.760312586310539,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 169900
-    },
-    {
-      "epoch": 2.7604750532079088,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 169910
-    },
-    {
-      "epoch": 2.7606375201052784,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2813,
-      "step": 169920
-    },
-    {
-      "epoch": 2.760799987002648,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 169930
-    },
-    {
-      "epoch": 2.7609624539000177,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2652,
-      "step": 169940
-    },
-    {
-      "epoch": 2.7611249207973874,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2736,
-      "step": 169950
-    },
-    {
-      "epoch": 2.7612873876947575,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 169960
-    },
-    {
-      "epoch": 2.7614498545921267,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 169970
-    },
-    {
-      "epoch": 2.7616123214894968,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 169980
-    },
-    {
-      "epoch": 2.761774788386866,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 169990
-    },
-    {
-      "epoch": 2.761937255284236,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 170000
-    },
-    {
-      "epoch": 2.7620997221816053,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 170010
-    },
-    {
-      "epoch": 2.7622621890789754,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3031,
-      "step": 170020
-    },
-    {
-      "epoch": 2.7624246559763446,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2793,
-      "step": 170030
-    },
-    {
-      "epoch": 2.7625871228737147,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 170040
-    },
-    {
-      "epoch": 2.762749589771084,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 170050
-    },
-    {
-      "epoch": 2.762912056668454,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 170060
-    },
-    {
-      "epoch": 2.7630745235658236,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 170070
-    },
-    {
-      "epoch": 2.7632369904631933,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2633,
-      "step": 170080
-    },
-    {
-      "epoch": 2.763399457360563,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2665,
-      "step": 170090
-    },
-    {
-      "epoch": 2.7635619242579326,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2677,
-      "step": 170100
-    },
-    {
-      "epoch": 2.763724391155302,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 170110
-    },
-    {
-      "epoch": 2.763886858052672,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 170120
-    },
-    {
-      "epoch": 2.7640493249500415,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.277,
-      "step": 170130
-    },
-    {
-      "epoch": 2.764211791847411,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 170140
-    },
-    {
-      "epoch": 2.764374258744781,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 170150
-    },
-    {
-      "epoch": 2.7645367256421505,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2566,
-      "step": 170160
-    },
-    {
-      "epoch": 2.76469919253952,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2494,
-      "step": 170170
-    },
-    {
-      "epoch": 2.7648616594368898,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 170180
-    },
-    {
-      "epoch": 2.7650241263342594,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.242,
-      "step": 170190
-    },
-    {
-      "epoch": 2.765186593231629,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 170200
-    },
-    {
-      "epoch": 2.7653490601289987,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 170210
-    },
-    {
-      "epoch": 2.7655115270263684,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 170220
-    },
-    {
-      "epoch": 2.765673993923738,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.299,
-      "step": 170230
-    },
-    {
-      "epoch": 2.7658364608211077,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 170240
-    },
-    {
-      "epoch": 2.7659989277184773,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 170250
-    },
-    {
-      "epoch": 2.766161394615847,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2921,
-      "step": 170260
-    },
-    {
-      "epoch": 2.7663238615132166,
-      "grad_norm": 2.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 170270
-    },
-    {
-      "epoch": 2.7664863284105863,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 170280
-    },
-    {
-      "epoch": 2.766648795307956,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 170290
-    },
-    {
-      "epoch": 2.7668112622053256,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 170300
-    },
-    {
-      "epoch": 2.7669737291026952,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2895,
-      "step": 170310
-    },
-    {
-      "epoch": 2.767136196000065,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.3049,
-      "step": 170320
-    },
-    {
-      "epoch": 2.7672986628974345,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 170330
-    },
-    {
-      "epoch": 2.767461129794804,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2764,
-      "step": 170340
-    },
-    {
-      "epoch": 2.767623596692174,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 170350
-    },
-    {
-      "epoch": 2.7677860635895435,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 170360
-    },
-    {
-      "epoch": 2.767948530486913,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 170370
-    },
-    {
-      "epoch": 2.768110997384283,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2884,
-      "step": 170380
-    },
-    {
-      "epoch": 2.768273464281653,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 170390
-    },
-    {
-      "epoch": 2.768435931179022,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 170400
-    },
-    {
-      "epoch": 2.768598398076392,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 170410
-    },
-    {
-      "epoch": 2.7687608649737614,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 170420
-    },
-    {
-      "epoch": 2.7689233318711315,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 170430
-    },
-    {
-      "epoch": 2.7690857987685007,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 170440
-    },
-    {
-      "epoch": 2.769248265665871,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2696,
-      "step": 170450
-    },
-    {
-      "epoch": 2.76941073256324,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.309,
-      "step": 170460
-    },
-    {
-      "epoch": 2.76957319946061,
-      "grad_norm": 2.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.3084,
-      "step": 170470
-    },
-    {
-      "epoch": 2.7697356663579793,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 170480
-    },
-    {
-      "epoch": 2.7698981332553494,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2989,
-      "step": 170490
-    },
-    {
-      "epoch": 2.770060600152719,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.3009,
-      "step": 170500
-    },
-    {
-      "epoch": 2.7702230670500887,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 170510
-    },
-    {
-      "epoch": 2.7703855339474583,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 170520
-    },
-    {
-      "epoch": 2.770548000844828,
-      "grad_norm": 2.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 170530
-    },
-    {
-      "epoch": 2.7707104677421976,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2633,
-      "step": 170540
-    },
-    {
-      "epoch": 2.7708729346395673,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 170550
-    },
-    {
-      "epoch": 2.771035401536937,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 170560
-    },
-    {
-      "epoch": 2.7711978684343066,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.263,
-      "step": 170570
-    },
-    {
-      "epoch": 2.7713603353316763,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 170580
-    },
-    {
-      "epoch": 2.771522802229046,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2522,
-      "step": 170590
-    },
-    {
-      "epoch": 2.7716852691264156,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 170600
-    },
-    {
-      "epoch": 2.771847736023785,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 170610
-    },
-    {
-      "epoch": 2.772010202921155,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.3053,
-      "step": 170620
-    },
-    {
-      "epoch": 2.7721726698185245,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 170630
-    },
-    {
-      "epoch": 2.772335136715894,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.302,
-      "step": 170640
-    },
-    {
-      "epoch": 2.772497603613264,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 170650
-    },
-    {
-      "epoch": 2.7726600705106335,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 170660
-    },
-    {
-      "epoch": 2.772822537408003,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2955,
-      "step": 170670
-    },
-    {
-      "epoch": 2.7729850043053728,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 170680
-    },
-    {
-      "epoch": 2.7731474712027424,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2886,
-      "step": 170690
-    },
-    {
-      "epoch": 2.773309938100112,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 170700
-    },
-    {
-      "epoch": 2.7734724049974817,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2513,
-      "step": 170710
-    },
-    {
-      "epoch": 2.7736348718948514,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 170720
-    },
-    {
-      "epoch": 2.773797338792221,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 170730
-    },
-    {
-      "epoch": 2.7739598056895907,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2957,
-      "step": 170740
-    },
-    {
-      "epoch": 2.7741222725869603,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2964,
-      "step": 170750
-    },
-    {
-      "epoch": 2.77428473948433,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 170760
-    },
-    {
-      "epoch": 2.7744472063816996,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 170770
-    },
-    {
-      "epoch": 2.7746096732790693,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 170780
-    },
-    {
-      "epoch": 2.774772140176439,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 170790
-    },
-    {
-      "epoch": 2.7749346070738086,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 170800
-    },
-    {
-      "epoch": 2.775097073971178,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 170810
-    },
-    {
-      "epoch": 2.775259540868548,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.305,
-      "step": 170820
-    },
-    {
-      "epoch": 2.7754220077659175,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2992,
-      "step": 170830
-    },
-    {
-      "epoch": 2.7755844746632876,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 170840
-    },
-    {
-      "epoch": 2.775746941560657,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 170850
-    },
-    {
-      "epoch": 2.775909408458027,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 170860
-    },
-    {
-      "epoch": 2.776071875355396,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2738,
-      "step": 170870
-    },
-    {
-      "epoch": 2.776234342252766,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 170880
-    },
-    {
-      "epoch": 2.7763968091501354,
-      "grad_norm": 2.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 170890
-    },
-    {
-      "epoch": 2.7765592760475055,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.3043,
-      "step": 170900
-    },
-    {
-      "epoch": 2.7767217429448747,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 170910
-    },
-    {
-      "epoch": 2.776884209842245,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 170920
-    },
-    {
-      "epoch": 2.777046676739614,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3045,
-      "step": 170930
-    },
-    {
-      "epoch": 2.777209143636984,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 170940
-    },
-    {
-      "epoch": 2.7773716105343538,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 170950
-    },
-    {
-      "epoch": 2.7775340774317234,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.3114,
-      "step": 170960
-    },
-    {
-      "epoch": 2.777696544329093,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 170970
-    },
-    {
-      "epoch": 2.7778590112264627,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 170980
-    },
-    {
-      "epoch": 2.7780214781238324,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2852,
-      "step": 170990
-    },
-    {
-      "epoch": 2.778183945021202,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 171000
-    },
-    {
-      "epoch": 2.7783464119185717,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2675,
-      "step": 171010
-    },
-    {
-      "epoch": 2.7785088788159413,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3016,
-      "step": 171020
-    },
-    {
-      "epoch": 2.778671345713311,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 171030
-    },
-    {
-      "epoch": 2.7788338126106806,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 171040
-    },
-    {
-      "epoch": 2.7789962795080503,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 171050
-    },
-    {
-      "epoch": 2.77915874640542,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 171060
-    },
-    {
-      "epoch": 2.7793212133027896,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3016,
-      "step": 171070
-    },
-    {
-      "epoch": 2.7794836802001592,
-      "grad_norm": 2.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2811,
-      "step": 171080
-    },
-    {
-      "epoch": 2.779646147097529,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 171090
-    },
-    {
-      "epoch": 2.7798086139948985,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 171100
-    },
-    {
-      "epoch": 2.779971080892268,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 171110
-    },
-    {
-      "epoch": 2.780133547789638,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.309,
-      "step": 171120
-    },
-    {
-      "epoch": 2.7802960146870075,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 171130
-    },
-    {
-      "epoch": 2.780458481584377,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 171140
-    },
-    {
-      "epoch": 2.780620948481747,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.3013,
-      "step": 171150
-    },
-    {
-      "epoch": 2.7807834153791164,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2878,
-      "step": 171160
-    },
-    {
-      "epoch": 2.780945882276486,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 171170
-    },
-    {
-      "epoch": 2.7811083491738557,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 171180
-    },
-    {
-      "epoch": 2.7812708160712254,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 171190
-    },
-    {
-      "epoch": 2.781433282968595,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 171200
-    },
-    {
-      "epoch": 2.7815957498659647,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3087,
-      "step": 171210
-    },
-    {
-      "epoch": 2.7817582167633343,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2902,
-      "step": 171220
-    },
-    {
-      "epoch": 2.781920683660704,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 171230
-    },
-    {
-      "epoch": 2.7820831505580736,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 171240
-    },
-    {
-      "epoch": 2.7822456174554433,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 171250
-    },
-    {
-      "epoch": 2.782408084352813,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 171260
-    },
-    {
-      "epoch": 2.782570551250183,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2601,
-      "step": 171270
-    },
-    {
-      "epoch": 2.7827330181475523,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 171280
-    },
-    {
-      "epoch": 2.7828954850449223,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 171290
-    },
-    {
-      "epoch": 2.7830579519422916,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 171300
-    },
-    {
-      "epoch": 2.7832204188396616,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.314,
-      "step": 171310
-    },
-    {
-      "epoch": 2.783382885737031,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.298,
-      "step": 171320
-    },
-    {
-      "epoch": 2.783545352634401,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2988,
-      "step": 171330
-    },
-    {
-      "epoch": 2.78370781953177,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 171340
-    },
-    {
-      "epoch": 2.7838702864291403,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 171350
-    },
-    {
-      "epoch": 2.7840327533265095,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2506,
-      "step": 171360
-    },
-    {
-      "epoch": 2.7841952202238796,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 171370
-    },
-    {
-      "epoch": 2.784357687121249,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2906,
-      "step": 171380
-    },
-    {
-      "epoch": 2.784520154018619,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2895,
-      "step": 171390
-    },
-    {
-      "epoch": 2.7846826209159885,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 171400
-    },
-    {
-      "epoch": 2.784845087813358,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 171410
-    },
-    {
-      "epoch": 2.785007554710728,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 171420
-    },
-    {
-      "epoch": 2.7851700216080975,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 171430
-    },
-    {
-      "epoch": 2.785332488505467,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2717,
-      "step": 171440
-    },
-    {
-      "epoch": 2.7854949554028368,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 171450
-    },
-    {
-      "epoch": 2.7856574223002064,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2674,
-      "step": 171460
-    },
-    {
-      "epoch": 2.785819889197576,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.274,
-      "step": 171470
-    },
-    {
-      "epoch": 2.7859823560949457,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 171480
-    },
-    {
-      "epoch": 2.7861448229923154,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2596,
-      "step": 171490
-    },
-    {
-      "epoch": 2.786307289889685,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 171500
-    },
-    {
-      "epoch": 2.7864697567870547,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 171510
-    },
-    {
-      "epoch": 2.7866322236844243,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3159,
-      "step": 171520
-    },
-    {
-      "epoch": 2.786794690581794,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2852,
-      "step": 171530
-    },
-    {
-      "epoch": 2.7869571574791636,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3089,
-      "step": 171540
-    },
-    {
-      "epoch": 2.7871196243765333,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 171550
-    },
-    {
-      "epoch": 2.787282091273903,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 171560
-    },
-    {
-      "epoch": 2.7874445581712726,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 171570
-    },
-    {
-      "epoch": 2.787607025068642,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3037,
-      "step": 171580
-    },
-    {
-      "epoch": 2.787769491966012,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 171590
-    },
-    {
-      "epoch": 2.7879319588633815,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 171600
-    },
-    {
-      "epoch": 2.788094425760751,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 171610
-    },
-    {
-      "epoch": 2.788256892658121,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2933,
-      "step": 171620
-    },
-    {
-      "epoch": 2.7884193595554905,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 171630
-    },
-    {
-      "epoch": 2.78858182645286,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 171640
-    },
-    {
-      "epoch": 2.7887442933502298,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 171650
-    },
-    {
-      "epoch": 2.7889067602475994,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2906,
-      "step": 171660
-    },
-    {
-      "epoch": 2.789069227144969,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 171670
-    },
-    {
-      "epoch": 2.7892316940423387,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 171680
-    },
-    {
-      "epoch": 2.7893941609397084,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 171690
-    },
-    {
-      "epoch": 2.789556627837078,
-      "grad_norm": 2.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2639,
-      "step": 171700
-    },
-    {
-      "epoch": 2.7897190947344477,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2573,
-      "step": 171710
-    },
-    {
-      "epoch": 2.7898815616318178,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 171720
-    },
-    {
-      "epoch": 2.790044028529187,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3,
-      "step": 171730
-    },
-    {
-      "epoch": 2.790206495426557,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 171740
-    },
-    {
-      "epoch": 2.7903689623239263,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 171750
-    },
-    {
-      "epoch": 2.7905314292212964,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 171760
-    },
-    {
-      "epoch": 2.7906938961186656,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2467,
-      "step": 171770
-    },
-    {
-      "epoch": 2.7908563630160357,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2459,
-      "step": 171780
-    },
-    {
-      "epoch": 2.791018829913405,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 171790
-    },
-    {
-      "epoch": 2.791181296810775,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2573,
-      "step": 171800
-    },
-    {
-      "epoch": 2.7913437637081446,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2807,
-      "step": 171810
-    },
-    {
-      "epoch": 2.7915062306055143,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 171820
-    },
-    {
-      "epoch": 2.791668697502884,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 171830
-    },
-    {
-      "epoch": 2.7918311644002536,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 171840
-    },
-    {
-      "epoch": 2.7919936312976232,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2626,
-      "step": 171850
-    },
-    {
-      "epoch": 2.792156098194993,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2656,
-      "step": 171860
-    },
-    {
-      "epoch": 2.7923185650923625,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2542,
-      "step": 171870
-    },
-    {
-      "epoch": 2.792481031989732,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 171880
-    },
-    {
-      "epoch": 2.792643498887102,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2371,
-      "step": 171890
-    },
-    {
-      "epoch": 2.7928059657844715,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2583,
-      "step": 171900
-    },
-    {
-      "epoch": 2.792968432681841,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2439,
-      "step": 171910
-    },
-    {
-      "epoch": 2.793130899579211,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2467,
-      "step": 171920
-    },
-    {
-      "epoch": 2.7932933664765804,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 171930
-    },
-    {
-      "epoch": 2.79345583337395,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 171940
-    },
-    {
-      "epoch": 2.7936183002713197,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2721,
-      "step": 171950
-    },
-    {
-      "epoch": 2.7937807671686894,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2633,
-      "step": 171960
-    },
-    {
-      "epoch": 2.793943234066059,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 171970
-    },
-    {
-      "epoch": 2.7941057009634287,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 171980
-    },
-    {
-      "epoch": 2.7942681678607983,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 171990
-    },
-    {
-      "epoch": 2.794430634758168,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 172000
-    },
-    {
-      "epoch": 2.7945931016555376,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 172010
-    },
-    {
-      "epoch": 2.7947555685529073,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 172020
-    },
-    {
-      "epoch": 2.794918035450277,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 172030
-    },
-    {
-      "epoch": 2.7950805023476466,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 172040
-    },
-    {
-      "epoch": 2.7952429692450163,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 172050
-    },
-    {
-      "epoch": 2.795405436142386,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3043,
-      "step": 172060
-    },
-    {
-      "epoch": 2.7955679030397556,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 172070
-    },
-    {
-      "epoch": 2.795730369937125,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2916,
-      "step": 172080
-    },
-    {
-      "epoch": 2.795892836834495,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 172090
-    },
-    {
-      "epoch": 2.7960553037318645,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3072,
-      "step": 172100
-    },
-    {
-      "epoch": 2.796217770629234,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 172110
-    },
-    {
-      "epoch": 2.796380237526604,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3077,
-      "step": 172120
-    },
-    {
-      "epoch": 2.7965427044239735,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.3126,
-      "step": 172130
-    },
-    {
-      "epoch": 2.796705171321343,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2921,
-      "step": 172140
-    },
-    {
-      "epoch": 2.796867638218713,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.294,
-      "step": 172150
-    },
-    {
-      "epoch": 2.7970301051160824,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2874,
-      "step": 172160
-    },
-    {
-      "epoch": 2.7971925720134525,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 172170
-    },
-    {
-      "epoch": 2.7973550389108217,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 172180
-    },
-    {
-      "epoch": 2.797517505808192,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 172190
-    },
-    {
-      "epoch": 2.797679972705561,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 172200
-    },
-    {
-      "epoch": 2.797842439602931,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 172210
-    },
-    {
-      "epoch": 2.7980049065003003,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 172220
-    },
-    {
-      "epoch": 2.7981673733976704,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 172230
-    },
-    {
-      "epoch": 2.7983298402950396,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 172240
-    },
-    {
-      "epoch": 2.7984923071924097,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 172250
-    },
-    {
-      "epoch": 2.7986547740897794,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 172260
-    },
-    {
-      "epoch": 2.798817240987149,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 172270
-    },
-    {
-      "epoch": 2.7989797078845187,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 172280
-    },
-    {
-      "epoch": 2.7991421747818883,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3091,
-      "step": 172290
-    },
-    {
-      "epoch": 2.799304641679258,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 172300
-    },
-    {
-      "epoch": 2.7994671085766276,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2595,
-      "step": 172310
-    },
-    {
-      "epoch": 2.7996295754739973,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2776,
-      "step": 172320
-    },
-    {
-      "epoch": 2.799792042371367,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 172330
-    },
-    {
-      "epoch": 2.7999545092687366,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2625,
-      "step": 172340
-    },
-    {
-      "epoch": 2.800116976166106,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2566,
-      "step": 172350
-    },
-    {
-      "epoch": 2.800279443063476,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 172360
-    },
-    {
-      "epoch": 2.8004419099608455,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 172370
-    },
-    {
-      "epoch": 2.800604376858215,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.31,
-      "step": 172380
-    },
-    {
-      "epoch": 2.800766843755585,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 172390
-    },
-    {
-      "epoch": 2.8009293106529545,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2592,
-      "step": 172400
-    },
-    {
-      "epoch": 2.801091777550324,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2524,
-      "step": 172410
-    },
-    {
-      "epoch": 2.8012542444476938,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2781,
-      "step": 172420
-    },
-    {
-      "epoch": 2.8014167113450634,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 172430
-    },
-    {
-      "epoch": 2.801579178242433,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2644,
-      "step": 172440
-    },
-    {
-      "epoch": 2.8017416451398027,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2659,
-      "step": 172450
-    },
-    {
-      "epoch": 2.8019041120371724,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2635,
-      "step": 172460
-    },
-    {
-      "epoch": 2.802066578934542,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 172470
-    },
-    {
-      "epoch": 2.8022290458319117,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 172480
-    },
-    {
-      "epoch": 2.8023915127292813,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 172490
-    },
-    {
-      "epoch": 2.802553979626651,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 172500
-    },
-    {
-      "epoch": 2.8027164465240206,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.262,
-      "step": 172510
-    },
-    {
-      "epoch": 2.8028789134213903,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 172520
-    },
-    {
-      "epoch": 2.80304138031876,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2863,
-      "step": 172530
-    },
-    {
-      "epoch": 2.8032038472161296,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 172540
-    },
-    {
-      "epoch": 2.8033663141134992,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 172550
-    },
-    {
-      "epoch": 2.803528781010869,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 172560
-    },
-    {
-      "epoch": 2.8036912479082385,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 172570
-    },
-    {
-      "epoch": 2.8038537148056086,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 172580
-    },
-    {
-      "epoch": 2.804016181702978,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 172590
-    },
-    {
-      "epoch": 2.804178648600348,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 172600
-    },
-    {
-      "epoch": 2.804341115497717,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 172610
-    },
-    {
-      "epoch": 2.8045035823950872,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 172620
-    },
-    {
-      "epoch": 2.8046660492924564,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2629,
-      "step": 172630
-    },
-    {
-      "epoch": 2.8048285161898265,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 172640
-    },
-    {
-      "epoch": 2.8049909830871957,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 172650
-    },
-    {
-      "epoch": 2.805153449984566,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 172660
-    },
-    {
-      "epoch": 2.805315916881935,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2649,
-      "step": 172670
-    },
-    {
-      "epoch": 2.805478383779305,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2675,
-      "step": 172680
-    },
-    {
-      "epoch": 2.805640850676675,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2539,
-      "step": 172690
-    },
-    {
-      "epoch": 2.8058033175740444,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3148,
-      "step": 172700
-    },
-    {
-      "epoch": 2.805965784471414,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2824,
-      "step": 172710
-    },
-    {
-      "epoch": 2.8061282513687837,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 172720
-    },
-    {
-      "epoch": 2.8062907182661534,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 172730
-    },
-    {
-      "epoch": 2.806453185163523,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.3079,
-      "step": 172740
-    },
-    {
-      "epoch": 2.8066156520608927,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 172750
-    },
-    {
-      "epoch": 2.8067781189582623,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2683,
-      "step": 172760
-    },
-    {
-      "epoch": 2.806940585855632,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 172770
-    },
-    {
-      "epoch": 2.8071030527530016,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2943,
-      "step": 172780
-    },
-    {
-      "epoch": 2.8072655196503713,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 172790
-    },
-    {
-      "epoch": 2.807427986547741,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 172800
-    },
-    {
-      "epoch": 2.8075904534451106,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 172810
-    },
-    {
-      "epoch": 2.8077529203424803,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2823,
-      "step": 172820
-    },
-    {
-      "epoch": 2.80791538723985,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 172830
-    },
-    {
-      "epoch": 2.8080778541372196,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3008,
-      "step": 172840
-    },
-    {
-      "epoch": 2.808240321034589,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 172850
-    },
-    {
-      "epoch": 2.808402787931959,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 172860
-    },
-    {
-      "epoch": 2.8085652548293285,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 172870
-    },
-    {
-      "epoch": 2.808727721726698,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 172880
-    },
-    {
-      "epoch": 2.808890188624068,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 172890
-    },
-    {
-      "epoch": 2.8090526555214375,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2984,
-      "step": 172900
-    },
-    {
-      "epoch": 2.809215122418807,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 172910
-    },
-    {
-      "epoch": 2.8093775893161768,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 172920
-    },
-    {
-      "epoch": 2.8095400562135464,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 172930
-    },
-    {
-      "epoch": 2.809702523110916,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 172940
-    },
-    {
-      "epoch": 2.8098649900082857,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 172950
-    },
-    {
-      "epoch": 2.8100274569056554,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3013,
-      "step": 172960
-    },
-    {
-      "epoch": 2.810189923803025,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2714,
-      "step": 172970
-    },
-    {
-      "epoch": 2.8103523907003947,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2705,
-      "step": 172980
-    },
-    {
-      "epoch": 2.8105148575977643,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.3045,
-      "step": 172990
-    },
-    {
-      "epoch": 2.810677324495134,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2982,
-      "step": 173000
-    },
-    {
-      "epoch": 2.8108397913925036,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 173010
-    },
-    {
-      "epoch": 2.8110022582898733,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 173020
-    },
-    {
-      "epoch": 2.8111647251872434,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.302,
-      "step": 173030
-    },
-    {
-      "epoch": 2.8113271920846126,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 173040
-    },
-    {
-      "epoch": 2.8114896589819827,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 173050
-    },
-    {
-      "epoch": 2.811652125879352,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 173060
-    },
-    {
-      "epoch": 2.811814592776722,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 173070
-    },
-    {
-      "epoch": 2.811977059674091,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2609,
-      "step": 173080
-    },
-    {
-      "epoch": 2.8121395265714613,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 173090
-    },
-    {
-      "epoch": 2.8123019934688305,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2969,
-      "step": 173100
-    },
-    {
-      "epoch": 2.8124644603662006,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 173110
-    },
-    {
-      "epoch": 2.8126269272635698,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.302,
-      "step": 173120
-    },
-    {
-      "epoch": 2.81278939416094,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 173130
-    },
-    {
-      "epoch": 2.8129518610583095,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.3117,
-      "step": 173140
-    },
-    {
-      "epoch": 2.813114327955679,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3009,
-      "step": 173150
-    },
-    {
-      "epoch": 2.813276794853049,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3027,
-      "step": 173160
-    },
-    {
-      "epoch": 2.8134392617504185,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2989,
-      "step": 173170
-    },
-    {
-      "epoch": 2.813601728647788,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 173180
-    },
-    {
-      "epoch": 2.8137641955451578,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2698,
-      "step": 173190
-    },
-    {
-      "epoch": 2.8139266624425274,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 173200
-    },
-    {
-      "epoch": 2.814089129339897,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 173210
-    },
-    {
-      "epoch": 2.8142515962372667,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2991,
-      "step": 173220
-    },
-    {
-      "epoch": 2.8144140631346364,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2978,
-      "step": 173230
-    },
-    {
-      "epoch": 2.814576530032006,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 173240
-    },
-    {
-      "epoch": 2.8147389969293757,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 173250
-    },
-    {
-      "epoch": 2.8149014638267453,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2978,
-      "step": 173260
-    },
-    {
-      "epoch": 2.815063930724115,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 173270
-    },
-    {
-      "epoch": 2.8152263976214846,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 173280
-    },
-    {
-      "epoch": 2.8153888645188543,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 173290
-    },
-    {
-      "epoch": 2.815551331416224,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2852,
-      "step": 173300
-    },
-    {
-      "epoch": 2.8157137983135936,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 173310
-    },
-    {
-      "epoch": 2.8158762652109632,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 173320
-    },
-    {
-      "epoch": 2.816038732108333,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 173330
-    },
-    {
-      "epoch": 2.8162011990057025,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 173340
-    },
-    {
-      "epoch": 2.816363665903072,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3083,
-      "step": 173350
-    },
-    {
-      "epoch": 2.816526132800442,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 173360
-    },
-    {
-      "epoch": 2.8166885996978115,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 173370
-    },
-    {
-      "epoch": 2.816851066595181,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 173380
-    },
-    {
-      "epoch": 2.817013533492551,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 173390
-    },
-    {
-      "epoch": 2.8171760003899204,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.275,
-      "step": 173400
-    },
-    {
-      "epoch": 2.81733846728729,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 173410
-    },
-    {
-      "epoch": 2.8175009341846597,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2602,
-      "step": 173420
-    },
-    {
-      "epoch": 2.8176634010820294,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 173430
-    },
-    {
-      "epoch": 2.817825867979399,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3007,
-      "step": 173440
-    },
-    {
-      "epoch": 2.8179883348767687,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3124,
-      "step": 173450
-    },
-    {
-      "epoch": 2.818150801774139,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 173460
-    },
-    {
-      "epoch": 2.818313268671508,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.276,
-      "step": 173470
-    },
-    {
-      "epoch": 2.818475735568878,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2712,
-      "step": 173480
-    },
-    {
-      "epoch": 2.8186382024662473,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 173490
-    },
-    {
-      "epoch": 2.8188006693636174,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 173500
-    },
-    {
-      "epoch": 2.8189631362609866,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 173510
-    },
-    {
-      "epoch": 2.8191256031583567,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 173520
-    },
-    {
-      "epoch": 2.819288070055726,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.3069,
-      "step": 173530
-    },
-    {
-      "epoch": 2.819450536953096,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 173540
-    },
-    {
-      "epoch": 2.819613003850465,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 173550
-    },
-    {
-      "epoch": 2.8197754707478353,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 173560
-    },
-    {
-      "epoch": 2.819937937645205,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2882,
-      "step": 173570
-    },
-    {
-      "epoch": 2.8201004045425746,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3003,
-      "step": 173580
-    },
-    {
-      "epoch": 2.8202628714399443,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 173590
-    },
-    {
-      "epoch": 2.820425338337314,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 173600
-    },
-    {
-      "epoch": 2.8205878052346836,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 173610
-    },
-    {
-      "epoch": 2.820750272132053,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2648,
-      "step": 173620
-    },
-    {
-      "epoch": 2.820912739029423,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 173630
-    },
-    {
-      "epoch": 2.8210752059267925,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 173640
-    },
-    {
-      "epoch": 2.821237672824162,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 173650
-    },
-    {
-      "epoch": 2.821400139721532,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 173660
-    },
-    {
-      "epoch": 2.8215626066189015,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 173670
-    },
-    {
-      "epoch": 2.821725073516271,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 173680
-    },
-    {
-      "epoch": 2.8218875404136408,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.264,
-      "step": 173690
-    },
-    {
-      "epoch": 2.8220500073110104,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 173700
-    },
-    {
-      "epoch": 2.82221247420838,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 173710
-    },
-    {
-      "epoch": 2.8223749411057497,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 173720
-    },
-    {
-      "epoch": 2.8225374080031194,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3108,
-      "step": 173730
-    },
-    {
-      "epoch": 2.822699874900489,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 173740
-    },
-    {
-      "epoch": 2.8228623417978587,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 173750
-    },
-    {
-      "epoch": 2.8230248086952283,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 173760
-    },
-    {
-      "epoch": 2.823187275592598,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 173770
-    },
-    {
-      "epoch": 2.8233497424899676,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 173780
-    },
-    {
-      "epoch": 2.8235122093873373,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2939,
-      "step": 173790
-    },
-    {
-      "epoch": 2.823674676284707,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 173800
-    },
-    {
-      "epoch": 2.8238371431820766,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 173810
-    },
-    {
-      "epoch": 2.823999610079446,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3143,
-      "step": 173820
-    },
-    {
-      "epoch": 2.824162076976816,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 173830
-    },
-    {
-      "epoch": 2.8243245438741855,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 173840
-    },
-    {
-      "epoch": 2.824487010771555,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 173850
-    },
-    {
-      "epoch": 2.824649477668925,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2836,
-      "step": 173860
-    },
-    {
-      "epoch": 2.8248119445662945,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 173870
-    },
-    {
-      "epoch": 2.824974411463664,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 173880
-    },
-    {
-      "epoch": 2.8251368783610338,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2988,
-      "step": 173890
-    },
-    {
-      "epoch": 2.8252993452584034,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 173900
-    },
-    {
-      "epoch": 2.8254618121557735,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2956,
-      "step": 173910
-    },
-    {
-      "epoch": 2.8256242790531427,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 173920
-    },
-    {
-      "epoch": 2.825786745950513,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 173930
-    },
-    {
-      "epoch": 2.825949212847882,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 173940
-    },
-    {
-      "epoch": 2.826111679745252,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2751,
-      "step": 173950
-    },
-    {
-      "epoch": 2.8262741466426213,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 173960
-    },
-    {
-      "epoch": 2.8264366135399914,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2988,
-      "step": 173970
-    },
-    {
-      "epoch": 2.8265990804373606,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 173980
-    },
-    {
-      "epoch": 2.8267615473347307,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 173990
-    },
-    {
-      "epoch": 2.8269240142321,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2921,
-      "step": 174000
-    },
-    {
-      "epoch": 2.82708648112947,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 174010
-    },
-    {
-      "epoch": 2.8272489480268397,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 174020
-    },
-    {
-      "epoch": 2.8274114149242093,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 174030
-    },
-    {
-      "epoch": 2.827573881821579,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 174040
-    },
-    {
-      "epoch": 2.8277363487189486,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3086,
-      "step": 174050
-    },
-    {
-      "epoch": 2.8278988156163183,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2801,
-      "step": 174060
-    },
-    {
-      "epoch": 2.828061282513688,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3248,
-      "step": 174070
-    },
-    {
-      "epoch": 2.8282237494110576,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 174080
-    },
-    {
-      "epoch": 2.8283862163084272,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2593,
-      "step": 174090
-    },
-    {
-      "epoch": 2.828548683205797,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2647,
-      "step": 174100
-    },
-    {
-      "epoch": 2.8287111501031665,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 174110
-    },
-    {
-      "epoch": 2.828873617000536,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2689,
-      "step": 174120
-    },
-    {
-      "epoch": 2.829036083897906,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 174130
-    },
-    {
-      "epoch": 2.8291985507952755,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 174140
-    },
-    {
-      "epoch": 2.829361017692645,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2643,
-      "step": 174150
-    },
-    {
-      "epoch": 2.829523484590015,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 174160
-    },
-    {
-      "epoch": 2.8296859514873844,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2965,
-      "step": 174170
-    },
-    {
-      "epoch": 2.829848418384754,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 174180
-    },
-    {
-      "epoch": 2.8300108852821237,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.3114,
-      "step": 174190
-    },
-    {
-      "epoch": 2.8301733521794934,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 174200
-    },
-    {
-      "epoch": 2.830335819076863,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 174210
-    },
-    {
-      "epoch": 2.8304982859742327,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 174220
-    },
-    {
-      "epoch": 2.8306607528716023,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 174230
-    },
-    {
-      "epoch": 2.830823219768972,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2884,
-      "step": 174240
-    },
-    {
-      "epoch": 2.8309856866663416,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 174250
-    },
-    {
-      "epoch": 2.8311481535637113,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 174260
-    },
-    {
-      "epoch": 2.831310620461081,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 174270
-    },
-    {
-      "epoch": 2.8314730873584506,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 174280
-    },
-    {
-      "epoch": 2.8316355542558203,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 174290
-    },
-    {
-      "epoch": 2.83179802115319,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2695,
-      "step": 174300
-    },
-    {
-      "epoch": 2.8319604880505596,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 174310
-    },
-    {
-      "epoch": 2.832122954947929,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 174320
-    },
-    {
-      "epoch": 2.832285421845299,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 174330
-    },
-    {
-      "epoch": 2.832447888742669,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 174340
-    },
-    {
-      "epoch": 2.832610355640038,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 174350
-    },
-    {
-      "epoch": 2.8327728225374083,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2688,
-      "step": 174360
-    },
-    {
-      "epoch": 2.8329352894347775,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 174370
-    },
-    {
-      "epoch": 2.8330977563321476,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3137,
-      "step": 174380
-    },
-    {
-      "epoch": 2.8332602232295168,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 174390
-    },
-    {
-      "epoch": 2.833422690126887,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2803,
-      "step": 174400
-    },
-    {
-      "epoch": 2.833585157024256,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 174410
-    },
-    {
-      "epoch": 2.833747623921626,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2651,
-      "step": 174420
-    },
-    {
-      "epoch": 2.8339100908189954,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 174430
-    },
-    {
-      "epoch": 2.8340725577163655,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 174440
-    },
-    {
-      "epoch": 2.834235024613735,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2515,
-      "step": 174450
-    },
-    {
-      "epoch": 2.8343974915111048,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 174460
-    },
-    {
-      "epoch": 2.8345599584084744,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.293,
-      "step": 174470
-    },
-    {
-      "epoch": 2.834722425305844,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2987,
-      "step": 174480
-    },
-    {
-      "epoch": 2.8348848922032137,
-      "grad_norm": 2.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 174490
-    },
-    {
-      "epoch": 2.8350473591005834,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2996,
-      "step": 174500
-    },
-    {
-      "epoch": 2.835209825997953,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 174510
-    },
-    {
-      "epoch": 2.8353722928953227,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3036,
-      "step": 174520
-    },
-    {
-      "epoch": 2.8355347597926923,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 174530
-    },
-    {
-      "epoch": 2.835697226690062,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 174540
-    },
-    {
-      "epoch": 2.8358596935874316,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 174550
-    },
-    {
-      "epoch": 2.8360221604848013,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2927,
-      "step": 174560
-    },
-    {
-      "epoch": 2.836184627382171,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 174570
-    },
-    {
-      "epoch": 2.8363470942795406,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 174580
-    },
-    {
-      "epoch": 2.83650956117691,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 174590
-    },
-    {
-      "epoch": 2.83667202807428,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 174600
-    },
-    {
-      "epoch": 2.8368344949716495,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 174610
-    },
-    {
-      "epoch": 2.836996961869019,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2708,
-      "step": 174620
-    },
-    {
-      "epoch": 2.837159428766389,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3007,
-      "step": 174630
-    },
-    {
-      "epoch": 2.8373218956637585,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 174640
-    },
-    {
-      "epoch": 2.837484362561128,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3035,
-      "step": 174650
-    },
-    {
-      "epoch": 2.8376468294584978,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 174660
-    },
-    {
-      "epoch": 2.8378092963558674,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.298,
-      "step": 174670
-    },
-    {
-      "epoch": 2.837971763253237,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 174680
-    },
-    {
-      "epoch": 2.8381342301506067,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 174690
-    },
-    {
-      "epoch": 2.8382966970479764,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3079,
-      "step": 174700
-    },
-    {
-      "epoch": 2.838459163945346,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 174710
-    },
-    {
-      "epoch": 2.8386216308427157,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 174720
-    },
-    {
-      "epoch": 2.8387840977400853,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2946,
-      "step": 174730
-    },
-    {
-      "epoch": 2.838946564637455,
-      "grad_norm": 3.390625,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 174740
-    },
-    {
-      "epoch": 2.8391090315348246,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.291,
-      "step": 174750
-    },
-    {
-      "epoch": 2.8392714984321943,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2927,
-      "step": 174760
-    },
-    {
-      "epoch": 2.839433965329564,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.3227,
-      "step": 174770
-    },
-    {
-      "epoch": 2.8395964322269336,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2981,
-      "step": 174780
-    },
-    {
-      "epoch": 2.8397588991243037,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 174790
-    },
-    {
-      "epoch": 2.839921366021673,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2697,
-      "step": 174800
-    },
-    {
-      "epoch": 2.840083832919043,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2774,
-      "step": 174810
-    },
-    {
-      "epoch": 2.840246299816412,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 174820
-    },
-    {
-      "epoch": 2.8404087667137823,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3072,
-      "step": 174830
-    },
-    {
-      "epoch": 2.8405712336111515,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3037,
-      "step": 174840
-    },
-    {
-      "epoch": 2.8407337005085216,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 174850
-    },
-    {
-      "epoch": 2.840896167405891,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2897,
-      "step": 174860
-    },
-    {
-      "epoch": 2.841058634303261,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3134,
-      "step": 174870
-    },
-    {
-      "epoch": 2.84122110120063,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 174880
-    },
-    {
-      "epoch": 2.841383568098,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 174890
-    },
-    {
-      "epoch": 2.84154603499537,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2978,
-      "step": 174900
-    },
-    {
-      "epoch": 2.8417085018927395,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 174910
-    },
-    {
-      "epoch": 2.841870968790109,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2979,
-      "step": 174920
-    },
-    {
-      "epoch": 2.842033435687479,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 174930
-    },
-    {
-      "epoch": 2.8421959025848484,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2921,
-      "step": 174940
-    },
-    {
-      "epoch": 2.842358369482218,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 174950
-    },
-    {
-      "epoch": 2.8425208363795877,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 174960
-    },
-    {
-      "epoch": 2.8426833032769574,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 174970
-    },
-    {
-      "epoch": 2.842845770174327,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3128,
-      "step": 174980
-    },
-    {
-      "epoch": 2.8430082370716967,
-      "grad_norm": 2.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 174990
-    },
-    {
-      "epoch": 2.8431707039690663,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 175000
-    },
-    {
-      "epoch": 2.843333170866436,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2949,
-      "step": 175010
-    },
-    {
-      "epoch": 2.8434956377638057,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2642,
-      "step": 175020
-    },
-    {
-      "epoch": 2.8436581046611753,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 175030
-    },
-    {
-      "epoch": 2.843820571558545,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 175040
-    },
-    {
-      "epoch": 2.8439830384559146,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2707,
-      "step": 175050
-    },
-    {
-      "epoch": 2.8441455053532843,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 175060
-    },
-    {
-      "epoch": 2.844307972250654,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 175070
-    },
-    {
-      "epoch": 2.8444704391480236,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 175080
-    },
-    {
-      "epoch": 2.844632906045393,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 175090
-    },
-    {
-      "epoch": 2.844795372942763,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.261,
-      "step": 175100
-    },
-    {
-      "epoch": 2.8449578398401325,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2781,
-      "step": 175110
-    },
-    {
-      "epoch": 2.845120306737502,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2863,
-      "step": 175120
-    },
-    {
-      "epoch": 2.845282773634872,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 175130
-    },
-    {
-      "epoch": 2.8454452405322415,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 175140
-    },
-    {
-      "epoch": 2.845607707429611,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2581,
-      "step": 175150
-    },
-    {
-      "epoch": 2.8457701743269808,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 175160
-    },
-    {
-      "epoch": 2.8459326412243504,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2634,
-      "step": 175170
-    },
-    {
-      "epoch": 2.84609510812172,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 175180
-    },
-    {
-      "epoch": 2.8462575750190897,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 175190
-    },
-    {
-      "epoch": 2.8464200419164594,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 175200
-    },
-    {
-      "epoch": 2.846582508813829,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 175210
-    },
-    {
-      "epoch": 2.846744975711199,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 175220
-    },
-    {
-      "epoch": 2.8469074426085683,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 175230
-    },
-    {
-      "epoch": 2.8470699095059384,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.274,
-      "step": 175240
-    },
-    {
-      "epoch": 2.8472323764033076,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.3048,
-      "step": 175250
-    },
-    {
-      "epoch": 2.8473948433006777,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 175260
-    },
-    {
-      "epoch": 2.847557310198047,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2583,
-      "step": 175270
-    },
-    {
-      "epoch": 2.847719777095417,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 175280
-    },
-    {
-      "epoch": 2.847882243992786,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2719,
-      "step": 175290
-    },
-    {
-      "epoch": 2.8480447108901563,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 175300
-    },
-    {
-      "epoch": 2.8482071777875255,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2878,
-      "step": 175310
-    },
-    {
-      "epoch": 2.8483696446848956,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 175320
-    },
-    {
-      "epoch": 2.8485321115822653,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.3065,
-      "step": 175330
-    },
-    {
-      "epoch": 2.848694578479635,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 175340
-    },
-    {
-      "epoch": 2.8488570453770046,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3072,
-      "step": 175350
-    },
-    {
-      "epoch": 2.849019512274374,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3084,
-      "step": 175360
-    },
-    {
-      "epoch": 2.849181979171744,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.3043,
-      "step": 175370
-    },
-    {
-      "epoch": 2.8493444460691135,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2937,
-      "step": 175380
-    },
-    {
-      "epoch": 2.849506912966483,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3006,
-      "step": 175390
-    },
-    {
-      "epoch": 2.849669379863853,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 175400
-    },
-    {
-      "epoch": 2.8498318467612225,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2995,
-      "step": 175410
-    },
-    {
-      "epoch": 2.849994313658592,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2772,
-      "step": 175420
-    },
-    {
-      "epoch": 2.8501567805559618,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 175430
-    },
-    {
-      "epoch": 2.8503192474533314,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 175440
-    },
-    {
-      "epoch": 2.850481714350701,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 175450
-    },
-    {
-      "epoch": 2.8506441812480707,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 175460
-    },
-    {
-      "epoch": 2.8508066481454404,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 175470
-    },
-    {
-      "epoch": 2.85096911504281,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 175480
-    },
-    {
-      "epoch": 2.8511315819401797,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 175490
-    },
-    {
-      "epoch": 2.8512940488375493,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 175500
-    },
-    {
-      "epoch": 2.851456515734919,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 175510
-    },
-    {
-      "epoch": 2.8516189826322886,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 175520
-    },
-    {
-      "epoch": 2.8517814495296583,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 175530
-    },
-    {
-      "epoch": 2.851943916427028,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 175540
-    },
-    {
-      "epoch": 2.8521063833243976,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.305,
-      "step": 175550
-    },
-    {
-      "epoch": 2.8522688502217672,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2767,
-      "step": 175560
-    },
-    {
-      "epoch": 2.852431317119137,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 175570
-    },
-    {
-      "epoch": 2.8525937840165065,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2662,
-      "step": 175580
-    },
-    {
-      "epoch": 2.852756250913876,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 175590
-    },
-    {
-      "epoch": 2.852918717811246,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2679,
-      "step": 175600
-    },
-    {
-      "epoch": 2.8530811847086155,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 175610
-    },
-    {
-      "epoch": 2.853243651605985,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 175620
-    },
-    {
-      "epoch": 2.853406118503355,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 175630
-    },
-    {
-      "epoch": 2.8535685854007244,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2905,
-      "step": 175640
-    },
-    {
-      "epoch": 2.8537310522980945,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 175650
-    },
-    {
-      "epoch": 2.8538935191954637,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3165,
-      "step": 175660
-    },
-    {
-      "epoch": 2.854055986092834,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2632,
-      "step": 175670
-    },
-    {
-      "epoch": 2.854218452990203,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 175680
-    },
-    {
-      "epoch": 2.854380919887573,
-      "grad_norm": 2.625,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 175690
-    },
-    {
-      "epoch": 2.8545433867849423,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 175700
-    },
-    {
-      "epoch": 2.8547058536823124,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2679,
-      "step": 175710
-    },
-    {
-      "epoch": 2.8548683205796817,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 175720
-    },
-    {
-      "epoch": 2.8550307874770517,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3031,
-      "step": 175730
-    },
-    {
-      "epoch": 2.855193254374421,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3082,
-      "step": 175740
-    },
-    {
-      "epoch": 2.855355721271791,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 175750
-    },
-    {
-      "epoch": 2.8555181881691607,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 175760
-    },
-    {
-      "epoch": 2.8556806550665303,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.324,
-      "step": 175770
-    },
-    {
-      "epoch": 2.8558431219639,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3101,
-      "step": 175780
-    },
-    {
-      "epoch": 2.8560055888612697,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3004,
-      "step": 175790
-    },
-    {
-      "epoch": 2.8561680557586393,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 175800
-    },
-    {
-      "epoch": 2.856330522656009,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 175810
-    },
-    {
-      "epoch": 2.8564929895533786,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 175820
-    },
-    {
-      "epoch": 2.8566554564507483,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 175830
-    },
-    {
-      "epoch": 2.856817923348118,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 175840
-    },
-    {
-      "epoch": 2.8569803902454876,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2978,
-      "step": 175850
-    },
-    {
-      "epoch": 2.857142857142857,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2586,
-      "step": 175860
-    },
-    {
-      "epoch": 2.857305324040227,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 175870
-    },
-    {
-      "epoch": 2.8574677909375965,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 175880
-    },
-    {
-      "epoch": 2.857630257834966,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2776,
-      "step": 175890
-    },
-    {
-      "epoch": 2.857792724732336,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 175900
-    },
-    {
-      "epoch": 2.8579551916297055,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 175910
-    },
-    {
-      "epoch": 2.858117658527075,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 175920
-    },
-    {
-      "epoch": 2.8582801254244448,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 175930
-    },
-    {
-      "epoch": 2.8584425923218144,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.311,
-      "step": 175940
-    },
-    {
-      "epoch": 2.858605059219184,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 175950
-    },
-    {
-      "epoch": 2.8587675261165537,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 175960
-    },
-    {
-      "epoch": 2.8589299930139234,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 175970
-    },
-    {
-      "epoch": 2.859092459911293,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 175980
-    },
-    {
-      "epoch": 2.8592549268086627,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 175990
-    },
-    {
-      "epoch": 2.8594173937060323,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 176000
-    },
-    {
-      "epoch": 2.859579860603402,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 176010
-    },
-    {
-      "epoch": 2.8597423275007716,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 176020
-    },
-    {
-      "epoch": 2.8599047943981413,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 176030
-    },
-    {
-      "epoch": 2.860067261295511,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 176040
-    },
-    {
-      "epoch": 2.8602297281928806,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 176050
-    },
-    {
-      "epoch": 2.8603921950902502,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 176060
-    },
-    {
-      "epoch": 2.86055466198762,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 176070
-    },
-    {
-      "epoch": 2.8607171288849895,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 176080
-    },
-    {
-      "epoch": 2.860879595782359,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 176090
-    },
-    {
-      "epoch": 2.8610420626797293,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 176100
-    },
-    {
-      "epoch": 2.8612045295770985,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2516,
-      "step": 176110
-    },
-    {
-      "epoch": 2.8613669964744686,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2902,
-      "step": 176120
-    },
-    {
-      "epoch": 2.861529463371838,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 176130
-    },
-    {
-      "epoch": 2.861691930269208,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.265,
-      "step": 176140
-    },
-    {
-      "epoch": 2.861854397166577,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 176150
-    },
-    {
-      "epoch": 2.862016864063947,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2832,
-      "step": 176160
-    },
-    {
-      "epoch": 2.8621793309613164,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3186,
-      "step": 176170
-    },
-    {
-      "epoch": 2.8623417978586865,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3034,
-      "step": 176180
-    },
-    {
-      "epoch": 2.8625042647560557,
-      "grad_norm": 2.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2993,
-      "step": 176190
-    },
-    {
-      "epoch": 2.8626667316534258,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.3002,
-      "step": 176200
-    },
-    {
-      "epoch": 2.8628291985507954,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 176210
-    },
-    {
-      "epoch": 2.862991665448165,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 176220
-    },
-    {
-      "epoch": 2.8631541323455347,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 176230
-    },
-    {
-      "epoch": 2.8633165992429044,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2691,
-      "step": 176240
-    },
-    {
-      "epoch": 2.863479066140274,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3081,
-      "step": 176250
-    },
-    {
-      "epoch": 2.8636415330376437,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2962,
-      "step": 176260
-    },
-    {
-      "epoch": 2.8638039999350133,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 176270
-    },
-    {
-      "epoch": 2.863966466832383,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2894,
-      "step": 176280
-    },
-    {
-      "epoch": 2.8641289337297526,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 176290
-    },
-    {
-      "epoch": 2.8642914006271223,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2566,
-      "step": 176300
-    },
-    {
-      "epoch": 2.864453867524492,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3083,
-      "step": 176310
-    },
-    {
-      "epoch": 2.8646163344218616,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2909,
-      "step": 176320
-    },
-    {
-      "epoch": 2.8647788013192312,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 176330
-    },
-    {
-      "epoch": 2.864941268216601,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 176340
-    },
-    {
-      "epoch": 2.8651037351139705,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2897,
-      "step": 176350
-    },
-    {
-      "epoch": 2.86526620201134,
-      "grad_norm": 2.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2948,
-      "step": 176360
-    },
-    {
-      "epoch": 2.86542866890871,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 176370
-    },
-    {
-      "epoch": 2.8655911358060795,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 176380
-    },
-    {
-      "epoch": 2.865753602703449,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3067,
-      "step": 176390
-    },
-    {
-      "epoch": 2.865916069600819,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 176400
-    },
-    {
-      "epoch": 2.8660785364981884,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2999,
-      "step": 176410
-    },
-    {
-      "epoch": 2.866241003395558,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2722,
-      "step": 176420
-    },
-    {
-      "epoch": 2.8664034702929277,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 176430
-    },
-    {
-      "epoch": 2.8665659371902974,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2799,
-      "step": 176440
-    },
-    {
-      "epoch": 2.866728404087667,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 176450
-    },
-    {
-      "epoch": 2.8668908709850367,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.288,
-      "step": 176460
-    },
-    {
-      "epoch": 2.8670533378824063,
-      "grad_norm": 2.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2965,
-      "step": 176470
-    },
-    {
-      "epoch": 2.867215804779776,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 176480
-    },
-    {
-      "epoch": 2.8673782716771457,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3231,
-      "step": 176490
-    },
-    {
-      "epoch": 2.8675407385745153,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3116,
-      "step": 176500
-    },
-    {
-      "epoch": 2.867703205471885,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 176510
-    },
-    {
-      "epoch": 2.8678656723692546,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 176520
-    },
-    {
-      "epoch": 2.8680281392666247,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 176530
-    },
-    {
-      "epoch": 2.868190606163994,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 176540
-    },
-    {
-      "epoch": 2.868353073061364,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 176550
-    },
-    {
-      "epoch": 2.868515539958733,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 176560
-    },
-    {
-      "epoch": 2.8686780068561033,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 176570
-    },
-    {
-      "epoch": 2.8688404737534725,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 176580
-    },
-    {
-      "epoch": 2.8690029406508426,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2612,
-      "step": 176590
-    },
-    {
-      "epoch": 2.869165407548212,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 176600
-    },
-    {
-      "epoch": 2.869327874445582,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 176610
-    },
-    {
-      "epoch": 2.869490341342951,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 176620
-    },
-    {
-      "epoch": 2.869652808240321,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2839,
-      "step": 176630
-    },
-    {
-      "epoch": 2.869815275137691,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 176640
-    },
-    {
-      "epoch": 2.8699777420350605,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 176650
-    },
-    {
-      "epoch": 2.87014020893243,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3016,
-      "step": 176660
-    },
-    {
-      "epoch": 2.8703026758298,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3067,
-      "step": 176670
-    },
-    {
-      "epoch": 2.8704651427271695,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3072,
-      "step": 176680
-    },
-    {
-      "epoch": 2.870627609624539,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.307,
-      "step": 176690
-    },
-    {
-      "epoch": 2.8707900765219088,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3093,
-      "step": 176700
-    },
-    {
-      "epoch": 2.8709525434192784,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2999,
-      "step": 176710
-    },
-    {
-      "epoch": 2.871115010316648,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2897,
-      "step": 176720
-    },
-    {
-      "epoch": 2.8712774772140177,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 176730
-    },
-    {
-      "epoch": 2.8714399441113874,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 176740
-    },
-    {
-      "epoch": 2.871602411008757,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2972,
-      "step": 176750
-    },
-    {
-      "epoch": 2.8717648779061267,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2731,
-      "step": 176760
-    },
-    {
-      "epoch": 2.8719273448034963,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2875,
-      "step": 176770
-    },
-    {
-      "epoch": 2.872089811700866,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2992,
-      "step": 176780
-    },
-    {
-      "epoch": 2.8722522785982356,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 176790
-    },
-    {
-      "epoch": 2.8724147454956053,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 176800
-    },
-    {
-      "epoch": 2.872577212392975,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 176810
-    },
-    {
-      "epoch": 2.8727396792903446,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3248,
-      "step": 176820
-    },
-    {
-      "epoch": 2.8729021461877142,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3012,
-      "step": 176830
-    },
-    {
-      "epoch": 2.873064613085084,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2868,
-      "step": 176840
-    },
-    {
-      "epoch": 2.8732270799824535,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 176850
-    },
-    {
-      "epoch": 2.873389546879823,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2736,
-      "step": 176860
-    },
-    {
-      "epoch": 2.873552013777193,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 176870
-    },
-    {
-      "epoch": 2.8737144806745625,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3034,
-      "step": 176880
-    },
-    {
-      "epoch": 2.873876947571932,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 176890
-    },
-    {
-      "epoch": 2.874039414469302,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 176900
-    },
-    {
-      "epoch": 2.8742018813666714,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3053,
-      "step": 176910
-    },
-    {
-      "epoch": 2.874364348264041,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.3109,
-      "step": 176920
-    },
-    {
-      "epoch": 2.8745268151614107,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2918,
-      "step": 176930
-    },
-    {
-      "epoch": 2.8746892820587804,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2743,
-      "step": 176940
-    },
-    {
-      "epoch": 2.87485174895615,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 176950
-    },
-    {
-      "epoch": 2.8750142158535197,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 176960
-    },
-    {
-      "epoch": 2.8751766827508893,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 176970
-    },
-    {
-      "epoch": 2.8753391496482594,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3041,
-      "step": 176980
-    },
-    {
-      "epoch": 2.8755016165456286,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 176990
-    },
-    {
-      "epoch": 2.8756640834429987,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3188,
-      "step": 177000
-    },
-    {
-      "epoch": 2.875826550340368,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 177010
-    },
-    {
-      "epoch": 2.875989017237738,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 177020
-    },
-    {
-      "epoch": 2.8761514841351072,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 177030
-    },
-    {
-      "epoch": 2.8763139510324773,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 177040
-    },
-    {
-      "epoch": 2.8764764179298465,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 177050
-    },
-    {
-      "epoch": 2.8766388848272166,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 177060
-    },
-    {
-      "epoch": 2.876801351724586,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 177070
-    },
-    {
-      "epoch": 2.876963818621956,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2758,
-      "step": 177080
-    },
-    {
-      "epoch": 2.8771262855193256,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 177090
-    },
-    {
-      "epoch": 2.8772887524166952,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 177100
-    },
-    {
-      "epoch": 2.877451219314065,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 177110
-    },
-    {
-      "epoch": 2.8776136862114345,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 177120
-    },
-    {
-      "epoch": 2.877776153108804,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3085,
-      "step": 177130
-    },
-    {
-      "epoch": 2.877938620006174,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 177140
-    },
-    {
-      "epoch": 2.8781010869035435,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 177150
-    },
-    {
-      "epoch": 2.878263553800913,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2875,
-      "step": 177160
-    },
-    {
-      "epoch": 2.878426020698283,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.307,
-      "step": 177170
-    },
-    {
-      "epoch": 2.8785884875956524,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 177180
-    },
-    {
-      "epoch": 2.878750954493022,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 177190
-    },
-    {
-      "epoch": 2.8789134213903917,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 177200
-    },
-    {
-      "epoch": 2.8790758882877614,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 177210
-    },
-    {
-      "epoch": 2.879238355185131,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2965,
-      "step": 177220
-    },
-    {
-      "epoch": 2.8794008220825007,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 177230
-    },
-    {
-      "epoch": 2.8795632889798704,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 177240
-    },
-    {
-      "epoch": 2.87972575587724,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3028,
-      "step": 177250
-    },
-    {
-      "epoch": 2.8798882227746097,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 177260
-    },
-    {
-      "epoch": 2.8800506896719793,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2785,
-      "step": 177270
-    },
-    {
-      "epoch": 2.880213156569349,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 177280
-    },
-    {
-      "epoch": 2.8803756234667186,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 177290
-    },
-    {
-      "epoch": 2.8805380903640883,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 177300
-    },
-    {
-      "epoch": 2.880700557261458,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3002,
-      "step": 177310
-    },
-    {
-      "epoch": 2.8808630241588276,
-      "grad_norm": 3.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2709,
-      "step": 177320
-    },
-    {
-      "epoch": 2.881025491056197,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2927,
-      "step": 177330
-    },
-    {
-      "epoch": 2.881187957953567,
-      "grad_norm": 3.109375,
-      "learning_rate": 5e-05,
-      "loss": 0.2618,
-      "step": 177340
-    },
-    {
-      "epoch": 2.8813504248509365,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 177350
-    },
-    {
-      "epoch": 2.881512891748306,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2533,
-      "step": 177360
-    },
-    {
-      "epoch": 2.881675358645676,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2539,
-      "step": 177370
-    },
-    {
-      "epoch": 2.8818378255430455,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 177380
-    },
-    {
-      "epoch": 2.882000292440415,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2921,
-      "step": 177390
-    },
-    {
-      "epoch": 2.8821627593377848,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2958,
-      "step": 177400
-    },
-    {
-      "epoch": 2.882325226235155,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 177410
-    },
-    {
-      "epoch": 2.882487693132524,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 177420
-    },
-    {
-      "epoch": 2.882650160029894,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2957,
-      "step": 177430
-    },
-    {
-      "epoch": 2.8828126269272634,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 177440
-    },
-    {
-      "epoch": 2.8829750938246335,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2695,
-      "step": 177450
-    },
-    {
-      "epoch": 2.8831375607220027,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 177460
-    },
-    {
-      "epoch": 2.8833000276193728,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 177470
-    },
-    {
-      "epoch": 2.883462494516742,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2682,
-      "step": 177480
-    },
-    {
-      "epoch": 2.883624961414112,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 177490
-    },
-    {
-      "epoch": 2.8837874283114813,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 177500
-    },
-    {
-      "epoch": 2.8839498952088514,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2956,
-      "step": 177510
-    },
-    {
-      "epoch": 2.884112362106221,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3107,
-      "step": 177520
-    },
-    {
-      "epoch": 2.8842748290035907,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.3073,
-      "step": 177530
-    },
-    {
-      "epoch": 2.8844372959009603,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 177540
-    },
-    {
-      "epoch": 2.88459976279833,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2982,
-      "step": 177550
-    },
-    {
-      "epoch": 2.8847622296956996,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 177560
-    },
-    {
-      "epoch": 2.8849246965930693,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 177570
-    },
-    {
-      "epoch": 2.885087163490439,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 177580
-    },
-    {
-      "epoch": 2.8852496303878086,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2496,
-      "step": 177590
-    },
-    {
-      "epoch": 2.8854120972851782,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2781,
-      "step": 177600
-    },
-    {
-      "epoch": 2.885574564182548,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 177610
-    },
-    {
-      "epoch": 2.8857370310799175,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3,
-      "step": 177620
-    },
-    {
-      "epoch": 2.885899497977287,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2907,
-      "step": 177630
-    },
-    {
-      "epoch": 2.886061964874657,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 177640
-    },
-    {
-      "epoch": 2.8862244317720265,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 177650
-    },
-    {
-      "epoch": 2.886386898669396,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.3246,
-      "step": 177660
-    },
-    {
-      "epoch": 2.886549365566766,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 177670
-    },
-    {
-      "epoch": 2.8867118324641354,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 177680
-    },
-    {
-      "epoch": 2.886874299361505,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2547,
-      "step": 177690
-    },
-    {
-      "epoch": 2.8870367662588747,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 177700
-    },
-    {
-      "epoch": 2.8871992331562444,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2735,
-      "step": 177710
-    },
-    {
-      "epoch": 2.887361700053614,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2687,
-      "step": 177720
-    },
-    {
-      "epoch": 2.8875241669509837,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 177730
-    },
-    {
-      "epoch": 2.8876866338483533,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 177740
-    },
-    {
-      "epoch": 2.887849100745723,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 177750
-    },
-    {
-      "epoch": 2.8880115676430926,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 177760
-    },
-    {
-      "epoch": 2.8881740345404623,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 177770
-    },
-    {
-      "epoch": 2.888336501437832,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 177780
-    },
-    {
-      "epoch": 2.8884989683352016,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 177790
-    },
-    {
-      "epoch": 2.8886614352325712,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 177800
-    },
-    {
-      "epoch": 2.888823902129941,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3032,
-      "step": 177810
-    },
-    {
-      "epoch": 2.8889863690273105,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2753,
-      "step": 177820
-    },
-    {
-      "epoch": 2.88914883592468,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2943,
-      "step": 177830
-    },
-    {
-      "epoch": 2.88931130282205,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 177840
-    },
-    {
-      "epoch": 2.8894737697194195,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 177850
-    },
-    {
-      "epoch": 2.8896362366167896,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 177860
-    },
-    {
-      "epoch": 2.889798703514159,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2505,
-      "step": 177870
-    },
-    {
-      "epoch": 2.889961170411529,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2667,
-      "step": 177880
-    },
-    {
-      "epoch": 2.890123637308898,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2672,
-      "step": 177890
-    },
-    {
-      "epoch": 2.890286104206268,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2705,
-      "step": 177900
-    },
-    {
-      "epoch": 2.8904485711036374,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 177910
-    },
-    {
-      "epoch": 2.8906110380010075,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3026,
-      "step": 177920
-    },
-    {
-      "epoch": 2.8907735048983767,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2989,
-      "step": 177930
-    },
-    {
-      "epoch": 2.890935971795747,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2948,
-      "step": 177940
-    },
-    {
-      "epoch": 2.891098438693116,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 177950
-    },
-    {
-      "epoch": 2.891260905590486,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2565,
-      "step": 177960
-    },
-    {
-      "epoch": 2.8914233724878557,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2771,
-      "step": 177970
-    },
-    {
-      "epoch": 2.8915858393852254,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2704,
-      "step": 177980
-    },
-    {
-      "epoch": 2.891748306282595,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 177990
-    },
-    {
-      "epoch": 2.8919107731799647,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 178000
-    },
-    {
-      "epoch": 2.8920732400773344,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2999,
-      "step": 178010
-    },
-    {
-      "epoch": 2.892235706974704,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 178020
-    },
-    {
-      "epoch": 2.8923981738720737,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 178030
-    },
-    {
-      "epoch": 2.8925606407694433,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 178040
-    },
-    {
-      "epoch": 2.892723107666813,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.3126,
-      "step": 178050
-    },
-    {
-      "epoch": 2.8928855745641826,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 178060
-    },
-    {
-      "epoch": 2.8930480414615523,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2927,
-      "step": 178070
-    },
-    {
-      "epoch": 2.893210508358922,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2845,
-      "step": 178080
-    },
-    {
-      "epoch": 2.8933729752562916,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3015,
-      "step": 178090
-    },
-    {
-      "epoch": 2.893535442153661,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 178100
-    },
-    {
-      "epoch": 2.893697909051031,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 178110
-    },
-    {
-      "epoch": 2.8938603759484005,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3048,
-      "step": 178120
-    },
-    {
-      "epoch": 2.89402284284577,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.3216,
-      "step": 178130
-    },
-    {
-      "epoch": 2.89418530974314,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 178140
-    },
-    {
-      "epoch": 2.8943477766405095,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2971,
-      "step": 178150
-    },
-    {
-      "epoch": 2.894510243537879,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 178160
-    },
-    {
-      "epoch": 2.8946727104352488,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3039,
-      "step": 178170
-    },
-    {
-      "epoch": 2.8948351773326184,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 178180
-    },
-    {
-      "epoch": 2.894997644229988,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2924,
-      "step": 178190
-    },
-    {
-      "epoch": 2.8951601111273577,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2766,
-      "step": 178200
-    },
-    {
-      "epoch": 2.8953225780247274,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 178210
-    },
-    {
-      "epoch": 2.895485044922097,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2622,
-      "step": 178220
-    },
-    {
-      "epoch": 2.8956475118194667,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 178230
-    },
-    {
-      "epoch": 2.8958099787168363,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 178240
-    },
-    {
-      "epoch": 2.895972445614206,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 178250
-    },
-    {
-      "epoch": 2.8961349125115756,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 178260
-    },
-    {
-      "epoch": 2.8962973794089453,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.264,
-      "step": 178270
-    },
-    {
-      "epoch": 2.896459846306315,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 178280
-    },
-    {
-      "epoch": 2.896622313203685,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 178290
-    },
-    {
-      "epoch": 2.8967847801010542,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 178300
-    },
-    {
-      "epoch": 2.8969472469984243,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2817,
-      "step": 178310
-    },
-    {
-      "epoch": 2.8971097138957935,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2821,
-      "step": 178320
-    },
-    {
-      "epoch": 2.8972721807931636,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 178330
-    },
-    {
-      "epoch": 2.897434647690533,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 178340
-    },
-    {
-      "epoch": 2.897597114587903,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 178350
-    },
-    {
-      "epoch": 2.897759581485272,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.3039,
-      "step": 178360
-    },
-    {
-      "epoch": 2.8979220483826422,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 178370
-    },
-    {
-      "epoch": 2.8980845152800114,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 178380
-    },
-    {
-      "epoch": 2.8982469821773815,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2633,
-      "step": 178390
-    },
-    {
-      "epoch": 2.898409449074751,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 178400
-    },
-    {
-      "epoch": 2.898571915972121,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 178410
-    },
-    {
-      "epoch": 2.8987343828694905,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2721,
-      "step": 178420
-    },
-    {
-      "epoch": 2.89889684976686,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 178430
-    },
-    {
-      "epoch": 2.89905931666423,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 178440
-    },
-    {
-      "epoch": 2.8992217835615994,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 178450
-    },
-    {
-      "epoch": 2.899384250458969,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 178460
-    },
-    {
-      "epoch": 2.8995467173563387,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 178470
-    },
-    {
-      "epoch": 2.8997091842537084,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 178480
-    },
-    {
-      "epoch": 2.899871651151078,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 178490
-    },
-    {
-      "epoch": 2.9000341180484477,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2676,
-      "step": 178500
-    },
-    {
-      "epoch": 2.9001965849458173,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2579,
-      "step": 178510
-    },
-    {
-      "epoch": 2.900359051843187,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2613,
-      "step": 178520
-    },
-    {
-      "epoch": 2.9005215187405566,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 178530
-    },
-    {
-      "epoch": 2.9006839856379263,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 178540
-    },
-    {
-      "epoch": 2.900846452535296,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3066,
-      "step": 178550
-    },
-    {
-      "epoch": 2.9010089194326656,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 178560
-    },
-    {
-      "epoch": 2.9011713863300352,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2617,
-      "step": 178570
-    },
-    {
-      "epoch": 2.901333853227405,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 178580
-    },
-    {
-      "epoch": 2.9014963201247745,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2916,
-      "step": 178590
-    },
-    {
-      "epoch": 2.901658787022144,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.258,
-      "step": 178600
-    },
-    {
-      "epoch": 2.901821253919514,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2597,
-      "step": 178610
-    },
-    {
-      "epoch": 2.9019837208168835,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2933,
-      "step": 178620
-    },
-    {
-      "epoch": 2.902146187714253,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2715,
-      "step": 178630
-    },
-    {
-      "epoch": 2.902308654611623,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2706,
-      "step": 178640
-    },
-    {
-      "epoch": 2.9024711215089924,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 178650
-    },
-    {
-      "epoch": 2.902633588406362,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3215,
-      "step": 178660
-    },
-    {
-      "epoch": 2.9027960553037317,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3118,
-      "step": 178670
-    },
-    {
-      "epoch": 2.9029585222011014,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 178680
-    },
-    {
-      "epoch": 2.903120989098471,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 178690
-    },
-    {
-      "epoch": 2.9032834559958407,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2647,
-      "step": 178700
-    },
-    {
-      "epoch": 2.9034459228932104,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2685,
-      "step": 178710
-    },
-    {
-      "epoch": 2.9036083897905804,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2671,
-      "step": 178720
-    },
-    {
-      "epoch": 2.9037708566879497,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 178730
-    },
-    {
-      "epoch": 2.9039333235853197,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 178740
-    },
-    {
-      "epoch": 2.904095790482689,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.269,
-      "step": 178750
-    },
-    {
-      "epoch": 2.904258257380059,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2671,
-      "step": 178760
-    },
-    {
-      "epoch": 2.9044207242774283,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 178770
-    },
-    {
-      "epoch": 2.9045831911747984,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 178780
-    },
-    {
-      "epoch": 2.9047456580721676,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.268,
-      "step": 178790
-    },
-    {
-      "epoch": 2.9049081249695377,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2996,
-      "step": 178800
-    },
-    {
-      "epoch": 2.905070591866907,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 178810
-    },
-    {
-      "epoch": 2.905233058764277,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 178820
-    },
-    {
-      "epoch": 2.9053955256616466,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2948,
-      "step": 178830
-    },
-    {
-      "epoch": 2.9055579925590163,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3077,
-      "step": 178840
-    },
-    {
-      "epoch": 2.905720459456386,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 178850
-    },
-    {
-      "epoch": 2.9058829263537556,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 178860
-    },
-    {
-      "epoch": 2.906045393251125,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2782,
-      "step": 178870
-    },
-    {
-      "epoch": 2.906207860148495,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 178880
-    },
-    {
-      "epoch": 2.9063703270458645,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 178890
-    },
-    {
-      "epoch": 2.906532793943234,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 178900
-    },
-    {
-      "epoch": 2.906695260840604,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2815,
-      "step": 178910
-    },
-    {
-      "epoch": 2.9068577277379735,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 178920
-    },
-    {
-      "epoch": 2.907020194635343,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2623,
-      "step": 178930
-    },
-    {
-      "epoch": 2.9071826615327128,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2519,
-      "step": 178940
-    },
-    {
-      "epoch": 2.9073451284300824,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 178950
-    },
-    {
-      "epoch": 2.907507595327452,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 178960
-    },
-    {
-      "epoch": 2.9076700622248217,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 178970
-    },
-    {
-      "epoch": 2.9078325291221914,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 178980
-    },
-    {
-      "epoch": 2.907994996019561,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 178990
-    },
-    {
-      "epoch": 2.9081574629169307,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.3102,
-      "step": 179000
-    },
-    {
-      "epoch": 2.9083199298143003,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 179010
-    },
-    {
-      "epoch": 2.90848239671167,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2913,
-      "step": 179020
-    },
-    {
-      "epoch": 2.9086448636090396,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.308,
-      "step": 179030
-    },
-    {
-      "epoch": 2.9088073305064093,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3094,
-      "step": 179040
-    },
-    {
-      "epoch": 2.908969797403779,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 179050
-    },
-    {
-      "epoch": 2.9091322643011486,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 179060
-    },
-    {
-      "epoch": 2.9092947311985182,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.314,
-      "step": 179070
-    },
-    {
-      "epoch": 2.909457198095888,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 179080
-    },
-    {
-      "epoch": 2.9096196649932575,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 179090
-    },
-    {
-      "epoch": 2.909782131890627,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2971,
-      "step": 179100
-    },
-    {
-      "epoch": 2.909944598787997,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 179110
-    },
-    {
-      "epoch": 2.9101070656853665,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 179120
-    },
-    {
-      "epoch": 2.910269532582736,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 179130
-    },
-    {
-      "epoch": 2.910431999480106,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2981,
-      "step": 179140
-    },
-    {
-      "epoch": 2.9105944663774754,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 179150
-    },
-    {
-      "epoch": 2.910756933274845,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 179160
-    },
-    {
-      "epoch": 2.910919400172215,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 179170
-    },
-    {
-      "epoch": 2.9110818670695844,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2922,
-      "step": 179180
-    },
-    {
-      "epoch": 2.9112443339669545,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2673,
-      "step": 179190
-    },
-    {
-      "epoch": 2.9114068008643237,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.3179,
-      "step": 179200
-    },
-    {
-      "epoch": 2.911569267761694,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 179210
-    },
-    {
-      "epoch": 2.911731734659063,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 179220
-    },
-    {
-      "epoch": 2.911894201556433,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 179230
-    },
-    {
-      "epoch": 2.9120566684538023,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.295,
-      "step": 179240
-    },
-    {
-      "epoch": 2.9122191353511724,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 179250
-    },
-    {
-      "epoch": 2.9123816022485416,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3158,
-      "step": 179260
-    },
-    {
-      "epoch": 2.9125440691459117,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.3036,
-      "step": 179270
-    },
-    {
-      "epoch": 2.9127065360432813,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 179280
-    },
-    {
-      "epoch": 2.912869002940651,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 179290
-    },
-    {
-      "epoch": 2.9130314698380206,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2576,
-      "step": 179300
-    },
-    {
-      "epoch": 2.9131939367353903,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2849,
-      "step": 179310
-    },
-    {
-      "epoch": 2.91335640363276,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3024,
-      "step": 179320
-    },
-    {
-      "epoch": 2.9135188705301296,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 179330
-    },
-    {
-      "epoch": 2.9136813374274992,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3024,
-      "step": 179340
-    },
-    {
-      "epoch": 2.913843804324869,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 179350
-    },
-    {
-      "epoch": 2.9140062712222385,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.297,
-      "step": 179360
-    },
-    {
-      "epoch": 2.914168738119608,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2961,
-      "step": 179370
-    },
-    {
-      "epoch": 2.914331205016978,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 179380
-    },
-    {
-      "epoch": 2.9144936719143475,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 179390
-    },
-    {
-      "epoch": 2.914656138811717,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 179400
-    },
-    {
-      "epoch": 2.914818605709087,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 179410
-    },
-    {
-      "epoch": 2.9149810726064564,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 179420
-    },
-    {
-      "epoch": 2.915143539503826,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 179430
-    },
-    {
-      "epoch": 2.9153060064011957,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 179440
-    },
-    {
-      "epoch": 2.9154684732985654,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2698,
-      "step": 179450
-    },
-    {
-      "epoch": 2.915630940195935,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2417,
-      "step": 179460
-    },
-    {
-      "epoch": 2.9157934070933047,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2664,
-      "step": 179470
-    },
-    {
-      "epoch": 2.9159558739906744,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2968,
-      "step": 179480
-    },
-    {
-      "epoch": 2.916118340888044,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 179490
-    },
-    {
-      "epoch": 2.9162808077854137,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 179500
-    },
-    {
-      "epoch": 2.9164432746827833,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3015,
-      "step": 179510
-    },
-    {
-      "epoch": 2.916605741580153,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2719,
-      "step": 179520
-    },
-    {
-      "epoch": 2.9167682084775226,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.3065,
-      "step": 179530
-    },
-    {
-      "epoch": 2.9169306753748923,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.286,
-      "step": 179540
-    },
-    {
-      "epoch": 2.917093142272262,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.277,
-      "step": 179550
-    },
-    {
-      "epoch": 2.9172556091696316,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 179560
-    },
-    {
-      "epoch": 2.917418076067001,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2985,
-      "step": 179570
-    },
-    {
-      "epoch": 2.917580542964371,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 179580
-    },
-    {
-      "epoch": 2.9177430098617405,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2867,
-      "step": 179590
-    },
-    {
-      "epoch": 2.9179054767591106,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2741,
-      "step": 179600
-    },
-    {
-      "epoch": 2.91806794365648,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 179610
-    },
-    {
-      "epoch": 2.91823041055385,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2838,
-      "step": 179620
-    },
-    {
-      "epoch": 2.918392877451219,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 179630
-    },
-    {
-      "epoch": 2.918555344348589,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 179640
-    },
-    {
-      "epoch": 2.9187178112459584,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 179650
-    },
-    {
-      "epoch": 2.9188802781433285,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2829,
-      "step": 179660
-    },
-    {
-      "epoch": 2.9190427450406977,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 179670
-    },
-    {
-      "epoch": 2.919205211938068,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2727,
-      "step": 179680
-    },
-    {
-      "epoch": 2.919367678835437,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3087,
-      "step": 179690
-    },
-    {
-      "epoch": 2.919530145732807,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 179700
-    },
-    {
-      "epoch": 2.9196926126301768,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2762,
-      "step": 179710
-    },
-    {
-      "epoch": 2.9198550795275464,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 179720
-    },
-    {
-      "epoch": 2.920017546424916,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 179730
-    },
-    {
-      "epoch": 2.9201800133222857,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3056,
-      "step": 179740
-    },
-    {
-      "epoch": 2.9203424802196554,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2948,
-      "step": 179750
-    },
-    {
-      "epoch": 2.920504947117025,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 179760
-    },
-    {
-      "epoch": 2.9206674140143947,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 179770
-    },
-    {
-      "epoch": 2.9208298809117643,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2828,
-      "step": 179780
-    },
-    {
-      "epoch": 2.920992347809134,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3086,
-      "step": 179790
-    },
-    {
-      "epoch": 2.9211548147065036,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2633,
-      "step": 179800
-    },
-    {
-      "epoch": 2.9213172816038733,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 179810
-    },
-    {
-      "epoch": 2.921479748501243,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 179820
-    },
-    {
-      "epoch": 2.9216422153986126,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 179830
-    },
-    {
-      "epoch": 2.9218046822959822,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 179840
-    },
-    {
-      "epoch": 2.921967149193352,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2808,
-      "step": 179850
-    },
-    {
-      "epoch": 2.9221296160907215,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 179860
-    },
-    {
-      "epoch": 2.922292082988091,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3243,
-      "step": 179870
-    },
-    {
-      "epoch": 2.922454549885461,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.3028,
-      "step": 179880
-    },
-    {
-      "epoch": 2.9226170167828305,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 179890
-    },
-    {
-      "epoch": 2.9227794836802,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3171,
-      "step": 179900
-    },
-    {
-      "epoch": 2.92294195057757,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2775,
-      "step": 179910
-    },
-    {
-      "epoch": 2.9231044174749394,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2613,
-      "step": 179920
-    },
-    {
-      "epoch": 2.923266884372309,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 179930
-    },
-    {
-      "epoch": 2.9234293512696787,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 179940
-    },
-    {
-      "epoch": 2.9235918181670484,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 179950
-    },
-    {
-      "epoch": 2.923754285064418,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2915,
-      "step": 179960
-    },
-    {
-      "epoch": 2.9239167519617877,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2793,
-      "step": 179970
-    },
-    {
-      "epoch": 2.9240792188591573,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.277,
-      "step": 179980
-    },
-    {
-      "epoch": 2.924241685756527,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 179990
-    },
-    {
-      "epoch": 2.9244041526538966,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 180000
-    },
-    {
-      "epoch": 2.9245666195512663,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 180010
-    },
-    {
-      "epoch": 2.924729086448636,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.3133,
-      "step": 180020
-    },
-    {
-      "epoch": 2.9248915533460056,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.291,
-      "step": 180030
-    },
-    {
-      "epoch": 2.9250540202433752,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3114,
-      "step": 180040
-    },
-    {
-      "epoch": 2.9252164871407453,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.3109,
-      "step": 180050
-    },
-    {
-      "epoch": 2.9253789540381145,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2972,
-      "step": 180060
-    },
-    {
-      "epoch": 2.9255414209354846,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 180070
-    },
-    {
-      "epoch": 2.925703887832854,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 180080
-    },
-    {
-      "epoch": 2.925866354730224,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 180090
-    },
-    {
-      "epoch": 2.926028821627593,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 180100
-    },
-    {
-      "epoch": 2.9261912885249632,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 180110
-    },
-    {
-      "epoch": 2.9263537554223324,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2712,
-      "step": 180120
-    },
-    {
-      "epoch": 2.9265162223197025,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 180130
-    },
-    {
-      "epoch": 2.9266786892170717,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 180140
-    },
-    {
-      "epoch": 2.926841156114442,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.3096,
-      "step": 180150
-    },
-    {
-      "epoch": 2.9270036230118115,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 180160
-    },
-    {
-      "epoch": 2.927166089909181,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 180170
-    },
-    {
-      "epoch": 2.927328556806551,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 180180
-    },
-    {
-      "epoch": 2.9274910237039204,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.284,
-      "step": 180190
-    },
-    {
-      "epoch": 2.92765349060129,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 180200
-    },
-    {
-      "epoch": 2.9278159574986597,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 180210
-    },
-    {
-      "epoch": 2.9279784243960294,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3019,
-      "step": 180220
-    },
-    {
-      "epoch": 2.928140891293399,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.3295,
-      "step": 180230
-    },
-    {
-      "epoch": 2.9283033581907687,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 180240
-    },
-    {
-      "epoch": 2.9284658250881384,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 180250
-    },
-    {
-      "epoch": 2.928628291985508,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.3308,
-      "step": 180260
-    },
-    {
-      "epoch": 2.9287907588828777,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.3073,
-      "step": 180270
-    },
-    {
-      "epoch": 2.9289532257802473,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2791,
-      "step": 180280
-    },
-    {
-      "epoch": 2.929115692677617,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2908,
-      "step": 180290
-    },
-    {
-      "epoch": 2.9292781595749866,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 180300
-    },
-    {
-      "epoch": 2.9294406264723563,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3132,
-      "step": 180310
-    },
-    {
-      "epoch": 2.929603093369726,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 180320
-    },
-    {
-      "epoch": 2.9297655602670956,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2878,
-      "step": 180330
-    },
-    {
-      "epoch": 2.929928027164465,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 180340
-    },
-    {
-      "epoch": 2.930090494061835,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 180350
-    },
-    {
-      "epoch": 2.9302529609592045,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 180360
-    },
-    {
-      "epoch": 2.930415427856574,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 180370
-    },
-    {
-      "epoch": 2.930577894753944,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 180380
-    },
-    {
-      "epoch": 2.9307403616513135,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 180390
-    },
-    {
-      "epoch": 2.930902828548683,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.306,
-      "step": 180400
-    },
-    {
-      "epoch": 2.9310652954460528,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 180410
-    },
-    {
-      "epoch": 2.9312277623434224,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 180420
-    },
-    {
-      "epoch": 2.931390229240792,
-      "grad_norm": 2.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 180430
-    },
-    {
-      "epoch": 2.9315526961381617,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 180440
-    },
-    {
-      "epoch": 2.9317151630355314,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2964,
-      "step": 180450
-    },
-    {
-      "epoch": 2.931877629932901,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 180460
-    },
-    {
-      "epoch": 2.9320400968302707,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.3012,
-      "step": 180470
-    },
-    {
-      "epoch": 2.9322025637276408,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2957,
-      "step": 180480
-    },
-    {
-      "epoch": 2.93236503062501,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.3088,
-      "step": 180490
-    },
-    {
-      "epoch": 2.93252749752238,
-      "grad_norm": 2.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2956,
-      "step": 180500
-    },
-    {
-      "epoch": 2.9326899644197493,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 180510
-    },
-    {
-      "epoch": 2.9328524313171194,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.3094,
-      "step": 180520
-    },
-    {
-      "epoch": 2.9330148982144886,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2783,
-      "step": 180530
-    },
-    {
-      "epoch": 2.9331773651118587,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2711,
-      "step": 180540
-    },
-    {
-      "epoch": 2.933339832009228,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 180550
-    },
-    {
-      "epoch": 2.933502298906598,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2855,
-      "step": 180560
-    },
-    {
-      "epoch": 2.933664765803967,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 180570
-    },
-    {
-      "epoch": 2.9338272327013373,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 180580
-    },
-    {
-      "epoch": 2.933989699598707,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 180590
-    },
-    {
-      "epoch": 2.9341521664960766,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2787,
-      "step": 180600
-    },
-    {
-      "epoch": 2.9343146333934462,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 180610
-    },
-    {
-      "epoch": 2.934477100290816,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2692,
-      "step": 180620
-    },
-    {
-      "epoch": 2.9346395671881855,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 180630
-    },
-    {
-      "epoch": 2.934802034085555,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.283,
-      "step": 180640
-    },
-    {
-      "epoch": 2.934964500982925,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2645,
-      "step": 180650
-    },
-    {
-      "epoch": 2.9351269678802945,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3052,
-      "step": 180660
-    },
-    {
-      "epoch": 2.935289434777664,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.308,
-      "step": 180670
-    },
-    {
-      "epoch": 2.935451901675034,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2789,
-      "step": 180680
-    },
-    {
-      "epoch": 2.9356143685724034,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3029,
-      "step": 180690
-    },
-    {
-      "epoch": 2.935776835469773,
-      "grad_norm": 3.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 180700
-    },
-    {
-      "epoch": 2.9359393023671427,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 180710
-    },
-    {
-      "epoch": 2.9361017692645124,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 180720
-    },
-    {
-      "epoch": 2.936264236161882,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2619,
-      "step": 180730
-    },
-    {
-      "epoch": 2.9364267030592517,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 180740
-    },
-    {
-      "epoch": 2.9365891699566213,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 180750
-    },
-    {
-      "epoch": 2.936751636853991,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 180760
-    },
-    {
-      "epoch": 2.9369141037513606,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 180770
-    },
-    {
-      "epoch": 2.9370765706487303,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 180780
-    },
-    {
-      "epoch": 2.9372390375461,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 180790
-    },
-    {
-      "epoch": 2.9374015044434696,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2831,
-      "step": 180800
-    },
-    {
-      "epoch": 2.9375639713408392,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 180810
-    },
-    {
-      "epoch": 2.937726438238209,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2976,
-      "step": 180820
-    },
-    {
-      "epoch": 2.9378889051355785,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3246,
-      "step": 180830
-    },
-    {
-      "epoch": 2.938051372032948,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.3013,
-      "step": 180840
-    },
-    {
-      "epoch": 2.938213838930318,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 180850
-    },
-    {
-      "epoch": 2.9383763058276875,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 180860
-    },
-    {
-      "epoch": 2.938538772725057,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 180870
-    },
-    {
-      "epoch": 2.938701239622427,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2552,
-      "step": 180880
-    },
-    {
-      "epoch": 2.9388637065197964,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 180890
-    },
-    {
-      "epoch": 2.939026173417166,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 180900
-    },
-    {
-      "epoch": 2.9391886403145357,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 180910
-    },
-    {
-      "epoch": 2.9393511072119054,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2668,
-      "step": 180920
-    },
-    {
-      "epoch": 2.9395135741092755,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2841,
-      "step": 180930
-    },
-    {
-      "epoch": 2.9396760410066447,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 180940
-    },
-    {
-      "epoch": 2.939838507904015,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 180950
-    },
-    {
-      "epoch": 2.940000974801384,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2869,
-      "step": 180960
-    },
-    {
-      "epoch": 2.940163441698754,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2798,
-      "step": 180970
-    },
-    {
-      "epoch": 2.9403259085961233,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2527,
-      "step": 180980
-    },
-    {
-      "epoch": 2.9404883754934934,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 180990
-    },
-    {
-      "epoch": 2.9406508423908626,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2588,
-      "step": 181000
-    },
-    {
-      "epoch": 2.9408133092882327,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2837,
-      "step": 181010
-    },
-    {
-      "epoch": 2.940975776185602,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3033,
-      "step": 181020
-    },
-    {
-      "epoch": 2.941138243082972,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 181030
-    },
-    {
-      "epoch": 2.9413007099803417,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 181040
-    },
-    {
-      "epoch": 2.9414631768777113,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2955,
-      "step": 181050
-    },
-    {
-      "epoch": 2.941625643775081,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 181060
-    },
-    {
-      "epoch": 2.9417881106724506,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2943,
-      "step": 181070
-    },
-    {
-      "epoch": 2.9419505775698203,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3028,
-      "step": 181080
-    },
-    {
-      "epoch": 2.94211304446719,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3085,
-      "step": 181090
-    },
-    {
-      "epoch": 2.9422755113645596,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 181100
-    },
-    {
-      "epoch": 2.942437978261929,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2879,
-      "step": 181110
-    },
-    {
-      "epoch": 2.942600445159299,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 181120
-    },
-    {
-      "epoch": 2.9427629120566685,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2843,
-      "step": 181130
-    },
-    {
-      "epoch": 2.942925378954038,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.3116,
-      "step": 181140
-    },
-    {
-      "epoch": 2.943087845851408,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.304,
-      "step": 181150
-    },
-    {
-      "epoch": 2.9432503127487775,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.306,
-      "step": 181160
-    },
-    {
-      "epoch": 2.943412779646147,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 181170
-    },
-    {
-      "epoch": 2.9435752465435168,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2969,
-      "step": 181180
-    },
-    {
-      "epoch": 2.9437377134408864,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3002,
-      "step": 181190
-    },
-    {
-      "epoch": 2.943900180338256,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3049,
-      "step": 181200
-    },
-    {
-      "epoch": 2.9440626472356257,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 181210
-    },
-    {
-      "epoch": 2.9442251141329954,
-      "grad_norm": 3.015625,
-      "learning_rate": 5e-05,
-      "loss": 0.3076,
-      "step": 181220
-    },
-    {
-      "epoch": 2.944387581030365,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 181230
-    },
-    {
-      "epoch": 2.9445500479277347,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 181240
-    },
-    {
-      "epoch": 2.9447125148251043,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.266,
-      "step": 181250
-    },
-    {
-      "epoch": 2.944874981722474,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 181260
-    },
-    {
-      "epoch": 2.9450374486198436,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2991,
-      "step": 181270
-    },
-    {
-      "epoch": 2.9451999155172133,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 181280
-    },
-    {
-      "epoch": 2.945362382414583,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2926,
-      "step": 181290
-    },
-    {
-      "epoch": 2.9455248493119526,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2992,
-      "step": 181300
-    },
-    {
-      "epoch": 2.9456873162093222,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3021,
-      "step": 181310
-    },
-    {
-      "epoch": 2.945849783106692,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 181320
-    },
-    {
-      "epoch": 2.9460122500040615,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 181330
-    },
-    {
-      "epoch": 2.946174716901431,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 181340
-    },
-    {
-      "epoch": 2.946337183798801,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 181350
-    },
-    {
-      "epoch": 2.946499650696171,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 181360
-    },
-    {
-      "epoch": 2.94666211759354,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 181370
-    },
-    {
-      "epoch": 2.9468245844909102,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 181380
-    },
-    {
-      "epoch": 2.9469870513882794,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2747,
-      "step": 181390
-    },
-    {
-      "epoch": 2.9471495182856495,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2661,
-      "step": 181400
-    },
-    {
-      "epoch": 2.9473119851830187,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2985,
-      "step": 181410
-    },
-    {
-      "epoch": 2.947474452080389,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 181420
-    },
-    {
-      "epoch": 2.947636918977758,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2936,
-      "step": 181430
-    },
-    {
-      "epoch": 2.947799385875128,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3037,
-      "step": 181440
-    },
-    {
-      "epoch": 2.9479618527724973,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3037,
-      "step": 181450
-    },
-    {
-      "epoch": 2.9481243196698674,
-      "grad_norm": 2.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2917,
-      "step": 181460
-    },
-    {
-      "epoch": 2.948286786567237,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 181470
-    },
-    {
-      "epoch": 2.9484492534646067,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 181480
-    },
-    {
-      "epoch": 2.9486117203619764,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 181490
-    },
-    {
-      "epoch": 2.948774187259346,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2606,
-      "step": 181500
-    },
-    {
-      "epoch": 2.9489366541567157,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 181510
-    },
-    {
-      "epoch": 2.9490991210540853,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 181520
-    },
-    {
-      "epoch": 2.949261587951455,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 181530
-    },
-    {
-      "epoch": 2.9494240548488246,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2778,
-      "step": 181540
-    },
-    {
-      "epoch": 2.9495865217461943,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 181550
-    },
-    {
-      "epoch": 2.949748988643564,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.3049,
-      "step": 181560
-    },
-    {
-      "epoch": 2.9499114555409336,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.302,
-      "step": 181570
-    },
-    {
-      "epoch": 2.9500739224383032,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2989,
-      "step": 181580
-    },
-    {
-      "epoch": 2.950236389335673,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.3095,
-      "step": 181590
-    },
-    {
-      "epoch": 2.9503988562330425,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 181600
-    },
-    {
-      "epoch": 2.950561323130412,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3104,
-      "step": 181610
-    },
-    {
-      "epoch": 2.950723790027782,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2996,
-      "step": 181620
-    },
-    {
-      "epoch": 2.9508862569251515,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 181630
-    },
-    {
-      "epoch": 2.951048723822521,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3181,
-      "step": 181640
-    },
-    {
-      "epoch": 2.951211190719891,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3118,
-      "step": 181650
-    },
-    {
-      "epoch": 2.9513736576172604,
-      "grad_norm": 3.046875,
-      "learning_rate": 5e-05,
-      "loss": 0.298,
-      "step": 181660
-    },
-    {
-      "epoch": 2.95153612451463,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.34,
-      "step": 181670
-    },
-    {
-      "epoch": 2.9516985914119998,
-      "grad_norm": 3.078125,
-      "learning_rate": 5e-05,
-      "loss": 0.2884,
-      "step": 181680
-    },
-    {
-      "epoch": 2.9518610583093694,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 181690
-    },
-    {
-      "epoch": 2.952023525206739,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3057,
-      "step": 181700
-    },
-    {
-      "epoch": 2.9521859921041087,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 181710
-    },
-    {
-      "epoch": 2.9523484590014784,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2853,
-      "step": 181720
-    },
-    {
-      "epoch": 2.952510925898848,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 181730
-    },
-    {
-      "epoch": 2.9526733927962177,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2997,
-      "step": 181740
-    },
-    {
-      "epoch": 2.9528358596935873,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 181750
-    },
-    {
-      "epoch": 2.952998326590957,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2835,
-      "step": 181760
-    },
-    {
-      "epoch": 2.9531607934883266,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2887,
-      "step": 181770
-    },
-    {
-      "epoch": 2.9533232603856963,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2903,
-      "step": 181780
-    },
-    {
-      "epoch": 2.953485727283066,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2757,
-      "step": 181790
-    },
-    {
-      "epoch": 2.9536481941804356,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 181800
-    },
-    {
-      "epoch": 2.9538106610778057,
-      "grad_norm": 2.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2655,
-      "step": 181810
-    },
-    {
-      "epoch": 2.953973127975175,
-      "grad_norm": 3.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2794,
-      "step": 181820
-    },
-    {
-      "epoch": 2.954135594872545,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2923,
-      "step": 181830
-    },
-    {
-      "epoch": 2.954298061769914,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.296,
-      "step": 181840
-    },
-    {
-      "epoch": 2.9544605286672843,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2647,
-      "step": 181850
-    },
-    {
-      "epoch": 2.9546229955646535,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2826,
-      "step": 181860
-    },
-    {
-      "epoch": 2.9547854624620236,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2637,
-      "step": 181870
-    },
-    {
-      "epoch": 2.9549479293593928,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3008,
-      "step": 181880
-    },
-    {
-      "epoch": 2.955110396256763,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 181890
-    },
-    {
-      "epoch": 2.955272863154132,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3069,
-      "step": 181900
-    },
-    {
-      "epoch": 2.955435330051502,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 181910
-    },
-    {
-      "epoch": 2.955597796948872,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.3004,
-      "step": 181920
-    },
-    {
-      "epoch": 2.9557602638462415,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2866,
-      "step": 181930
-    },
-    {
-      "epoch": 2.955922730743611,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3084,
-      "step": 181940
-    },
-    {
-      "epoch": 2.9560851976409808,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2725,
-      "step": 181950
-    },
-    {
-      "epoch": 2.9562476645383504,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3103,
-      "step": 181960
-    },
-    {
-      "epoch": 2.95641013143572,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.297,
-      "step": 181970
-    },
-    {
-      "epoch": 2.9565725983330897,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 181980
-    },
-    {
-      "epoch": 2.9567350652304594,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 181990
-    },
-    {
-      "epoch": 2.956897532127829,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2994,
-      "step": 182000
-    },
-    {
-      "epoch": 2.9570599990251987,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 182010
-    },
-    {
-      "epoch": 2.9572224659225683,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2816,
-      "step": 182020
-    },
-    {
-      "epoch": 2.957384932819938,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 182030
-    },
-    {
-      "epoch": 2.9575473997173076,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2636,
-      "step": 182040
-    },
-    {
-      "epoch": 2.9577098666146773,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2902,
-      "step": 182050
-    },
-    {
-      "epoch": 2.957872333512047,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 182060
-    },
-    {
-      "epoch": 2.9580348004094166,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2818,
-      "step": 182070
-    },
-    {
-      "epoch": 2.9581972673067862,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2786,
-      "step": 182080
-    },
-    {
-      "epoch": 2.958359734204156,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2957,
-      "step": 182090
-    },
-    {
-      "epoch": 2.9585222011015255,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3131,
-      "step": 182100
-    },
-    {
-      "epoch": 2.958684667998895,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.312,
-      "step": 182110
-    },
-    {
-      "epoch": 2.958847134896265,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2709,
-      "step": 182120
-    },
-    {
-      "epoch": 2.9590096017936345,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3011,
-      "step": 182130
-    },
-    {
-      "epoch": 2.959172068691004,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.3,
-      "step": 182140
-    },
-    {
-      "epoch": 2.959334535588374,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2962,
-      "step": 182150
-    },
-    {
-      "epoch": 2.9594970024857434,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2898,
-      "step": 182160
-    },
-    {
-      "epoch": 2.959659469383113,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.289,
-      "step": 182170
-    },
-    {
-      "epoch": 2.9598219362804827,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2814,
-      "step": 182180
-    },
-    {
-      "epoch": 2.9599844031778524,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 182190
-    },
-    {
-      "epoch": 2.960146870075222,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2861,
-      "step": 182200
-    },
-    {
-      "epoch": 2.9603093369725917,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2983,
-      "step": 182210
-    },
-    {
-      "epoch": 2.9604718038699613,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.3023,
-      "step": 182220
-    },
-    {
-      "epoch": 2.960634270767331,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3124,
-      "step": 182230
-    },
-    {
-      "epoch": 2.960796737664701,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 182240
-    },
-    {
-      "epoch": 2.9609592045620703,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 182250
-    },
-    {
-      "epoch": 2.9611216714594404,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2779,
-      "step": 182260
-    },
-    {
-      "epoch": 2.9612841383568096,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2646,
-      "step": 182270
-    },
-    {
-      "epoch": 2.9614466052541797,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2625,
-      "step": 182280
-    },
-    {
-      "epoch": 2.961609072151549,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 182290
-    },
-    {
-      "epoch": 2.961771539048919,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.301,
-      "step": 182300
-    },
-    {
-      "epoch": 2.961934005946288,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.3086,
-      "step": 182310
-    },
-    {
-      "epoch": 2.9620964728436583,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 182320
-    },
-    {
-      "epoch": 2.9622589397410275,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 182330
-    },
-    {
-      "epoch": 2.9624214066383976,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2804,
-      "step": 182340
-    },
-    {
-      "epoch": 2.9625838735357672,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2784,
-      "step": 182350
-    },
-    {
-      "epoch": 2.962746340433137,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2948,
-      "step": 182360
-    },
-    {
-      "epoch": 2.9629088073305065,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2761,
-      "step": 182370
-    },
-    {
-      "epoch": 2.963071274227876,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2769,
-      "step": 182380
-    },
-    {
-      "epoch": 2.963233741125246,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2965,
-      "step": 182390
-    },
-    {
-      "epoch": 2.9633962080226155,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 182400
-    },
-    {
-      "epoch": 2.963558674919985,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2856,
-      "step": 182410
-    },
-    {
-      "epoch": 2.963721141817355,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 182420
-    },
-    {
-      "epoch": 2.9638836087147244,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2729,
-      "step": 182430
-    },
-    {
-      "epoch": 2.964046075612094,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2697,
-      "step": 182440
-    },
-    {
-      "epoch": 2.9642085425094638,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.271,
-      "step": 182450
-    },
-    {
-      "epoch": 2.9643710094068334,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2925,
-      "step": 182460
-    },
-    {
-      "epoch": 2.964533476304203,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 182470
-    },
-    {
-      "epoch": 2.9646959432015727,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2952,
-      "step": 182480
-    },
-    {
-      "epoch": 2.9648584100989424,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2754,
-      "step": 182490
-    },
-    {
-      "epoch": 2.965020876996312,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2984,
-      "step": 182500
-    },
-    {
-      "epoch": 2.9651833438936817,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.267,
-      "step": 182510
-    },
-    {
-      "epoch": 2.9653458107910513,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2911,
-      "step": 182520
-    },
-    {
-      "epoch": 2.965508277688421,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2928,
-      "step": 182530
-    },
-    {
-      "epoch": 2.9656707445857906,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3064,
-      "step": 182540
-    },
-    {
-      "epoch": 2.9658332114831603,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2987,
-      "step": 182550
-    },
-    {
-      "epoch": 2.96599567838053,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 182560
-    },
-    {
-      "epoch": 2.9661581452778996,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 182570
-    },
-    {
-      "epoch": 2.966320612175269,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2825,
-      "step": 182580
-    },
-    {
-      "epoch": 2.966483079072639,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2693,
-      "step": 182590
-    },
-    {
-      "epoch": 2.9666455459700085,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2677,
-      "step": 182600
-    },
-    {
-      "epoch": 2.966808012867378,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2891,
-      "step": 182610
-    },
-    {
-      "epoch": 2.966970479764748,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2736,
-      "step": 182620
-    },
-    {
-      "epoch": 2.9671329466621175,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3045,
-      "step": 182630
-    },
-    {
-      "epoch": 2.967295413559487,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2986,
-      "step": 182640
-    },
-    {
-      "epoch": 2.9674578804568568,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2969,
-      "step": 182650
-    },
-    {
-      "epoch": 2.9676203473542264,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2739,
-      "step": 182660
-    },
-    {
-      "epoch": 2.9677828142515965,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 182670
-    },
-    {
-      "epoch": 2.9679452811489657,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2748,
-      "step": 182680
-    },
-    {
-      "epoch": 2.968107748046336,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 182690
-    },
-    {
-      "epoch": 2.968270214943705,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 182700
-    },
-    {
-      "epoch": 2.968432681841075,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 182710
-    },
-    {
-      "epoch": 2.9685951487384443,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 182720
-    },
-    {
-      "epoch": 2.9687576156358144,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2966,
-      "step": 182730
-    },
-    {
-      "epoch": 2.9689200825331836,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.294,
-      "step": 182740
-    },
-    {
-      "epoch": 2.9690825494305537,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.3018,
-      "step": 182750
-    },
-    {
-      "epoch": 2.969245016327923,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2671,
-      "step": 182760
-    },
-    {
-      "epoch": 2.969407483225293,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.305,
-      "step": 182770
-    },
-    {
-      "epoch": 2.9695699501226627,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2889,
-      "step": 182780
-    },
-    {
-      "epoch": 2.9697324170200323,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2805,
-      "step": 182790
-    },
-    {
-      "epoch": 2.969894883917402,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 182800
-    },
-    {
-      "epoch": 2.9700573508147716,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2568,
-      "step": 182810
-    },
-    {
-      "epoch": 2.9702198177121413,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 182820
-    },
-    {
-      "epoch": 2.970382284609511,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.279,
-      "step": 182830
-    },
-    {
-      "epoch": 2.9705447515068806,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2571,
-      "step": 182840
-    },
-    {
-      "epoch": 2.9707072184042502,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2896,
-      "step": 182850
-    },
-    {
-      "epoch": 2.97086968530162,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2985,
-      "step": 182860
-    },
-    {
-      "epoch": 2.9710321521989895,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 182870
-    },
-    {
-      "epoch": 2.971194619096359,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2967,
-      "step": 182880
-    },
-    {
-      "epoch": 2.971357085993729,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 182890
-    },
-    {
-      "epoch": 2.9715195528910985,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 182900
-    },
-    {
-      "epoch": 2.971682019788468,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2899,
-      "step": 182910
-    },
-    {
-      "epoch": 2.971844486685838,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 182920
-    },
-    {
-      "epoch": 2.9720069535832074,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.3066,
-      "step": 182930
-    },
-    {
-      "epoch": 2.972169420480577,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 182940
-    },
-    {
-      "epoch": 2.9723318873779467,
-      "grad_norm": 2.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 182950
-    },
-    {
-      "epoch": 2.9724943542753164,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 182960
-    },
-    {
-      "epoch": 2.972656821172686,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 182970
-    },
-    {
-      "epoch": 2.9728192880700557,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2947,
-      "step": 182980
-    },
-    {
-      "epoch": 2.9729817549674253,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 182990
-    },
-    {
-      "epoch": 2.973144221864795,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 183000
-    },
-    {
-      "epoch": 2.9733066887621646,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2819,
-      "step": 183010
-    },
-    {
-      "epoch": 2.9734691556595343,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2742,
-      "step": 183020
-    },
-    {
-      "epoch": 2.973631622556904,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2733,
-      "step": 183030
-    },
-    {
-      "epoch": 2.9737940894542736,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2734,
-      "step": 183040
-    },
-    {
-      "epoch": 2.9739565563516432,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2755,
-      "step": 183050
-    },
-    {
-      "epoch": 2.974119023249013,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 183060
-    },
-    {
-      "epoch": 2.9742814901463825,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 183070
-    },
-    {
-      "epoch": 2.974443957043752,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 183080
-    },
-    {
-      "epoch": 2.974606423941122,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2792,
-      "step": 183090
-    },
-    {
-      "epoch": 2.9747688908384915,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2905,
-      "step": 183100
-    },
-    {
-      "epoch": 2.974931357735861,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 183110
-    },
-    {
-      "epoch": 2.9750938246332312,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2929,
-      "step": 183120
-    },
-    {
-      "epoch": 2.9752562915306004,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2726,
-      "step": 183130
-    },
-    {
-      "epoch": 2.9754187584279705,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2946,
-      "step": 183140
-    },
-    {
-      "epoch": 2.9755812253253398,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.3005,
-      "step": 183150
-    },
-    {
-      "epoch": 2.97574369222271,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2877,
-      "step": 183160
-    },
-    {
-      "epoch": 2.975906159120079,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2796,
-      "step": 183170
-    },
-    {
-      "epoch": 2.976068626017449,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2851,
-      "step": 183180
-    },
-    {
-      "epoch": 2.9762310929148184,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.3022,
-      "step": 183190
-    },
-    {
-      "epoch": 2.9763935598121884,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.3141,
-      "step": 183200
-    },
-    {
-      "epoch": 2.9765560267095577,
-      "grad_norm": 3.359375,
-      "learning_rate": 5e-05,
-      "loss": 0.2951,
-      "step": 183210
-    },
-    {
-      "epoch": 2.9767184936069278,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3069,
-      "step": 183220
-    },
-    {
-      "epoch": 2.9768809605042974,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.3045,
-      "step": 183230
-    },
-    {
-      "epoch": 2.977043427401667,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.3139,
-      "step": 183240
-    },
-    {
-      "epoch": 2.9772058942990367,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2953,
-      "step": 183250
-    },
-    {
-      "epoch": 2.9773683611964064,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2892,
-      "step": 183260
-    },
-    {
-      "epoch": 2.977530828093776,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2858,
-      "step": 183270
-    },
-    {
-      "epoch": 2.9776932949911457,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2989,
-      "step": 183280
-    },
-    {
-      "epoch": 2.9778557618885153,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.303,
-      "step": 183290
-    },
-    {
-      "epoch": 2.978018228785885,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2942,
-      "step": 183300
-    },
-    {
-      "epoch": 2.9781806956832546,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.273,
-      "step": 183310
-    },
-    {
-      "epoch": 2.9783431625806243,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.3118,
-      "step": 183320
-    },
-    {
-      "epoch": 2.978505629477994,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2962,
-      "step": 183330
-    },
-    {
-      "epoch": 2.9786680963753636,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2591,
-      "step": 183340
-    },
-    {
-      "epoch": 2.978830563272733,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2933,
-      "step": 183350
-    },
-    {
-      "epoch": 2.978993030170103,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 183360
-    },
-    {
-      "epoch": 2.9791554970674725,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 183370
-    },
-    {
-      "epoch": 2.979317963964842,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.272,
-      "step": 183380
-    },
-    {
-      "epoch": 2.979480430862212,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2933,
-      "step": 183390
-    },
-    {
-      "epoch": 2.9796428977595815,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2719,
-      "step": 183400
-    },
-    {
-      "epoch": 2.979805364656951,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 183410
-    },
-    {
-      "epoch": 2.9799678315543208,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 183420
-    },
-    {
-      "epoch": 2.9801302984516904,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 183430
-    },
-    {
-      "epoch": 2.98029276534906,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2827,
-      "step": 183440
-    },
-    {
-      "epoch": 2.9804552322464297,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2941,
-      "step": 183450
-    },
-    {
-      "epoch": 2.9806176991437994,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2701,
-      "step": 183460
-    },
-    {
-      "epoch": 2.980780166041169,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 183470
-    },
-    {
-      "epoch": 2.9809426329385387,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2881,
-      "step": 183480
-    },
-    {
-      "epoch": 2.9811050998359083,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2975,
-      "step": 183490
-    },
-    {
-      "epoch": 2.981267566733278,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2854,
-      "step": 183500
-    },
-    {
-      "epoch": 2.9814300336306476,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2765,
-      "step": 183510
-    },
-    {
-      "epoch": 2.9815925005280173,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2904,
-      "step": 183520
-    },
-    {
-      "epoch": 2.981754967425387,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.3058,
-      "step": 183530
-    },
-    {
-      "epoch": 2.9819174343227566,
-      "grad_norm": 3.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2844,
-      "step": 183540
-    },
-    {
-      "epoch": 2.9820799012201267,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2998,
-      "step": 183550
-    },
-    {
-      "epoch": 2.982242368117496,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3088,
-      "step": 183560
-    },
-    {
-      "epoch": 2.982404835014866,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.3122,
-      "step": 183570
-    },
-    {
-      "epoch": 2.982567301912235,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.299,
-      "step": 183580
-    },
-    {
-      "epoch": 2.9827297688096053,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.287,
-      "step": 183590
-    },
-    {
-      "epoch": 2.9828922357069745,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 183600
-    },
-    {
-      "epoch": 2.9830547026043446,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 183610
-    },
-    {
-      "epoch": 2.983217169501714,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.285,
-      "step": 183620
-    },
-    {
-      "epoch": 2.983379636399084,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.3067,
-      "step": 183630
-    },
-    {
-      "epoch": 2.983542103296453,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3323,
-      "step": 183640
-    },
-    {
-      "epoch": 2.983704570193823,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.3015,
-      "step": 183650
-    },
-    {
-      "epoch": 2.983867037091193,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2981,
-      "step": 183660
-    },
-    {
-      "epoch": 2.9840295039885625,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2862,
-      "step": 183670
-    },
-    {
-      "epoch": 2.984191970885932,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 183680
-    },
-    {
-      "epoch": 2.984354437783302,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2888,
-      "step": 183690
-    },
-    {
-      "epoch": 2.9845169046806714,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2883,
-      "step": 183700
-    },
-    {
-      "epoch": 2.984679371578041,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.292,
-      "step": 183710
-    },
-    {
-      "epoch": 2.9848418384754107,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2864,
-      "step": 183720
-    },
-    {
-      "epoch": 2.9850043053727804,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2699,
-      "step": 183730
-    },
-    {
-      "epoch": 2.98516677227015,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.2542,
-      "step": 183740
-    },
-    {
-      "epoch": 2.9853292391675197,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2694,
-      "step": 183750
-    },
-    {
-      "epoch": 2.9854917060648893,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2812,
-      "step": 183760
-    },
-    {
-      "epoch": 2.985654172962259,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2763,
-      "step": 183770
-    },
-    {
-      "epoch": 2.9858166398596286,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2893,
-      "step": 183780
-    },
-    {
-      "epoch": 2.9859791067569983,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2945,
-      "step": 183790
-    },
-    {
-      "epoch": 2.986141573654368,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.2797,
-      "step": 183800
-    },
-    {
-      "epoch": 2.9863040405517376,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.3036,
-      "step": 183810
-    },
-    {
-      "epoch": 2.9864665074491072,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2865,
-      "step": 183820
-    },
-    {
-      "epoch": 2.986628974346477,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2746,
-      "step": 183830
-    },
-    {
-      "epoch": 2.9867914412438465,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2577,
-      "step": 183840
-    },
-    {
-      "epoch": 2.986953908141216,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2833,
-      "step": 183850
-    },
-    {
-      "epoch": 2.987116375038586,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 183860
-    },
-    {
-      "epoch": 2.9872788419359555,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 183870
-    },
-    {
-      "epoch": 2.987441308833325,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2963,
-      "step": 183880
-    },
-    {
-      "epoch": 2.987603775730695,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2724,
-      "step": 183890
-    },
-    {
-      "epoch": 2.9877662426280645,
-      "grad_norm": 3.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.261,
-      "step": 183900
-    },
-    {
-      "epoch": 2.987928709525434,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2859,
-      "step": 183910
-    },
-    {
-      "epoch": 2.9880911764228038,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 183920
-    },
-    {
-      "epoch": 2.9882536433201734,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2872,
-      "step": 183930
-    },
-    {
-      "epoch": 2.988416110217543,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.3034,
-      "step": 183940
-    },
-    {
-      "epoch": 2.9885785771149127,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2884,
-      "step": 183950
-    },
-    {
-      "epoch": 2.9887410440122824,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.281,
-      "step": 183960
-    },
-    {
-      "epoch": 2.988903510909652,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 183970
-    },
-    {
-      "epoch": 2.9890659778070217,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2846,
-      "step": 183980
-    },
-    {
-      "epoch": 2.9892284447043913,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.278,
-      "step": 183990
-    },
-    {
-      "epoch": 2.9893909116017614,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2932,
-      "step": 184000
-    },
-    {
-      "epoch": 2.9895533784991306,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.282,
-      "step": 184010
-    },
-    {
-      "epoch": 2.9897158453965007,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2994,
-      "step": 184020
-    },
-    {
-      "epoch": 2.98987831229387,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2977,
-      "step": 184030
-    },
-    {
-      "epoch": 2.99004077919124,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.3124,
-      "step": 184040
-    },
-    {
-      "epoch": 2.990203246088609,
-      "grad_norm": 3.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2773,
-      "step": 184050
-    },
-    {
-      "epoch": 2.9903657129859793,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2979,
-      "step": 184060
-    },
-    {
-      "epoch": 2.9905281798833485,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2914,
-      "step": 184070
-    },
-    {
-      "epoch": 2.9906906467807186,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2703,
-      "step": 184080
-    },
-    {
-      "epoch": 2.990853113678088,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2806,
-      "step": 184090
-    },
-    {
-      "epoch": 2.991015580575458,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2776,
-      "step": 184100
-    },
-    {
-      "epoch": 2.9911780474728276,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2744,
-      "step": 184110
-    },
-    {
-      "epoch": 2.991340514370197,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2658,
-      "step": 184120
-    },
-    {
-      "epoch": 2.991502981267567,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.2988,
-      "step": 184130
-    },
-    {
-      "epoch": 2.9916654481649365,
-      "grad_norm": 2.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2566,
-      "step": 184140
-    },
-    {
-      "epoch": 2.991827915062306,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2885,
-      "step": 184150
-    },
-    {
-      "epoch": 2.991990381959676,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.3113,
-      "step": 184160
-    },
-    {
-      "epoch": 2.9921528488570455,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.3027,
-      "step": 184170
-    },
-    {
-      "epoch": 2.992315315754415,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.3023,
-      "step": 184180
-    },
-    {
-      "epoch": 2.9924777826517848,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2912,
-      "step": 184190
-    },
-    {
-      "epoch": 2.9926402495491544,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2768,
-      "step": 184200
-    },
-    {
-      "epoch": 2.992802716446524,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2901,
-      "step": 184210
-    },
-    {
-      "epoch": 2.9929651833438937,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2737,
-      "step": 184220
-    },
-    {
-      "epoch": 2.9931276502412634,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2834,
-      "step": 184230
-    },
-    {
-      "epoch": 2.993290117138633,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 184240
-    },
-    {
-      "epoch": 2.9934525840360027,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2848,
-      "step": 184250
-    },
-    {
-      "epoch": 2.9936150509333723,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2756,
-      "step": 184260
-    },
-    {
-      "epoch": 2.993777517830742,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2749,
-      "step": 184270
-    },
-    {
-      "epoch": 2.9939399847281116,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.3093,
-      "step": 184280
-    },
-    {
-      "epoch": 2.9941024516254813,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2876,
-      "step": 184290
-    },
-    {
-      "epoch": 2.994264918522851,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 184300
-    },
-    {
-      "epoch": 2.9944273854202206,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2788,
-      "step": 184310
-    },
-    {
-      "epoch": 2.9945898523175902,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2981,
-      "step": 184320
-    },
-    {
-      "epoch": 2.99475231921496,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2857,
-      "step": 184330
-    },
-    {
-      "epoch": 2.9949147861123295,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2988,
-      "step": 184340
-    },
-    {
-      "epoch": 2.995077253009699,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2973,
-      "step": 184350
-    },
-    {
-      "epoch": 2.995239719907069,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.325,
-      "step": 184360
-    },
-    {
-      "epoch": 2.9954021868044385,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2802,
-      "step": 184370
-    },
-    {
-      "epoch": 2.995564653701808,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2842,
-      "step": 184380
-    },
-    {
-      "epoch": 2.995727120599178,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2954,
-      "step": 184390
-    },
-    {
-      "epoch": 2.9958895874965474,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.3046,
-      "step": 184400
-    },
-    {
-      "epoch": 2.996052054393917,
-      "grad_norm": 3.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 184410
-    },
-    {
-      "epoch": 2.9962145212912867,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2759,
-      "step": 184420
-    },
-    {
-      "epoch": 2.996376988188657,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2946,
-      "step": 184430
-    },
-    {
-      "epoch": 2.996539455086026,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2776,
-      "step": 184440
-    },
-    {
-      "epoch": 2.996701921983396,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2777,
-      "step": 184450
-    },
-    {
-      "epoch": 2.9968643888807653,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.3109,
-      "step": 184460
-    },
-    {
-      "epoch": 2.9970268557781354,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.3018,
-      "step": 184470
-    },
-    {
-      "epoch": 2.9971893226755046,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.3073,
-      "step": 184480
-    },
-    {
-      "epoch": 2.9973517895728747,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2919,
-      "step": 184490
-    },
-    {
-      "epoch": 2.997514256470244,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2752,
-      "step": 184500
-    },
-    {
-      "epoch": 2.997676723367614,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2916,
-      "step": 184510
-    },
-    {
-      "epoch": 2.9978391902649832,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.3137,
-      "step": 184520
-    },
-    {
-      "epoch": 2.9980016571623533,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2822,
-      "step": 184530
-    },
-    {
-      "epoch": 2.998164124059723,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.3014,
-      "step": 184540
-    },
-    {
-      "epoch": 2.9983265909570926,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2847,
-      "step": 184550
-    },
-    {
-      "epoch": 2.9984890578544623,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2935,
-      "step": 184560
-    },
-    {
-      "epoch": 2.998651524751832,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 184570
-    },
-    {
-      "epoch": 2.9988139916492016,
-      "grad_norm": 3.078125,
-      "learning_rate": 5e-05,
-      "loss": 0.3002,
-      "step": 184580
-    },
-    {
-      "epoch": 2.9989764585465712,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2698,
-      "step": 184590
-    },
-    {
-      "epoch": 2.999138925443941,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.3076,
-      "step": 184600
-    },
-    {
-      "epoch": 2.9993013923413105,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2944,
-      "step": 184610
-    },
-    {
-      "epoch": 2.99946385923868,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2795,
-      "step": 184620
-    },
-    {
-      "epoch": 2.99962632613605,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.2957,
-      "step": 184630
-    },
-    {
-      "epoch": 2.9997887930334195,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.2677,
-      "step": 184640
-    },
-    {
-      "epoch": 2.999951259930789,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.28,
-      "step": 184650
-    },
-    {
-      "epoch": 3.000113726828159,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2652,
-      "step": 184660
-    },
-    {
-      "epoch": 3.0002761937255285,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2732,
-      "step": 184670
-    },
-    {
-      "epoch": 3.000438660622898,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2873,
-      "step": 184680
-    },
-    {
-      "epoch": 3.0006011275202678,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2446,
-      "step": 184690
-    },
-    {
-      "epoch": 3.0007635944176374,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2459,
-      "step": 184700
-    },
-    {
-      "epoch": 3.000926061315007,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2465,
-      "step": 184710
-    },
-    {
-      "epoch": 3.0010885282123767,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.243,
-      "step": 184720
-    },
-    {
-      "epoch": 3.0012509951097464,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2521,
-      "step": 184730
-    },
-    {
-      "epoch": 3.001413462007116,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2414,
-      "step": 184740
-    },
-    {
-      "epoch": 3.0015759289044857,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2532,
-      "step": 184750
-    },
-    {
-      "epoch": 3.0017383958018553,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2532,
-      "step": 184760
-    },
-    {
-      "epoch": 3.001900862699225,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2608,
-      "step": 184770
-    },
-    {
-      "epoch": 3.0020633295965946,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.256,
-      "step": 184780
-    },
-    {
-      "epoch": 3.0022257964939643,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.2501,
-      "step": 184790
-    },
-    {
-      "epoch": 3.002388263391334,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2809,
-      "step": 184800
-    },
-    {
-      "epoch": 3.0025507302887036,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2614,
-      "step": 184810
-    },
-    {
-      "epoch": 3.002713197186073,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.2455,
-      "step": 184820
-    },
-    {
-      "epoch": 3.002875664083443,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2509,
-      "step": 184830
-    },
-    {
-      "epoch": 3.0030381309808125,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2519,
-      "step": 184840
-    },
-    {
-      "epoch": 3.003200597878182,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2745,
-      "step": 184850
-    },
-    {
-      "epoch": 3.003363064775552,
-      "grad_norm": 3.046875,
-      "learning_rate": 5e-05,
-      "loss": 0.2716,
-      "step": 184860
-    },
-    {
-      "epoch": 3.0035255316729215,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2572,
-      "step": 184870
-    },
-    {
-      "epoch": 3.003687998570291,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2393,
-      "step": 184880
-    },
-    {
-      "epoch": 3.0038504654676608,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.254,
-      "step": 184890
-    },
-    {
-      "epoch": 3.0040129323650304,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2601,
-      "step": 184900
-    },
-    {
-      "epoch": 3.0041753992624,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.27,
-      "step": 184910
-    },
-    {
-      "epoch": 3.00433786615977,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2653,
-      "step": 184920
-    },
-    {
-      "epoch": 3.00450033305714,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.25,
-      "step": 184930
-    },
-    {
-      "epoch": 3.0046627999545095,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2871,
-      "step": 184940
-    },
-    {
-      "epoch": 3.004825266851879,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2507,
-      "step": 184950
-    },
-    {
-      "epoch": 3.0049877337492488,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2723,
-      "step": 184960
-    },
-    {
-      "epoch": 3.0051502006466184,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2566,
-      "step": 184970
-    },
-    {
-      "epoch": 3.005312667543988,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2329,
-      "step": 184980
-    },
-    {
-      "epoch": 3.0054751344413577,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2476,
-      "step": 184990
-    },
-    {
-      "epoch": 3.0056376013387274,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2465,
-      "step": 185000
-    },
-    {
-      "epoch": 3.005800068236097,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2384,
-      "step": 185010
-    },
-    {
-      "epoch": 3.0059625351334667,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2406,
-      "step": 185020
-    },
-    {
-      "epoch": 3.0061250020308363,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2249,
-      "step": 185030
-    },
-    {
-      "epoch": 3.006287468928206,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2326,
-      "step": 185040
-    },
-    {
-      "epoch": 3.0064499358255756,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 185050
-    },
-    {
-      "epoch": 3.0066124027229453,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2535,
-      "step": 185060
-    },
-    {
-      "epoch": 3.006774869620315,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2278,
-      "step": 185070
-    },
-    {
-      "epoch": 3.0069373365176846,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2321,
-      "step": 185080
-    },
-    {
-      "epoch": 3.0070998034150542,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2537,
-      "step": 185090
-    },
-    {
-      "epoch": 3.007262270312424,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2334,
-      "step": 185100
-    },
-    {
-      "epoch": 3.0074247372097935,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2246,
-      "step": 185110
-    },
-    {
-      "epoch": 3.007587204107163,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2297,
-      "step": 185120
-    },
-    {
-      "epoch": 3.007749671004533,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2349,
-      "step": 185130
-    },
-    {
-      "epoch": 3.0079121379019025,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2557,
-      "step": 185140
-    },
-    {
-      "epoch": 3.008074604799272,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2584,
-      "step": 185150
-    },
-    {
-      "epoch": 3.008237071696642,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2401,
-      "step": 185160
-    },
-    {
-      "epoch": 3.0083995385940114,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 185170
-    },
-    {
-      "epoch": 3.008562005491381,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2564,
-      "step": 185180
-    },
-    {
-      "epoch": 3.0087244723887507,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2489,
-      "step": 185190
-    },
-    {
-      "epoch": 3.0088869392861204,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2527,
-      "step": 185200
-    },
-    {
-      "epoch": 3.00904940618349,
-      "grad_norm": 3.484375,
-      "learning_rate": 5e-05,
-      "loss": 0.2482,
-      "step": 185210
-    },
-    {
-      "epoch": 3.0092118730808597,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2514,
-      "step": 185220
-    },
-    {
-      "epoch": 3.0093743399782293,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.245,
-      "step": 185230
-    },
-    {
-      "epoch": 3.009536806875599,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2299,
-      "step": 185240
-    },
-    {
-      "epoch": 3.0096992737729686,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2341,
-      "step": 185250
-    },
-    {
-      "epoch": 3.0098617406703383,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2482,
-      "step": 185260
-    },
-    {
-      "epoch": 3.010024207567708,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2392,
-      "step": 185270
-    },
-    {
-      "epoch": 3.0101866744650776,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2465,
-      "step": 185280
-    },
-    {
-      "epoch": 3.0103491413624472,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 185290
-    },
-    {
-      "epoch": 3.010511608259817,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 185300
-    },
-    {
-      "epoch": 3.0106740751571865,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2281,
-      "step": 185310
-    },
-    {
-      "epoch": 3.010836542054556,
-      "grad_norm": 3.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2223,
-      "step": 185320
-    },
-    {
-      "epoch": 3.010999008951926,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2463,
-      "step": 185330
-    },
-    {
-      "epoch": 3.0111614758492955,
-      "grad_norm": 3.234375,
-      "learning_rate": 5e-05,
-      "loss": 0.2227,
-      "step": 185340
-    },
-    {
-      "epoch": 3.011323942746665,
-      "grad_norm": 3.0,
-      "learning_rate": 5e-05,
-      "loss": 0.216,
-      "step": 185350
-    },
-    {
-      "epoch": 3.0114864096440352,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 185360
-    },
-    {
-      "epoch": 3.011648876541405,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 185370
-    },
-    {
-      "epoch": 3.0118113434387745,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2166,
-      "step": 185380
-    },
-    {
-      "epoch": 3.011973810336144,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2226,
-      "step": 185390
-    },
-    {
-      "epoch": 3.012136277233514,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 185400
-    },
-    {
-      "epoch": 3.0122987441308835,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 185410
-    },
-    {
-      "epoch": 3.012461211028253,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2368,
-      "step": 185420
-    },
-    {
-      "epoch": 3.012623677925623,
-      "grad_norm": 3.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2335,
-      "step": 185430
-    },
-    {
-      "epoch": 3.0127861448229925,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2282,
-      "step": 185440
-    },
-    {
-      "epoch": 3.012948611720362,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2324,
-      "step": 185450
-    },
-    {
-      "epoch": 3.0131110786177318,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2272,
-      "step": 185460
-    },
-    {
-      "epoch": 3.0132735455151014,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 185470
-    },
-    {
-      "epoch": 3.013436012412471,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2304,
-      "step": 185480
-    },
-    {
-      "epoch": 3.0135984793098407,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2217,
-      "step": 185490
-    },
-    {
-      "epoch": 3.0137609462072104,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 185500
-    },
-    {
-      "epoch": 3.01392341310458,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2379,
-      "step": 185510
-    },
-    {
-      "epoch": 3.0140858800019497,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2389,
-      "step": 185520
-    },
-    {
-      "epoch": 3.0142483468993193,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2283,
-      "step": 185530
-    },
-    {
-      "epoch": 3.014410813796689,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2317,
-      "step": 185540
-    },
-    {
-      "epoch": 3.0145732806940586,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2259,
-      "step": 185550
-    },
-    {
-      "epoch": 3.0147357475914283,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 185560
-    },
-    {
-      "epoch": 3.014898214488798,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2337,
-      "step": 185570
-    },
-    {
-      "epoch": 3.0150606813861676,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 185580
-    },
-    {
-      "epoch": 3.015223148283537,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 185590
-    },
-    {
-      "epoch": 3.015385615180907,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 185600
-    },
-    {
-      "epoch": 3.0155480820782765,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2494,
-      "step": 185610
-    },
-    {
-      "epoch": 3.015710548975646,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2318,
-      "step": 185620
-    },
-    {
-      "epoch": 3.015873015873016,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.2319,
-      "step": 185630
-    },
-    {
-      "epoch": 3.0160354827703855,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 185640
-    },
-    {
-      "epoch": 3.016197949667755,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 185650
-    },
-    {
-      "epoch": 3.0163604165651248,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2256,
-      "step": 185660
-    },
-    {
-      "epoch": 3.0165228834624944,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2255,
-      "step": 185670
-    },
-    {
-      "epoch": 3.016685350359864,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.238,
-      "step": 185680
-    },
-    {
-      "epoch": 3.0168478172572337,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2479,
-      "step": 185690
-    },
-    {
-      "epoch": 3.0170102841546034,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2331,
-      "step": 185700
-    },
-    {
-      "epoch": 3.017172751051973,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2132,
-      "step": 185710
-    },
-    {
-      "epoch": 3.0173352179493427,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 185720
-    },
-    {
-      "epoch": 3.0174976848467123,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 185730
-    },
-    {
-      "epoch": 3.017660151744082,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 185740
-    },
-    {
-      "epoch": 3.0178226186414516,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2131,
-      "step": 185750
-    },
-    {
-      "epoch": 3.0179850855388213,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1976,
-      "step": 185760
-    },
-    {
-      "epoch": 3.018147552436191,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 185770
-    },
-    {
-      "epoch": 3.0183100193335606,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 185780
-    },
-    {
-      "epoch": 3.0184724862309302,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 185790
-    },
-    {
-      "epoch": 3.0186349531283003,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 185800
-    },
-    {
-      "epoch": 3.01879742002567,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 185810
-    },
-    {
-      "epoch": 3.0189598869230396,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 185820
-    },
-    {
-      "epoch": 3.0191223538204093,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1801,
-      "step": 185830
-    },
-    {
-      "epoch": 3.019284820717779,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 185840
-    },
-    {
-      "epoch": 3.0194472876151486,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2334,
-      "step": 185850
-    },
-    {
-      "epoch": 3.0196097545125182,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2267,
-      "step": 185860
-    },
-    {
-      "epoch": 3.019772221409888,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2386,
-      "step": 185870
-    },
-    {
-      "epoch": 3.0199346883072575,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2082,
-      "step": 185880
-    },
-    {
-      "epoch": 3.020097155204627,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 185890
-    },
-    {
-      "epoch": 3.020259622101997,
-      "grad_norm": 2.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 185900
-    },
-    {
-      "epoch": 3.0204220889993665,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 185910
-    },
-    {
-      "epoch": 3.020584555896736,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 185920
-    },
-    {
-      "epoch": 3.020747022794106,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 185930
-    },
-    {
-      "epoch": 3.0209094896914754,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2126,
-      "step": 185940
-    },
-    {
-      "epoch": 3.021071956588845,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2011,
-      "step": 185950
-    },
-    {
-      "epoch": 3.0212344234862147,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 185960
-    },
-    {
-      "epoch": 3.0213968903835844,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 185970
-    },
-    {
-      "epoch": 3.021559357280954,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 185980
-    },
-    {
-      "epoch": 3.0217218241783237,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 185990
-    },
-    {
-      "epoch": 3.0218842910756933,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 186000
-    },
-    {
-      "epoch": 3.022046757973063,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 186010
-    },
-    {
-      "epoch": 3.0222092248704326,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 186020
-    },
-    {
-      "epoch": 3.0223716917678023,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 186030
-    },
-    {
-      "epoch": 3.022534158665172,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2114,
-      "step": 186040
-    },
-    {
-      "epoch": 3.0226966255625416,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 186050
-    },
-    {
-      "epoch": 3.0228590924599112,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.213,
-      "step": 186060
-    },
-    {
-      "epoch": 3.023021559357281,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 186070
-    },
-    {
-      "epoch": 3.0231840262546505,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2124,
-      "step": 186080
-    },
-    {
-      "epoch": 3.02334649315202,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2231,
-      "step": 186090
-    },
-    {
-      "epoch": 3.02350896004939,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2086,
-      "step": 186100
-    },
-    {
-      "epoch": 3.0236714269467595,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 186110
-    },
-    {
-      "epoch": 3.023833893844129,
-      "grad_norm": 3.046875,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 186120
-    },
-    {
-      "epoch": 3.023996360741499,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2169,
-      "step": 186130
-    },
-    {
-      "epoch": 3.0241588276388685,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2085,
-      "step": 186140
-    },
-    {
-      "epoch": 3.024321294536238,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 186150
-    },
-    {
-      "epoch": 3.0244837614336078,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 186160
-    },
-    {
-      "epoch": 3.0246462283309774,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 186170
-    },
-    {
-      "epoch": 3.024808695228347,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 186180
-    },
-    {
-      "epoch": 3.0249711621257167,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 186190
-    },
-    {
-      "epoch": 3.0251336290230864,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1817,
-      "step": 186200
-    },
-    {
-      "epoch": 3.025296095920456,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1803,
-      "step": 186210
-    },
-    {
-      "epoch": 3.0254585628178257,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 186220
-    },
-    {
-      "epoch": 3.0256210297151953,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 186230
-    },
-    {
-      "epoch": 3.0257834966125654,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 186240
-    },
-    {
-      "epoch": 3.025945963509935,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 186250
-    },
-    {
-      "epoch": 3.0261084304073047,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 186260
-    },
-    {
-      "epoch": 3.0262708973046744,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 186270
-    },
-    {
-      "epoch": 3.026433364202044,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2144,
-      "step": 186280
-    },
-    {
-      "epoch": 3.0265958310994137,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 186290
-    },
-    {
-      "epoch": 3.0267582979967833,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 186300
-    },
-    {
-      "epoch": 3.026920764894153,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 186310
-    },
-    {
-      "epoch": 3.0270832317915226,
-      "grad_norm": 3.171875,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 186320
-    },
-    {
-      "epoch": 3.0272456986888923,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 186330
-    },
-    {
-      "epoch": 3.027408165586262,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.179,
-      "step": 186340
-    },
-    {
-      "epoch": 3.0275706324836316,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1682,
-      "step": 186350
-    },
-    {
-      "epoch": 3.027733099381001,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 186360
-    },
-    {
-      "epoch": 3.027895566278371,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 186370
-    },
-    {
-      "epoch": 3.0280580331757405,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1681,
-      "step": 186380
-    },
-    {
-      "epoch": 3.02822050007311,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1896,
-      "step": 186390
-    },
-    {
-      "epoch": 3.02838296697048,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1731,
-      "step": 186400
-    },
-    {
-      "epoch": 3.0285454338678495,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1658,
-      "step": 186410
-    },
-    {
-      "epoch": 3.028707900765219,
-      "grad_norm": 2.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 186420
-    },
-    {
-      "epoch": 3.0288703676625888,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1642,
-      "step": 186430
-    },
-    {
-      "epoch": 3.0290328345599584,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 186440
-    },
-    {
-      "epoch": 3.029195301457328,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 186450
-    },
-    {
-      "epoch": 3.0293577683546977,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 186460
-    },
-    {
-      "epoch": 3.0295202352520674,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 186470
-    },
-    {
-      "epoch": 3.029682702149437,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1959,
-      "step": 186480
-    },
-    {
-      "epoch": 3.0298451690468067,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 186490
-    },
-    {
-      "epoch": 3.0300076359441763,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 186500
-    },
-    {
-      "epoch": 3.030170102841546,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2053,
-      "step": 186510
-    },
-    {
-      "epoch": 3.0303325697389156,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 186520
-    },
-    {
-      "epoch": 3.0304950366362853,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 186530
-    },
-    {
-      "epoch": 3.030657503533655,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 186540
-    },
-    {
-      "epoch": 3.0308199704310246,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 186550
-    },
-    {
-      "epoch": 3.0309824373283942,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 186560
-    },
-    {
-      "epoch": 3.031144904225764,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2133,
-      "step": 186570
-    },
-    {
-      "epoch": 3.0313073711231335,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 186580
-    },
-    {
-      "epoch": 3.031469838020503,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1698,
-      "step": 186590
-    },
-    {
-      "epoch": 3.031632304917873,
-      "grad_norm": 7.75,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 186600
-    },
-    {
-      "epoch": 3.0317947718152425,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 186610
-    },
-    {
-      "epoch": 3.031957238712612,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.176,
-      "step": 186620
-    },
-    {
-      "epoch": 3.032119705609982,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 186630
-    },
-    {
-      "epoch": 3.0322821725073514,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 186640
-    },
-    {
-      "epoch": 3.032444639404721,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 186650
-    },
-    {
-      "epoch": 3.0326071063020907,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 186660
-    },
-    {
-      "epoch": 3.0327695731994604,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 186670
-    },
-    {
-      "epoch": 3.0329320400968305,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 186680
-    },
-    {
-      "epoch": 3.0330945069942,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 186690
-    },
-    {
-      "epoch": 3.03325697389157,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 186700
-    },
-    {
-      "epoch": 3.0334194407889394,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 186710
-    },
-    {
-      "epoch": 3.033581907686309,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 186720
-    },
-    {
-      "epoch": 3.0337443745836787,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1978,
-      "step": 186730
-    },
-    {
-      "epoch": 3.0339068414810484,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 186740
-    },
-    {
-      "epoch": 3.034069308378418,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1662,
-      "step": 186750
-    },
-    {
-      "epoch": 3.0342317752757877,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 186760
-    },
-    {
-      "epoch": 3.0343942421731573,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 186770
-    },
-    {
-      "epoch": 3.034556709070527,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 186780
-    },
-    {
-      "epoch": 3.0347191759678966,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 186790
-    },
-    {
-      "epoch": 3.0348816428652663,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 186800
-    },
-    {
-      "epoch": 3.035044109762636,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 186810
-    },
-    {
-      "epoch": 3.0352065766600056,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 186820
-    },
-    {
-      "epoch": 3.0353690435573752,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 186830
-    },
-    {
-      "epoch": 3.035531510454745,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 186840
-    },
-    {
-      "epoch": 3.0356939773521145,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1707,
-      "step": 186850
-    },
-    {
-      "epoch": 3.035856444249484,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1743,
-      "step": 186860
-    },
-    {
-      "epoch": 3.036018911146854,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1885,
-      "step": 186870
-    },
-    {
-      "epoch": 3.0361813780442235,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 186880
-    },
-    {
-      "epoch": 3.036343844941593,
-      "grad_norm": 3.515625,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 186890
-    },
-    {
-      "epoch": 3.036506311838963,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 186900
-    },
-    {
-      "epoch": 3.0366687787363325,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 186910
-    },
-    {
-      "epoch": 3.036831245633702,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.186,
-      "step": 186920
-    },
-    {
-      "epoch": 3.0369937125310718,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 186930
-    },
-    {
-      "epoch": 3.0371561794284414,
-      "grad_norm": 2.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.1618,
-      "step": 186940
-    },
-    {
-      "epoch": 3.037318646325811,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1711,
-      "step": 186950
-    },
-    {
-      "epoch": 3.0374811132231807,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 186960
-    },
-    {
-      "epoch": 3.0376435801205504,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.17,
-      "step": 186970
-    },
-    {
-      "epoch": 3.03780604701792,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1648,
-      "step": 186980
-    },
-    {
-      "epoch": 3.0379685139152897,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1959,
-      "step": 186990
-    },
-    {
-      "epoch": 3.0381309808126593,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 187000
-    },
-    {
-      "epoch": 3.038293447710029,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 187010
-    },
-    {
-      "epoch": 3.0384559146073986,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1796,
-      "step": 187020
-    },
-    {
-      "epoch": 3.0386183815047683,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.216,
-      "step": 187030
-    },
-    {
-      "epoch": 3.038780848402138,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 187040
-    },
-    {
-      "epoch": 3.0389433152995076,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 187050
-    },
-    {
-      "epoch": 3.039105782196877,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 187060
-    },
-    {
-      "epoch": 3.039268249094247,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 187070
-    },
-    {
-      "epoch": 3.0394307159916165,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 187080
-    },
-    {
-      "epoch": 3.039593182888986,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 187090
-    },
-    {
-      "epoch": 3.039755649786356,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 187100
-    },
-    {
-      "epoch": 3.039918116683726,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 187110
-    },
-    {
-      "epoch": 3.0400805835810956,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 187120
-    },
-    {
-      "epoch": 3.040243050478465,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 187130
-    },
-    {
-      "epoch": 3.040405517375835,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1809,
-      "step": 187140
-    },
-    {
-      "epoch": 3.0405679842732045,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 187150
-    },
-    {
-      "epoch": 3.040730451170574,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1978,
-      "step": 187160
-    },
-    {
-      "epoch": 3.040892918067944,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 187170
-    },
-    {
-      "epoch": 3.0410553849653135,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2058,
-      "step": 187180
-    },
-    {
-      "epoch": 3.041217851862683,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 187190
-    },
-    {
-      "epoch": 3.0413803187600528,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.1939,
-      "step": 187200
-    },
-    {
-      "epoch": 3.0415427856574224,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.176,
-      "step": 187210
-    },
-    {
-      "epoch": 3.041705252554792,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 187220
-    },
-    {
-      "epoch": 3.0418677194521617,
-      "grad_norm": 3.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1671,
-      "step": 187230
-    },
-    {
-      "epoch": 3.0420301863495314,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 187240
-    },
-    {
-      "epoch": 3.042192653246901,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 187250
-    },
-    {
-      "epoch": 3.0423551201442707,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 187260
-    },
-    {
-      "epoch": 3.0425175870416403,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 187270
-    },
-    {
-      "epoch": 3.04268005393901,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2101,
-      "step": 187280
-    },
-    {
-      "epoch": 3.0428425208363796,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 187290
-    },
-    {
-      "epoch": 3.0430049877337493,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 187300
-    },
-    {
-      "epoch": 3.043167454631119,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 187310
-    },
-    {
-      "epoch": 3.0433299215284886,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 187320
-    },
-    {
-      "epoch": 3.0434923884258582,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1983,
-      "step": 187330
-    },
-    {
-      "epoch": 3.043654855323228,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 187340
-    },
-    {
-      "epoch": 3.0438173222205975,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 187350
-    },
-    {
-      "epoch": 3.043979789117967,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1758,
-      "step": 187360
-    },
-    {
-      "epoch": 3.044142256015337,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1785,
-      "step": 187370
-    },
-    {
-      "epoch": 3.0443047229127065,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1749,
-      "step": 187380
-    },
-    {
-      "epoch": 3.044467189810076,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1722,
-      "step": 187390
-    },
-    {
-      "epoch": 3.044629656707446,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 187400
-    },
-    {
-      "epoch": 3.0447921236048154,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 187410
-    },
-    {
-      "epoch": 3.044954590502185,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1545,
-      "step": 187420
-    },
-    {
-      "epoch": 3.0451170573995547,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.168,
-      "step": 187430
-    },
-    {
-      "epoch": 3.0452795242969244,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1807,
-      "step": 187440
-    },
-    {
-      "epoch": 3.045441991194294,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 187450
-    },
-    {
-      "epoch": 3.0456044580916637,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 187460
-    },
-    {
-      "epoch": 3.0457669249890333,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 187470
-    },
-    {
-      "epoch": 3.045929391886403,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.229,
-      "step": 187480
-    },
-    {
-      "epoch": 3.0460918587837726,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 187490
-    },
-    {
-      "epoch": 3.0462543256811423,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 187500
-    },
-    {
-      "epoch": 3.046416792578512,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 187510
-    },
-    {
-      "epoch": 3.0465792594758816,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 187520
-    },
-    {
-      "epoch": 3.0467417263732512,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 187530
-    },
-    {
-      "epoch": 3.046904193270621,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 187540
-    },
-    {
-      "epoch": 3.0470666601679905,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 187550
-    },
-    {
-      "epoch": 3.0472291270653606,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 187560
-    },
-    {
-      "epoch": 3.0473915939627303,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 187570
-    },
-    {
-      "epoch": 3.0475540608601,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1653,
-      "step": 187580
-    },
-    {
-      "epoch": 3.0477165277574696,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 187590
-    },
-    {
-      "epoch": 3.0478789946548392,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 187600
-    },
-    {
-      "epoch": 3.048041461552209,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 187610
-    },
-    {
-      "epoch": 3.0482039284495785,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 187620
-    },
-    {
-      "epoch": 3.048366395346948,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 187630
-    },
-    {
-      "epoch": 3.048528862244318,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 187640
-    },
-    {
-      "epoch": 3.0486913291416875,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 187650
-    },
-    {
-      "epoch": 3.048853796039057,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 187660
-    },
-    {
-      "epoch": 3.049016262936427,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 187670
-    },
-    {
-      "epoch": 3.0491787298337965,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1693,
-      "step": 187680
-    },
-    {
-      "epoch": 3.049341196731166,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1807,
-      "step": 187690
-    },
-    {
-      "epoch": 3.0495036636285358,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.1554,
-      "step": 187700
-    },
-    {
-      "epoch": 3.0496661305259054,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1675,
-      "step": 187710
-    },
-    {
-      "epoch": 3.049828597423275,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.181,
-      "step": 187720
-    },
-    {
-      "epoch": 3.0499910643206447,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.167,
-      "step": 187730
-    },
-    {
-      "epoch": 3.0501535312180144,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.159,
-      "step": 187740
-    },
-    {
-      "epoch": 3.050315998115384,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1645,
-      "step": 187750
-    },
-    {
-      "epoch": 3.0504784650127537,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.181,
-      "step": 187760
-    },
-    {
-      "epoch": 3.0506409319101233,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 187770
-    },
-    {
-      "epoch": 3.050803398807493,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 187780
-    },
-    {
-      "epoch": 3.0509658657048626,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1809,
-      "step": 187790
-    },
-    {
-      "epoch": 3.0511283326022323,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1667,
-      "step": 187800
-    },
-    {
-      "epoch": 3.051290799499602,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.191,
-      "step": 187810
-    },
-    {
-      "epoch": 3.0514532663969716,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 187820
-    },
-    {
-      "epoch": 3.051615733294341,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1927,
-      "step": 187830
-    },
-    {
-      "epoch": 3.051778200191711,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 187840
-    },
-    {
-      "epoch": 3.0519406670890805,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 187850
-    },
-    {
-      "epoch": 3.05210313398645,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1664,
-      "step": 187860
-    },
-    {
-      "epoch": 3.05226560088382,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1679,
-      "step": 187870
-    },
-    {
-      "epoch": 3.0524280677811895,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.165,
-      "step": 187880
-    },
-    {
-      "epoch": 3.052590534678559,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 187890
-    },
-    {
-      "epoch": 3.0527530015759288,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 187900
-    },
-    {
-      "epoch": 3.0529154684732984,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 187910
-    },
-    {
-      "epoch": 3.053077935370668,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1845,
-      "step": 187920
-    },
-    {
-      "epoch": 3.0532404022680377,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 187930
-    },
-    {
-      "epoch": 3.0534028691654074,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 187940
-    },
-    {
-      "epoch": 3.053565336062777,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 187950
-    },
-    {
-      "epoch": 3.0537278029601467,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 187960
-    },
-    {
-      "epoch": 3.0538902698575163,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 187970
-    },
-    {
-      "epoch": 3.054052736754886,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 187980
-    },
-    {
-      "epoch": 3.054215203652256,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 187990
-    },
-    {
-      "epoch": 3.0543776705496257,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 188000
-    },
-    {
-      "epoch": 3.0545401374469954,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 188010
-    },
-    {
-      "epoch": 3.054702604344365,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 188020
-    },
-    {
-      "epoch": 3.0548650712417347,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 188030
-    },
-    {
-      "epoch": 3.0550275381391043,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 188040
-    },
-    {
-      "epoch": 3.055190005036474,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 188050
-    },
-    {
-      "epoch": 3.0553524719338436,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 188060
-    },
-    {
-      "epoch": 3.0555149388312133,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 188070
-    },
-    {
-      "epoch": 3.055677405728583,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 188080
-    },
-    {
-      "epoch": 3.0558398726259526,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1885,
-      "step": 188090
-    },
-    {
-      "epoch": 3.0560023395233222,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 188100
-    },
-    {
-      "epoch": 3.056164806420692,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1666,
-      "step": 188110
-    },
-    {
-      "epoch": 3.0563272733180615,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1735,
-      "step": 188120
-    },
-    {
-      "epoch": 3.056489740215431,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 188130
-    },
-    {
-      "epoch": 3.056652207112801,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 188140
-    },
-    {
-      "epoch": 3.0568146740101705,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 188150
-    },
-    {
-      "epoch": 3.05697714090754,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 188160
-    },
-    {
-      "epoch": 3.05713960780491,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 188170
-    },
-    {
-      "epoch": 3.0573020747022794,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1633,
-      "step": 188180
-    },
-    {
-      "epoch": 3.057464541599649,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1696,
-      "step": 188190
-    },
-    {
-      "epoch": 3.0576270084970187,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 188200
-    },
-    {
-      "epoch": 3.0577894753943884,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.179,
-      "step": 188210
-    },
-    {
-      "epoch": 3.057951942291758,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1861,
-      "step": 188220
-    },
-    {
-      "epoch": 3.0581144091891277,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1638,
-      "step": 188230
-    },
-    {
-      "epoch": 3.0582768760864973,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1437,
-      "step": 188240
-    },
-    {
-      "epoch": 3.058439342983867,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1671,
-      "step": 188250
-    },
-    {
-      "epoch": 3.0586018098812366,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.1733,
-      "step": 188260
-    },
-    {
-      "epoch": 3.0587642767786063,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.1665,
-      "step": 188270
-    },
-    {
-      "epoch": 3.058926743675976,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1658,
-      "step": 188280
-    },
-    {
-      "epoch": 3.0590892105733456,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1831,
-      "step": 188290
-    },
-    {
-      "epoch": 3.0592516774707152,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 188300
-    },
-    {
-      "epoch": 3.059414144368085,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1823,
-      "step": 188310
-    },
-    {
-      "epoch": 3.0595766112654545,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 188320
-    },
-    {
-      "epoch": 3.059739078162824,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 188330
-    },
-    {
-      "epoch": 3.059901545060194,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.175,
-      "step": 188340
-    },
-    {
-      "epoch": 3.0600640119575635,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 188350
-    },
-    {
-      "epoch": 3.060226478854933,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.152,
-      "step": 188360
-    },
-    {
-      "epoch": 3.060388945752303,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1668,
-      "step": 188370
-    },
-    {
-      "epoch": 3.0605514126496725,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 188380
-    },
-    {
-      "epoch": 3.060713879547042,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1667,
-      "step": 188390
-    },
-    {
-      "epoch": 3.0608763464444118,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1763,
-      "step": 188400
-    },
-    {
-      "epoch": 3.0610388133417814,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 188410
-    },
-    {
-      "epoch": 3.061201280239151,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 188420
-    },
-    {
-      "epoch": 3.061363747136521,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1691,
-      "step": 188430
-    },
-    {
-      "epoch": 3.061526214033891,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 188440
-    },
-    {
-      "epoch": 3.0616886809312605,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1626,
-      "step": 188450
-    },
-    {
-      "epoch": 3.06185114782863,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1735,
-      "step": 188460
-    },
-    {
-      "epoch": 3.0620136147259998,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 188470
-    },
-    {
-      "epoch": 3.0621760816233694,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1636,
-      "step": 188480
-    },
-    {
-      "epoch": 3.062338548520739,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 188490
-    },
-    {
-      "epoch": 3.0625010154181087,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 188500
-    },
-    {
-      "epoch": 3.0626634823154784,
-      "grad_norm": 3.328125,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 188510
-    },
-    {
-      "epoch": 3.062825949212848,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1572,
-      "step": 188520
-    },
-    {
-      "epoch": 3.0629884161102177,
-      "grad_norm": 3.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1646,
-      "step": 188530
-    },
-    {
-      "epoch": 3.0631508830075873,
-      "grad_norm": 2.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.1437,
-      "step": 188540
-    },
-    {
-      "epoch": 3.063313349904957,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.1505,
-      "step": 188550
-    },
-    {
-      "epoch": 3.0634758168023266,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1426,
-      "step": 188560
-    },
-    {
-      "epoch": 3.0636382836996963,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1589,
-      "step": 188570
-    },
-    {
-      "epoch": 3.063800750597066,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 188580
-    },
-    {
-      "epoch": 3.0639632174944356,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 188590
-    },
-    {
-      "epoch": 3.064125684391805,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 188600
-    },
-    {
-      "epoch": 3.064288151289175,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 188610
-    },
-    {
-      "epoch": 3.0644506181865445,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 188620
-    },
-    {
-      "epoch": 3.064613085083914,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1689,
-      "step": 188630
-    },
-    {
-      "epoch": 3.064775551981284,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 188640
-    },
-    {
-      "epoch": 3.0649380188786535,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 188650
-    },
-    {
-      "epoch": 3.065100485776023,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1927,
-      "step": 188660
-    },
-    {
-      "epoch": 3.0652629526733928,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 188670
-    },
-    {
-      "epoch": 3.0654254195707624,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1818,
-      "step": 188680
-    },
-    {
-      "epoch": 3.065587886468132,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 188690
-    },
-    {
-      "epoch": 3.0657503533655017,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 188700
-    },
-    {
-      "epoch": 3.0659128202628714,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1707,
-      "step": 188710
-    },
-    {
-      "epoch": 3.066075287160241,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 188720
-    },
-    {
-      "epoch": 3.0662377540576107,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 188730
-    },
-    {
-      "epoch": 3.0664002209549803,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 188740
-    },
-    {
-      "epoch": 3.06656268785235,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 188750
-    },
-    {
-      "epoch": 3.0667251547497196,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1839,
-      "step": 188760
-    },
-    {
-      "epoch": 3.0668876216470893,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1757,
-      "step": 188770
-    },
-    {
-      "epoch": 3.067050088544459,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1794,
-      "step": 188780
-    },
-    {
-      "epoch": 3.0672125554418286,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2259,
-      "step": 188790
-    },
-    {
-      "epoch": 3.0673750223391982,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1939,
-      "step": 188800
-    },
-    {
-      "epoch": 3.067537489236568,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1719,
-      "step": 188810
-    },
-    {
-      "epoch": 3.0676999561339375,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1681,
-      "step": 188820
-    },
-    {
-      "epoch": 3.067862423031307,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1839,
-      "step": 188830
-    },
-    {
-      "epoch": 3.068024889928677,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 188840
-    },
-    {
-      "epoch": 3.0681873568260465,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 188850
-    },
-    {
-      "epoch": 3.068349823723416,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 188860
-    },
-    {
-      "epoch": 3.0685122906207862,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.178,
-      "step": 188870
-    },
-    {
-      "epoch": 3.068674757518156,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 188880
-    },
-    {
-      "epoch": 3.0688372244155255,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 188890
-    },
-    {
-      "epoch": 3.068999691312895,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1567,
-      "step": 188900
-    },
-    {
-      "epoch": 3.069162158210265,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1691,
-      "step": 188910
-    },
-    {
-      "epoch": 3.0693246251076345,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1807,
-      "step": 188920
-    },
-    {
-      "epoch": 3.069487092005004,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1734,
-      "step": 188930
-    },
-    {
-      "epoch": 3.069649558902374,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 188940
-    },
-    {
-      "epoch": 3.0698120257997434,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 188950
-    },
-    {
-      "epoch": 3.069974492697113,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1527,
-      "step": 188960
-    },
-    {
-      "epoch": 3.0701369595944827,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 188970
-    },
-    {
-      "epoch": 3.0702994264918524,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 188980
-    },
-    {
-      "epoch": 3.070461893389222,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 188990
-    },
-    {
-      "epoch": 3.0706243602865917,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 189000
-    },
-    {
-      "epoch": 3.0707868271839613,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 189010
-    },
-    {
-      "epoch": 3.070949294081331,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1795,
-      "step": 189020
-    },
-    {
-      "epoch": 3.0711117609787006,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 189030
-    },
-    {
-      "epoch": 3.0712742278760703,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1712,
-      "step": 189040
-    },
-    {
-      "epoch": 3.07143669477344,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 189050
-    },
-    {
-      "epoch": 3.0715991616708096,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1717,
-      "step": 189060
-    },
-    {
-      "epoch": 3.0717616285681792,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 189070
-    },
-    {
-      "epoch": 3.071924095465549,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 189080
-    },
-    {
-      "epoch": 3.0720865623629185,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 189090
-    },
-    {
-      "epoch": 3.072249029260288,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 189100
-    },
-    {
-      "epoch": 3.072411496157658,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 189110
-    },
-    {
-      "epoch": 3.0725739630550275,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1739,
-      "step": 189120
-    },
-    {
-      "epoch": 3.072736429952397,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1586,
-      "step": 189130
-    },
-    {
-      "epoch": 3.072898896849767,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1692,
-      "step": 189140
-    },
-    {
-      "epoch": 3.0730613637471365,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 189150
-    },
-    {
-      "epoch": 3.073223830644506,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 189160
-    },
-    {
-      "epoch": 3.0733862975418758,
-      "grad_norm": 3.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1845,
-      "step": 189170
-    },
-    {
-      "epoch": 3.0735487644392454,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1631,
-      "step": 189180
-    },
-    {
-      "epoch": 3.073711231336615,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 189190
-    },
-    {
-      "epoch": 3.0738736982339847,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1719,
-      "step": 189200
-    },
-    {
-      "epoch": 3.0740361651313544,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 189210
-    },
-    {
-      "epoch": 3.074198632028724,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 189220
-    },
-    {
-      "epoch": 3.0743610989260937,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.1705,
-      "step": 189230
-    },
-    {
-      "epoch": 3.0745235658234633,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1761,
-      "step": 189240
-    },
-    {
-      "epoch": 3.074686032720833,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 189250
-    },
-    {
-      "epoch": 3.0748484996182026,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 189260
-    },
-    {
-      "epoch": 3.0750109665155723,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 189270
-    },
-    {
-      "epoch": 3.075173433412942,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1664,
-      "step": 189280
-    },
-    {
-      "epoch": 3.0753359003103116,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1602,
-      "step": 189290
-    },
-    {
-      "epoch": 3.0754983672076817,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1663,
-      "step": 189300
-    },
-    {
-      "epoch": 3.075660834105051,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 189310
-    },
-    {
-      "epoch": 3.075823301002421,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1606,
-      "step": 189320
-    },
-    {
-      "epoch": 3.0759857678997906,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1698,
-      "step": 189330
-    },
-    {
-      "epoch": 3.0761482347971603,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 189340
-    },
-    {
-      "epoch": 3.07631070169453,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1706,
-      "step": 189350
-    },
-    {
-      "epoch": 3.0764731685918996,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1503,
-      "step": 189360
-    },
-    {
-      "epoch": 3.076635635489269,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1551,
-      "step": 189370
-    },
-    {
-      "epoch": 3.076798102386639,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 189380
-    },
-    {
-      "epoch": 3.0769605692840085,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1632,
-      "step": 189390
-    },
-    {
-      "epoch": 3.077123036181378,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 189400
-    },
-    {
-      "epoch": 3.077285503078748,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 189410
-    },
-    {
-      "epoch": 3.0774479699761175,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1767,
-      "step": 189420
-    },
-    {
-      "epoch": 3.077610436873487,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 189430
-    },
-    {
-      "epoch": 3.0777729037708568,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 189440
-    },
-    {
-      "epoch": 3.0779353706682264,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1764,
-      "step": 189450
-    },
-    {
-      "epoch": 3.078097837565596,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2044,
-      "step": 189460
-    },
-    {
-      "epoch": 3.0782603044629657,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 189470
-    },
-    {
-      "epoch": 3.0784227713603354,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2176,
-      "step": 189480
-    },
-    {
-      "epoch": 3.078585238257705,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 189490
-    },
-    {
-      "epoch": 3.0787477051550747,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 189500
-    },
-    {
-      "epoch": 3.0789101720524443,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1786,
-      "step": 189510
-    },
-    {
-      "epoch": 3.079072638949814,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2159,
-      "step": 189520
-    },
-    {
-      "epoch": 3.0792351058471836,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 189530
-    },
-    {
-      "epoch": 3.0793975727445533,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 189540
-    },
-    {
-      "epoch": 3.079560039641923,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 189550
-    },
-    {
-      "epoch": 3.0797225065392926,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 189560
-    },
-    {
-      "epoch": 3.0798849734366622,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 189570
-    },
-    {
-      "epoch": 3.080047440334032,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.188,
-      "step": 189580
-    },
-    {
-      "epoch": 3.0802099072314015,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.1732,
-      "step": 189590
-    },
-    {
-      "epoch": 3.080372374128771,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 189600
-    },
-    {
-      "epoch": 3.080534841026141,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 189610
-    },
-    {
-      "epoch": 3.0806973079235105,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 189620
-    },
-    {
-      "epoch": 3.08085977482088,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 189630
-    },
-    {
-      "epoch": 3.08102224171825,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1649,
-      "step": 189640
-    },
-    {
-      "epoch": 3.0811847086156194,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.173,
-      "step": 189650
-    },
-    {
-      "epoch": 3.081347175512989,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.1648,
-      "step": 189660
-    },
-    {
-      "epoch": 3.0815096424103587,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1685,
-      "step": 189670
-    },
-    {
-      "epoch": 3.0816721093077284,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 189680
-    },
-    {
-      "epoch": 3.081834576205098,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1748,
-      "step": 189690
-    },
-    {
-      "epoch": 3.0819970431024677,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1577,
-      "step": 189700
-    },
-    {
-      "epoch": 3.0821595099998373,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.15,
-      "step": 189710
-    },
-    {
-      "epoch": 3.082321976897207,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1467,
-      "step": 189720
-    },
-    {
-      "epoch": 3.0824844437945766,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.159,
-      "step": 189730
-    },
-    {
-      "epoch": 3.0826469106919463,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1634,
-      "step": 189740
-    },
-    {
-      "epoch": 3.0828093775893164,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1711,
-      "step": 189750
-    },
-    {
-      "epoch": 3.082971844486686,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 189760
-    },
-    {
-      "epoch": 3.0831343113840557,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1823,
-      "step": 189770
-    },
-    {
-      "epoch": 3.0832967782814253,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1735,
-      "step": 189780
-    },
-    {
-      "epoch": 3.083459245178795,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 189790
-    },
-    {
-      "epoch": 3.0836217120761646,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1693,
-      "step": 189800
-    },
-    {
-      "epoch": 3.0837841789735343,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1598,
-      "step": 189810
-    },
-    {
-      "epoch": 3.083946645870904,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1697,
-      "step": 189820
-    },
-    {
-      "epoch": 3.0841091127682736,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1778,
-      "step": 189830
-    },
-    {
-      "epoch": 3.0842715796656432,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 189840
-    },
-    {
-      "epoch": 3.084434046563013,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 189850
-    },
-    {
-      "epoch": 3.0845965134603825,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1605,
-      "step": 189860
-    },
-    {
-      "epoch": 3.084758980357752,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1761,
-      "step": 189870
-    },
-    {
-      "epoch": 3.084921447255122,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1786,
-      "step": 189880
-    },
-    {
-      "epoch": 3.0850839141524915,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 189890
-    },
-    {
-      "epoch": 3.085246381049861,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 189900
-    },
-    {
-      "epoch": 3.085408847947231,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 189910
-    },
-    {
-      "epoch": 3.0855713148446005,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1758,
-      "step": 189920
-    },
-    {
-      "epoch": 3.08573378174197,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 189930
-    },
-    {
-      "epoch": 3.0858962486393398,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 189940
-    },
-    {
-      "epoch": 3.0860587155367094,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.175,
-      "step": 189950
-    },
-    {
-      "epoch": 3.086221182434079,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.1687,
-      "step": 189960
-    },
-    {
-      "epoch": 3.0863836493314487,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 189970
-    },
-    {
-      "epoch": 3.0865461162288184,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 189980
-    },
-    {
-      "epoch": 3.086708583126188,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1782,
-      "step": 189990
-    },
-    {
-      "epoch": 3.0868710500235577,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1894,
-      "step": 190000
-    },
-    {
-      "epoch": 3.0870335169209273,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 190010
-    },
-    {
-      "epoch": 3.087195983818297,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 190020
-    },
-    {
-      "epoch": 3.0873584507156666,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2164,
-      "step": 190030
-    },
-    {
-      "epoch": 3.0875209176130363,
-      "grad_norm": 3.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 190040
-    },
-    {
-      "epoch": 3.087683384510406,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1748,
-      "step": 190050
-    },
-    {
-      "epoch": 3.0878458514077756,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1681,
-      "step": 190060
-    },
-    {
-      "epoch": 3.088008318305145,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 190070
-    },
-    {
-      "epoch": 3.088170785202515,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 190080
-    },
-    {
-      "epoch": 3.0883332520998845,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 190090
-    },
-    {
-      "epoch": 3.088495718997254,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.157,
-      "step": 190100
-    },
-    {
-      "epoch": 3.088658185894624,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1904,
-      "step": 190110
-    },
-    {
-      "epoch": 3.0888206527919935,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 190120
-    },
-    {
-      "epoch": 3.088983119689363,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2067,
-      "step": 190130
-    },
-    {
-      "epoch": 3.0891455865867328,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 190140
-    },
-    {
-      "epoch": 3.0893080534841024,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 190150
-    },
-    {
-      "epoch": 3.089470520381472,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 190160
-    },
-    {
-      "epoch": 3.0896329872788417,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 190170
-    },
-    {
-      "epoch": 3.089795454176212,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 190180
-    },
-    {
-      "epoch": 3.0899579210735815,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.1602,
-      "step": 190190
-    },
-    {
-      "epoch": 3.090120387970951,
-      "grad_norm": 3.265625,
-      "learning_rate": 5e-05,
-      "loss": 0.1524,
-      "step": 190200
-    },
-    {
-      "epoch": 3.0902828548683208,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1566,
-      "step": 190210
-    },
-    {
-      "epoch": 3.0904453217656904,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1425,
-      "step": 190220
-    },
-    {
-      "epoch": 3.09060778866306,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1593,
-      "step": 190230
-    },
-    {
-      "epoch": 3.0907702555604297,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2008,
-      "step": 190240
-    },
-    {
-      "epoch": 3.0909327224577994,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1823,
-      "step": 190250
-    },
-    {
-      "epoch": 3.091095189355169,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 190260
-    },
-    {
-      "epoch": 3.0912576562525387,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 190270
-    },
-    {
-      "epoch": 3.0914201231499083,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 190280
-    },
-    {
-      "epoch": 3.091582590047278,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 190290
-    },
-    {
-      "epoch": 3.0917450569446476,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 190300
-    },
-    {
-      "epoch": 3.0919075238420173,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2084,
-      "step": 190310
-    },
-    {
-      "epoch": 3.092069990739387,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 190320
-    },
-    {
-      "epoch": 3.0922324576367566,
-      "grad_norm": 3.609375,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 190330
-    },
-    {
-      "epoch": 3.0923949245341262,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 190340
-    },
-    {
-      "epoch": 3.092557391431496,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1702,
-      "step": 190350
-    },
-    {
-      "epoch": 3.0927198583288655,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1544,
-      "step": 190360
-    },
-    {
-      "epoch": 3.092882325226235,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1425,
-      "step": 190370
-    },
-    {
-      "epoch": 3.093044792123605,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1558,
-      "step": 190380
-    },
-    {
-      "epoch": 3.0932072590209745,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1683,
-      "step": 190390
-    },
-    {
-      "epoch": 3.093369725918344,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 190400
-    },
-    {
-      "epoch": 3.093532192815714,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1685,
-      "step": 190410
-    },
-    {
-      "epoch": 3.0936946597130834,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.163,
-      "step": 190420
-    },
-    {
-      "epoch": 3.093857126610453,
-      "grad_norm": 3.296875,
-      "learning_rate": 5e-05,
-      "loss": 0.1699,
-      "step": 190430
-    },
-    {
-      "epoch": 3.0940195935078227,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 190440
-    },
-    {
-      "epoch": 3.0941820604051924,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1814,
-      "step": 190450
-    },
-    {
-      "epoch": 3.094344527302562,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.184,
-      "step": 190460
-    },
-    {
-      "epoch": 3.0945069941999317,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1927,
-      "step": 190470
-    },
-    {
-      "epoch": 3.0946694610973013,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 190480
-    },
-    {
-      "epoch": 3.094831927994671,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.191,
-      "step": 190490
-    },
-    {
-      "epoch": 3.0949943948920406,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 190500
-    },
-    {
-      "epoch": 3.0951568617894103,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 190510
-    },
-    {
-      "epoch": 3.09531932868678,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1782,
-      "step": 190520
-    },
-    {
-      "epoch": 3.0954817955841496,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 190530
-    },
-    {
-      "epoch": 3.0956442624815192,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 190540
-    },
-    {
-      "epoch": 3.095806729378889,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 190550
-    },
-    {
-      "epoch": 3.0959691962762586,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 190560
-    },
-    {
-      "epoch": 3.096131663173628,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1795,
-      "step": 190570
-    },
-    {
-      "epoch": 3.096294130070998,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1889,
-      "step": 190580
-    },
-    {
-      "epoch": 3.0964565969683675,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 190590
-    },
-    {
-      "epoch": 3.096619063865737,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 190600
-    },
-    {
-      "epoch": 3.096781530763107,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1681,
-      "step": 190610
-    },
-    {
-      "epoch": 3.0969439976604765,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1653,
-      "step": 190620
-    },
-    {
-      "epoch": 3.0971064645578466,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 190630
-    },
-    {
-      "epoch": 3.097268931455216,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1757,
-      "step": 190640
-    },
-    {
-      "epoch": 3.097431398352586,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 190650
-    },
-    {
-      "epoch": 3.0975938652499555,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 190660
-    },
-    {
-      "epoch": 3.097756332147325,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1681,
-      "step": 190670
-    },
-    {
-      "epoch": 3.097918799044695,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 190680
-    },
-    {
-      "epoch": 3.0980812659420645,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1717,
-      "step": 190690
-    },
-    {
-      "epoch": 3.098243732839434,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1954,
-      "step": 190700
-    },
-    {
-      "epoch": 3.0984061997368038,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 190710
-    },
-    {
-      "epoch": 3.0985686666341734,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 190720
-    },
-    {
-      "epoch": 3.098731133531543,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 190730
-    },
-    {
-      "epoch": 3.0988936004289127,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 190740
-    },
-    {
-      "epoch": 3.0990560673262824,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 190750
-    },
-    {
-      "epoch": 3.099218534223652,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.18,
-      "step": 190760
-    },
-    {
-      "epoch": 3.0993810011210217,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 190770
-    },
-    {
-      "epoch": 3.0995434680183913,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1828,
-      "step": 190780
-    },
-    {
-      "epoch": 3.099705934915761,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1603,
-      "step": 190790
-    },
-    {
-      "epoch": 3.0998684018131306,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 190800
-    },
-    {
-      "epoch": 3.1000308687105003,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 190810
-    },
-    {
-      "epoch": 3.10019333560787,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 190820
-    },
-    {
-      "epoch": 3.1003558025052396,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 190830
-    },
-    {
-      "epoch": 3.100518269402609,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2133,
-      "step": 190840
-    },
-    {
-      "epoch": 3.100680736299979,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 190850
-    },
-    {
-      "epoch": 3.1008432031973485,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1839,
-      "step": 190860
-    },
-    {
-      "epoch": 3.101005670094718,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 190870
-    },
-    {
-      "epoch": 3.101168136992088,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1683,
-      "step": 190880
-    },
-    {
-      "epoch": 3.1013306038894575,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 190890
-    },
-    {
-      "epoch": 3.101493070786827,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 190900
-    },
-    {
-      "epoch": 3.1016555376841968,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 190910
-    },
-    {
-      "epoch": 3.1018180045815664,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1627,
-      "step": 190920
-    },
-    {
-      "epoch": 3.101980471478936,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 190930
-    },
-    {
-      "epoch": 3.1021429383763057,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1774,
-      "step": 190940
-    },
-    {
-      "epoch": 3.1023054052736754,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.17,
-      "step": 190950
-    },
-    {
-      "epoch": 3.102467872171045,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 190960
-    },
-    {
-      "epoch": 3.1026303390684147,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 190970
-    },
-    {
-      "epoch": 3.1027928059657843,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 190980
-    },
-    {
-      "epoch": 3.102955272863154,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 190990
-    },
-    {
-      "epoch": 3.1031177397605236,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 191000
-    },
-    {
-      "epoch": 3.1032802066578933,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1715,
-      "step": 191010
-    },
-    {
-      "epoch": 3.103442673555263,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2072,
-      "step": 191020
-    },
-    {
-      "epoch": 3.1036051404526326,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 191030
-    },
-    {
-      "epoch": 3.1037676073500022,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 191040
-    },
-    {
-      "epoch": 3.103930074247372,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1684,
-      "step": 191050
-    },
-    {
-      "epoch": 3.104092541144742,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 191060
-    },
-    {
-      "epoch": 3.1042550080421116,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1809,
-      "step": 191070
-    },
-    {
-      "epoch": 3.1044174749394813,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 191080
-    },
-    {
-      "epoch": 3.104579941836851,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1704,
-      "step": 191090
-    },
-    {
-      "epoch": 3.1047424087342206,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1904,
-      "step": 191100
-    },
-    {
-      "epoch": 3.1049048756315902,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1761,
-      "step": 191110
-    },
-    {
-      "epoch": 3.10506734252896,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1731,
-      "step": 191120
-    },
-    {
-      "epoch": 3.1052298094263295,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 191130
-    },
-    {
-      "epoch": 3.105392276323699,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 191140
-    },
-    {
-      "epoch": 3.105554743221069,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 191150
-    },
-    {
-      "epoch": 3.1057172101184385,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 191160
-    },
-    {
-      "epoch": 3.105879677015808,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 191170
-    },
-    {
-      "epoch": 3.106042143913178,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 191180
-    },
-    {
-      "epoch": 3.1062046108105474,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1722,
-      "step": 191190
-    },
-    {
-      "epoch": 3.106367077707917,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1561,
-      "step": 191200
-    },
-    {
-      "epoch": 3.1065295446052867,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 191210
-    },
-    {
-      "epoch": 3.1066920115026564,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.183,
-      "step": 191220
-    },
-    {
-      "epoch": 3.106854478400026,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 191230
-    },
-    {
-      "epoch": 3.1070169452973957,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 191240
-    },
-    {
-      "epoch": 3.1071794121947653,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1757,
-      "step": 191250
-    },
-    {
-      "epoch": 3.107341879092135,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 191260
-    },
-    {
-      "epoch": 3.1075043459895046,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1728,
-      "step": 191270
-    },
-    {
-      "epoch": 3.1076668128868743,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 191280
-    },
-    {
-      "epoch": 3.107829279784244,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 191290
-    },
-    {
-      "epoch": 3.1079917466816136,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 191300
-    },
-    {
-      "epoch": 3.1081542135789832,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1679,
-      "step": 191310
-    },
-    {
-      "epoch": 3.108316680476353,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1786,
-      "step": 191320
-    },
-    {
-      "epoch": 3.1084791473737226,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 191330
-    },
-    {
-      "epoch": 3.108641614271092,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1508,
-      "step": 191340
-    },
-    {
-      "epoch": 3.108804081168462,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1731,
-      "step": 191350
-    },
-    {
-      "epoch": 3.1089665480658315,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 191360
-    },
-    {
-      "epoch": 3.109129014963201,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1713,
-      "step": 191370
-    },
-    {
-      "epoch": 3.109291481860571,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 191380
-    },
-    {
-      "epoch": 3.1094539487579405,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 191390
-    },
-    {
-      "epoch": 3.10961641565531,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1737,
-      "step": 191400
-    },
-    {
-      "epoch": 3.1097788825526798,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 191410
-    },
-    {
-      "epoch": 3.1099413494500494,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 191420
-    },
-    {
-      "epoch": 3.110103816347419,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 191430
-    },
-    {
-      "epoch": 3.1102662832447887,
-      "grad_norm": 3.140625,
-      "learning_rate": 5e-05,
-      "loss": 0.1678,
-      "step": 191440
-    },
-    {
-      "epoch": 3.1104287501421584,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 191450
-    },
-    {
-      "epoch": 3.110591217039528,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 191460
-    },
-    {
-      "epoch": 3.1107536839368977,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.187,
-      "step": 191470
-    },
-    {
-      "epoch": 3.1109161508342673,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1728,
-      "step": 191480
-    },
-    {
-      "epoch": 3.111078617731637,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 191490
-    },
-    {
-      "epoch": 3.1112410846290066,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.199,
-      "step": 191500
-    },
-    {
-      "epoch": 3.1114035515263767,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 191510
-    },
-    {
-      "epoch": 3.1115660184237464,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1669,
-      "step": 191520
-    },
-    {
-      "epoch": 3.111728485321116,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 191530
-    },
-    {
-      "epoch": 3.1118909522184857,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 191540
-    },
-    {
-      "epoch": 3.1120534191158553,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 191550
-    },
-    {
-      "epoch": 3.112215886013225,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 191560
-    },
-    {
-      "epoch": 3.1123783529105946,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 191570
-    },
-    {
-      "epoch": 3.1125408198079643,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 191580
-    },
-    {
-      "epoch": 3.112703286705334,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 191590
-    },
-    {
-      "epoch": 3.1128657536027036,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 191600
-    },
-    {
-      "epoch": 3.113028220500073,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 191610
-    },
-    {
-      "epoch": 3.113190687397443,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 191620
-    },
-    {
-      "epoch": 3.1133531542948125,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 191630
-    },
-    {
-      "epoch": 3.113515621192182,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.188,
-      "step": 191640
-    },
-    {
-      "epoch": 3.113678088089552,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 191650
-    },
-    {
-      "epoch": 3.1138405549869215,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1927,
-      "step": 191660
-    },
-    {
-      "epoch": 3.114003021884291,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1697,
-      "step": 191670
-    },
-    {
-      "epoch": 3.1141654887816608,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 191680
-    },
-    {
-      "epoch": 3.1143279556790304,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 191690
-    },
-    {
-      "epoch": 3.1144904225764,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 191700
-    },
-    {
-      "epoch": 3.1146528894737697,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1904,
-      "step": 191710
-    },
-    {
-      "epoch": 3.1148153563711394,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1778,
-      "step": 191720
-    },
-    {
-      "epoch": 3.114977823268509,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 191730
-    },
-    {
-      "epoch": 3.1151402901658787,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 191740
-    },
-    {
-      "epoch": 3.1153027570632483,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 191750
-    },
-    {
-      "epoch": 3.115465223960618,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1735,
-      "step": 191760
-    },
-    {
-      "epoch": 3.1156276908579876,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1633,
-      "step": 191770
-    },
-    {
-      "epoch": 3.1157901577553573,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1716,
-      "step": 191780
-    },
-    {
-      "epoch": 3.115952624652727,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1579,
-      "step": 191790
-    },
-    {
-      "epoch": 3.1161150915500966,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1706,
-      "step": 191800
-    },
-    {
-      "epoch": 3.1162775584474662,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1642,
-      "step": 191810
-    },
-    {
-      "epoch": 3.116440025344836,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1634,
-      "step": 191820
-    },
-    {
-      "epoch": 3.1166024922422055,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1719,
-      "step": 191830
-    },
-    {
-      "epoch": 3.116764959139575,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1675,
-      "step": 191840
-    },
-    {
-      "epoch": 3.116927426036945,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1724,
-      "step": 191850
-    },
-    {
-      "epoch": 3.1170898929343145,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 191860
-    },
-    {
-      "epoch": 3.117252359831684,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 191870
-    },
-    {
-      "epoch": 3.117414826729054,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 191880
-    },
-    {
-      "epoch": 3.1175772936264234,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1767,
-      "step": 191890
-    },
-    {
-      "epoch": 3.117739760523793,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 191900
-    },
-    {
-      "epoch": 3.1179022274211627,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 191910
-    },
-    {
-      "epoch": 3.1180646943185324,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 191920
-    },
-    {
-      "epoch": 3.118227161215902,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 191930
-    },
-    {
-      "epoch": 3.118389628113272,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 191940
-    },
-    {
-      "epoch": 3.118552095010642,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 191950
-    },
-    {
-      "epoch": 3.1187145619080114,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 191960
-    },
-    {
-      "epoch": 3.118877028805381,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 191970
-    },
-    {
-      "epoch": 3.1190394957027507,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 191980
-    },
-    {
-      "epoch": 3.1192019626001204,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 191990
-    },
-    {
-      "epoch": 3.11936442949749,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 192000
-    },
-    {
-      "epoch": 3.1195268963948597,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1734,
-      "step": 192010
-    },
-    {
-      "epoch": 3.1196893632922293,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1889,
-      "step": 192020
-    },
-    {
-      "epoch": 3.119851830189599,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1817,
-      "step": 192030
-    },
-    {
-      "epoch": 3.1200142970869686,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 192040
-    },
-    {
-      "epoch": 3.1201767639843383,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1706,
-      "step": 192050
-    },
-    {
-      "epoch": 3.120339230881708,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.152,
-      "step": 192060
-    },
-    {
-      "epoch": 3.1205016977790776,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1466,
-      "step": 192070
-    },
-    {
-      "epoch": 3.1206641646764472,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1487,
-      "step": 192080
-    },
-    {
-      "epoch": 3.120826631573817,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1446,
-      "step": 192090
-    },
-    {
-      "epoch": 3.1209890984711866,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1692,
-      "step": 192100
-    },
-    {
-      "epoch": 3.121151565368556,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 192110
-    },
-    {
-      "epoch": 3.121314032265926,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1558,
-      "step": 192120
-    },
-    {
-      "epoch": 3.1214764991632955,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1906,
-      "step": 192130
-    },
-    {
-      "epoch": 3.121638966060665,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 192140
-    },
-    {
-      "epoch": 3.121801432958035,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 192150
-    },
-    {
-      "epoch": 3.1219638998554045,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 192160
-    },
-    {
-      "epoch": 3.122126366752774,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 192170
-    },
-    {
-      "epoch": 3.1222888336501438,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1713,
-      "step": 192180
-    },
-    {
-      "epoch": 3.1224513005475134,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 192190
-    },
-    {
-      "epoch": 3.122613767444883,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 192200
-    },
-    {
-      "epoch": 3.1227762343422527,
-      "grad_norm": 8.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 192210
-    },
-    {
-      "epoch": 3.1229387012396224,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1764,
-      "step": 192220
-    },
-    {
-      "epoch": 3.123101168136992,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.186,
-      "step": 192230
-    },
-    {
-      "epoch": 3.1232636350343617,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1769,
-      "step": 192240
-    },
-    {
-      "epoch": 3.1234261019317313,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1581,
-      "step": 192250
-    },
-    {
-      "epoch": 3.123588568829101,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 192260
-    },
-    {
-      "epoch": 3.1237510357264706,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1781,
-      "step": 192270
-    },
-    {
-      "epoch": 3.1239135026238403,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 192280
-    },
-    {
-      "epoch": 3.12407596952121,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 192290
-    },
-    {
-      "epoch": 3.1242384364185796,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 192300
-    },
-    {
-      "epoch": 3.124400903315949,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 192310
-    },
-    {
-      "epoch": 3.124563370213319,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 192320
-    },
-    {
-      "epoch": 3.1247258371106885,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 192330
-    },
-    {
-      "epoch": 3.124888304008058,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 192340
-    },
-    {
-      "epoch": 3.125050770905428,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 192350
-    },
-    {
-      "epoch": 3.1252132378027975,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 192360
-    },
-    {
-      "epoch": 3.1253757047001676,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 192370
-    },
-    {
-      "epoch": 3.1255381715975368,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 192380
-    },
-    {
-      "epoch": 3.125700638494907,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 192390
-    },
-    {
-      "epoch": 3.1258631053922765,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 192400
-    },
-    {
-      "epoch": 3.126025572289646,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 192410
-    },
-    {
-      "epoch": 3.126188039187016,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 192420
-    },
-    {
-      "epoch": 3.1263505060843855,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 192430
-    },
-    {
-      "epoch": 3.126512972981755,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 192440
-    },
-    {
-      "epoch": 3.1266754398791248,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 192450
-    },
-    {
-      "epoch": 3.1268379067764944,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.14,
-      "step": 192460
-    },
-    {
-      "epoch": 3.127000373673864,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1489,
-      "step": 192470
-    },
-    {
-      "epoch": 3.1271628405712337,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1694,
-      "step": 192480
-    },
-    {
-      "epoch": 3.1273253074686034,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1701,
-      "step": 192490
-    },
-    {
-      "epoch": 3.127487774365973,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.17,
-      "step": 192500
-    },
-    {
-      "epoch": 3.1276502412633427,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 192510
-    },
-    {
-      "epoch": 3.1278127081607123,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 192520
-    },
-    {
-      "epoch": 3.127975175058082,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1828,
-      "step": 192530
-    },
-    {
-      "epoch": 3.1281376419554516,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 192540
-    },
-    {
-      "epoch": 3.1283001088528213,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1631,
-      "step": 192550
-    },
-    {
-      "epoch": 3.128462575750191,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1649,
-      "step": 192560
-    },
-    {
-      "epoch": 3.1286250426475606,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 192570
-    },
-    {
-      "epoch": 3.1287875095449302,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 192580
-    },
-    {
-      "epoch": 3.1289499764423,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 192590
-    },
-    {
-      "epoch": 3.1291124433396695,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 192600
-    },
-    {
-      "epoch": 3.129274910237039,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1542,
-      "step": 192610
-    },
-    {
-      "epoch": 3.129437377134409,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1546,
-      "step": 192620
-    },
-    {
-      "epoch": 3.1295998440317785,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1648,
-      "step": 192630
-    },
-    {
-      "epoch": 3.129762310929148,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 192640
-    },
-    {
-      "epoch": 3.129924777826518,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1603,
-      "step": 192650
-    },
-    {
-      "epoch": 3.1300872447238874,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1831,
-      "step": 192660
-    },
-    {
-      "epoch": 3.130249711621257,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1723,
-      "step": 192670
-    },
-    {
-      "epoch": 3.1304121785186267,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1943,
-      "step": 192680
-    },
-    {
-      "epoch": 3.1305746454159964,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 192690
-    },
-    {
-      "epoch": 3.130737112313366,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 192700
-    },
-    {
-      "epoch": 3.1308995792107357,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 192710
-    },
-    {
-      "epoch": 3.1310620461081053,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1804,
-      "step": 192720
-    },
-    {
-      "epoch": 3.131224513005475,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 192730
-    },
-    {
-      "epoch": 3.1313869799028446,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 192740
-    },
-    {
-      "epoch": 3.1315494468002143,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 192750
-    },
-    {
-      "epoch": 3.131711913697584,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1521,
-      "step": 192760
-    },
-    {
-      "epoch": 3.1318743805949536,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1574,
-      "step": 192770
-    },
-    {
-      "epoch": 3.1320368474923233,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1632,
-      "step": 192780
-    },
-    {
-      "epoch": 3.132199314389693,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 192790
-    },
-    {
-      "epoch": 3.132361781287063,
-      "grad_norm": 3.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1673,
-      "step": 192800
-    },
-    {
-      "epoch": 3.132524248184432,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1584,
-      "step": 192810
-    },
-    {
-      "epoch": 3.1326867150818023,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.167,
-      "step": 192820
-    },
-    {
-      "epoch": 3.132849181979172,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1846,
-      "step": 192830
-    },
-    {
-      "epoch": 3.1330116488765416,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 192840
-    },
-    {
-      "epoch": 3.1331741157739113,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 192850
-    },
-    {
-      "epoch": 3.133336582671281,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 192860
-    },
-    {
-      "epoch": 3.1334990495686506,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 192870
-    },
-    {
-      "epoch": 3.13366151646602,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 192880
-    },
-    {
-      "epoch": 3.13382398336339,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1662,
-      "step": 192890
-    },
-    {
-      "epoch": 3.1339864502607595,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 192900
-    },
-    {
-      "epoch": 3.134148917158129,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 192910
-    },
-    {
-      "epoch": 3.134311384055499,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 192920
-    },
-    {
-      "epoch": 3.1344738509528685,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 192930
-    },
-    {
-      "epoch": 3.134636317850238,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 192940
-    },
-    {
-      "epoch": 3.1347987847476078,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 192950
-    },
-    {
-      "epoch": 3.1349612516449774,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 192960
-    },
-    {
-      "epoch": 3.135123718542347,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 192970
-    },
-    {
-      "epoch": 3.1352861854397167,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1726,
-      "step": 192980
-    },
-    {
-      "epoch": 3.1354486523370864,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 192990
-    },
-    {
-      "epoch": 3.135611119234456,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 193000
-    },
-    {
-      "epoch": 3.1357735861318257,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1644,
-      "step": 193010
-    },
-    {
-      "epoch": 3.1359360530291953,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 193020
-    },
-    {
-      "epoch": 3.136098519926565,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1804,
-      "step": 193030
-    },
-    {
-      "epoch": 3.1362609868239346,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 193040
-    },
-    {
-      "epoch": 3.1364234537213043,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1721,
-      "step": 193050
-    },
-    {
-      "epoch": 3.136585920618674,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 193060
-    },
-    {
-      "epoch": 3.1367483875160436,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 193070
-    },
-    {
-      "epoch": 3.136910854413413,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 193080
-    },
-    {
-      "epoch": 3.137073321310783,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 193090
-    },
-    {
-      "epoch": 3.1372357882081525,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2187,
-      "step": 193100
-    },
-    {
-      "epoch": 3.137398255105522,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 193110
-    },
-    {
-      "epoch": 3.137560722002892,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 193120
-    },
-    {
-      "epoch": 3.1377231889002615,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.18,
-      "step": 193130
-    },
-    {
-      "epoch": 3.137885655797631,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 193140
-    },
-    {
-      "epoch": 3.1380481226950008,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 193150
-    },
-    {
-      "epoch": 3.1382105895923704,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1641,
-      "step": 193160
-    },
-    {
-      "epoch": 3.13837305648974,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1675,
-      "step": 193170
-    },
-    {
-      "epoch": 3.1385355233871097,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.157,
-      "step": 193180
-    },
-    {
-      "epoch": 3.1386979902844794,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1659,
-      "step": 193190
-    },
-    {
-      "epoch": 3.138860457181849,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1906,
-      "step": 193200
-    },
-    {
-      "epoch": 3.1390229240792187,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 193210
-    },
-    {
-      "epoch": 3.1391853909765883,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 193220
-    },
-    {
-      "epoch": 3.139347857873958,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.1717,
-      "step": 193230
-    },
-    {
-      "epoch": 3.1395103247713276,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 193240
-    },
-    {
-      "epoch": 3.1396727916686977,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 193250
-    },
-    {
-      "epoch": 3.139835258566067,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 193260
-    },
-    {
-      "epoch": 3.139997725463437,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1659,
-      "step": 193270
-    },
-    {
-      "epoch": 3.1401601923608067,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 193280
-    },
-    {
-      "epoch": 3.1403226592581763,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 193290
-    },
-    {
-      "epoch": 3.140485126155546,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1697,
-      "step": 193300
-    },
-    {
-      "epoch": 3.1406475930529156,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1739,
-      "step": 193310
-    },
-    {
-      "epoch": 3.1408100599502853,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 193320
-    },
-    {
-      "epoch": 3.140972526847655,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 193330
-    },
-    {
-      "epoch": 3.1411349937450246,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 193340
-    },
-    {
-      "epoch": 3.1412974606423942,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1804,
-      "step": 193350
-    },
-    {
-      "epoch": 3.141459927539764,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 193360
-    },
-    {
-      "epoch": 3.1416223944371335,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 193370
-    },
-    {
-      "epoch": 3.141784861334503,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 193380
-    },
-    {
-      "epoch": 3.141947328231873,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 193390
-    },
-    {
-      "epoch": 3.1421097951292425,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1939,
-      "step": 193400
-    },
-    {
-      "epoch": 3.142272262026612,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1624,
-      "step": 193410
-    },
-    {
-      "epoch": 3.142434728923982,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1846,
-      "step": 193420
-    },
-    {
-      "epoch": 3.1425971958213514,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1611,
-      "step": 193430
-    },
-    {
-      "epoch": 3.142759662718721,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.157,
-      "step": 193440
-    },
-    {
-      "epoch": 3.1429221296160907,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 193450
-    },
-    {
-      "epoch": 3.1430845965134604,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1486,
-      "step": 193460
-    },
-    {
-      "epoch": 3.14324706341083,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1747,
-      "step": 193470
-    },
-    {
-      "epoch": 3.1434095303081997,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 193480
-    },
-    {
-      "epoch": 3.1435719972055693,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1743,
-      "step": 193490
-    },
-    {
-      "epoch": 3.143734464102939,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1936,
-      "step": 193500
-    },
-    {
-      "epoch": 3.1438969310003086,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1549,
-      "step": 193510
-    },
-    {
-      "epoch": 3.1440593978976783,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1592,
-      "step": 193520
-    },
-    {
-      "epoch": 3.144221864795048,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1682,
-      "step": 193530
-    },
-    {
-      "epoch": 3.1443843316924176,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 193540
-    },
-    {
-      "epoch": 3.1445467985897873,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 193550
-    },
-    {
-      "epoch": 3.144709265487157,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2358,
-      "step": 193560
-    },
-    {
-      "epoch": 3.1448717323845266,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 193570
-    },
-    {
-      "epoch": 3.145034199281896,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2115,
-      "step": 193580
-    },
-    {
-      "epoch": 3.145196666179266,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 193590
-    },
-    {
-      "epoch": 3.1453591330766355,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 193600
-    },
-    {
-      "epoch": 3.145521599974005,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1811,
-      "step": 193610
-    },
-    {
-      "epoch": 3.145684066871375,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 193620
-    },
-    {
-      "epoch": 3.1458465337687445,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1835,
-      "step": 193630
-    },
-    {
-      "epoch": 3.146009000666114,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.1592,
-      "step": 193640
-    },
-    {
-      "epoch": 3.1461714675634838,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1499,
-      "step": 193650
-    },
-    {
-      "epoch": 3.1463339344608534,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 193660
-    },
-    {
-      "epoch": 3.146496401358223,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 193670
-    },
-    {
-      "epoch": 3.146658868255593,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1514,
-      "step": 193680
-    },
-    {
-      "epoch": 3.1468213351529624,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1712,
-      "step": 193690
-    },
-    {
-      "epoch": 3.1469838020503325,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 193700
-    },
-    {
-      "epoch": 3.147146268947702,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 193710
-    },
-    {
-      "epoch": 3.1473087358450718,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 193720
-    },
-    {
-      "epoch": 3.1474712027424414,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1702,
-      "step": 193730
-    },
-    {
-      "epoch": 3.147633669639811,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 193740
-    },
-    {
-      "epoch": 3.1477961365371807,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1481,
-      "step": 193750
-    },
-    {
-      "epoch": 3.1479586034345504,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 193760
-    },
-    {
-      "epoch": 3.14812107033192,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1616,
-      "step": 193770
-    },
-    {
-      "epoch": 3.1482835372292897,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1894,
-      "step": 193780
-    },
-    {
-      "epoch": 3.1484460041266593,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 193790
-    },
-    {
-      "epoch": 3.148608471024029,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 193800
-    },
-    {
-      "epoch": 3.1487709379213986,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 193810
-    },
-    {
-      "epoch": 3.1489334048187683,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 193820
-    },
-    {
-      "epoch": 3.149095871716138,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 193830
-    },
-    {
-      "epoch": 3.1492583386135076,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1724,
-      "step": 193840
-    },
-    {
-      "epoch": 3.149420805510877,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1895,
-      "step": 193850
-    },
-    {
-      "epoch": 3.149583272408247,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1753,
-      "step": 193860
-    },
-    {
-      "epoch": 3.1497457393056165,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 193870
-    },
-    {
-      "epoch": 3.149908206202986,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 193880
-    },
-    {
-      "epoch": 3.150070673100356,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 193890
-    },
-    {
-      "epoch": 3.1502331399977255,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.165,
-      "step": 193900
-    },
-    {
-      "epoch": 3.150395606895095,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1794,
-      "step": 193910
-    },
-    {
-      "epoch": 3.1505580737924648,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 193920
-    },
-    {
-      "epoch": 3.1507205406898344,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 193930
-    },
-    {
-      "epoch": 3.150883007587204,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1751,
-      "step": 193940
-    },
-    {
-      "epoch": 3.1510454744845737,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1943,
-      "step": 193950
-    },
-    {
-      "epoch": 3.1512079413819434,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 193960
-    },
-    {
-      "epoch": 3.151370408279313,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1747,
-      "step": 193970
-    },
-    {
-      "epoch": 3.1515328751766827,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1605,
-      "step": 193980
-    },
-    {
-      "epoch": 3.1516953420740523,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1677,
-      "step": 193990
-    },
-    {
-      "epoch": 3.151857808971422,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1605,
-      "step": 194000
-    },
-    {
-      "epoch": 3.1520202758687916,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 194010
-    },
-    {
-      "epoch": 3.1521827427661613,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1977,
-      "step": 194020
-    },
-    {
-      "epoch": 3.152345209663531,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 194030
-    },
-    {
-      "epoch": 3.1525076765609006,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1895,
-      "step": 194040
-    },
-    {
-      "epoch": 3.1526701434582702,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 194050
-    },
-    {
-      "epoch": 3.15283261035564,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 194060
-    },
-    {
-      "epoch": 3.1529950772530095,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.166,
-      "step": 194070
-    },
-    {
-      "epoch": 3.153157544150379,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 194080
-    },
-    {
-      "epoch": 3.153320011047749,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 194090
-    },
-    {
-      "epoch": 3.1534824779451185,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 194100
-    },
-    {
-      "epoch": 3.153644944842488,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1623,
-      "step": 194110
-    },
-    {
-      "epoch": 3.153807411739858,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 194120
-    },
-    {
-      "epoch": 3.153969878637228,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 194130
-    },
-    {
-      "epoch": 3.154132345534597,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.18,
-      "step": 194140
-    },
-    {
-      "epoch": 3.154294812431967,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 194150
-    },
-    {
-      "epoch": 3.154457279329337,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 194160
-    },
-    {
-      "epoch": 3.1546197462267065,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 194170
-    },
-    {
-      "epoch": 3.154782213124076,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1682,
-      "step": 194180
-    },
-    {
-      "epoch": 3.154944680021446,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.156,
-      "step": 194190
-    },
-    {
-      "epoch": 3.1551071469188154,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1493,
-      "step": 194200
-    },
-    {
-      "epoch": 3.155269613816185,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1709,
-      "step": 194210
-    },
-    {
-      "epoch": 3.1554320807135547,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1816,
-      "step": 194220
-    },
-    {
-      "epoch": 3.1555945476109244,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 194230
-    },
-    {
-      "epoch": 3.155757014508294,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 194240
-    },
-    {
-      "epoch": 3.1559194814056637,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 194250
-    },
-    {
-      "epoch": 3.1560819483030333,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 194260
-    },
-    {
-      "epoch": 3.156244415200403,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 194270
-    },
-    {
-      "epoch": 3.1564068820977726,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 194280
-    },
-    {
-      "epoch": 3.1565693489951423,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1889,
-      "step": 194290
-    },
-    {
-      "epoch": 3.156731815892512,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 194300
-    },
-    {
-      "epoch": 3.1568942827898816,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1599,
-      "step": 194310
-    },
-    {
-      "epoch": 3.1570567496872513,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 194320
-    },
-    {
-      "epoch": 3.157219216584621,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1863,
-      "step": 194330
-    },
-    {
-      "epoch": 3.1573816834819906,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 194340
-    },
-    {
-      "epoch": 3.15754415037936,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 194350
-    },
-    {
-      "epoch": 3.15770661727673,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1894,
-      "step": 194360
-    },
-    {
-      "epoch": 3.1578690841740995,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 194370
-    },
-    {
-      "epoch": 3.158031551071469,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 194380
-    },
-    {
-      "epoch": 3.158194017968839,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 194390
-    },
-    {
-      "epoch": 3.1583564848662085,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 194400
-    },
-    {
-      "epoch": 3.158518951763578,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 194410
-    },
-    {
-      "epoch": 3.1586814186609478,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 194420
-    },
-    {
-      "epoch": 3.1588438855583174,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 194430
-    },
-    {
-      "epoch": 3.159006352455687,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1816,
-      "step": 194440
-    },
-    {
-      "epoch": 3.1591688193530567,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1731,
-      "step": 194450
-    },
-    {
-      "epoch": 3.1593312862504264,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1896,
-      "step": 194460
-    },
-    {
-      "epoch": 3.159493753147796,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1699,
-      "step": 194470
-    },
-    {
-      "epoch": 3.1596562200451657,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 194480
-    },
-    {
-      "epoch": 3.1598186869425353,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 194490
-    },
-    {
-      "epoch": 3.159981153839905,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.181,
-      "step": 194500
-    },
-    {
-      "epoch": 3.1601436207372746,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 194510
-    },
-    {
-      "epoch": 3.1603060876346443,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 194520
-    },
-    {
-      "epoch": 3.160468554532014,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1959,
-      "step": 194530
-    },
-    {
-      "epoch": 3.1606310214293836,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 194540
-    },
-    {
-      "epoch": 3.160793488326753,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.18,
-      "step": 194550
-    },
-    {
-      "epoch": 3.1609559552241233,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 194560
-    },
-    {
-      "epoch": 3.1611184221214925,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 194570
-    },
-    {
-      "epoch": 3.1612808890188626,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 194580
-    },
-    {
-      "epoch": 3.1614433559162323,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 194590
-    },
-    {
-      "epoch": 3.161605822813602,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.1796,
-      "step": 194600
-    },
-    {
-      "epoch": 3.1617682897109716,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 194610
-    },
-    {
-      "epoch": 3.161930756608341,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1479,
-      "step": 194620
-    },
-    {
-      "epoch": 3.162093223505711,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1517,
-      "step": 194630
-    },
-    {
-      "epoch": 3.1622556904030805,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1763,
-      "step": 194640
-    },
-    {
-      "epoch": 3.16241815730045,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 194650
-    },
-    {
-      "epoch": 3.16258062419782,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1667,
-      "step": 194660
-    },
-    {
-      "epoch": 3.1627430910951895,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 194670
-    },
-    {
-      "epoch": 3.162905557992559,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1725,
-      "step": 194680
-    },
-    {
-      "epoch": 3.1630680248899288,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 194690
-    },
-    {
-      "epoch": 3.1632304917872984,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.1677,
-      "step": 194700
-    },
-    {
-      "epoch": 3.163392958684668,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 194710
-    },
-    {
-      "epoch": 3.1635554255820377,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1671,
-      "step": 194720
-    },
-    {
-      "epoch": 3.1637178924794074,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 194730
-    },
-    {
-      "epoch": 3.163880359376777,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 194740
-    },
-    {
-      "epoch": 3.1640428262741467,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.165,
-      "step": 194750
-    },
-    {
-      "epoch": 3.1642052931715163,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1689,
-      "step": 194760
-    },
-    {
-      "epoch": 3.164367760068886,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 194770
-    },
-    {
-      "epoch": 3.1645302269662556,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1493,
-      "step": 194780
-    },
-    {
-      "epoch": 3.1646926938636253,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.165,
-      "step": 194790
-    },
-    {
-      "epoch": 3.164855160760995,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1566,
-      "step": 194800
-    },
-    {
-      "epoch": 3.1650176276583646,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1586,
-      "step": 194810
-    },
-    {
-      "epoch": 3.1651800945557342,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 194820
-    },
-    {
-      "epoch": 3.165342561453104,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 194830
-    },
-    {
-      "epoch": 3.1655050283504735,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 194840
-    },
-    {
-      "epoch": 3.165667495247843,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.1837,
-      "step": 194850
-    },
-    {
-      "epoch": 3.165829962145213,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1734,
-      "step": 194860
-    },
-    {
-      "epoch": 3.1659924290425825,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 194870
-    },
-    {
-      "epoch": 3.166154895939952,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1738,
-      "step": 194880
-    },
-    {
-      "epoch": 3.166317362837322,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1802,
-      "step": 194890
-    },
-    {
-      "epoch": 3.1664798297346914,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 194900
-    },
-    {
-      "epoch": 3.166642296632061,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 194910
-    },
-    {
-      "epoch": 3.1668047635294307,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 194920
-    },
-    {
-      "epoch": 3.1669672304268004,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1706,
-      "step": 194930
-    },
-    {
-      "epoch": 3.16712969732417,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 194940
-    },
-    {
-      "epoch": 3.1672921642215397,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 194950
-    },
-    {
-      "epoch": 3.1674546311189093,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1751,
-      "step": 194960
-    },
-    {
-      "epoch": 3.167617098016279,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1574,
-      "step": 194970
-    },
-    {
-      "epoch": 3.1677795649136486,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 194980
-    },
-    {
-      "epoch": 3.1679420318110183,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 194990
-    },
-    {
-      "epoch": 3.168104498708388,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 195000
-    },
-    {
-      "epoch": 3.168266965605758,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 195010
-    },
-    {
-      "epoch": 3.1684294325031273,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 195020
-    },
-    {
-      "epoch": 3.1685918994004973,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 195030
-    },
-    {
-      "epoch": 3.168754366297867,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 195040
-    },
-    {
-      "epoch": 3.1689168331952366,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.164,
-      "step": 195050
-    },
-    {
-      "epoch": 3.1690793000926063,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1927,
-      "step": 195060
-    },
-    {
-      "epoch": 3.169241766989976,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1667,
-      "step": 195070
-    },
-    {
-      "epoch": 3.1694042338873456,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1817,
-      "step": 195080
-    },
-    {
-      "epoch": 3.1695667007847153,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 195090
-    },
-    {
-      "epoch": 3.169729167682085,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1699,
-      "step": 195100
-    },
-    {
-      "epoch": 3.1698916345794546,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1691,
-      "step": 195110
-    },
-    {
-      "epoch": 3.170054101476824,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 195120
-    },
-    {
-      "epoch": 3.170216568374194,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 195130
-    },
-    {
-      "epoch": 3.1703790352715635,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1706,
-      "step": 195140
-    },
-    {
-      "epoch": 3.170541502168933,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 195150
-    },
-    {
-      "epoch": 3.170703969066303,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 195160
-    },
-    {
-      "epoch": 3.1708664359636725,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1755,
-      "step": 195170
-    },
-    {
-      "epoch": 3.171028902861042,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1614,
-      "step": 195180
-    },
-    {
-      "epoch": 3.1711913697584118,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1602,
-      "step": 195190
-    },
-    {
-      "epoch": 3.1713538366557814,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1549,
-      "step": 195200
-    },
-    {
-      "epoch": 3.171516303553151,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 195210
-    },
-    {
-      "epoch": 3.1716787704505207,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 195220
-    },
-    {
-      "epoch": 3.1718412373478904,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 195230
-    },
-    {
-      "epoch": 3.17200370424526,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1936,
-      "step": 195240
-    },
-    {
-      "epoch": 3.1721661711426297,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 195250
-    },
-    {
-      "epoch": 3.1723286380399993,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 195260
-    },
-    {
-      "epoch": 3.172491104937369,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1877,
-      "step": 195270
-    },
-    {
-      "epoch": 3.1726535718347386,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 195280
-    },
-    {
-      "epoch": 3.1728160387321083,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 195290
-    },
-    {
-      "epoch": 3.172978505629478,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 195300
-    },
-    {
-      "epoch": 3.1731409725268476,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1659,
-      "step": 195310
-    },
-    {
-      "epoch": 3.173303439424217,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 195320
-    },
-    {
-      "epoch": 3.173465906321587,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 195330
-    },
-    {
-      "epoch": 3.1736283732189565,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 195340
-    },
-    {
-      "epoch": 3.173790840116326,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 195350
-    },
-    {
-      "epoch": 3.173953307013696,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 195360
-    },
-    {
-      "epoch": 3.1741157739110655,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1631,
-      "step": 195370
-    },
-    {
-      "epoch": 3.174278240808435,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1795,
-      "step": 195380
-    },
-    {
-      "epoch": 3.1744407077058048,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 195390
-    },
-    {
-      "epoch": 3.1746031746031744,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.181,
-      "step": 195400
-    },
-    {
-      "epoch": 3.174765641500544,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 195410
-    },
-    {
-      "epoch": 3.1749281083979137,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 195420
-    },
-    {
-      "epoch": 3.1750905752952834,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 195430
-    },
-    {
-      "epoch": 3.1752530421926535,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1699,
-      "step": 195440
-    },
-    {
-      "epoch": 3.1754155090900227,
-      "grad_norm": 7.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2004,
-      "step": 195450
-    },
-    {
-      "epoch": 3.1755779759873928,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 195460
-    },
-    {
-      "epoch": 3.1757404428847624,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 195470
-    },
-    {
-      "epoch": 3.175902909782132,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 195480
-    },
-    {
-      "epoch": 3.1760653766795017,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1594,
-      "step": 195490
-    },
-    {
-      "epoch": 3.1762278435768714,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.1584,
-      "step": 195500
-    },
-    {
-      "epoch": 3.176390310474241,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1631,
-      "step": 195510
-    },
-    {
-      "epoch": 3.1765527773716107,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 195520
-    },
-    {
-      "epoch": 3.1767152442689803,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 195530
-    },
-    {
-      "epoch": 3.17687771116635,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1725,
-      "step": 195540
-    },
-    {
-      "epoch": 3.1770401780637196,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1669,
-      "step": 195550
-    },
-    {
-      "epoch": 3.1772026449610893,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1557,
-      "step": 195560
-    },
-    {
-      "epoch": 3.177365111858459,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1525,
-      "step": 195570
-    },
-    {
-      "epoch": 3.1775275787558286,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1388,
-      "step": 195580
-    },
-    {
-      "epoch": 3.1776900456531982,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1462,
-      "step": 195590
-    },
-    {
-      "epoch": 3.177852512550568,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1353,
-      "step": 195600
-    },
-    {
-      "epoch": 3.1780149794479375,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1626,
-      "step": 195610
-    },
-    {
-      "epoch": 3.178177446345307,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1615,
-      "step": 195620
-    },
-    {
-      "epoch": 3.178339913242677,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1696,
-      "step": 195630
-    },
-    {
-      "epoch": 3.1785023801400465,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 195640
-    },
-    {
-      "epoch": 3.178664847037416,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 195650
-    },
-    {
-      "epoch": 3.178827313934786,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 195660
-    },
-    {
-      "epoch": 3.1789897808321554,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 195670
-    },
-    {
-      "epoch": 3.179152247729525,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1802,
-      "step": 195680
-    },
-    {
-      "epoch": 3.1793147146268947,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 195690
-    },
-    {
-      "epoch": 3.1794771815242644,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 195700
-    },
-    {
-      "epoch": 3.179639648421634,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 195710
-    },
-    {
-      "epoch": 3.1798021153190037,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 195720
-    },
-    {
-      "epoch": 3.1799645822163733,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 195730
-    },
-    {
-      "epoch": 3.180127049113743,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 195740
-    },
-    {
-      "epoch": 3.1802895160111126,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 195750
-    },
-    {
-      "epoch": 3.1804519829084823,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 195760
-    },
-    {
-      "epoch": 3.180614449805852,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1809,
-      "step": 195770
-    },
-    {
-      "epoch": 3.1807769167032216,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 195780
-    },
-    {
-      "epoch": 3.1809393836005913,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 195790
-    },
-    {
-      "epoch": 3.181101850497961,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 195800
-    },
-    {
-      "epoch": 3.1812643173953306,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 195810
-    },
-    {
-      "epoch": 3.1814267842927,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 195820
-    },
-    {
-      "epoch": 3.18158925119007,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1817,
-      "step": 195830
-    },
-    {
-      "epoch": 3.1817517180874395,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1736,
-      "step": 195840
-    },
-    {
-      "epoch": 3.181914184984809,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1727,
-      "step": 195850
-    },
-    {
-      "epoch": 3.182076651882179,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1575,
-      "step": 195860
-    },
-    {
-      "epoch": 3.1822391187795485,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1673,
-      "step": 195870
-    },
-    {
-      "epoch": 3.182401585676918,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 195880
-    },
-    {
-      "epoch": 3.182564052574288,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1796,
-      "step": 195890
-    },
-    {
-      "epoch": 3.182726519471658,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 195900
-    },
-    {
-      "epoch": 3.1828889863690275,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 195910
-    },
-    {
-      "epoch": 3.183051453266397,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1749,
-      "step": 195920
-    },
-    {
-      "epoch": 3.183213920163767,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1717,
-      "step": 195930
-    },
-    {
-      "epoch": 3.1833763870611365,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.175,
-      "step": 195940
-    },
-    {
-      "epoch": 3.183538853958506,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 195950
-    },
-    {
-      "epoch": 3.1837013208558758,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 195960
-    },
-    {
-      "epoch": 3.1838637877532454,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 195970
-    },
-    {
-      "epoch": 3.184026254650615,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 195980
-    },
-    {
-      "epoch": 3.1841887215479847,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1734,
-      "step": 195990
-    },
-    {
-      "epoch": 3.1843511884453544,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 196000
-    },
-    {
-      "epoch": 3.184513655342724,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 196010
-    },
-    {
-      "epoch": 3.1846761222400937,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 196020
-    },
-    {
-      "epoch": 3.1848385891374633,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 196030
-    },
-    {
-      "epoch": 3.185001056034833,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 196040
-    },
-    {
-      "epoch": 3.1851635229322026,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 196050
-    },
-    {
-      "epoch": 3.1853259898295723,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2209,
-      "step": 196060
-    },
-    {
-      "epoch": 3.185488456726942,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 196070
-    },
-    {
-      "epoch": 3.1856509236243116,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 196080
-    },
-    {
-      "epoch": 3.185813390521681,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 196090
-    },
-    {
-      "epoch": 3.185975857419051,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 196100
-    },
-    {
-      "epoch": 3.1861383243164205,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 196110
-    },
-    {
-      "epoch": 3.18630079121379,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1889,
-      "step": 196120
-    },
-    {
-      "epoch": 3.18646325811116,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 196130
-    },
-    {
-      "epoch": 3.1866257250085295,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 196140
-    },
-    {
-      "epoch": 3.186788191905899,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 196150
-    },
-    {
-      "epoch": 3.1869506588032688,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2134,
-      "step": 196160
-    },
-    {
-      "epoch": 3.1871131257006384,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 196170
-    },
-    {
-      "epoch": 3.187275592598008,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2126,
-      "step": 196180
-    },
-    {
-      "epoch": 3.1874380594953777,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2232,
-      "step": 196190
-    },
-    {
-      "epoch": 3.1876005263927474,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 196200
-    },
-    {
-      "epoch": 3.187762993290117,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 196210
-    },
-    {
-      "epoch": 3.1879254601874867,
-      "grad_norm": 3.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 196220
-    },
-    {
-      "epoch": 3.1880879270848563,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 196230
-    },
-    {
-      "epoch": 3.188250393982226,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 196240
-    },
-    {
-      "epoch": 3.1884128608795956,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1764,
-      "step": 196250
-    },
-    {
-      "epoch": 3.1885753277769653,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 196260
-    },
-    {
-      "epoch": 3.188737794674335,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 196270
-    },
-    {
-      "epoch": 3.1889002615717046,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 196280
-    },
-    {
-      "epoch": 3.1890627284690742,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 196290
-    },
-    {
-      "epoch": 3.189225195366444,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 196300
-    },
-    {
-      "epoch": 3.1893876622638135,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1716,
-      "step": 196310
-    },
-    {
-      "epoch": 3.1895501291611836,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 196320
-    },
-    {
-      "epoch": 3.189712596058553,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1624,
-      "step": 196330
-    },
-    {
-      "epoch": 3.189875062955923,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 196340
-    },
-    {
-      "epoch": 3.1900375298532926,
-      "grad_norm": 3.578125,
-      "learning_rate": 5e-05,
-      "loss": 0.1393,
-      "step": 196350
-    },
-    {
-      "epoch": 3.1901999967506622,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1583,
-      "step": 196360
-    },
-    {
-      "epoch": 3.190362463648032,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1666,
-      "step": 196370
-    },
-    {
-      "epoch": 3.1905249305454015,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 196380
-    },
-    {
-      "epoch": 3.190687397442771,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1904,
-      "step": 196390
-    },
-    {
-      "epoch": 3.190849864340141,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1688,
-      "step": 196400
-    },
-    {
-      "epoch": 3.1910123312375105,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 196410
-    },
-    {
-      "epoch": 3.19117479813488,
-      "grad_norm": 7.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 196420
-    },
-    {
-      "epoch": 3.19133726503225,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1786,
-      "step": 196430
-    },
-    {
-      "epoch": 3.1914997319296194,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.1996,
-      "step": 196440
-    },
-    {
-      "epoch": 3.191662198826989,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 196450
-    },
-    {
-      "epoch": 3.1918246657243587,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 196460
-    },
-    {
-      "epoch": 3.1919871326217284,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1764,
-      "step": 196470
-    },
-    {
-      "epoch": 3.192149599519098,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 196480
-    },
-    {
-      "epoch": 3.1923120664164677,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 196490
-    },
-    {
-      "epoch": 3.1924745333138373,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 196500
-    },
-    {
-      "epoch": 3.192637000211207,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.198,
-      "step": 196510
-    },
-    {
-      "epoch": 3.1927994671085766,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 196520
-    },
-    {
-      "epoch": 3.1929619340059463,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 196530
-    },
-    {
-      "epoch": 3.193124400903316,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1778,
-      "step": 196540
-    },
-    {
-      "epoch": 3.1932868678006856,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1731,
-      "step": 196550
-    },
-    {
-      "epoch": 3.1934493346980553,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 196560
-    },
-    {
-      "epoch": 3.193611801595425,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 196570
-    },
-    {
-      "epoch": 3.1937742684927946,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 196580
-    },
-    {
-      "epoch": 3.193936735390164,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 196590
-    },
-    {
-      "epoch": 3.194099202287534,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 196600
-    },
-    {
-      "epoch": 3.1942616691849035,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 196610
-    },
-    {
-      "epoch": 3.194424136082273,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1885,
-      "step": 196620
-    },
-    {
-      "epoch": 3.194586602979643,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.176,
-      "step": 196630
-    },
-    {
-      "epoch": 3.1947490698770125,
-      "grad_norm": 7.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 196640
-    },
-    {
-      "epoch": 3.194911536774382,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 196650
-    },
-    {
-      "epoch": 3.1950740036717518,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 196660
-    },
-    {
-      "epoch": 3.1952364705691214,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 196670
-    },
-    {
-      "epoch": 3.195398937466491,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2177,
-      "step": 196680
-    },
-    {
-      "epoch": 3.1955614043638607,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 196690
-    },
-    {
-      "epoch": 3.1957238712612304,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2058,
-      "step": 196700
-    },
-    {
-      "epoch": 3.1958863381586,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 196710
-    },
-    {
-      "epoch": 3.1960488050559697,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 196720
-    },
-    {
-      "epoch": 3.1962112719533393,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1959,
-      "step": 196730
-    },
-    {
-      "epoch": 3.196373738850709,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 196740
-    },
-    {
-      "epoch": 3.196536205748079,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2191,
-      "step": 196750
-    },
-    {
-      "epoch": 3.1966986726454483,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 196760
-    },
-    {
-      "epoch": 3.1968611395428184,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 196770
-    },
-    {
-      "epoch": 3.197023606440188,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1705,
-      "step": 196780
-    },
-    {
-      "epoch": 3.1971860733375577,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1613,
-      "step": 196790
-    },
-    {
-      "epoch": 3.1973485402349273,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1557,
-      "step": 196800
-    },
-    {
-      "epoch": 3.197511007132297,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.156,
-      "step": 196810
-    },
-    {
-      "epoch": 3.1976734740296666,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1565,
-      "step": 196820
-    },
-    {
-      "epoch": 3.1978359409270363,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 196830
-    },
-    {
-      "epoch": 3.197998407824406,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1718,
-      "step": 196840
-    },
-    {
-      "epoch": 3.1981608747217756,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 196850
-    },
-    {
-      "epoch": 3.198323341619145,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 196860
-    },
-    {
-      "epoch": 3.198485808516515,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1996,
-      "step": 196870
-    },
-    {
-      "epoch": 3.1986482754138845,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 196880
-    },
-    {
-      "epoch": 3.198810742311254,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1837,
-      "step": 196890
-    },
-    {
-      "epoch": 3.198973209208624,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 196900
-    },
-    {
-      "epoch": 3.1991356761059935,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.2121,
-      "step": 196910
-    },
-    {
-      "epoch": 3.199298143003363,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 196920
-    },
-    {
-      "epoch": 3.1994606099007328,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 196930
-    },
-    {
-      "epoch": 3.1996230767981024,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 196940
-    },
-    {
-      "epoch": 3.199785543695472,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 196950
-    },
-    {
-      "epoch": 3.1999480105928417,
-      "grad_norm": 7.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1998,
-      "step": 196960
-    },
-    {
-      "epoch": 3.2001104774902114,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1738,
-      "step": 196970
-    },
-    {
-      "epoch": 3.200272944387581,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1775,
-      "step": 196980
-    },
-    {
-      "epoch": 3.2004354112849507,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 196990
-    },
-    {
-      "epoch": 3.2005978781823203,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 197000
-    },
-    {
-      "epoch": 3.20076034507969,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 197010
-    },
-    {
-      "epoch": 3.2009228119770596,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1939,
-      "step": 197020
-    },
-    {
-      "epoch": 3.2010852788744293,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1801,
-      "step": 197030
-    },
-    {
-      "epoch": 3.201247745771799,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 197040
-    },
-    {
-      "epoch": 3.2014102126691686,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 197050
-    },
-    {
-      "epoch": 3.2015726795665382,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1845,
-      "step": 197060
-    },
-    {
-      "epoch": 3.201735146463908,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 197070
-    },
-    {
-      "epoch": 3.2018976133612775,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.188,
-      "step": 197080
-    },
-    {
-      "epoch": 3.202060080258647,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 197090
-    },
-    {
-      "epoch": 3.202222547156017,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.191,
-      "step": 197100
-    },
-    {
-      "epoch": 3.2023850140533865,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1866,
-      "step": 197110
-    },
-    {
-      "epoch": 3.202547480950756,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 197120
-    },
-    {
-      "epoch": 3.202709947848126,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1681,
-      "step": 197130
-    },
-    {
-      "epoch": 3.2028724147454954,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.17,
-      "step": 197140
-    },
-    {
-      "epoch": 3.203034881642865,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1676,
-      "step": 197150
-    },
-    {
-      "epoch": 3.2031973485402347,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 197160
-    },
-    {
-      "epoch": 3.2033598154376044,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1711,
-      "step": 197170
-    },
-    {
-      "epoch": 3.203522282334974,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1673,
-      "step": 197180
-    },
-    {
-      "epoch": 3.2036847492323437,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1694,
-      "step": 197190
-    },
-    {
-      "epoch": 3.203847216129714,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1704,
-      "step": 197200
-    },
-    {
-      "epoch": 3.204009683027083,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1641,
-      "step": 197210
-    },
-    {
-      "epoch": 3.204172149924453,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 197220
-    },
-    {
-      "epoch": 3.2043346168218227,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1936,
-      "step": 197230
-    },
-    {
-      "epoch": 3.2044970837191924,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 197240
-    },
-    {
-      "epoch": 3.204659550616562,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 197250
-    },
-    {
-      "epoch": 3.2048220175139317,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 197260
-    },
-    {
-      "epoch": 3.2049844844113013,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 197270
-    },
-    {
-      "epoch": 3.205146951308671,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1726,
-      "step": 197280
-    },
-    {
-      "epoch": 3.2053094182060407,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 197290
-    },
-    {
-      "epoch": 3.2054718851034103,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1676,
-      "step": 197300
-    },
-    {
-      "epoch": 3.20563435200078,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 197310
-    },
-    {
-      "epoch": 3.2057968188981496,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 197320
-    },
-    {
-      "epoch": 3.2059592857955193,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 197330
-    },
-    {
-      "epoch": 3.206121752692889,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 197340
-    },
-    {
-      "epoch": 3.2062842195902586,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 197350
-    },
-    {
-      "epoch": 3.206446686487628,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 197360
-    },
-    {
-      "epoch": 3.206609153384998,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 197370
-    },
-    {
-      "epoch": 3.2067716202823675,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1959,
-      "step": 197380
-    },
-    {
-      "epoch": 3.206934087179737,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 197390
-    },
-    {
-      "epoch": 3.207096554077107,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 197400
-    },
-    {
-      "epoch": 3.2072590209744765,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 197410
-    },
-    {
-      "epoch": 3.207421487871846,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1649,
-      "step": 197420
-    },
-    {
-      "epoch": 3.2075839547692158,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1749,
-      "step": 197430
-    },
-    {
-      "epoch": 3.2077464216665854,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1568,
-      "step": 197440
-    },
-    {
-      "epoch": 3.207908888563955,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 197450
-    },
-    {
-      "epoch": 3.2080713554613247,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 197460
-    },
-    {
-      "epoch": 3.2082338223586944,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 197470
-    },
-    {
-      "epoch": 3.208396289256064,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 197480
-    },
-    {
-      "epoch": 3.2085587561534337,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1731,
-      "step": 197490
-    },
-    {
-      "epoch": 3.2087212230508033,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 197500
-    },
-    {
-      "epoch": 3.208883689948173,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2021,
-      "step": 197510
-    },
-    {
-      "epoch": 3.2090461568455426,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 197520
-    },
-    {
-      "epoch": 3.2092086237429123,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 197530
-    },
-    {
-      "epoch": 3.209371090640282,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 197540
-    },
-    {
-      "epoch": 3.2095335575376516,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 197550
-    },
-    {
-      "epoch": 3.2096960244350212,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1694,
-      "step": 197560
-    },
-    {
-      "epoch": 3.209858491332391,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1545,
-      "step": 197570
-    },
-    {
-      "epoch": 3.2100209582297605,
-      "grad_norm": 8.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 197580
-    },
-    {
-      "epoch": 3.21018342512713,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1718,
-      "step": 197590
-    },
-    {
-      "epoch": 3.2103458920245,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.186,
-      "step": 197600
-    },
-    {
-      "epoch": 3.2105083589218695,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 197610
-    },
-    {
-      "epoch": 3.210670825819239,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1664,
-      "step": 197620
-    },
-    {
-      "epoch": 3.210833292716609,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 197630
-    },
-    {
-      "epoch": 3.2109957596139784,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1627,
-      "step": 197640
-    },
-    {
-      "epoch": 3.2111582265113485,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 197650
-    },
-    {
-      "epoch": 3.211320693408718,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1669,
-      "step": 197660
-    },
-    {
-      "epoch": 3.211483160306088,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 197670
-    },
-    {
-      "epoch": 3.2116456272034575,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1629,
-      "step": 197680
-    },
-    {
-      "epoch": 3.211808094100827,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1701,
-      "step": 197690
-    },
-    {
-      "epoch": 3.2119705609981968,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.173,
-      "step": 197700
-    },
-    {
-      "epoch": 3.2121330278955664,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1512,
-      "step": 197710
-    },
-    {
-      "epoch": 3.212295494792936,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1717,
-      "step": 197720
-    },
-    {
-      "epoch": 3.2124579616903057,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 197730
-    },
-    {
-      "epoch": 3.2126204285876754,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1713,
-      "step": 197740
-    },
-    {
-      "epoch": 3.212782895485045,
-      "grad_norm": 3.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1632,
-      "step": 197750
-    },
-    {
-      "epoch": 3.2129453623824147,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1575,
-      "step": 197760
-    },
-    {
-      "epoch": 3.2131078292797843,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1741,
-      "step": 197770
-    },
-    {
-      "epoch": 3.213270296177154,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1677,
-      "step": 197780
-    },
-    {
-      "epoch": 3.2134327630745236,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 197790
-    },
-    {
-      "epoch": 3.2135952299718933,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 197800
-    },
-    {
-      "epoch": 3.213757696869263,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 197810
-    },
-    {
-      "epoch": 3.2139201637666326,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 197820
-    },
-    {
-      "epoch": 3.2140826306640022,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 197830
-    },
-    {
-      "epoch": 3.214245097561372,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 197840
-    },
-    {
-      "epoch": 3.2144075644587415,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1896,
-      "step": 197850
-    },
-    {
-      "epoch": 3.214570031356111,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 197860
-    },
-    {
-      "epoch": 3.214732498253481,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1774,
-      "step": 197870
-    },
-    {
-      "epoch": 3.2148949651508505,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 197880
-    },
-    {
-      "epoch": 3.21505743204822,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 197890
-    },
-    {
-      "epoch": 3.21521989894559,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 197900
-    },
-    {
-      "epoch": 3.2153823658429594,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1619,
-      "step": 197910
-    },
-    {
-      "epoch": 3.215544832740329,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1729,
-      "step": 197920
-    },
-    {
-      "epoch": 3.2157072996376987,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1704,
-      "step": 197930
-    },
-    {
-      "epoch": 3.2158697665350684,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 197940
-    },
-    {
-      "epoch": 3.216032233432438,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 197950
-    },
-    {
-      "epoch": 3.2161947003298077,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 197960
-    },
-    {
-      "epoch": 3.2163571672271773,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1715,
-      "step": 197970
-    },
-    {
-      "epoch": 3.216519634124547,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1735,
-      "step": 197980
-    },
-    {
-      "epoch": 3.2166821010219167,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 197990
-    },
-    {
-      "epoch": 3.2168445679192863,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 198000
-    },
-    {
-      "epoch": 3.217007034816656,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 198010
-    },
-    {
-      "epoch": 3.2171695017140256,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 198020
-    },
-    {
-      "epoch": 3.2173319686113953,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 198030
-    },
-    {
-      "epoch": 3.217494435508765,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 198040
-    },
-    {
-      "epoch": 3.2176569024061346,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 198050
-    },
-    {
-      "epoch": 3.217819369303504,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.186,
-      "step": 198060
-    },
-    {
-      "epoch": 3.217981836200874,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2005,
-      "step": 198070
-    },
-    {
-      "epoch": 3.218144303098244,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1991,
-      "step": 198080
-    },
-    {
-      "epoch": 3.218306769995613,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 198090
-    },
-    {
-      "epoch": 3.2184692368929833,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 198100
-    },
-    {
-      "epoch": 3.218631703790353,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.158,
-      "step": 198110
-    },
-    {
-      "epoch": 3.2187941706877226,
-      "grad_norm": 7.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 198120
-    },
-    {
-      "epoch": 3.218956637585092,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 198130
-    },
-    {
-      "epoch": 3.219119104482462,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 198140
-    },
-    {
-      "epoch": 3.2192815713798315,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1682,
-      "step": 198150
-    },
-    {
-      "epoch": 3.219444038277201,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 198160
-    },
-    {
-      "epoch": 3.219606505174571,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 198170
-    },
-    {
-      "epoch": 3.2197689720719405,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1727,
-      "step": 198180
-    },
-    {
-      "epoch": 3.21993143896931,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 198190
-    },
-    {
-      "epoch": 3.2200939058666798,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 198200
-    },
-    {
-      "epoch": 3.2202563727640494,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 198210
-    },
-    {
-      "epoch": 3.220418839661419,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 198220
-    },
-    {
-      "epoch": 3.2205813065587887,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1475,
-      "step": 198230
-    },
-    {
-      "epoch": 3.2207437734561584,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1434,
-      "step": 198240
-    },
-    {
-      "epoch": 3.220906240353528,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 198250
-    },
-    {
-      "epoch": 3.2210687072508977,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 198260
-    },
-    {
-      "epoch": 3.2212311741482673,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 198270
-    },
-    {
-      "epoch": 3.221393641045637,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 198280
-    },
-    {
-      "epoch": 3.2215561079430066,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 198290
-    },
-    {
-      "epoch": 3.2217185748403763,
-      "grad_norm": 7.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 198300
-    },
-    {
-      "epoch": 3.221881041737746,
-      "grad_norm": 8.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2166,
-      "step": 198310
-    },
-    {
-      "epoch": 3.2220435086351156,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 198320
-    },
-    {
-      "epoch": 3.2222059755324852,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 198330
-    },
-    {
-      "epoch": 3.222368442429855,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 198340
-    },
-    {
-      "epoch": 3.2225309093272245,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1747,
-      "step": 198350
-    },
-    {
-      "epoch": 3.222693376224594,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 198360
-    },
-    {
-      "epoch": 3.222855843121964,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1527,
-      "step": 198370
-    },
-    {
-      "epoch": 3.2230183100193335,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1534,
-      "step": 198380
-    },
-    {
-      "epoch": 3.223180776916703,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1694,
-      "step": 198390
-    },
-    {
-      "epoch": 3.223343243814073,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 198400
-    },
-    {
-      "epoch": 3.2235057107114424,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 198410
-    },
-    {
-      "epoch": 3.223668177608812,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 198420
-    },
-    {
-      "epoch": 3.2238306445061817,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 198430
-    },
-    {
-      "epoch": 3.2239931114035514,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 198440
-    },
-    {
-      "epoch": 3.224155578300921,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1712,
-      "step": 198450
-    },
-    {
-      "epoch": 3.2243180451982907,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 198460
-    },
-    {
-      "epoch": 3.2244805120956603,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 198470
-    },
-    {
-      "epoch": 3.22464297899303,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 198480
-    },
-    {
-      "epoch": 3.2248054458903996,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 198490
-    },
-    {
-      "epoch": 3.2249679127877693,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 198500
-    },
-    {
-      "epoch": 3.2251303796851394,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.187,
-      "step": 198510
-    },
-    {
-      "epoch": 3.2252928465825086,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 198520
-    },
-    {
-      "epoch": 3.2254553134798787,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 198530
-    },
-    {
-      "epoch": 3.2256177803772483,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 198540
-    },
-    {
-      "epoch": 3.225780247274618,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 198550
-    },
-    {
-      "epoch": 3.2259427141719876,
-      "grad_norm": 3.828125,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 198560
-    },
-    {
-      "epoch": 3.2261051810693573,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 198570
-    },
-    {
-      "epoch": 3.226267647966727,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 198580
-    },
-    {
-      "epoch": 3.2264301148640966,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 198590
-    },
-    {
-      "epoch": 3.2265925817614662,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 198600
-    },
-    {
-      "epoch": 3.226755048658836,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 198610
-    },
-    {
-      "epoch": 3.2269175155562055,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1936,
-      "step": 198620
-    },
-    {
-      "epoch": 3.227079982453575,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 198630
-    },
-    {
-      "epoch": 3.227242449350945,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 198640
-    },
-    {
-      "epoch": 3.2274049162483145,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1691,
-      "step": 198650
-    },
-    {
-      "epoch": 3.227567383145684,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 198660
-    },
-    {
-      "epoch": 3.227729850043054,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 198670
-    },
-    {
-      "epoch": 3.2278923169404234,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 198680
-    },
-    {
-      "epoch": 3.228054783837793,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 198690
-    },
-    {
-      "epoch": 3.2282172507351627,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.18,
-      "step": 198700
-    },
-    {
-      "epoch": 3.2283797176325324,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1687,
-      "step": 198710
-    },
-    {
-      "epoch": 3.228542184529902,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1816,
-      "step": 198720
-    },
-    {
-      "epoch": 3.2287046514272717,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 198730
-    },
-    {
-      "epoch": 3.2288671183246413,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2143,
-      "step": 198740
-    },
-    {
-      "epoch": 3.229029585222011,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 198750
-    },
-    {
-      "epoch": 3.2291920521193807,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 198760
-    },
-    {
-      "epoch": 3.2293545190167503,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1719,
-      "step": 198770
-    },
-    {
-      "epoch": 3.22951698591412,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1574,
-      "step": 198780
-    },
-    {
-      "epoch": 3.2296794528114896,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1713,
-      "step": 198790
-    },
-    {
-      "epoch": 3.2298419197088593,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 198800
-    },
-    {
-      "epoch": 3.230004386606229,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 198810
-    },
-    {
-      "epoch": 3.2301668535035986,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1788,
-      "step": 198820
-    },
-    {
-      "epoch": 3.230329320400968,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 198830
-    },
-    {
-      "epoch": 3.230491787298338,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2212,
-      "step": 198840
-    },
-    {
-      "epoch": 3.2306542541957075,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 198850
-    },
-    {
-      "epoch": 3.230816721093077,
-      "grad_norm": 9.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 198860
-    },
-    {
-      "epoch": 3.230979187990447,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 198870
-    },
-    {
-      "epoch": 3.2311416548878165,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 198880
-    },
-    {
-      "epoch": 3.231304121785186,
-      "grad_norm": 3.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1693,
-      "step": 198890
-    },
-    {
-      "epoch": 3.2314665886825558,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1592,
-      "step": 198900
-    },
-    {
-      "epoch": 3.2316290555799254,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.169,
-      "step": 198910
-    },
-    {
-      "epoch": 3.231791522477295,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1701,
-      "step": 198920
-    },
-    {
-      "epoch": 3.2319539893746647,
-      "grad_norm": 7.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 198930
-    },
-    {
-      "epoch": 3.2321164562720344,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 198940
-    },
-    {
-      "epoch": 3.232278923169404,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1904,
-      "step": 198950
-    },
-    {
-      "epoch": 3.232441390066774,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 198960
-    },
-    {
-      "epoch": 3.2326038569641438,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1624,
-      "step": 198970
-    },
-    {
-      "epoch": 3.2327663238615134,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1709,
-      "step": 198980
-    },
-    {
-      "epoch": 3.232928790758883,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 198990
-    },
-    {
-      "epoch": 3.2330912576562527,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1877,
-      "step": 199000
-    },
-    {
-      "epoch": 3.2332537245536224,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1667,
-      "step": 199010
-    },
-    {
-      "epoch": 3.233416191450992,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 199020
-    },
-    {
-      "epoch": 3.2335786583483617,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1794,
-      "step": 199030
-    },
-    {
-      "epoch": 3.2337411252457313,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1639,
-      "step": 199040
-    },
-    {
-      "epoch": 3.233903592143101,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 199050
-    },
-    {
-      "epoch": 3.2340660590404706,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 199060
-    },
-    {
-      "epoch": 3.2342285259378403,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2101,
-      "step": 199070
-    },
-    {
-      "epoch": 3.23439099283521,
-      "grad_norm": 7.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 199080
-    },
-    {
-      "epoch": 3.2345534597325796,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 199090
-    },
-    {
-      "epoch": 3.2347159266299492,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 199100
-    },
-    {
-      "epoch": 3.234878393527319,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 199110
-    },
-    {
-      "epoch": 3.2350408604246885,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1887,
-      "step": 199120
-    },
-    {
-      "epoch": 3.235203327322058,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1631,
-      "step": 199130
-    },
-    {
-      "epoch": 3.235365794219428,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 199140
-    },
-    {
-      "epoch": 3.2355282611167975,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 199150
-    },
-    {
-      "epoch": 3.235690728014167,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1584,
-      "step": 199160
-    },
-    {
-      "epoch": 3.235853194911537,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1814,
-      "step": 199170
-    },
-    {
-      "epoch": 3.2360156618089064,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 199180
-    },
-    {
-      "epoch": 3.236178128706276,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 199190
-    },
-    {
-      "epoch": 3.2363405956036457,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1659,
-      "step": 199200
-    },
-    {
-      "epoch": 3.2365030625010154,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 199210
-    },
-    {
-      "epoch": 3.236665529398385,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 199220
-    },
-    {
-      "epoch": 3.2368279962957547,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1802,
-      "step": 199230
-    },
-    {
-      "epoch": 3.2369904631931243,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1781,
-      "step": 199240
-    },
-    {
-      "epoch": 3.237152930090494,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1839,
-      "step": 199250
-    },
-    {
-      "epoch": 3.2373153969878636,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1666,
-      "step": 199260
-    },
-    {
-      "epoch": 3.2374778638852333,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1936,
-      "step": 199270
-    },
-    {
-      "epoch": 3.237640330782603,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 199280
-    },
-    {
-      "epoch": 3.2378027976799726,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 199290
-    },
-    {
-      "epoch": 3.2379652645773422,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1763,
-      "step": 199300
-    },
-    {
-      "epoch": 3.238127731474712,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1704,
-      "step": 199310
-    },
-    {
-      "epoch": 3.2382901983720815,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 199320
-    },
-    {
-      "epoch": 3.238452665269451,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.167,
-      "step": 199330
-    },
-    {
-      "epoch": 3.238615132166821,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1602,
-      "step": 199340
-    },
-    {
-      "epoch": 3.2387775990641905,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.152,
-      "step": 199350
-    },
-    {
-      "epoch": 3.23894006596156,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1692,
-      "step": 199360
-    },
-    {
-      "epoch": 3.23910253285893,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.15,
-      "step": 199370
-    },
-    {
-      "epoch": 3.2392649997562994,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.148,
-      "step": 199380
-    },
-    {
-      "epoch": 3.2394274666536695,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1646,
-      "step": 199390
-    },
-    {
-      "epoch": 3.2395899335510387,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 199400
-    },
-    {
-      "epoch": 3.239752400448409,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 199410
-    },
-    {
-      "epoch": 3.2399148673457785,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1615,
-      "step": 199420
-    },
-    {
-      "epoch": 3.240077334243148,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 199430
-    },
-    {
-      "epoch": 3.240239801140518,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 199440
-    },
-    {
-      "epoch": 3.2404022680378874,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 199450
-    },
-    {
-      "epoch": 3.240564734935257,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1675,
-      "step": 199460
-    },
-    {
-      "epoch": 3.2407272018326267,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 199470
-    },
-    {
-      "epoch": 3.2408896687299964,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 199480
-    },
-    {
-      "epoch": 3.241052135627366,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 199490
-    },
-    {
-      "epoch": 3.2412146025247357,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 199500
-    },
-    {
-      "epoch": 3.2413770694221054,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1749,
-      "step": 199510
-    },
-    {
-      "epoch": 3.241539536319475,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 199520
-    },
-    {
-      "epoch": 3.2417020032168447,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2176,
-      "step": 199530
-    },
-    {
-      "epoch": 3.2418644701142143,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 199540
-    },
-    {
-      "epoch": 3.242026937011584,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 199550
-    },
-    {
-      "epoch": 3.2421894039089536,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 199560
-    },
-    {
-      "epoch": 3.2423518708063233,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 199570
-    },
-    {
-      "epoch": 3.242514337703693,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 199580
-    },
-    {
-      "epoch": 3.2426768046010626,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 199590
-    },
-    {
-      "epoch": 3.242839271498432,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 199600
-    },
-    {
-      "epoch": 3.243001738395802,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1687,
-      "step": 199610
-    },
-    {
-      "epoch": 3.2431642052931715,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1553,
-      "step": 199620
-    },
-    {
-      "epoch": 3.243326672190541,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 199630
-    },
-    {
-      "epoch": 3.243489139087911,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 199640
-    },
-    {
-      "epoch": 3.2436516059852805,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1699,
-      "step": 199650
-    },
-    {
-      "epoch": 3.24381407288265,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.175,
-      "step": 199660
-    },
-    {
-      "epoch": 3.2439765397800198,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1536,
-      "step": 199670
-    },
-    {
-      "epoch": 3.2441390066773894,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.175,
-      "step": 199680
-    },
-    {
-      "epoch": 3.244301473574759,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.173,
-      "step": 199690
-    },
-    {
-      "epoch": 3.2444639404721287,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 199700
-    },
-    {
-      "epoch": 3.2446264073694984,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 199710
-    },
-    {
-      "epoch": 3.244788874266868,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1671,
-      "step": 199720
-    },
-    {
-      "epoch": 3.2449513411642377,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 199730
-    },
-    {
-      "epoch": 3.2451138080616073,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 199740
-    },
-    {
-      "epoch": 3.245276274958977,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 199750
-    },
-    {
-      "epoch": 3.2454387418563466,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 199760
-    },
-    {
-      "epoch": 3.2456012087537163,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 199770
-    },
-    {
-      "epoch": 3.245763675651086,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 199780
-    },
-    {
-      "epoch": 3.2459261425484556,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 199790
-    },
-    {
-      "epoch": 3.2460886094458252,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1851,
-      "step": 199800
-    },
-    {
-      "epoch": 3.246251076343195,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 199810
-    },
-    {
-      "epoch": 3.246413543240565,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1766,
-      "step": 199820
-    },
-    {
-      "epoch": 3.246576010137934,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 199830
-    },
-    {
-      "epoch": 3.2467384770353043,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 199840
-    },
-    {
-      "epoch": 3.246900943932674,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 199850
-    },
-    {
-      "epoch": 3.2470634108300436,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1781,
-      "step": 199860
-    },
-    {
-      "epoch": 3.2472258777274132,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 199870
-    },
-    {
-      "epoch": 3.247388344624783,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1709,
-      "step": 199880
-    },
-    {
-      "epoch": 3.2475508115221525,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1639,
-      "step": 199890
-    },
-    {
-      "epoch": 3.247713278419522,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.158,
-      "step": 199900
-    },
-    {
-      "epoch": 3.247875745316892,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2008,
-      "step": 199910
-    },
-    {
-      "epoch": 3.2480382122142615,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 199920
-    },
-    {
-      "epoch": 3.248200679111631,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 199930
-    },
-    {
-      "epoch": 3.248363146009001,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 199940
-    },
-    {
-      "epoch": 3.2485256129063704,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 199950
-    },
-    {
-      "epoch": 3.24868807980374,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 199960
-    },
-    {
-      "epoch": 3.2488505467011097,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 199970
-    },
-    {
-      "epoch": 3.2490130135984794,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 199980
-    },
-    {
-      "epoch": 3.249175480495849,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 199990
-    },
-    {
-      "epoch": 3.2493379473932187,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1906,
-      "step": 200000
-    },
-    {
-      "epoch": 3.2495004142905883,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 200010
-    },
-    {
-      "epoch": 3.249662881187958,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 200020
-    },
-    {
-      "epoch": 3.2498253480853276,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1591,
-      "step": 200030
-    },
-    {
-      "epoch": 3.2499878149826973,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.156,
-      "step": 200040
-    },
-    {
-      "epoch": 3.250150281880067,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1679,
-      "step": 200050
-    },
-    {
-      "epoch": 3.2503127487774366,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1722,
-      "step": 200060
-    },
-    {
-      "epoch": 3.2504752156748062,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 200070
-    },
-    {
-      "epoch": 3.250637682572176,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1769,
-      "step": 200080
-    },
-    {
-      "epoch": 3.2508001494695455,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1814,
-      "step": 200090
-    },
-    {
-      "epoch": 3.250962616366915,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1658,
-      "step": 200100
-    },
-    {
-      "epoch": 3.251125083264285,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1491,
-      "step": 200110
-    },
-    {
-      "epoch": 3.2512875501616545,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1802,
-      "step": 200120
-    },
-    {
-      "epoch": 3.251450017059024,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1737,
-      "step": 200130
-    },
-    {
-      "epoch": 3.251612483956394,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1564,
-      "step": 200140
-    },
-    {
-      "epoch": 3.2517749508537634,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1635,
-      "step": 200150
-    },
-    {
-      "epoch": 3.251937417751133,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1525,
-      "step": 200160
-    },
-    {
-      "epoch": 3.2520998846485027,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 200170
-    },
-    {
-      "epoch": 3.2522623515458724,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.173,
-      "step": 200180
-    },
-    {
-      "epoch": 3.252424818443242,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 200190
-    },
-    {
-      "epoch": 3.2525872853406117,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 200200
-    },
-    {
-      "epoch": 3.2527497522379814,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 200210
-    },
-    {
-      "epoch": 3.252912219135351,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1722,
-      "step": 200220
-    },
-    {
-      "epoch": 3.2530746860327207,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1551,
-      "step": 200230
-    },
-    {
-      "epoch": 3.2532371529300903,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1688,
-      "step": 200240
-    },
-    {
-      "epoch": 3.2533996198274604,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1659,
-      "step": 200250
-    },
-    {
-      "epoch": 3.2535620867248296,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.178,
-      "step": 200260
-    },
-    {
-      "epoch": 3.2537245536221997,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 200270
-    },
-    {
-      "epoch": 3.253887020519569,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 200280
-    },
-    {
-      "epoch": 3.254049487416939,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1739,
-      "step": 200290
-    },
-    {
-      "epoch": 3.2542119543143087,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 200300
-    },
-    {
-      "epoch": 3.2543744212116783,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 200310
-    },
-    {
-      "epoch": 3.254536888109048,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 200320
-    },
-    {
-      "epoch": 3.2546993550064176,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2141,
-      "step": 200330
-    },
-    {
-      "epoch": 3.2548618219037873,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 200340
-    },
-    {
-      "epoch": 3.255024288801157,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1811,
-      "step": 200350
-    },
-    {
-      "epoch": 3.2551867556985266,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1705,
-      "step": 200360
-    },
-    {
-      "epoch": 3.255349222595896,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 200370
-    },
-    {
-      "epoch": 3.255511689493266,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1871,
-      "step": 200380
-    },
-    {
-      "epoch": 3.2556741563906355,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 200390
-    },
-    {
-      "epoch": 3.255836623288005,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.198,
-      "step": 200400
-    },
-    {
-      "epoch": 3.255999090185375,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 200410
-    },
-    {
-      "epoch": 3.2561615570827445,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 200420
-    },
-    {
-      "epoch": 3.256324023980114,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 200430
-    },
-    {
-      "epoch": 3.2564864908774838,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 200440
-    },
-    {
-      "epoch": 3.2566489577748534,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 200450
-    },
-    {
-      "epoch": 3.256811424672223,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 200460
-    },
-    {
-      "epoch": 3.2569738915695927,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 200470
-    },
-    {
-      "epoch": 3.2571363584669624,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 200480
-    },
-    {
-      "epoch": 3.257298825364332,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 200490
-    },
-    {
-      "epoch": 3.2574612922617017,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 200500
-    },
-    {
-      "epoch": 3.2576237591590713,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1787,
-      "step": 200510
-    },
-    {
-      "epoch": 3.257786226056441,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 200520
-    },
-    {
-      "epoch": 3.2579486929538106,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2277,
-      "step": 200530
-    },
-    {
-      "epoch": 3.2581111598511803,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2276,
-      "step": 200540
-    },
-    {
-      "epoch": 3.25827362674855,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 200550
-    },
-    {
-      "epoch": 3.2584360936459196,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 200560
-    },
-    {
-      "epoch": 3.2585985605432892,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1887,
-      "step": 200570
-    },
-    {
-      "epoch": 3.258761027440659,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1845,
-      "step": 200580
-    },
-    {
-      "epoch": 3.2589234943380285,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 200590
-    },
-    {
-      "epoch": 3.259085961235398,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 200600
-    },
-    {
-      "epoch": 3.259248428132768,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1846,
-      "step": 200610
-    },
-    {
-      "epoch": 3.2594108950301375,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 200620
-    },
-    {
-      "epoch": 3.259573361927507,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2258,
-      "step": 200630
-    },
-    {
-      "epoch": 3.259735828824877,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2005,
-      "step": 200640
-    },
-    {
-      "epoch": 3.2598982957222464,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 200650
-    },
-    {
-      "epoch": 3.260060762619616,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 200660
-    },
-    {
-      "epoch": 3.2602232295169857,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 200670
-    },
-    {
-      "epoch": 3.2603856964143554,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 200680
-    },
-    {
-      "epoch": 3.260548163311725,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1684,
-      "step": 200690
-    },
-    {
-      "epoch": 3.260710630209095,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 200700
-    },
-    {
-      "epoch": 3.2608730971064643,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 200710
-    },
-    {
-      "epoch": 3.2610355640038344,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 200720
-    },
-    {
-      "epoch": 3.2611980309012036,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 200730
-    },
-    {
-      "epoch": 3.2613604977985737,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1691,
-      "step": 200740
-    },
-    {
-      "epoch": 3.2615229646959434,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1794,
-      "step": 200750
-    },
-    {
-      "epoch": 3.261685431593313,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 200760
-    },
-    {
-      "epoch": 3.2618478984906827,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 200770
-    },
-    {
-      "epoch": 3.2620103653880523,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 200780
-    },
-    {
-      "epoch": 3.262172832285422,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 200790
-    },
-    {
-      "epoch": 3.2623352991827916,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1734,
-      "step": 200800
-    },
-    {
-      "epoch": 3.2624977660801613,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1828,
-      "step": 200810
-    },
-    {
-      "epoch": 3.262660232977531,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 200820
-    },
-    {
-      "epoch": 3.2628226998749006,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2114,
-      "step": 200830
-    },
-    {
-      "epoch": 3.2629851667722702,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 200840
-    },
-    {
-      "epoch": 3.26314763366964,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 200850
-    },
-    {
-      "epoch": 3.2633101005670095,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1952,
-      "step": 200860
-    },
-    {
-      "epoch": 3.263472567464379,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 200870
-    },
-    {
-      "epoch": 3.263635034361749,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 200880
-    },
-    {
-      "epoch": 3.2637975012591185,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 200890
-    },
-    {
-      "epoch": 3.263959968156488,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 200900
-    },
-    {
-      "epoch": 3.264122435053858,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 200910
-    },
-    {
-      "epoch": 3.2642849019512274,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 200920
-    },
-    {
-      "epoch": 3.264447368848597,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 200930
-    },
-    {
-      "epoch": 3.2646098357459667,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2067,
-      "step": 200940
-    },
-    {
-      "epoch": 3.2647723026433364,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.221,
-      "step": 200950
-    },
-    {
-      "epoch": 3.264934769540706,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 200960
-    },
-    {
-      "epoch": 3.2650972364380757,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 200970
-    },
-    {
-      "epoch": 3.2652597033354454,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 200980
-    },
-    {
-      "epoch": 3.265422170232815,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 200990
-    },
-    {
-      "epoch": 3.2655846371301847,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 201000
-    },
-    {
-      "epoch": 3.2657471040275543,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 201010
-    },
-    {
-      "epoch": 3.265909570924924,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 201020
-    },
-    {
-      "epoch": 3.2660720378222936,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 201030
-    },
-    {
-      "epoch": 3.2662345047196633,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1493,
-      "step": 201040
-    },
-    {
-      "epoch": 3.266396971617033,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1664,
-      "step": 201050
-    },
-    {
-      "epoch": 3.2665594385144026,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1712,
-      "step": 201060
-    },
-    {
-      "epoch": 3.266721905411772,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 201070
-    },
-    {
-      "epoch": 3.266884372309142,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 201080
-    },
-    {
-      "epoch": 3.2670468392065115,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 201090
-    },
-    {
-      "epoch": 3.267209306103881,
-      "grad_norm": 8.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 201100
-    },
-    {
-      "epoch": 3.267371773001251,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 201110
-    },
-    {
-      "epoch": 3.2675342398986205,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 201120
-    },
-    {
-      "epoch": 3.2676967067959906,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 201130
-    },
-    {
-      "epoch": 3.2678591736933598,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 201140
-    },
-    {
-      "epoch": 3.26802164059073,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 201150
-    },
-    {
-      "epoch": 3.268184107488099,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 201160
-    },
-    {
-      "epoch": 3.268346574385469,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 201170
-    },
-    {
-      "epoch": 3.268509041282839,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 201180
-    },
-    {
-      "epoch": 3.2686715081802085,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1644,
-      "step": 201190
-    },
-    {
-      "epoch": 3.268833975077578,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 201200
-    },
-    {
-      "epoch": 3.2689964419749478,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1618,
-      "step": 201210
-    },
-    {
-      "epoch": 3.2691589088723174,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1782,
-      "step": 201220
-    },
-    {
-      "epoch": 3.269321375769687,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.173,
-      "step": 201230
-    },
-    {
-      "epoch": 3.2694838426670567,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 201240
-    },
-    {
-      "epoch": 3.2696463095644264,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 201250
-    },
-    {
-      "epoch": 3.269808776461796,
-      "grad_norm": 3.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1684,
-      "step": 201260
-    },
-    {
-      "epoch": 3.2699712433591657,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 201270
-    },
-    {
-      "epoch": 3.2701337102565353,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1664,
-      "step": 201280
-    },
-    {
-      "epoch": 3.270296177153905,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1804,
-      "step": 201290
-    },
-    {
-      "epoch": 3.2704586440512746,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.164,
-      "step": 201300
-    },
-    {
-      "epoch": 3.2706211109486443,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1726,
-      "step": 201310
-    },
-    {
-      "epoch": 3.270783577846014,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 201320
-    },
-    {
-      "epoch": 3.2709460447433836,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1814,
-      "step": 201330
-    },
-    {
-      "epoch": 3.2711085116407532,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1764,
-      "step": 201340
-    },
-    {
-      "epoch": 3.271270978538123,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1816,
-      "step": 201350
-    },
-    {
-      "epoch": 3.2714334454354925,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1627,
-      "step": 201360
-    },
-    {
-      "epoch": 3.271595912332862,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1671,
-      "step": 201370
-    },
-    {
-      "epoch": 3.271758379230232,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 201380
-    },
-    {
-      "epoch": 3.2719208461276015,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1748,
-      "step": 201390
-    },
-    {
-      "epoch": 3.272083313024971,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1729,
-      "step": 201400
-    },
-    {
-      "epoch": 3.272245779922341,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 201410
-    },
-    {
-      "epoch": 3.2724082468197104,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1662,
-      "step": 201420
-    },
-    {
-      "epoch": 3.27257071371708,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 201430
-    },
-    {
-      "epoch": 3.2727331806144497,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1635,
-      "step": 201440
-    },
-    {
-      "epoch": 3.2728956475118194,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1721,
-      "step": 201450
-    },
-    {
-      "epoch": 3.273058114409189,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 201460
-    },
-    {
-      "epoch": 3.2732205813065587,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 201470
-    },
-    {
-      "epoch": 3.2733830482039283,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 201480
-    },
-    {
-      "epoch": 3.273545515101298,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 201490
-    },
-    {
-      "epoch": 3.2737079819986676,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 201500
-    },
-    {
-      "epoch": 3.2738704488960373,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1726,
-      "step": 201510
-    },
-    {
-      "epoch": 3.274032915793407,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 201520
-    },
-    {
-      "epoch": 3.2741953826907766,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 201530
-    },
-    {
-      "epoch": 3.2743578495881462,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 201540
-    },
-    {
-      "epoch": 3.274520316485516,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 201550
-    },
-    {
-      "epoch": 3.274682783382886,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 201560
-    },
-    {
-      "epoch": 3.274845250280255,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 201570
-    },
-    {
-      "epoch": 3.2750077171776253,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 201580
-    },
-    {
-      "epoch": 3.2751701840749945,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1818,
-      "step": 201590
-    },
-    {
-      "epoch": 3.2753326509723646,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2053,
-      "step": 201600
-    },
-    {
-      "epoch": 3.275495117869734,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2107,
-      "step": 201610
-    },
-    {
-      "epoch": 3.275657584767104,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 201620
-    },
-    {
-      "epoch": 3.2758200516644735,
-      "grad_norm": 3.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 201630
-    },
-    {
-      "epoch": 3.275982518561843,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 201640
-    },
-    {
-      "epoch": 3.276144985459213,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 201650
-    },
-    {
-      "epoch": 3.2763074523565825,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 201660
-    },
-    {
-      "epoch": 3.276469919253952,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 201670
-    },
-    {
-      "epoch": 3.276632386151322,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1787,
-      "step": 201680
-    },
-    {
-      "epoch": 3.2767948530486914,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 201690
-    },
-    {
-      "epoch": 3.276957319946061,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.187,
-      "step": 201700
-    },
-    {
-      "epoch": 3.2771197868434307,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 201710
-    },
-    {
-      "epoch": 3.2772822537408004,
-      "grad_norm": 3.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 201720
-    },
-    {
-      "epoch": 3.27744472063817,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 201730
-    },
-    {
-      "epoch": 3.2776071875355397,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 201740
-    },
-    {
-      "epoch": 3.2777696544329094,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 201750
-    },
-    {
-      "epoch": 3.277932121330279,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2266,
-      "step": 201760
-    },
-    {
-      "epoch": 3.2780945882276487,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 201770
-    },
-    {
-      "epoch": 3.2782570551250183,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2254,
-      "step": 201780
-    },
-    {
-      "epoch": 3.278419522022388,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 201790
-    },
-    {
-      "epoch": 3.2785819889197576,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2101,
-      "step": 201800
-    },
-    {
-      "epoch": 3.2787444558171273,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2172,
-      "step": 201810
-    },
-    {
-      "epoch": 3.278906922714497,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2144,
-      "step": 201820
-    },
-    {
-      "epoch": 3.2790693896118666,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 201830
-    },
-    {
-      "epoch": 3.279231856509236,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 201840
-    },
-    {
-      "epoch": 3.279394323406606,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 201850
-    },
-    {
-      "epoch": 3.2795567903039755,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1718,
-      "step": 201860
-    },
-    {
-      "epoch": 3.279719257201345,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 201870
-    },
-    {
-      "epoch": 3.279881724098715,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1658,
-      "step": 201880
-    },
-    {
-      "epoch": 3.2800441909960845,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2005,
-      "step": 201890
-    },
-    {
-      "epoch": 3.280206657893454,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1721,
-      "step": 201900
-    },
-    {
-      "epoch": 3.2803691247908238,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 201910
-    },
-    {
-      "epoch": 3.2805315916881934,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1683,
-      "step": 201920
-    },
-    {
-      "epoch": 3.280694058585563,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 201930
-    },
-    {
-      "epoch": 3.2808565254829327,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 201940
-    },
-    {
-      "epoch": 3.2810189923803024,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 201950
-    },
-    {
-      "epoch": 3.281181459277672,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1769,
-      "step": 201960
-    },
-    {
-      "epoch": 3.2813439261750417,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 201970
-    },
-    {
-      "epoch": 3.2815063930724113,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 201980
-    },
-    {
-      "epoch": 3.281668859969781,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 201990
-    },
-    {
-      "epoch": 3.2818313268671506,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1871,
-      "step": 202000
-    },
-    {
-      "epoch": 3.2819937937645207,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 202010
-    },
-    {
-      "epoch": 3.28215626066189,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 202020
-    },
-    {
-      "epoch": 3.28231872755926,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 202030
-    },
-    {
-      "epoch": 3.2824811944566292,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1787,
-      "step": 202040
-    },
-    {
-      "epoch": 3.2826436613539993,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 202050
-    },
-    {
-      "epoch": 3.282806128251369,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 202060
-    },
-    {
-      "epoch": 3.2829685951487386,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1863,
-      "step": 202070
-    },
-    {
-      "epoch": 3.2831310620461083,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1723,
-      "step": 202080
-    },
-    {
-      "epoch": 3.283293528943478,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 202090
-    },
-    {
-      "epoch": 3.2834559958408476,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 202100
-    },
-    {
-      "epoch": 3.2836184627382172,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 202110
-    },
-    {
-      "epoch": 3.283780929635587,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 202120
-    },
-    {
-      "epoch": 3.2839433965329565,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 202130
-    },
-    {
-      "epoch": 3.284105863430326,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1782,
-      "step": 202140
-    },
-    {
-      "epoch": 3.284268330327696,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 202150
-    },
-    {
-      "epoch": 3.2844307972250655,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 202160
-    },
-    {
-      "epoch": 3.284593264122435,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 202170
-    },
-    {
-      "epoch": 3.284755731019805,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1866,
-      "step": 202180
-    },
-    {
-      "epoch": 3.2849181979171744,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 202190
-    },
-    {
-      "epoch": 3.285080664814544,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 202200
-    },
-    {
-      "epoch": 3.2852431317119137,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 202210
-    },
-    {
-      "epoch": 3.2854055986092834,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.168,
-      "step": 202220
-    },
-    {
-      "epoch": 3.285568065506653,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 202230
-    },
-    {
-      "epoch": 3.2857305324040227,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1837,
-      "step": 202240
-    },
-    {
-      "epoch": 3.2858929993013923,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 202250
-    },
-    {
-      "epoch": 3.286055466198762,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 202260
-    },
-    {
-      "epoch": 3.2862179330961316,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 202270
-    },
-    {
-      "epoch": 3.2863803999935013,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 202280
-    },
-    {
-      "epoch": 3.286542866890871,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 202290
-    },
-    {
-      "epoch": 3.2867053337882406,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 202300
-    },
-    {
-      "epoch": 3.2868678006856102,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 202310
-    },
-    {
-      "epoch": 3.28703026758298,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 202320
-    },
-    {
-      "epoch": 3.2871927344803495,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 202330
-    },
-    {
-      "epoch": 3.287355201377719,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1809,
-      "step": 202340
-    },
-    {
-      "epoch": 3.287517668275089,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1648,
-      "step": 202350
-    },
-    {
-      "epoch": 3.2876801351724585,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.1421,
-      "step": 202360
-    },
-    {
-      "epoch": 3.287842602069828,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1596,
-      "step": 202370
-    },
-    {
-      "epoch": 3.288005068967198,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1702,
-      "step": 202380
-    },
-    {
-      "epoch": 3.2881675358645674,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.169,
-      "step": 202390
-    },
-    {
-      "epoch": 3.288330002761937,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1758,
-      "step": 202400
-    },
-    {
-      "epoch": 3.2884924696593067,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 202410
-    },
-    {
-      "epoch": 3.2886549365566764,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 202420
-    },
-    {
-      "epoch": 3.288817403454046,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 202430
-    },
-    {
-      "epoch": 3.288979870351416,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1688,
-      "step": 202440
-    },
-    {
-      "epoch": 3.2891423372487854,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 202450
-    },
-    {
-      "epoch": 3.2893048041461554,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 202460
-    },
-    {
-      "epoch": 3.2894672710435247,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 202470
-    },
-    {
-      "epoch": 3.2896297379408947,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.191,
-      "step": 202480
-    },
-    {
-      "epoch": 3.289792204838264,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 202490
-    },
-    {
-      "epoch": 3.289954671735634,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 202500
-    },
-    {
-      "epoch": 3.2901171386330037,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 202510
-    },
-    {
-      "epoch": 3.2902796055303734,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 202520
-    },
-    {
-      "epoch": 3.290442072427743,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 202530
-    },
-    {
-      "epoch": 3.2906045393251127,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1885,
-      "step": 202540
-    },
-    {
-      "epoch": 3.2907670062224823,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 202550
-    },
-    {
-      "epoch": 3.290929473119852,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 202560
-    },
-    {
-      "epoch": 3.2910919400172216,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.186,
-      "step": 202570
-    },
-    {
-      "epoch": 3.2912544069145913,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 202580
-    },
-    {
-      "epoch": 3.291416873811961,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 202590
-    },
-    {
-      "epoch": 3.2915793407093306,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 202600
-    },
-    {
-      "epoch": 3.2917418076067,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 202610
-    },
-    {
-      "epoch": 3.29190427450407,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1751,
-      "step": 202620
-    },
-    {
-      "epoch": 3.2920667414014395,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.188,
-      "step": 202630
-    },
-    {
-      "epoch": 3.292229208298809,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 202640
-    },
-    {
-      "epoch": 3.292391675196179,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 202650
-    },
-    {
-      "epoch": 3.2925541420935485,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 202660
-    },
-    {
-      "epoch": 3.292716608990918,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 202670
-    },
-    {
-      "epoch": 3.2928790758882878,
-      "grad_norm": 7.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 202680
-    },
-    {
-      "epoch": 3.2930415427856574,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 202690
-    },
-    {
-      "epoch": 3.293204009683027,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 202700
-    },
-    {
-      "epoch": 3.2933664765803967,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 202710
-    },
-    {
-      "epoch": 3.2935289434777664,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 202720
-    },
-    {
-      "epoch": 3.293691410375136,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 202730
-    },
-    {
-      "epoch": 3.2938538772725057,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 202740
-    },
-    {
-      "epoch": 3.2940163441698753,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1727,
-      "step": 202750
-    },
-    {
-      "epoch": 3.294178811067245,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 202760
-    },
-    {
-      "epoch": 3.2943412779646146,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1514,
-      "step": 202770
-    },
-    {
-      "epoch": 3.2945037448619843,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1591,
-      "step": 202780
-    },
-    {
-      "epoch": 3.294666211759354,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.167,
-      "step": 202790
-    },
-    {
-      "epoch": 3.2948286786567236,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1529,
-      "step": 202800
-    },
-    {
-      "epoch": 3.2949911455540932,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1607,
-      "step": 202810
-    },
-    {
-      "epoch": 3.295153612451463,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1536,
-      "step": 202820
-    },
-    {
-      "epoch": 3.2953160793488325,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 202830
-    },
-    {
-      "epoch": 3.295478546246202,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 202840
-    },
-    {
-      "epoch": 3.295641013143572,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 202850
-    },
-    {
-      "epoch": 3.2958034800409415,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 202860
-    },
-    {
-      "epoch": 3.295965946938311,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 202870
-    },
-    {
-      "epoch": 3.296128413835681,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 202880
-    },
-    {
-      "epoch": 3.296290880733051,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 202890
-    },
-    {
-      "epoch": 3.29645334763042,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 202900
-    },
-    {
-      "epoch": 3.29661581452779,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 202910
-    },
-    {
-      "epoch": 3.2967782814251594,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 202920
-    },
-    {
-      "epoch": 3.2969407483225295,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 202930
-    },
-    {
-      "epoch": 3.297103215219899,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 202940
-    },
-    {
-      "epoch": 3.297265682117269,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 202950
-    },
-    {
-      "epoch": 3.2974281490146384,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 202960
-    },
-    {
-      "epoch": 3.297590615912008,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1639,
-      "step": 202970
-    },
-    {
-      "epoch": 3.2977530828093777,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 202980
-    },
-    {
-      "epoch": 3.2979155497067474,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 202990
-    },
-    {
-      "epoch": 3.298078016604117,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 203000
-    },
-    {
-      "epoch": 3.2982404835014867,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 203010
-    },
-    {
-      "epoch": 3.2984029503988563,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 203020
-    },
-    {
-      "epoch": 3.298565417296226,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 203030
-    },
-    {
-      "epoch": 3.2987278841935956,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2033,
-      "step": 203040
-    },
-    {
-      "epoch": 3.2988903510909653,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1728,
-      "step": 203050
-    },
-    {
-      "epoch": 3.299052817988335,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 203060
-    },
-    {
-      "epoch": 3.2992152848857046,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 203070
-    },
-    {
-      "epoch": 3.2993777517830742,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1652,
-      "step": 203080
-    },
-    {
-      "epoch": 3.299540218680444,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 203090
-    },
-    {
-      "epoch": 3.2997026855778135,
-      "grad_norm": 7.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1649,
-      "step": 203100
-    },
-    {
-      "epoch": 3.299865152475183,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 203110
-    },
-    {
-      "epoch": 3.300027619372553,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 203120
-    },
-    {
-      "epoch": 3.3001900862699225,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 203130
-    },
-    {
-      "epoch": 3.300352553167292,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 203140
-    },
-    {
-      "epoch": 3.300515020064662,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1748,
-      "step": 203150
-    },
-    {
-      "epoch": 3.3006774869620314,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1693,
-      "step": 203160
-    },
-    {
-      "epoch": 3.300839953859401,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 203170
-    },
-    {
-      "epoch": 3.3010024207567707,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 203180
-    },
-    {
-      "epoch": 3.3011648876541404,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 203190
-    },
-    {
-      "epoch": 3.30132735455151,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 203200
-    },
-    {
-      "epoch": 3.3014898214488797,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 203210
-    },
-    {
-      "epoch": 3.3016522883462494,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 203220
-    },
-    {
-      "epoch": 3.301814755243619,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 203230
-    },
-    {
-      "epoch": 3.3019772221409887,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1579,
-      "step": 203240
-    },
-    {
-      "epoch": 3.3021396890383583,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1692,
-      "step": 203250
-    },
-    {
-      "epoch": 3.302302155935728,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 203260
-    },
-    {
-      "epoch": 3.3024646228330976,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.159,
-      "step": 203270
-    },
-    {
-      "epoch": 3.3026270897304673,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 203280
-    },
-    {
-      "epoch": 3.302789556627837,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 203290
-    },
-    {
-      "epoch": 3.3029520235252066,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 203300
-    },
-    {
-      "epoch": 3.303114490422576,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 203310
-    },
-    {
-      "epoch": 3.3032769573199463,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 203320
-    },
-    {
-      "epoch": 3.3034394242173155,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 203330
-    },
-    {
-      "epoch": 3.3036018911146856,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 203340
-    },
-    {
-      "epoch": 3.303764358012055,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.217,
-      "step": 203350
-    },
-    {
-      "epoch": 3.303926824909425,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 203360
-    },
-    {
-      "epoch": 3.3040892918067946,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 203370
-    },
-    {
-      "epoch": 3.304251758704164,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 203380
-    },
-    {
-      "epoch": 3.304414225601534,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 203390
-    },
-    {
-      "epoch": 3.3045766924989035,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 203400
-    },
-    {
-      "epoch": 3.304739159396273,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2141,
-      "step": 203410
-    },
-    {
-      "epoch": 3.304901626293643,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 203420
-    },
-    {
-      "epoch": 3.3050640931910125,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2268,
-      "step": 203430
-    },
-    {
-      "epoch": 3.305226560088382,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2179,
-      "step": 203440
-    },
-    {
-      "epoch": 3.3053890269857518,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 203450
-    },
-    {
-      "epoch": 3.3055514938831214,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1958,
-      "step": 203460
-    },
-    {
-      "epoch": 3.305713960780491,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1901,
-      "step": 203470
-    },
-    {
-      "epoch": 3.3058764276778607,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 203480
-    },
-    {
-      "epoch": 3.3060388945752304,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1717,
-      "step": 203490
-    },
-    {
-      "epoch": 3.3062013614726,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1645,
-      "step": 203500
-    },
-    {
-      "epoch": 3.3063638283699697,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1515,
-      "step": 203510
-    },
-    {
-      "epoch": 3.3065262952673393,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1767,
-      "step": 203520
-    },
-    {
-      "epoch": 3.306688762164709,
-      "grad_norm": 3.703125,
-      "learning_rate": 5e-05,
-      "loss": 0.1685,
-      "step": 203530
-    },
-    {
-      "epoch": 3.3068512290620786,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 203540
-    },
-    {
-      "epoch": 3.3070136959594483,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 203550
-    },
-    {
-      "epoch": 3.307176162856818,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1611,
-      "step": 203560
-    },
-    {
-      "epoch": 3.3073386297541876,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1709,
-      "step": 203570
-    },
-    {
-      "epoch": 3.3075010966515572,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1685,
-      "step": 203580
-    },
-    {
-      "epoch": 3.307663563548927,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1658,
-      "step": 203590
-    },
-    {
-      "epoch": 3.3078260304462965,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 203600
-    },
-    {
-      "epoch": 3.307988497343666,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 203610
-    },
-    {
-      "epoch": 3.308150964241036,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 203620
-    },
-    {
-      "epoch": 3.3083134311384055,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2142,
-      "step": 203630
-    },
-    {
-      "epoch": 3.308475898035775,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 203640
-    },
-    {
-      "epoch": 3.308638364933145,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 203650
-    },
-    {
-      "epoch": 3.3088008318305144,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 203660
-    },
-    {
-      "epoch": 3.308963298727884,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 203670
-    },
-    {
-      "epoch": 3.3091257656252537,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1851,
-      "step": 203680
-    },
-    {
-      "epoch": 3.3092882325226234,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1736,
-      "step": 203690
-    },
-    {
-      "epoch": 3.309450699419993,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1769,
-      "step": 203700
-    },
-    {
-      "epoch": 3.3096131663173627,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.181,
-      "step": 203710
-    },
-    {
-      "epoch": 3.3097756332147323,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 203720
-    },
-    {
-      "epoch": 3.309938100112102,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2086,
-      "step": 203730
-    },
-    {
-      "epoch": 3.3101005670094716,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 203740
-    },
-    {
-      "epoch": 3.3102630339068413,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 203750
-    },
-    {
-      "epoch": 3.310425500804211,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 203760
-    },
-    {
-      "epoch": 3.310587967701581,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 203770
-    },
-    {
-      "epoch": 3.3107504345989502,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 203780
-    },
-    {
-      "epoch": 3.3109129014963203,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.183,
-      "step": 203790
-    },
-    {
-      "epoch": 3.3110753683936895,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2109,
-      "step": 203800
-    },
-    {
-      "epoch": 3.3112378352910596,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 203810
-    },
-    {
-      "epoch": 3.3114003021884293,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2206,
-      "step": 203820
-    },
-    {
-      "epoch": 3.311562769085799,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 203830
-    },
-    {
-      "epoch": 3.3117252359831686,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 203840
-    },
-    {
-      "epoch": 3.3118877028805382,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.199,
-      "step": 203850
-    },
-    {
-      "epoch": 3.312050169777908,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 203860
-    },
-    {
-      "epoch": 3.3122126366752775,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2154,
-      "step": 203870
-    },
-    {
-      "epoch": 3.312375103572647,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 203880
-    },
-    {
-      "epoch": 3.312537570470017,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2281,
-      "step": 203890
-    },
-    {
-      "epoch": 3.3127000373673865,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 203900
-    },
-    {
-      "epoch": 3.312862504264756,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 203910
-    },
-    {
-      "epoch": 3.313024971162126,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 203920
-    },
-    {
-      "epoch": 3.3131874380594954,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1657,
-      "step": 203930
-    },
-    {
-      "epoch": 3.313349904956865,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 203940
-    },
-    {
-      "epoch": 3.3135123718542348,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2011,
-      "step": 203950
-    },
-    {
-      "epoch": 3.3136748387516044,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 203960
-    },
-    {
-      "epoch": 3.313837305648974,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 203970
-    },
-    {
-      "epoch": 3.3139997725463437,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 203980
-    },
-    {
-      "epoch": 3.3141622394437134,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.184,
-      "step": 203990
-    },
-    {
-      "epoch": 3.314324706341083,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1724,
-      "step": 204000
-    },
-    {
-      "epoch": 3.3144871732384527,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1851,
-      "step": 204010
-    },
-    {
-      "epoch": 3.3146496401358223,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 204020
-    },
-    {
-      "epoch": 3.314812107033192,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1699,
-      "step": 204030
-    },
-    {
-      "epoch": 3.3149745739305616,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 204040
-    },
-    {
-      "epoch": 3.3151370408279313,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 204050
-    },
-    {
-      "epoch": 3.315299507725301,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 204060
-    },
-    {
-      "epoch": 3.3154619746226706,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 204070
-    },
-    {
-      "epoch": 3.31562444152004,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 204080
-    },
-    {
-      "epoch": 3.31578690841741,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 204090
-    },
-    {
-      "epoch": 3.3159493753147795,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 204100
-    },
-    {
-      "epoch": 3.316111842212149,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 204110
-    },
-    {
-      "epoch": 3.316274309109519,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 204120
-    },
-    {
-      "epoch": 3.3164367760068885,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 204130
-    },
-    {
-      "epoch": 3.316599242904258,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 204140
-    },
-    {
-      "epoch": 3.3167617098016278,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 204150
-    },
-    {
-      "epoch": 3.3169241766989974,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 204160
-    },
-    {
-      "epoch": 3.317086643596367,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2419,
-      "step": 204170
-    },
-    {
-      "epoch": 3.3172491104937367,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.216,
-      "step": 204180
-    },
-    {
-      "epoch": 3.3174115773911064,
-      "grad_norm": 3.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 204190
-    },
-    {
-      "epoch": 3.3175740442884765,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 204200
-    },
-    {
-      "epoch": 3.3177365111858457,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 204210
-    },
-    {
-      "epoch": 3.3178989780832158,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2072,
-      "step": 204220
-    },
-    {
-      "epoch": 3.318061444980585,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 204230
-    },
-    {
-      "epoch": 3.318223911877955,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 204240
-    },
-    {
-      "epoch": 3.3183863787753247,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 204250
-    },
-    {
-      "epoch": 3.3185488456726944,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 204260
-    },
-    {
-      "epoch": 3.318711312570064,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1959,
-      "step": 204270
-    },
-    {
-      "epoch": 3.3188737794674337,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2204,
-      "step": 204280
-    },
-    {
-      "epoch": 3.3190362463648033,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 204290
-    },
-    {
-      "epoch": 3.319198713262173,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 204300
-    },
-    {
-      "epoch": 3.3193611801595426,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 204310
-    },
-    {
-      "epoch": 3.3195236470569123,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.224,
-      "step": 204320
-    },
-    {
-      "epoch": 3.319686113954282,
-      "grad_norm": 7.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 204330
-    },
-    {
-      "epoch": 3.3198485808516516,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 204340
-    },
-    {
-      "epoch": 3.3200110477490212,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2059,
-      "step": 204350
-    },
-    {
-      "epoch": 3.320173514646391,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 204360
-    },
-    {
-      "epoch": 3.3203359815437605,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2169,
-      "step": 204370
-    },
-    {
-      "epoch": 3.32049844844113,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 204380
-    },
-    {
-      "epoch": 3.3206609153385,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 204390
-    },
-    {
-      "epoch": 3.3208233822358695,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 204400
-    },
-    {
-      "epoch": 3.320985849133239,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 204410
-    },
-    {
-      "epoch": 3.321148316030609,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 204420
-    },
-    {
-      "epoch": 3.3213107829279784,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 204430
-    },
-    {
-      "epoch": 3.321473249825348,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 204440
-    },
-    {
-      "epoch": 3.3216357167227177,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 204450
-    },
-    {
-      "epoch": 3.3217981836200874,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 204460
-    },
-    {
-      "epoch": 3.321960650517457,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 204470
-    },
-    {
-      "epoch": 3.3221231174148267,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 204480
-    },
-    {
-      "epoch": 3.3222855843121963,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 204490
-    },
-    {
-      "epoch": 3.322448051209566,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 204500
-    },
-    {
-      "epoch": 3.3226105181069356,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 204510
-    },
-    {
-      "epoch": 3.3227729850043053,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1625,
-      "step": 204520
-    },
-    {
-      "epoch": 3.322935451901675,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 204530
-    },
-    {
-      "epoch": 3.3230979187990446,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 204540
-    },
-    {
-      "epoch": 3.3232603856964142,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 204550
-    },
-    {
-      "epoch": 3.323422852593784,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 204560
-    },
-    {
-      "epoch": 3.3235853194911535,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 204570
-    },
-    {
-      "epoch": 3.323747786388523,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2249,
-      "step": 204580
-    },
-    {
-      "epoch": 3.323910253285893,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 204590
-    },
-    {
-      "epoch": 3.3240727201832625,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 204600
-    },
-    {
-      "epoch": 3.324235187080632,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 204610
-    },
-    {
-      "epoch": 3.324397653978002,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 204620
-    },
-    {
-      "epoch": 3.3245601208753714,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 204630
-    },
-    {
-      "epoch": 3.324722587772741,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 204640
-    },
-    {
-      "epoch": 3.324885054670111,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2131,
-      "step": 204650
-    },
-    {
-      "epoch": 3.3250475215674804,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 204660
-    },
-    {
-      "epoch": 3.3252099884648505,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 204670
-    },
-    {
-      "epoch": 3.3253724553622197,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 204680
-    },
-    {
-      "epoch": 3.32553492225959,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1756,
-      "step": 204690
-    },
-    {
-      "epoch": 3.3256973891569594,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1719,
-      "step": 204700
-    },
-    {
-      "epoch": 3.325859856054329,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 204710
-    },
-    {
-      "epoch": 3.3260223229516988,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 204720
-    },
-    {
-      "epoch": 3.3261847898490684,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1824,
-      "step": 204730
-    },
-    {
-      "epoch": 3.326347256746438,
-      "grad_norm": 3.796875,
-      "learning_rate": 5e-05,
-      "loss": 0.1664,
-      "step": 204740
-    },
-    {
-      "epoch": 3.3265097236438077,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1581,
-      "step": 204750
-    },
-    {
-      "epoch": 3.3266721905411774,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 204760
-    },
-    {
-      "epoch": 3.326834657438547,
-      "grad_norm": 7.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 204770
-    },
-    {
-      "epoch": 3.3269971243359167,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1826,
-      "step": 204780
-    },
-    {
-      "epoch": 3.3271595912332863,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 204790
-    },
-    {
-      "epoch": 3.327322058130656,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 204800
-    },
-    {
-      "epoch": 3.3274845250280256,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 204810
-    },
-    {
-      "epoch": 3.3276469919253953,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 204820
-    },
-    {
-      "epoch": 3.327809458822765,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 204830
-    },
-    {
-      "epoch": 3.3279719257201346,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 204840
-    },
-    {
-      "epoch": 3.328134392617504,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 204850
-    },
-    {
-      "epoch": 3.328296859514874,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1646,
-      "step": 204860
-    },
-    {
-      "epoch": 3.3284593264122435,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1719,
-      "step": 204870
-    },
-    {
-      "epoch": 3.328621793309613,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1656,
-      "step": 204880
-    },
-    {
-      "epoch": 3.328784260206983,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.17,
-      "step": 204890
-    },
-    {
-      "epoch": 3.3289467271043525,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1639,
-      "step": 204900
-    },
-    {
-      "epoch": 3.329109194001722,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1552,
-      "step": 204910
-    },
-    {
-      "epoch": 3.3292716608990918,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1829,
-      "step": 204920
-    },
-    {
-      "epoch": 3.3294341277964614,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1754,
-      "step": 204930
-    },
-    {
-      "epoch": 3.329596594693831,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2011,
-      "step": 204940
-    },
-    {
-      "epoch": 3.3297590615912007,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 204950
-    },
-    {
-      "epoch": 3.3299215284885704,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.221,
-      "step": 204960
-    },
-    {
-      "epoch": 3.33008399538594,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 204970
-    },
-    {
-      "epoch": 3.3302464622833097,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 204980
-    },
-    {
-      "epoch": 3.3304089291806793,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 204990
-    },
-    {
-      "epoch": 3.330571396078049,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 205000
-    },
-    {
-      "epoch": 3.3307338629754186,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.179,
-      "step": 205010
-    },
-    {
-      "epoch": 3.3308963298727883,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1715,
-      "step": 205020
-    },
-    {
-      "epoch": 3.331058796770158,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1578,
-      "step": 205030
-    },
-    {
-      "epoch": 3.3312212636675276,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1894,
-      "step": 205040
-    },
-    {
-      "epoch": 3.3313837305648972,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2107,
-      "step": 205050
-    },
-    {
-      "epoch": 3.331546197462267,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 205060
-    },
-    {
-      "epoch": 3.3317086643596365,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2067,
-      "step": 205070
-    },
-    {
-      "epoch": 3.3318711312570066,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.217,
-      "step": 205080
-    },
-    {
-      "epoch": 3.332033598154376,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 205090
-    },
-    {
-      "epoch": 3.332196065051746,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1875,
-      "step": 205100
-    },
-    {
-      "epoch": 3.332358531949115,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 205110
-    },
-    {
-      "epoch": 3.3325209988464852,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 205120
-    },
-    {
-      "epoch": 3.332683465743855,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 205130
-    },
-    {
-      "epoch": 3.3328459326412245,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 205140
-    },
-    {
-      "epoch": 3.333008399538594,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.1896,
-      "step": 205150
-    },
-    {
-      "epoch": 3.333170866435964,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 205160
-    },
-    {
-      "epoch": 3.3333333333333335,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1697,
-      "step": 205170
-    },
-    {
-      "epoch": 3.333495800230703,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1545,
-      "step": 205180
-    },
-    {
-      "epoch": 3.333658267128073,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1677,
-      "step": 205190
-    },
-    {
-      "epoch": 3.3338207340254424,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 205200
-    },
-    {
-      "epoch": 3.333983200922812,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 205210
-    },
-    {
-      "epoch": 3.3341456678201817,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 205220
-    },
-    {
-      "epoch": 3.3343081347175514,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 205230
-    },
-    {
-      "epoch": 3.334470601614921,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 205240
-    },
-    {
-      "epoch": 3.3346330685122907,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 205250
-    },
-    {
-      "epoch": 3.3347955354096603,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 205260
-    },
-    {
-      "epoch": 3.33495800230703,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 205270
-    },
-    {
-      "epoch": 3.3351204692043996,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 205280
-    },
-    {
-      "epoch": 3.3352829361017693,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2139,
-      "step": 205290
-    },
-    {
-      "epoch": 3.335445402999139,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2383,
-      "step": 205300
-    },
-    {
-      "epoch": 3.3356078698965086,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 205310
-    },
-    {
-      "epoch": 3.3357703367938782,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1939,
-      "step": 205320
-    },
-    {
-      "epoch": 3.335932803691248,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 205330
-    },
-    {
-      "epoch": 3.3360952705886175,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 205340
-    },
-    {
-      "epoch": 3.336257737485987,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 205350
-    },
-    {
-      "epoch": 3.336420204383357,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2076,
-      "step": 205360
-    },
-    {
-      "epoch": 3.3365826712807265,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 205370
-    },
-    {
-      "epoch": 3.336745138178096,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2144,
-      "step": 205380
-    },
-    {
-      "epoch": 3.336907605075466,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 205390
-    },
-    {
-      "epoch": 3.3370700719728354,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2155,
-      "step": 205400
-    },
-    {
-      "epoch": 3.337232538870205,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 205410
-    },
-    {
-      "epoch": 3.3373950057675748,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 205420
-    },
-    {
-      "epoch": 3.3375574726649444,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 205430
-    },
-    {
-      "epoch": 3.337719939562314,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 205440
-    },
-    {
-      "epoch": 3.3378824064596837,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 205450
-    },
-    {
-      "epoch": 3.3380448733570534,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2141,
-      "step": 205460
-    },
-    {
-      "epoch": 3.338207340254423,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 205470
-    },
-    {
-      "epoch": 3.3383698071517927,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.178,
-      "step": 205480
-    },
-    {
-      "epoch": 3.3385322740491623,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 205490
-    },
-    {
-      "epoch": 3.338694740946532,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 205500
-    },
-    {
-      "epoch": 3.338857207843902,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 205510
-    },
-    {
-      "epoch": 3.3390196747412713,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1653,
-      "step": 205520
-    },
-    {
-      "epoch": 3.3391821416386414,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 205530
-    },
-    {
-      "epoch": 3.3393446085360106,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 205540
-    },
-    {
-      "epoch": 3.3395070754333807,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 205550
-    },
-    {
-      "epoch": 3.33966954233075,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 205560
-    },
-    {
-      "epoch": 3.33983200922812,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 205570
-    },
-    {
-      "epoch": 3.3399944761254896,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 205580
-    },
-    {
-      "epoch": 3.3401569430228593,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 205590
-    },
-    {
-      "epoch": 3.340319409920229,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 205600
-    },
-    {
-      "epoch": 3.3404818768175986,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 205610
-    },
-    {
-      "epoch": 3.340644343714968,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 205620
-    },
-    {
-      "epoch": 3.340806810612338,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 205630
-    },
-    {
-      "epoch": 3.3409692775097075,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1744,
-      "step": 205640
-    },
-    {
-      "epoch": 3.341131744407077,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 205650
-    },
-    {
-      "epoch": 3.341294211304447,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 205660
-    },
-    {
-      "epoch": 3.3414566782018165,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 205670
-    },
-    {
-      "epoch": 3.341619145099186,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1723,
-      "step": 205680
-    },
-    {
-      "epoch": 3.3417816119965558,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 205690
-    },
-    {
-      "epoch": 3.3419440788939254,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1877,
-      "step": 205700
-    },
-    {
-      "epoch": 3.342106545791295,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.149,
-      "step": 205710
-    },
-    {
-      "epoch": 3.3422690126886647,
-      "grad_norm": 7.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1626,
-      "step": 205720
-    },
-    {
-      "epoch": 3.3424314795860344,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 205730
-    },
-    {
-      "epoch": 3.342593946483404,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 205740
-    },
-    {
-      "epoch": 3.3427564133807737,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1861,
-      "step": 205750
-    },
-    {
-      "epoch": 3.3429188802781433,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 205760
-    },
-    {
-      "epoch": 3.343081347175513,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1989,
-      "step": 205770
-    },
-    {
-      "epoch": 3.3432438140728826,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 205780
-    },
-    {
-      "epoch": 3.3434062809702523,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 205790
-    },
-    {
-      "epoch": 3.343568747867622,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 205800
-    },
-    {
-      "epoch": 3.3437312147649916,
-      "grad_norm": 7.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1786,
-      "step": 205810
-    },
-    {
-      "epoch": 3.3438936816623612,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 205820
-    },
-    {
-      "epoch": 3.344056148559731,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1712,
-      "step": 205830
-    },
-    {
-      "epoch": 3.3442186154571005,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 205840
-    },
-    {
-      "epoch": 3.34438108235447,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1697,
-      "step": 205850
-    },
-    {
-      "epoch": 3.34454354925184,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 205860
-    },
-    {
-      "epoch": 3.3447060161492095,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1989,
-      "step": 205870
-    },
-    {
-      "epoch": 3.344868483046579,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.176,
-      "step": 205880
-    },
-    {
-      "epoch": 3.345030949943949,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 205890
-    },
-    {
-      "epoch": 3.3451934168413184,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2008,
-      "step": 205900
-    },
-    {
-      "epoch": 3.345355883738688,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 205910
-    },
-    {
-      "epoch": 3.3455183506360577,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 205920
-    },
-    {
-      "epoch": 3.3456808175334274,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 205930
-    },
-    {
-      "epoch": 3.345843284430797,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1763,
-      "step": 205940
-    },
-    {
-      "epoch": 3.3460057513281667,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 205950
-    },
-    {
-      "epoch": 3.346168218225537,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 205960
-    },
-    {
-      "epoch": 3.346330685122906,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2044,
-      "step": 205970
-    },
-    {
-      "epoch": 3.346493152020276,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1831,
-      "step": 205980
-    },
-    {
-      "epoch": 3.3466556189176453,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 205990
-    },
-    {
-      "epoch": 3.3468180858150154,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 206000
-    },
-    {
-      "epoch": 3.346980552712385,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 206010
-    },
-    {
-      "epoch": 3.3471430196097547,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 206020
-    },
-    {
-      "epoch": 3.3473054865071243,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 206030
-    },
-    {
-      "epoch": 3.347467953404494,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 206040
-    },
-    {
-      "epoch": 3.3476304203018636,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 206050
-    },
-    {
-      "epoch": 3.3477928871992333,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 206060
-    },
-    {
-      "epoch": 3.347955354096603,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 206070
-    },
-    {
-      "epoch": 3.3481178209939726,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 206080
-    },
-    {
-      "epoch": 3.3482802878913422,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 206090
-    },
-    {
-      "epoch": 3.348442754788712,
-      "grad_norm": 3.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 206100
-    },
-    {
-      "epoch": 3.3486052216860815,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 206110
-    },
-    {
-      "epoch": 3.348767688583451,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2126,
-      "step": 206120
-    },
-    {
-      "epoch": 3.348930155480821,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2097,
-      "step": 206130
-    },
-    {
-      "epoch": 3.3490926223781905,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 206140
-    },
-    {
-      "epoch": 3.34925508927556,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 206150
-    },
-    {
-      "epoch": 3.34941755617293,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2312,
-      "step": 206160
-    },
-    {
-      "epoch": 3.3495800230702995,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 206170
-    },
-    {
-      "epoch": 3.349742489967669,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1733,
-      "step": 206180
-    },
-    {
-      "epoch": 3.3499049568650388,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1723,
-      "step": 206190
-    },
-    {
-      "epoch": 3.3500674237624084,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 206200
-    },
-    {
-      "epoch": 3.350229890659778,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 206210
-    },
-    {
-      "epoch": 3.3503923575571477,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 206220
-    },
-    {
-      "epoch": 3.3505548244545174,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 206230
-    },
-    {
-      "epoch": 3.350717291351887,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1732,
-      "step": 206240
-    },
-    {
-      "epoch": 3.3508797582492567,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.177,
-      "step": 206250
-    },
-    {
-      "epoch": 3.3510422251466263,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 206260
-    },
-    {
-      "epoch": 3.351204692043996,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 206270
-    },
-    {
-      "epoch": 3.3513671589413656,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 206280
-    },
-    {
-      "epoch": 3.3515296258387353,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1638,
-      "step": 206290
-    },
-    {
-      "epoch": 3.351692092736105,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 206300
-    },
-    {
-      "epoch": 3.3518545596334746,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 206310
-    },
-    {
-      "epoch": 3.352017026530844,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1795,
-      "step": 206320
-    },
-    {
-      "epoch": 3.352179493428214,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1839,
-      "step": 206330
-    },
-    {
-      "epoch": 3.3523419603255835,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 206340
-    },
-    {
-      "epoch": 3.352504427222953,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 206350
-    },
-    {
-      "epoch": 3.352666894120323,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 206360
-    },
-    {
-      "epoch": 3.3528293610176925,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 206370
-    },
-    {
-      "epoch": 3.352991827915062,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 206380
-    },
-    {
-      "epoch": 3.353154294812432,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 206390
-    },
-    {
-      "epoch": 3.3533167617098014,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1866,
-      "step": 206400
-    },
-    {
-      "epoch": 3.3534792286071715,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 206410
-    },
-    {
-      "epoch": 3.3536416955045407,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1721,
-      "step": 206420
-    },
-    {
-      "epoch": 3.353804162401911,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 206430
-    },
-    {
-      "epoch": 3.35396662929928,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 206440
-    },
-    {
-      "epoch": 3.35412909619665,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2042,
-      "step": 206450
-    },
-    {
-      "epoch": 3.3542915630940198,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1694,
-      "step": 206460
-    },
-    {
-      "epoch": 3.3544540299913894,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 206470
-    },
-    {
-      "epoch": 3.354616496888759,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 206480
-    },
-    {
-      "epoch": 3.3547789637861287,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 206490
-    },
-    {
-      "epoch": 3.3549414306834984,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 206500
-    },
-    {
-      "epoch": 3.355103897580868,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 206510
-    },
-    {
-      "epoch": 3.3552663644782377,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 206520
-    },
-    {
-      "epoch": 3.3554288313756073,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1828,
-      "step": 206530
-    },
-    {
-      "epoch": 3.355591298272977,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1795,
-      "step": 206540
-    },
-    {
-      "epoch": 3.3557537651703466,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1662,
-      "step": 206550
-    },
-    {
-      "epoch": 3.3559162320677163,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1735,
-      "step": 206560
-    },
-    {
-      "epoch": 3.356078698965086,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 206570
-    },
-    {
-      "epoch": 3.3562411658624556,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1751,
-      "step": 206580
-    },
-    {
-      "epoch": 3.3564036327598252,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 206590
-    },
-    {
-      "epoch": 3.356566099657195,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 206600
-    },
-    {
-      "epoch": 3.3567285665545645,
-      "grad_norm": 7.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 206610
-    },
-    {
-      "epoch": 3.356891033451934,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 206620
-    },
-    {
-      "epoch": 3.357053500349304,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.191,
-      "step": 206630
-    },
-    {
-      "epoch": 3.3572159672466735,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 206640
-    },
-    {
-      "epoch": 3.357378434144043,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 206650
-    },
-    {
-      "epoch": 3.357540901041413,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 206660
-    },
-    {
-      "epoch": 3.3577033679387824,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 206670
-    },
-    {
-      "epoch": 3.357865834836152,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 206680
-    },
-    {
-      "epoch": 3.3580283017335217,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 206690
-    },
-    {
-      "epoch": 3.3581907686308914,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2203,
-      "step": 206700
-    },
-    {
-      "epoch": 3.358353235528261,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 206710
-    },
-    {
-      "epoch": 3.3585157024256307,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 206720
-    },
-    {
-      "epoch": 3.3586781693230003,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 206730
-    },
-    {
-      "epoch": 3.35884063622037,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 206740
-    },
-    {
-      "epoch": 3.3590031031177396,
-      "grad_norm": 7.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 206750
-    },
-    {
-      "epoch": 3.3591655700151093,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 206760
-    },
-    {
-      "epoch": 3.359328036912479,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 206770
-    },
-    {
-      "epoch": 3.3594905038098486,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 206780
-    },
-    {
-      "epoch": 3.3596529707072182,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 206790
-    },
-    {
-      "epoch": 3.359815437604588,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 206800
-    },
-    {
-      "epoch": 3.3599779045019575,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 206810
-    },
-    {
-      "epoch": 3.360140371399327,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 206820
-    },
-    {
-      "epoch": 3.360302838296697,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 206830
-    },
-    {
-      "epoch": 3.360465305194067,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 206840
-    },
-    {
-      "epoch": 3.360627772091436,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 206850
-    },
-    {
-      "epoch": 3.3607902389888062,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 206860
-    },
-    {
-      "epoch": 3.3609527058861755,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 206870
-    },
-    {
-      "epoch": 3.3611151727835455,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 206880
-    },
-    {
-      "epoch": 3.361277639680915,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.179,
-      "step": 206890
-    },
-    {
-      "epoch": 3.361440106578285,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2067,
-      "step": 206900
-    },
-    {
-      "epoch": 3.3616025734756545,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 206910
-    },
-    {
-      "epoch": 3.361765040373024,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 206920
-    },
-    {
-      "epoch": 3.361927507270394,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 206930
-    },
-    {
-      "epoch": 3.3620899741677635,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 206940
-    },
-    {
-      "epoch": 3.362252441065133,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 206950
-    },
-    {
-      "epoch": 3.3624149079625028,
-      "grad_norm": 8.0,
-      "learning_rate": 5e-05,
-      "loss": 0.21,
-      "step": 206960
-    },
-    {
-      "epoch": 3.3625773748598724,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 206970
-    },
-    {
-      "epoch": 3.362739841757242,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2021,
-      "step": 206980
-    },
-    {
-      "epoch": 3.3629023086546117,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 206990
-    },
-    {
-      "epoch": 3.3630647755519814,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 207000
-    },
-    {
-      "epoch": 3.363227242449351,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 207010
-    },
-    {
-      "epoch": 3.3633897093467207,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 207020
-    },
-    {
-      "epoch": 3.3635521762440903,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2066,
-      "step": 207030
-    },
-    {
-      "epoch": 3.36371464314146,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 207040
-    },
-    {
-      "epoch": 3.3638771100388296,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 207050
-    },
-    {
-      "epoch": 3.3640395769361993,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1871,
-      "step": 207060
-    },
-    {
-      "epoch": 3.364202043833569,
-      "grad_norm": 9.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2084,
-      "step": 207070
-    },
-    {
-      "epoch": 3.3643645107309386,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 207080
-    },
-    {
-      "epoch": 3.364526977628308,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 207090
-    },
-    {
-      "epoch": 3.364689444525678,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 207100
-    },
-    {
-      "epoch": 3.3648519114230475,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 207110
-    },
-    {
-      "epoch": 3.365014378320417,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 207120
-    },
-    {
-      "epoch": 3.365176845217787,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 207130
-    },
-    {
-      "epoch": 3.3653393121151565,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 207140
-    },
-    {
-      "epoch": 3.365501779012526,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 207150
-    },
-    {
-      "epoch": 3.3656642459098958,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 207160
-    },
-    {
-      "epoch": 3.3658267128072654,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 207170
-    },
-    {
-      "epoch": 3.365989179704635,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 207180
-    },
-    {
-      "epoch": 3.3661516466020047,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1662,
-      "step": 207190
-    },
-    {
-      "epoch": 3.3663141134993744,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 207200
-    },
-    {
-      "epoch": 3.366476580396744,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 207210
-    },
-    {
-      "epoch": 3.3666390472941137,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 207220
-    },
-    {
-      "epoch": 3.3668015141914833,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 207230
-    },
-    {
-      "epoch": 3.366963981088853,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 207240
-    },
-    {
-      "epoch": 3.3671264479862226,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1736,
-      "step": 207250
-    },
-    {
-      "epoch": 3.3672889148835923,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 207260
-    },
-    {
-      "epoch": 3.3674513817809624,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 207270
-    },
-    {
-      "epoch": 3.3676138486783316,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 207280
-    },
-    {
-      "epoch": 3.3677763155757017,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 207290
-    },
-    {
-      "epoch": 3.367938782473071,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 207300
-    },
-    {
-      "epoch": 3.368101249370441,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 207310
-    },
-    {
-      "epoch": 3.3682637162678106,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 207320
-    },
-    {
-      "epoch": 3.3684261831651803,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1835,
-      "step": 207330
-    },
-    {
-      "epoch": 3.36858865006255,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 207340
-    },
-    {
-      "epoch": 3.3687511169599196,
-      "grad_norm": 7.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 207350
-    },
-    {
-      "epoch": 3.3689135838572892,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1839,
-      "step": 207360
-    },
-    {
-      "epoch": 3.369076050754659,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 207370
-    },
-    {
-      "epoch": 3.3692385176520285,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1687,
-      "step": 207380
-    },
-    {
-      "epoch": 3.369400984549398,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1683,
-      "step": 207390
-    },
-    {
-      "epoch": 3.369563451446768,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1742,
-      "step": 207400
-    },
-    {
-      "epoch": 3.3697259183441375,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 207410
-    },
-    {
-      "epoch": 3.369888385241507,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1727,
-      "step": 207420
-    },
-    {
-      "epoch": 3.370050852138877,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.167,
-      "step": 207430
-    },
-    {
-      "epoch": 3.3702133190362464,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 207440
-    },
-    {
-      "epoch": 3.370375785933616,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.154,
-      "step": 207450
-    },
-    {
-      "epoch": 3.3705382528309857,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1714,
-      "step": 207460
-    },
-    {
-      "epoch": 3.3707007197283554,
-      "grad_norm": 7.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1801,
-      "step": 207470
-    },
-    {
-      "epoch": 3.370863186625725,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1757,
-      "step": 207480
-    },
-    {
-      "epoch": 3.3710256535230947,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 207490
-    },
-    {
-      "epoch": 3.3711881204204643,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 207500
-    },
-    {
-      "epoch": 3.371350587317834,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2053,
-      "step": 207510
-    },
-    {
-      "epoch": 3.3715130542152036,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 207520
-    },
-    {
-      "epoch": 3.3716755211125733,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1871,
-      "step": 207530
-    },
-    {
-      "epoch": 3.371837988009943,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 207540
-    },
-    {
-      "epoch": 3.3720004549073126,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1861,
-      "step": 207550
-    },
-    {
-      "epoch": 3.3721629218046822,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 207560
-    },
-    {
-      "epoch": 3.372325388702052,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1816,
-      "step": 207570
-    },
-    {
-      "epoch": 3.3724878555994215,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 207580
-    },
-    {
-      "epoch": 3.372650322496791,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 207590
-    },
-    {
-      "epoch": 3.372812789394161,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.213,
-      "step": 207600
-    },
-    {
-      "epoch": 3.3729752562915305,
-      "grad_norm": 3.734375,
-      "learning_rate": 5e-05,
-      "loss": 0.1943,
-      "step": 207610
-    },
-    {
-      "epoch": 3.3731377231889,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 207620
-    },
-    {
-      "epoch": 3.37330019008627,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 207630
-    },
-    {
-      "epoch": 3.3734626569836395,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 207640
-    },
-    {
-      "epoch": 3.373625123881009,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1796,
-      "step": 207650
-    },
-    {
-      "epoch": 3.3737875907783788,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 207660
-    },
-    {
-      "epoch": 3.3739500576757484,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1782,
-      "step": 207670
-    },
-    {
-      "epoch": 3.374112524573118,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 207680
-    },
-    {
-      "epoch": 3.3742749914704877,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 207690
-    },
-    {
-      "epoch": 3.3744374583678574,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1885,
-      "step": 207700
-    },
-    {
-      "epoch": 3.374599925265227,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 207710
-    },
-    {
-      "epoch": 3.374762392162597,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 207720
-    },
-    {
-      "epoch": 3.3749248590599663,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1827,
-      "step": 207730
-    },
-    {
-      "epoch": 3.3750873259573364,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.183,
-      "step": 207740
-    },
-    {
-      "epoch": 3.3752497928547056,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 207750
-    },
-    {
-      "epoch": 3.3754122597520757,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 207760
-    },
-    {
-      "epoch": 3.3755747266494454,
-      "grad_norm": 3.640625,
-      "learning_rate": 5e-05,
-      "loss": 0.1958,
-      "step": 207770
-    },
-    {
-      "epoch": 3.375737193546815,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 207780
-    },
-    {
-      "epoch": 3.3758996604441847,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 207790
-    },
-    {
-      "epoch": 3.3760621273415543,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.179,
-      "step": 207800
-    },
-    {
-      "epoch": 3.376224594238924,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 207810
-    },
-    {
-      "epoch": 3.3763870611362936,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1601,
-      "step": 207820
-    },
-    {
-      "epoch": 3.3765495280336633,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1601,
-      "step": 207830
-    },
-    {
-      "epoch": 3.376711994931033,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 207840
-    },
-    {
-      "epoch": 3.3768744618284026,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 207850
-    },
-    {
-      "epoch": 3.377036928725772,
-      "grad_norm": 3.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 207860
-    },
-    {
-      "epoch": 3.377199395623142,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 207870
-    },
-    {
-      "epoch": 3.3773618625205115,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 207880
-    },
-    {
-      "epoch": 3.377524329417881,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2129,
-      "step": 207890
-    },
-    {
-      "epoch": 3.377686796315251,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 207900
-    },
-    {
-      "epoch": 3.3778492632126205,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 207910
-    },
-    {
-      "epoch": 3.37801173010999,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 207920
-    },
-    {
-      "epoch": 3.3781741970073598,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1895,
-      "step": 207930
-    },
-    {
-      "epoch": 3.3783366639047294,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 207940
-    },
-    {
-      "epoch": 3.378499130802099,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 207950
-    },
-    {
-      "epoch": 3.3786615976994687,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 207960
-    },
-    {
-      "epoch": 3.3788240645968384,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.211,
-      "step": 207970
-    },
-    {
-      "epoch": 3.378986531494208,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 207980
-    },
-    {
-      "epoch": 3.3791489983915777,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 207990
-    },
-    {
-      "epoch": 3.3793114652889473,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 208000
-    },
-    {
-      "epoch": 3.379473932186317,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2,
-      "step": 208010
-    },
-    {
-      "epoch": 3.3796363990836866,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2157,
-      "step": 208020
-    },
-    {
-      "epoch": 3.3797988659810563,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 208030
-    },
-    {
-      "epoch": 3.379961332878426,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 208040
-    },
-    {
-      "epoch": 3.3801237997757956,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1989,
-      "step": 208050
-    },
-    {
-      "epoch": 3.3802862666731652,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 208060
-    },
-    {
-      "epoch": 3.380448733570535,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2,
-      "step": 208070
-    },
-    {
-      "epoch": 3.3806112004679045,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 208080
-    },
-    {
-      "epoch": 3.380773667365274,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 208090
-    },
-    {
-      "epoch": 3.380936134262644,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 208100
-    },
-    {
-      "epoch": 3.3810986011600135,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 208110
-    },
-    {
-      "epoch": 3.381261068057383,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 208120
-    },
-    {
-      "epoch": 3.381423534954753,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 208130
-    },
-    {
-      "epoch": 3.3815860018521224,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.173,
-      "step": 208140
-    },
-    {
-      "epoch": 3.3817484687494925,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 208150
-    },
-    {
-      "epoch": 3.3819109356468617,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 208160
-    },
-    {
-      "epoch": 3.382073402544232,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 208170
-    },
-    {
-      "epoch": 3.382235869441601,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 208180
-    },
-    {
-      "epoch": 3.382398336338971,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 208190
-    },
-    {
-      "epoch": 3.382560803236341,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 208200
-    },
-    {
-      "epoch": 3.3827232701337104,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1818,
-      "step": 208210
-    },
-    {
-      "epoch": 3.38288573703108,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.165,
-      "step": 208220
-    },
-    {
-      "epoch": 3.3830482039284497,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 208230
-    },
-    {
-      "epoch": 3.3832106708258194,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1804,
-      "step": 208240
-    },
-    {
-      "epoch": 3.383373137723189,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1721,
-      "step": 208250
-    },
-    {
-      "epoch": 3.3835356046205587,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 208260
-    },
-    {
-      "epoch": 3.3836980715179283,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 208270
-    },
-    {
-      "epoch": 3.383860538415298,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 208280
-    },
-    {
-      "epoch": 3.3840230053126676,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 208290
-    },
-    {
-      "epoch": 3.3841854722100373,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 208300
-    },
-    {
-      "epoch": 3.384347939107407,
-      "grad_norm": 7.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 208310
-    },
-    {
-      "epoch": 3.3845104060047766,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 208320
-    },
-    {
-      "epoch": 3.3846728729021462,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 208330
-    },
-    {
-      "epoch": 3.384835339799516,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 208340
-    },
-    {
-      "epoch": 3.3849978066968855,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.224,
-      "step": 208350
-    },
-    {
-      "epoch": 3.385160273594255,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 208360
-    },
-    {
-      "epoch": 3.385322740491625,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 208370
-    },
-    {
-      "epoch": 3.3854852073889945,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 208380
-    },
-    {
-      "epoch": 3.385647674286364,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2134,
-      "step": 208390
-    },
-    {
-      "epoch": 3.385810141183734,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 208400
-    },
-    {
-      "epoch": 3.3859726080811035,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2076,
-      "step": 208410
-    },
-    {
-      "epoch": 3.386135074978473,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 208420
-    },
-    {
-      "epoch": 3.3862975418758428,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2382,
-      "step": 208430
-    },
-    {
-      "epoch": 3.3864600087732124,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2246,
-      "step": 208440
-    },
-    {
-      "epoch": 3.386622475670582,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.211,
-      "step": 208450
-    },
-    {
-      "epoch": 3.3867849425679517,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2332,
-      "step": 208460
-    },
-    {
-      "epoch": 3.3869474094653214,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 208470
-    },
-    {
-      "epoch": 3.387109876362691,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 208480
-    },
-    {
-      "epoch": 3.3872723432600607,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2253,
-      "step": 208490
-    },
-    {
-      "epoch": 3.3874348101574303,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1939,
-      "step": 208500
-    },
-    {
-      "epoch": 3.3875972770548,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 208510
-    },
-    {
-      "epoch": 3.3877597439521696,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 208520
-    },
-    {
-      "epoch": 3.3879222108495393,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2234,
-      "step": 208530
-    },
-    {
-      "epoch": 3.388084677746909,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 208540
-    },
-    {
-      "epoch": 3.3882471446442786,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1836,
-      "step": 208550
-    },
-    {
-      "epoch": 3.388409611541648,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 208560
-    },
-    {
-      "epoch": 3.388572078439018,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2231,
-      "step": 208570
-    },
-    {
-      "epoch": 3.388734545336388,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 208580
-    },
-    {
-      "epoch": 3.388897012233757,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 208590
-    },
-    {
-      "epoch": 3.3890594791311273,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 208600
-    },
-    {
-      "epoch": 3.3892219460284965,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2082,
-      "step": 208610
-    },
-    {
-      "epoch": 3.3893844129258666,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1952,
-      "step": 208620
-    },
-    {
-      "epoch": 3.3895468798232358,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 208630
-    },
-    {
-      "epoch": 3.389709346720606,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 208640
-    },
-    {
-      "epoch": 3.3898718136179755,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1803,
-      "step": 208650
-    },
-    {
-      "epoch": 3.390034280515345,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 208660
-    },
-    {
-      "epoch": 3.390196747412715,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1688,
-      "step": 208670
-    },
-    {
-      "epoch": 3.3903592143100845,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 208680
-    },
-    {
-      "epoch": 3.390521681207454,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 208690
-    },
-    {
-      "epoch": 3.3906841481048238,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 208700
-    },
-    {
-      "epoch": 3.3908466150021934,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 208710
-    },
-    {
-      "epoch": 3.391009081899563,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2316,
-      "step": 208720
-    },
-    {
-      "epoch": 3.3911715487969327,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 208730
-    },
-    {
-      "epoch": 3.3913340156943024,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2253,
-      "step": 208740
-    },
-    {
-      "epoch": 3.391496482591672,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 208750
-    },
-    {
-      "epoch": 3.3916589494890417,
-      "grad_norm": 3.765625,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 208760
-    },
-    {
-      "epoch": 3.3918214163864113,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 208770
-    },
-    {
-      "epoch": 3.391983883283781,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.163,
-      "step": 208780
-    },
-    {
-      "epoch": 3.3921463501811506,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 208790
-    },
-    {
-      "epoch": 3.3923088170785203,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1645,
-      "step": 208800
-    },
-    {
-      "epoch": 3.39247128397589,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1787,
-      "step": 208810
-    },
-    {
-      "epoch": 3.3926337508732596,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 208820
-    },
-    {
-      "epoch": 3.3927962177706292,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 208830
-    },
-    {
-      "epoch": 3.392958684667999,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2085,
-      "step": 208840
-    },
-    {
-      "epoch": 3.3931211515653685,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 208850
-    },
-    {
-      "epoch": 3.393283618462738,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.223,
-      "step": 208860
-    },
-    {
-      "epoch": 3.393446085360108,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 208870
-    },
-    {
-      "epoch": 3.3936085522574775,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 208880
-    },
-    {
-      "epoch": 3.393771019154847,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1816,
-      "step": 208890
-    },
-    {
-      "epoch": 3.393933486052217,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 208900
-    },
-    {
-      "epoch": 3.3940959529495864,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1784,
-      "step": 208910
-    },
-    {
-      "epoch": 3.394258419846956,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1698,
-      "step": 208920
-    },
-    {
-      "epoch": 3.3944208867443257,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 208930
-    },
-    {
-      "epoch": 3.3945833536416954,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 208940
-    },
-    {
-      "epoch": 3.394745820539065,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 208950
-    },
-    {
-      "epoch": 3.3949082874364347,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 208960
-    },
-    {
-      "epoch": 3.3950707543338043,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 208970
-    },
-    {
-      "epoch": 3.395233221231174,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 208980
-    },
-    {
-      "epoch": 3.3953956881285436,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 208990
-    },
-    {
-      "epoch": 3.3955581550259133,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 209000
-    },
-    {
-      "epoch": 3.395720621923283,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2097,
-      "step": 209010
-    },
-    {
-      "epoch": 3.3958830888206526,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2177,
-      "step": 209020
-    },
-    {
-      "epoch": 3.3960455557180227,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2222,
-      "step": 209030
-    },
-    {
-      "epoch": 3.396208022615392,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.207,
-      "step": 209040
-    },
-    {
-      "epoch": 3.396370489512762,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2254,
-      "step": 209050
-    },
-    {
-      "epoch": 3.396532956410131,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2186,
-      "step": 209060
-    },
-    {
-      "epoch": 3.3966954233075013,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2011,
-      "step": 209070
-    },
-    {
-      "epoch": 3.396857890204871,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 209080
-    },
-    {
-      "epoch": 3.3970203571022406,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2053,
-      "step": 209090
-    },
-    {
-      "epoch": 3.3971828239996102,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 209100
-    },
-    {
-      "epoch": 3.39734529089698,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2131,
-      "step": 209110
-    },
-    {
-      "epoch": 3.3975077577943495,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 209120
-    },
-    {
-      "epoch": 3.397670224691719,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2196,
-      "step": 209130
-    },
-    {
-      "epoch": 3.397832691589089,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2347,
-      "step": 209140
-    },
-    {
-      "epoch": 3.3979951584864585,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 209150
-    },
-    {
-      "epoch": 3.398157625383828,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 209160
-    },
-    {
-      "epoch": 3.398320092281198,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2154,
-      "step": 209170
-    },
-    {
-      "epoch": 3.3984825591785675,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 209180
-    },
-    {
-      "epoch": 3.398645026075937,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2114,
-      "step": 209190
-    },
-    {
-      "epoch": 3.3988074929733068,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1902,
-      "step": 209200
-    },
-    {
-      "epoch": 3.3989699598706764,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 209210
-    },
-    {
-      "epoch": 3.399132426768046,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 209220
-    },
-    {
-      "epoch": 3.3992948936654157,
-      "grad_norm": 7.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 209230
-    },
-    {
-      "epoch": 3.3994573605627854,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 209240
-    },
-    {
-      "epoch": 3.399619827460155,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2254,
-      "step": 209250
-    },
-    {
-      "epoch": 3.3997822943575247,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2179,
-      "step": 209260
-    },
-    {
-      "epoch": 3.3999447612548943,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2279,
-      "step": 209270
-    },
-    {
-      "epoch": 3.400107228152264,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 209280
-    },
-    {
-      "epoch": 3.4002696950496336,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 209290
-    },
-    {
-      "epoch": 3.4004321619470033,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2171,
-      "step": 209300
-    },
-    {
-      "epoch": 3.400594628844373,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 209310
-    },
-    {
-      "epoch": 3.4007570957417426,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 209320
-    },
-    {
-      "epoch": 3.400919562639112,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 209330
-    },
-    {
-      "epoch": 3.401082029536482,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2358,
-      "step": 209340
-    },
-    {
-      "epoch": 3.4012444964338515,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 209350
-    },
-    {
-      "epoch": 3.401406963331221,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2097,
-      "step": 209360
-    },
-    {
-      "epoch": 3.401569430228591,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 209370
-    },
-    {
-      "epoch": 3.4017318971259605,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 209380
-    },
-    {
-      "epoch": 3.40189436402333,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 209390
-    },
-    {
-      "epoch": 3.4020568309206998,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 209400
-    },
-    {
-      "epoch": 3.4022192978180694,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 209410
-    },
-    {
-      "epoch": 3.402381764715439,
-      "grad_norm": 7.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2146,
-      "step": 209420
-    },
-    {
-      "epoch": 3.4025442316128087,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 209430
-    },
-    {
-      "epoch": 3.4027066985101784,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2466,
-      "step": 209440
-    },
-    {
-      "epoch": 3.402869165407548,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 209450
-    },
-    {
-      "epoch": 3.403031632304918,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 209460
-    },
-    {
-      "epoch": 3.4031940992022873,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 209470
-    },
-    {
-      "epoch": 3.4033565660996574,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 209480
-    },
-    {
-      "epoch": 3.4035190329970266,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2046,
-      "step": 209490
-    },
-    {
-      "epoch": 3.4036814998943967,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1936,
-      "step": 209500
-    },
-    {
-      "epoch": 3.403843966791766,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2143,
-      "step": 209510
-    },
-    {
-      "epoch": 3.404006433689136,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 209520
-    },
-    {
-      "epoch": 3.4041689005865057,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 209530
-    },
-    {
-      "epoch": 3.4043313674838753,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 209540
-    },
-    {
-      "epoch": 3.404493834381245,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 209550
-    },
-    {
-      "epoch": 3.4046563012786146,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1978,
-      "step": 209560
-    },
-    {
-      "epoch": 3.4048187681759843,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 209570
-    },
-    {
-      "epoch": 3.404981235073354,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 209580
-    },
-    {
-      "epoch": 3.4051437019707236,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 209590
-    },
-    {
-      "epoch": 3.4053061688680932,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1751,
-      "step": 209600
-    },
-    {
-      "epoch": 3.405468635765463,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1676,
-      "step": 209610
-    },
-    {
-      "epoch": 3.4056311026628325,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 209620
-    },
-    {
-      "epoch": 3.405793569560202,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 209630
-    },
-    {
-      "epoch": 3.405956036457572,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 209640
-    },
-    {
-      "epoch": 3.4061185033549415,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 209650
-    },
-    {
-      "epoch": 3.406280970252311,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 209660
-    },
-    {
-      "epoch": 3.406443437149681,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 209670
-    },
-    {
-      "epoch": 3.4066059040470504,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 209680
-    },
-    {
-      "epoch": 3.40676837094442,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1887,
-      "step": 209690
-    },
-    {
-      "epoch": 3.4069308378417897,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 209700
-    },
-    {
-      "epoch": 3.4070933047391594,
-      "grad_norm": 7.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 209710
-    },
-    {
-      "epoch": 3.407255771636529,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1743,
-      "step": 209720
-    },
-    {
-      "epoch": 3.4074182385338987,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.199,
-      "step": 209730
-    },
-    {
-      "epoch": 3.4075807054312683,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 209740
-    },
-    {
-      "epoch": 3.407743172328638,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1711,
-      "step": 209750
-    },
-    {
-      "epoch": 3.4079056392260076,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1863,
-      "step": 209760
-    },
-    {
-      "epoch": 3.4080681061233773,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 209770
-    },
-    {
-      "epoch": 3.408230573020747,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 209780
-    },
-    {
-      "epoch": 3.4083930399181166,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1895,
-      "step": 209790
-    },
-    {
-      "epoch": 3.4085555068154862,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 209800
-    },
-    {
-      "epoch": 3.408717973712856,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 209810
-    },
-    {
-      "epoch": 3.4088804406102255,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 209820
-    },
-    {
-      "epoch": 3.409042907507595,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 209830
-    },
-    {
-      "epoch": 3.409205374404965,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 209840
-    },
-    {
-      "epoch": 3.4093678413023345,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2137,
-      "step": 209850
-    },
-    {
-      "epoch": 3.409530308199704,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2251,
-      "step": 209860
-    },
-    {
-      "epoch": 3.409692775097074,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.21,
-      "step": 209870
-    },
-    {
-      "epoch": 3.4098552419944435,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 209880
-    },
-    {
-      "epoch": 3.410017708891813,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 209890
-    },
-    {
-      "epoch": 3.4101801757891828,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 209900
-    },
-    {
-      "epoch": 3.410342642686553,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 209910
-    },
-    {
-      "epoch": 3.410505109583922,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1446,
-      "step": 209920
-    },
-    {
-      "epoch": 3.410667576481292,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1616,
-      "step": 209930
-    },
-    {
-      "epoch": 3.4108300433786614,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1766,
-      "step": 209940
-    },
-    {
-      "epoch": 3.4109925102760315,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 209950
-    },
-    {
-      "epoch": 3.411154977173401,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1711,
-      "step": 209960
-    },
-    {
-      "epoch": 3.4113174440707708,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1906,
-      "step": 209970
-    },
-    {
-      "epoch": 3.4114799109681404,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 209980
-    },
-    {
-      "epoch": 3.41164237786551,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 209990
-    },
-    {
-      "epoch": 3.4118048447628797,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.172,
-      "step": 210000
-    },
-    {
-      "epoch": 3.4119673116602494,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1834,
-      "step": 210010
-    },
-    {
-      "epoch": 3.412129778557619,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 210020
-    },
-    {
-      "epoch": 3.4122922454549887,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 210030
-    },
-    {
-      "epoch": 3.4124547123523583,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.186,
-      "step": 210040
-    },
-    {
-      "epoch": 3.412617179249728,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 210050
-    },
-    {
-      "epoch": 3.4127796461470976,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 210060
-    },
-    {
-      "epoch": 3.4129421130444673,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 210070
-    },
-    {
-      "epoch": 3.413104579941837,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1814,
-      "step": 210080
-    },
-    {
-      "epoch": 3.4132670468392066,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1869,
-      "step": 210090
-    },
-    {
-      "epoch": 3.413429513736576,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 210100
-    },
-    {
-      "epoch": 3.413591980633946,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1619,
-      "step": 210110
-    },
-    {
-      "epoch": 3.4137544475313155,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 210120
-    },
-    {
-      "epoch": 3.413916914428685,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1677,
-      "step": 210130
-    },
-    {
-      "epoch": 3.414079381326055,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 210140
-    },
-    {
-      "epoch": 3.4142418482234245,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1796,
-      "step": 210150
-    },
-    {
-      "epoch": 3.414404315120794,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1685,
-      "step": 210160
-    },
-    {
-      "epoch": 3.4145667820181638,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 210170
-    },
-    {
-      "epoch": 3.4147292489155334,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 210180
-    },
-    {
-      "epoch": 3.414891715812903,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 210190
-    },
-    {
-      "epoch": 3.4150541827102727,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 210200
-    },
-    {
-      "epoch": 3.4152166496076424,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1837,
-      "step": 210210
-    },
-    {
-      "epoch": 3.415379116505012,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 210220
-    },
-    {
-      "epoch": 3.4155415834023817,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1977,
-      "step": 210230
-    },
-    {
-      "epoch": 3.4157040502997513,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 210240
-    },
-    {
-      "epoch": 3.415866517197121,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 210250
-    },
-    {
-      "epoch": 3.4160289840944906,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 210260
-    },
-    {
-      "epoch": 3.4161914509918603,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2236,
-      "step": 210270
-    },
-    {
-      "epoch": 3.41635391788923,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 210280
-    },
-    {
-      "epoch": 3.4165163847865996,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 210290
-    },
-    {
-      "epoch": 3.4166788516839692,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 210300
-    },
-    {
-      "epoch": 3.416841318581339,
-      "grad_norm": 7.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 210310
-    },
-    {
-      "epoch": 3.4170037854787085,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 210320
-    },
-    {
-      "epoch": 3.417166252376078,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 210330
-    },
-    {
-      "epoch": 3.4173287192734483,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 210340
-    },
-    {
-      "epoch": 3.4174911861708175,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2059,
-      "step": 210350
-    },
-    {
-      "epoch": 3.4176536530681876,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1814,
-      "step": 210360
-    },
-    {
-      "epoch": 3.417816119965557,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 210370
-    },
-    {
-      "epoch": 3.417978586862927,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 210380
-    },
-    {
-      "epoch": 3.4181410537602965,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.221,
-      "step": 210390
-    },
-    {
-      "epoch": 3.418303520657666,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 210400
-    },
-    {
-      "epoch": 3.418465987555036,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1811,
-      "step": 210410
-    },
-    {
-      "epoch": 3.4186284544524055,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 210420
-    },
-    {
-      "epoch": 3.418790921349775,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 210430
-    },
-    {
-      "epoch": 3.418953388247145,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 210440
-    },
-    {
-      "epoch": 3.4191158551445144,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 210450
-    },
-    {
-      "epoch": 3.419278322041884,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 210460
-    },
-    {
-      "epoch": 3.4194407889392537,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2346,
-      "step": 210470
-    },
-    {
-      "epoch": 3.4196032558366234,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 210480
-    },
-    {
-      "epoch": 3.419765722733993,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 210490
-    },
-    {
-      "epoch": 3.4199281896313627,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 210500
-    },
-    {
-      "epoch": 3.4200906565287323,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 210510
-    },
-    {
-      "epoch": 3.420253123426102,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.207,
-      "step": 210520
-    },
-    {
-      "epoch": 3.4204155903234716,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 210530
-    },
-    {
-      "epoch": 3.4205780572208413,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 210540
-    },
-    {
-      "epoch": 3.420740524118211,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 210550
-    },
-    {
-      "epoch": 3.4209029910155806,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1711,
-      "step": 210560
-    },
-    {
-      "epoch": 3.4210654579129502,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1733,
-      "step": 210570
-    },
-    {
-      "epoch": 3.42122792481032,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 210580
-    },
-    {
-      "epoch": 3.4213903917076895,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1943,
-      "step": 210590
-    },
-    {
-      "epoch": 3.421552858605059,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 210600
-    },
-    {
-      "epoch": 3.421715325502429,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1728,
-      "step": 210610
-    },
-    {
-      "epoch": 3.4218777923997985,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1801,
-      "step": 210620
-    },
-    {
-      "epoch": 3.422040259297168,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1687,
-      "step": 210630
-    },
-    {
-      "epoch": 3.422202726194538,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1767,
-      "step": 210640
-    },
-    {
-      "epoch": 3.4223651930919075,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 210650
-    },
-    {
-      "epoch": 3.422527659989277,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1815,
-      "step": 210660
-    },
-    {
-      "epoch": 3.4226901268866468,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 210670
-    },
-    {
-      "epoch": 3.4228525937840164,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 210680
-    },
-    {
-      "epoch": 3.423015060681386,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 210690
-    },
-    {
-      "epoch": 3.4231775275787557,
-      "grad_norm": 3.890625,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 210700
-    },
-    {
-      "epoch": 3.4233399944761254,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 210710
-    },
-    {
-      "epoch": 3.423502461373495,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 210720
-    },
-    {
-      "epoch": 3.4236649282708647,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1666,
-      "step": 210730
-    },
-    {
-      "epoch": 3.4238273951682343,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 210740
-    },
-    {
-      "epoch": 3.423989862065604,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 210750
-    },
-    {
-      "epoch": 3.4241523289629736,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 210760
-    },
-    {
-      "epoch": 3.4243147958603433,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1607,
-      "step": 210770
-    },
-    {
-      "epoch": 3.424477262757713,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1562,
-      "step": 210780
-    },
-    {
-      "epoch": 3.424639729655083,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 210790
-    },
-    {
-      "epoch": 3.424802196552452,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 210800
-    },
-    {
-      "epoch": 3.4249646634498223,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 210810
-    },
-    {
-      "epoch": 3.4251271303471915,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 210820
-    },
-    {
-      "epoch": 3.4252895972445616,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2351,
-      "step": 210830
-    },
-    {
-      "epoch": 3.4254520641419313,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 210840
-    },
-    {
-      "epoch": 3.425614531039301,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 210850
-    },
-    {
-      "epoch": 3.4257769979366706,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 210860
-    },
-    {
-      "epoch": 3.42593946483404,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 210870
-    },
-    {
-      "epoch": 3.42610193173141,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2356,
-      "step": 210880
-    },
-    {
-      "epoch": 3.4262643986287795,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 210890
-    },
-    {
-      "epoch": 3.426426865526149,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 210900
-    },
-    {
-      "epoch": 3.426589332423519,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 210910
-    },
-    {
-      "epoch": 3.4267517993208885,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 210920
-    },
-    {
-      "epoch": 3.426914266218258,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 210930
-    },
-    {
-      "epoch": 3.4270767331156278,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 210940
-    },
-    {
-      "epoch": 3.4272392000129974,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1943,
-      "step": 210950
-    },
-    {
-      "epoch": 3.427401666910367,
-      "grad_norm": 7.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 210960
-    },
-    {
-      "epoch": 3.4275641338077367,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 210970
-    },
-    {
-      "epoch": 3.4277266007051064,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1898,
-      "step": 210980
-    },
-    {
-      "epoch": 3.427889067602476,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 210990
-    },
-    {
-      "epoch": 3.4280515344998457,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 211000
-    },
-    {
-      "epoch": 3.4282140013972153,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 211010
-    },
-    {
-      "epoch": 3.428376468294585,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 211020
-    },
-    {
-      "epoch": 3.4285389351919546,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1989,
-      "step": 211030
-    },
-    {
-      "epoch": 3.4287014020893243,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1837,
-      "step": 211040
-    },
-    {
-      "epoch": 3.428863868986694,
-      "grad_norm": 7.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 211050
-    },
-    {
-      "epoch": 3.4290263358840636,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1779,
-      "step": 211060
-    },
-    {
-      "epoch": 3.4291888027814332,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 211070
-    },
-    {
-      "epoch": 3.429351269678803,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.188,
-      "step": 211080
-    },
-    {
-      "epoch": 3.4295137365761725,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1947,
-      "step": 211090
-    },
-    {
-      "epoch": 3.429676203473542,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1767,
-      "step": 211100
-    },
-    {
-      "epoch": 3.429838670370912,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1807,
-      "step": 211110
-    },
-    {
-      "epoch": 3.4300011372682815,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 211120
-    },
-    {
-      "epoch": 3.430163604165651,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2004,
-      "step": 211130
-    },
-    {
-      "epoch": 3.430326071063021,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1728,
-      "step": 211140
-    },
-    {
-      "epoch": 3.4304885379603904,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1885,
-      "step": 211150
-    },
-    {
-      "epoch": 3.43065100485776,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 211160
-    },
-    {
-      "epoch": 3.4308134717551297,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 211170
-    },
-    {
-      "epoch": 3.4309759386524994,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 211180
-    },
-    {
-      "epoch": 3.431138405549869,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 211190
-    },
-    {
-      "epoch": 3.4313008724472387,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 211200
-    },
-    {
-      "epoch": 3.4314633393446083,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 211210
-    },
-    {
-      "epoch": 3.4316258062419784,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 211220
-    },
-    {
-      "epoch": 3.4317882731393476,
-      "grad_norm": 8.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2244,
-      "step": 211230
-    },
-    {
-      "epoch": 3.4319507400367177,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2248,
-      "step": 211240
-    },
-    {
-      "epoch": 3.432113206934087,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 211250
-    },
-    {
-      "epoch": 3.432275673831457,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 211260
-    },
-    {
-      "epoch": 3.4324381407288267,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 211270
-    },
-    {
-      "epoch": 3.4326006076261963,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2011,
-      "step": 211280
-    },
-    {
-      "epoch": 3.432763074523566,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2,
-      "step": 211290
-    },
-    {
-      "epoch": 3.4329255414209356,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 211300
-    },
-    {
-      "epoch": 3.4330880083183053,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1797,
-      "step": 211310
-    },
-    {
-      "epoch": 3.433250475215675,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1828,
-      "step": 211320
-    },
-    {
-      "epoch": 3.4334129421130446,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1875,
-      "step": 211330
-    },
-    {
-      "epoch": 3.4335754090104142,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2269,
-      "step": 211340
-    },
-    {
-      "epoch": 3.433737875907784,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 211350
-    },
-    {
-      "epoch": 3.4339003428051535,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 211360
-    },
-    {
-      "epoch": 3.434062809702523,
-      "grad_norm": 7.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2261,
-      "step": 211370
-    },
-    {
-      "epoch": 3.434225276599893,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2152,
-      "step": 211380
-    },
-    {
-      "epoch": 3.4343877434972625,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 211390
-    },
-    {
-      "epoch": 3.434550210394632,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 211400
-    },
-    {
-      "epoch": 3.434712677292002,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 211410
-    },
-    {
-      "epoch": 3.4348751441893715,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 211420
-    },
-    {
-      "epoch": 3.435037611086741,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 211430
-    },
-    {
-      "epoch": 3.4352000779841108,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 211440
-    },
-    {
-      "epoch": 3.4353625448814804,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1684,
-      "step": 211450
-    },
-    {
-      "epoch": 3.43552501177885,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 211460
-    },
-    {
-      "epoch": 3.4356874786762197,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 211470
-    },
-    {
-      "epoch": 3.4358499455735894,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1991,
-      "step": 211480
-    },
-    {
-      "epoch": 3.436012412470959,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 211490
-    },
-    {
-      "epoch": 3.4361748793683287,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2084,
-      "step": 211500
-    },
-    {
-      "epoch": 3.4363373462656983,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 211510
-    },
-    {
-      "epoch": 3.436499813163068,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 211520
-    },
-    {
-      "epoch": 3.4366622800604376,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 211530
-    },
-    {
-      "epoch": 3.4368247469578073,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1743,
-      "step": 211540
-    },
-    {
-      "epoch": 3.436987213855177,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2228,
-      "step": 211550
-    },
-    {
-      "epoch": 3.4371496807525466,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 211560
-    },
-    {
-      "epoch": 3.437312147649916,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1966,
-      "step": 211570
-    },
-    {
-      "epoch": 3.437474614547286,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2179,
-      "step": 211580
-    },
-    {
-      "epoch": 3.4376370814446555,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 211590
-    },
-    {
-      "epoch": 3.437799548342025,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 211600
-    },
-    {
-      "epoch": 3.437962015239395,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 211610
-    },
-    {
-      "epoch": 3.4381244821367645,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 211620
-    },
-    {
-      "epoch": 3.438286949034134,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 211630
-    },
-    {
-      "epoch": 3.4384494159315038,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 211640
-    },
-    {
-      "epoch": 3.4386118828288734,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 211650
-    },
-    {
-      "epoch": 3.438774349726243,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.178,
-      "step": 211660
-    },
-    {
-      "epoch": 3.438936816623613,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2166,
-      "step": 211670
-    },
-    {
-      "epoch": 3.4390992835209824,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2084,
-      "step": 211680
-    },
-    {
-      "epoch": 3.4392617504183525,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 211690
-    },
-    {
-      "epoch": 3.4394242173157217,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.188,
-      "step": 211700
-    },
-    {
-      "epoch": 3.4395866842130918,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 211710
-    },
-    {
-      "epoch": 3.4397491511104614,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 211720
-    },
-    {
-      "epoch": 3.439911618007831,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 211730
-    },
-    {
-      "epoch": 3.4400740849052007,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 211740
-    },
-    {
-      "epoch": 3.4402365518025704,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 211750
-    },
-    {
-      "epoch": 3.44039901869994,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2209,
-      "step": 211760
-    },
-    {
-      "epoch": 3.4405614855973097,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 211770
-    },
-    {
-      "epoch": 3.4407239524946793,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 211780
-    },
-    {
-      "epoch": 3.440886419392049,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2295,
-      "step": 211790
-    },
-    {
-      "epoch": 3.4410488862894186,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2445,
-      "step": 211800
-    },
-    {
-      "epoch": 3.4412113531867883,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 211810
-    },
-    {
-      "epoch": 3.441373820084158,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 211820
-    },
-    {
-      "epoch": 3.4415362869815276,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 211830
-    },
-    {
-      "epoch": 3.4416987538788972,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1906,
-      "step": 211840
-    },
-    {
-      "epoch": 3.441861220776267,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 211850
-    },
-    {
-      "epoch": 3.4420236876736365,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 211860
-    },
-    {
-      "epoch": 3.442186154571006,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 211870
-    },
-    {
-      "epoch": 3.442348621468376,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 211880
-    },
-    {
-      "epoch": 3.4425110883657455,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 211890
-    },
-    {
-      "epoch": 3.442673555263115,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 211900
-    },
-    {
-      "epoch": 3.442836022160485,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 211910
-    },
-    {
-      "epoch": 3.4429984890578544,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2169,
-      "step": 211920
-    },
-    {
-      "epoch": 3.443160955955224,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 211930
-    },
-    {
-      "epoch": 3.4433234228525937,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 211940
-    },
-    {
-      "epoch": 3.4434858897499634,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1977,
-      "step": 211950
-    },
-    {
-      "epoch": 3.443648356647333,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2082,
-      "step": 211960
-    },
-    {
-      "epoch": 3.4438108235447027,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 211970
-    },
-    {
-      "epoch": 3.4439732904420723,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 211980
-    },
-    {
-      "epoch": 3.444135757339442,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 211990
-    },
-    {
-      "epoch": 3.4442982242368116,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 212000
-    },
-    {
-      "epoch": 3.4444606911341813,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 212010
-    },
-    {
-      "epoch": 3.444623158031551,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 212020
-    },
-    {
-      "epoch": 3.4447856249289206,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.247,
-      "step": 212030
-    },
-    {
-      "epoch": 3.4449480918262902,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2201,
-      "step": 212040
-    },
-    {
-      "epoch": 3.44511055872366,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 212050
-    },
-    {
-      "epoch": 3.4452730256210296,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1983,
-      "step": 212060
-    },
-    {
-      "epoch": 3.445435492518399,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 212070
-    },
-    {
-      "epoch": 3.445597959415769,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 212080
-    },
-    {
-      "epoch": 3.4457604263131385,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 212090
-    },
-    {
-      "epoch": 3.4459228932105086,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 212100
-    },
-    {
-      "epoch": 3.446085360107878,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 212110
-    },
-    {
-      "epoch": 3.446247827005248,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 212120
-    },
-    {
-      "epoch": 3.446410293902617,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 212130
-    },
-    {
-      "epoch": 3.446572760799987,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1861,
-      "step": 212140
-    },
-    {
-      "epoch": 3.446735227697357,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 212150
-    },
-    {
-      "epoch": 3.4468976945947265,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 212160
-    },
-    {
-      "epoch": 3.447060161492096,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 212170
-    },
-    {
-      "epoch": 3.447222628389466,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2228,
-      "step": 212180
-    },
-    {
-      "epoch": 3.4473850952868355,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2228,
-      "step": 212190
-    },
-    {
-      "epoch": 3.447547562184205,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 212200
-    },
-    {
-      "epoch": 3.4477100290815748,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2187,
-      "step": 212210
-    },
-    {
-      "epoch": 3.4478724959789444,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2192,
-      "step": 212220
-    },
-    {
-      "epoch": 3.448034962876314,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 212230
-    },
-    {
-      "epoch": 3.4481974297736837,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2346,
-      "step": 212240
-    },
-    {
-      "epoch": 3.4483598966710534,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 212250
-    },
-    {
-      "epoch": 3.448522363568423,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2276,
-      "step": 212260
-    },
-    {
-      "epoch": 3.4486848304657927,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 212270
-    },
-    {
-      "epoch": 3.4488472973631623,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2066,
-      "step": 212280
-    },
-    {
-      "epoch": 3.449009764260532,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 212290
-    },
-    {
-      "epoch": 3.4491722311579016,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 212300
-    },
-    {
-      "epoch": 3.4493346980552713,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 212310
-    },
-    {
-      "epoch": 3.449497164952641,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 212320
-    },
-    {
-      "epoch": 3.4496596318500106,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 212330
-    },
-    {
-      "epoch": 3.44982209874738,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1794,
-      "step": 212340
-    },
-    {
-      "epoch": 3.44998456564475,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 212350
-    },
-    {
-      "epoch": 3.4501470325421195,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 212360
-    },
-    {
-      "epoch": 3.450309499439489,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 212370
-    },
-    {
-      "epoch": 3.450471966336859,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1618,
-      "step": 212380
-    },
-    {
-      "epoch": 3.4506344332342285,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.184,
-      "step": 212390
-    },
-    {
-      "epoch": 3.450796900131598,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1703,
-      "step": 212400
-    },
-    {
-      "epoch": 3.4509593670289678,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1646,
-      "step": 212410
-    },
-    {
-      "epoch": 3.4511218339263374,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 212420
-    },
-    {
-      "epoch": 3.451284300823707,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1651,
-      "step": 212430
-    },
-    {
-      "epoch": 3.4514467677210767,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1695,
-      "step": 212440
-    },
-    {
-      "epoch": 3.4516092346184464,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1532,
-      "step": 212450
-    },
-    {
-      "epoch": 3.451771701515816,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 212460
-    },
-    {
-      "epoch": 3.4519341684131857,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 212470
-    },
-    {
-      "epoch": 3.4520966353105553,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 212480
-    },
-    {
-      "epoch": 3.452259102207925,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1859,
-      "step": 212490
-    },
-    {
-      "epoch": 3.4524215691052946,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 212500
-    },
-    {
-      "epoch": 3.4525840360026643,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.211,
-      "step": 212510
-    },
-    {
-      "epoch": 3.452746502900034,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1853,
-      "step": 212520
-    },
-    {
-      "epoch": 3.452908969797404,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 212530
-    },
-    {
-      "epoch": 3.4530714366947732,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 212540
-    },
-    {
-      "epoch": 3.4532339035921433,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 212550
-    },
-    {
-      "epoch": 3.4533963704895125,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 212560
-    },
-    {
-      "epoch": 3.4535588373868826,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 212570
-    },
-    {
-      "epoch": 3.453721304284252,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 212580
-    },
-    {
-      "epoch": 3.453883771181622,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 212590
-    },
-    {
-      "epoch": 3.4540462380789916,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1758,
-      "step": 212600
-    },
-    {
-      "epoch": 3.4542087049763612,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2107,
-      "step": 212610
-    },
-    {
-      "epoch": 3.454371171873731,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1783,
-      "step": 212620
-    },
-    {
-      "epoch": 3.4545336387711005,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1977,
-      "step": 212630
-    },
-    {
-      "epoch": 3.45469610566847,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1893,
-      "step": 212640
-    },
-    {
-      "epoch": 3.45485857256584,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 212650
-    },
-    {
-      "epoch": 3.4550210394632095,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 212660
-    },
-    {
-      "epoch": 3.455183506360579,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.18,
-      "step": 212670
-    },
-    {
-      "epoch": 3.455345973257949,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 212680
-    },
-    {
-      "epoch": 3.4555084401553184,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 212690
-    },
-    {
-      "epoch": 3.455670907052688,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 212700
-    },
-    {
-      "epoch": 3.4558333739500577,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 212710
-    },
-    {
-      "epoch": 3.4559958408474274,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 212720
-    },
-    {
-      "epoch": 3.456158307744797,
-      "grad_norm": 7.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2294,
-      "step": 212730
-    },
-    {
-      "epoch": 3.4563207746421667,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2115,
-      "step": 212740
-    },
-    {
-      "epoch": 3.4564832415395363,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 212750
-    },
-    {
-      "epoch": 3.456645708436906,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2239,
-      "step": 212760
-    },
-    {
-      "epoch": 3.4568081753342756,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 212770
-    },
-    {
-      "epoch": 3.4569706422316453,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 212780
-    },
-    {
-      "epoch": 3.457133109129015,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 212790
-    },
-    {
-      "epoch": 3.4572955760263846,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 212800
-    },
-    {
-      "epoch": 3.4574580429237542,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 212810
-    },
-    {
-      "epoch": 3.457620509821124,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 212820
-    },
-    {
-      "epoch": 3.4577829767184936,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1804,
-      "step": 212830
-    },
-    {
-      "epoch": 3.457945443615863,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 212840
-    },
-    {
-      "epoch": 3.458107910513233,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 212850
-    },
-    {
-      "epoch": 3.4582703774106025,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 212860
-    },
-    {
-      "epoch": 3.458432844307972,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 212870
-    },
-    {
-      "epoch": 3.458595311205342,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 212880
-    },
-    {
-      "epoch": 3.4587577781027115,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1835,
-      "step": 212890
-    },
-    {
-      "epoch": 3.458920245000081,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 212900
-    },
-    {
-      "epoch": 3.4590827118974508,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1762,
-      "step": 212910
-    },
-    {
-      "epoch": 3.4592451787948204,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 212920
-    },
-    {
-      "epoch": 3.45940764569219,
-      "grad_norm": 7.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.182,
-      "step": 212930
-    },
-    {
-      "epoch": 3.4595701125895597,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1843,
-      "step": 212940
-    },
-    {
-      "epoch": 3.4597325794869294,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 212950
-    },
-    {
-      "epoch": 3.459895046384299,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 212960
-    },
-    {
-      "epoch": 3.4600575132816687,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 212970
-    },
-    {
-      "epoch": 3.4602199801790388,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 212980
-    },
-    {
-      "epoch": 3.460382447076408,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 212990
-    },
-    {
-      "epoch": 3.460544913973778,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 213000
-    },
-    {
-      "epoch": 3.4607073808711473,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 213010
-    },
-    {
-      "epoch": 3.4608698477685174,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2008,
-      "step": 213020
-    },
-    {
-      "epoch": 3.461032314665887,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 213030
-    },
-    {
-      "epoch": 3.4611947815632567,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 213040
-    },
-    {
-      "epoch": 3.4613572484606263,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.223,
-      "step": 213050
-    },
-    {
-      "epoch": 3.461519715357996,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 213060
-    },
-    {
-      "epoch": 3.4616821822553656,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 213070
-    },
-    {
-      "epoch": 3.4618446491527353,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.215,
-      "step": 213080
-    },
-    {
-      "epoch": 3.462007116050105,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 213090
-    },
-    {
-      "epoch": 3.4621695829474746,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 213100
-    },
-    {
-      "epoch": 3.462332049844844,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 213110
-    },
-    {
-      "epoch": 3.462494516742214,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 213120
-    },
-    {
-      "epoch": 3.4626569836395835,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 213130
-    },
-    {
-      "epoch": 3.462819450536953,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 213140
-    },
-    {
-      "epoch": 3.462981917434323,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 213150
-    },
-    {
-      "epoch": 3.4631443843316925,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1803,
-      "step": 213160
-    },
-    {
-      "epoch": 3.463306851229062,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1866,
-      "step": 213170
-    },
-    {
-      "epoch": 3.4634693181264318,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2058,
-      "step": 213180
-    },
-    {
-      "epoch": 3.4636317850238014,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 213190
-    },
-    {
-      "epoch": 3.463794251921171,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1818,
-      "step": 213200
-    },
-    {
-      "epoch": 3.4639567188185407,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2129,
-      "step": 213210
-    },
-    {
-      "epoch": 3.4641191857159104,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 213220
-    },
-    {
-      "epoch": 3.46428165261328,
-      "grad_norm": 7.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 213230
-    },
-    {
-      "epoch": 3.4644441195106497,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 213240
-    },
-    {
-      "epoch": 3.4646065864080193,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 213250
-    },
-    {
-      "epoch": 3.464769053305389,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1947,
-      "step": 213260
-    },
-    {
-      "epoch": 3.4649315202027586,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 213270
-    },
-    {
-      "epoch": 3.4650939871001283,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2331,
-      "step": 213280
-    },
-    {
-      "epoch": 3.465256453997498,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 213290
-    },
-    {
-      "epoch": 3.4654189208948676,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.183,
-      "step": 213300
-    },
-    {
-      "epoch": 3.4655813877922372,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 213310
-    },
-    {
-      "epoch": 3.465743854689607,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 213320
-    },
-    {
-      "epoch": 3.4659063215869765,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 213330
-    },
-    {
-      "epoch": 3.466068788484346,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 213340
-    },
-    {
-      "epoch": 3.466231255381716,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 213350
-    },
-    {
-      "epoch": 3.4663937222790855,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 213360
-    },
-    {
-      "epoch": 3.466556189176455,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 213370
-    },
-    {
-      "epoch": 3.466718656073825,
-      "grad_norm": 8.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1825,
-      "step": 213380
-    },
-    {
-      "epoch": 3.4668811229711944,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 213390
-    },
-    {
-      "epoch": 3.467043589868564,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 213400
-    },
-    {
-      "epoch": 3.467206056765934,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1841,
-      "step": 213410
-    },
-    {
-      "epoch": 3.4673685236633034,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 213420
-    },
-    {
-      "epoch": 3.4675309905606735,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2171,
-      "step": 213430
-    },
-    {
-      "epoch": 3.4676934574580427,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 213440
-    },
-    {
-      "epoch": 3.467855924355413,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2191,
-      "step": 213450
-    },
-    {
-      "epoch": 3.4680183912527824,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 213460
-    },
-    {
-      "epoch": 3.468180858150152,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2133,
-      "step": 213470
-    },
-    {
-      "epoch": 3.4683433250475217,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 213480
-    },
-    {
-      "epoch": 3.4685057919448914,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1861,
-      "step": 213490
-    },
-    {
-      "epoch": 3.468668258842261,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 213500
-    },
-    {
-      "epoch": 3.4688307257396307,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2141,
-      "step": 213510
-    },
-    {
-      "epoch": 3.4689931926370003,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 213520
-    },
-    {
-      "epoch": 3.46915565953437,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 213530
-    },
-    {
-      "epoch": 3.4693181264317396,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 213540
-    },
-    {
-      "epoch": 3.4694805933291093,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 213550
-    },
-    {
-      "epoch": 3.469643060226479,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 213560
-    },
-    {
-      "epoch": 3.4698055271238486,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 213570
-    },
-    {
-      "epoch": 3.4699679940212182,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 213580
-    },
-    {
-      "epoch": 3.470130460918588,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2189,
-      "step": 213590
-    },
-    {
-      "epoch": 3.4702929278159576,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2314,
-      "step": 213600
-    },
-    {
-      "epoch": 3.470455394713327,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 213610
-    },
-    {
-      "epoch": 3.470617861610697,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 213620
-    },
-    {
-      "epoch": 3.4707803285080665,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 213630
-    },
-    {
-      "epoch": 3.470942795405436,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 213640
-    },
-    {
-      "epoch": 3.471105262302806,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 213650
-    },
-    {
-      "epoch": 3.4712677292001755,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.202,
-      "step": 213660
-    },
-    {
-      "epoch": 3.471430196097545,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 213670
-    },
-    {
-      "epoch": 3.4715926629949148,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 213680
-    },
-    {
-      "epoch": 3.4717551298922844,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2218,
-      "step": 213690
-    },
-    {
-      "epoch": 3.471917596789654,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2212,
-      "step": 213700
-    },
-    {
-      "epoch": 3.4720800636870237,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1746,
-      "step": 213710
-    },
-    {
-      "epoch": 3.4722425305843934,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 213720
-    },
-    {
-      "epoch": 3.472404997481763,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2089,
-      "step": 213730
-    },
-    {
-      "epoch": 3.4725674643791327,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 213740
-    },
-    {
-      "epoch": 3.4727299312765023,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1716,
-      "step": 213750
-    },
-    {
-      "epoch": 3.472892398173872,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 213760
-    },
-    {
-      "epoch": 3.4730548650712416,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1758,
-      "step": 213770
-    },
-    {
-      "epoch": 3.4732173319686113,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 213780
-    },
-    {
-      "epoch": 3.473379798865981,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 213790
-    },
-    {
-      "epoch": 3.4735422657633506,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 213800
-    },
-    {
-      "epoch": 3.47370473266072,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2089,
-      "step": 213810
-    },
-    {
-      "epoch": 3.47386719955809,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 213820
-    },
-    {
-      "epoch": 3.4740296664554595,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 213830
-    },
-    {
-      "epoch": 3.474192133352829,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2409,
-      "step": 213840
-    },
-    {
-      "epoch": 3.474354600250199,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.215,
-      "step": 213850
-    },
-    {
-      "epoch": 3.474517067147569,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2244,
-      "step": 213860
-    },
-    {
-      "epoch": 3.474679534044938,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 213870
-    },
-    {
-      "epoch": 3.474842000942308,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2254,
-      "step": 213880
-    },
-    {
-      "epoch": 3.4750044678396774,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2044,
-      "step": 213890
-    },
-    {
-      "epoch": 3.4751669347370475,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2145,
-      "step": 213900
-    },
-    {
-      "epoch": 3.475329401634417,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2136,
-      "step": 213910
-    },
-    {
-      "epoch": 3.475491868531787,
-      "grad_norm": 8.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 213920
-    },
-    {
-      "epoch": 3.4756543354291565,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2316,
-      "step": 213930
-    },
-    {
-      "epoch": 3.475816802326526,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.199,
-      "step": 213940
-    },
-    {
-      "epoch": 3.4759792692238958,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2064,
-      "step": 213950
-    },
-    {
-      "epoch": 3.4761417361212654,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 213960
-    },
-    {
-      "epoch": 3.476304203018635,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2033,
-      "step": 213970
-    },
-    {
-      "epoch": 3.4764666699160047,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1871,
-      "step": 213980
-    },
-    {
-      "epoch": 3.4766291368133744,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 213990
-    },
-    {
-      "epoch": 3.476791603710744,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2072,
-      "step": 214000
-    },
-    {
-      "epoch": 3.4769540706081137,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 214010
-    },
-    {
-      "epoch": 3.4771165375054833,
-      "grad_norm": 8.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1989,
-      "step": 214020
-    },
-    {
-      "epoch": 3.477279004402853,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1895,
-      "step": 214030
-    },
-    {
-      "epoch": 3.4774414713002226,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 214040
-    },
-    {
-      "epoch": 3.4776039381975923,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.228,
-      "step": 214050
-    },
-    {
-      "epoch": 3.477766405094962,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 214060
-    },
-    {
-      "epoch": 3.4779288719923316,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2256,
-      "step": 214070
-    },
-    {
-      "epoch": 3.4780913388897012,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 214080
-    },
-    {
-      "epoch": 3.478253805787071,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 214090
-    },
-    {
-      "epoch": 3.4784162726844405,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 214100
-    },
-    {
-      "epoch": 3.47857873958181,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 214110
-    },
-    {
-      "epoch": 3.47874120647918,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 214120
-    },
-    {
-      "epoch": 3.4789036733765495,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 214130
-    },
-    {
-      "epoch": 3.479066140273919,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 214140
-    },
-    {
-      "epoch": 3.479228607171289,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 214150
-    },
-    {
-      "epoch": 3.4793910740686584,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 214160
-    },
-    {
-      "epoch": 3.479553540966028,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1624,
-      "step": 214170
-    },
-    {
-      "epoch": 3.4797160078633977,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1611,
-      "step": 214180
-    },
-    {
-      "epoch": 3.4798784747607674,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 214190
-    },
-    {
-      "epoch": 3.480040941658137,
-      "grad_norm": 3.421875,
-      "learning_rate": 5e-05,
-      "loss": 0.1728,
-      "step": 214200
-    },
-    {
-      "epoch": 3.4802034085555067,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 214210
-    },
-    {
-      "epoch": 3.4803658754528763,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 214220
-    },
-    {
-      "epoch": 3.480528342350246,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 214230
-    },
-    {
-      "epoch": 3.4806908092476156,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 214240
-    },
-    {
-      "epoch": 3.4808532761449853,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 214250
-    },
-    {
-      "epoch": 3.481015743042355,
-      "grad_norm": 7.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 214260
-    },
-    {
-      "epoch": 3.4811782099397246,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1998,
-      "step": 214270
-    },
-    {
-      "epoch": 3.4813406768370943,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2175,
-      "step": 214280
-    },
-    {
-      "epoch": 3.4815031437344643,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 214290
-    },
-    {
-      "epoch": 3.4816656106318336,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2032,
-      "step": 214300
-    },
-    {
-      "epoch": 3.4818280775292036,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.227,
-      "step": 214310
-    },
-    {
-      "epoch": 3.481990544426573,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2246,
-      "step": 214320
-    },
-    {
-      "epoch": 3.482153011323943,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2193,
-      "step": 214330
-    },
-    {
-      "epoch": 3.4823154782213126,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2146,
-      "step": 214340
-    },
-    {
-      "epoch": 3.4824779451186822,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 214350
-    },
-    {
-      "epoch": 3.482640412016052,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 214360
-    },
-    {
-      "epoch": 3.4828028789134216,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2142,
-      "step": 214370
-    },
-    {
-      "epoch": 3.482965345810791,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2268,
-      "step": 214380
-    },
-    {
-      "epoch": 3.483127812708161,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 214390
-    },
-    {
-      "epoch": 3.4832902796055305,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2339,
-      "step": 214400
-    },
-    {
-      "epoch": 3.4834527465029,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2374,
-      "step": 214410
-    },
-    {
-      "epoch": 3.48361521340027,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 214420
-    },
-    {
-      "epoch": 3.4837776802976395,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 214430
-    },
-    {
-      "epoch": 3.483940147195009,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 214440
-    },
-    {
-      "epoch": 3.4841026140923788,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2143,
-      "step": 214450
-    },
-    {
-      "epoch": 3.4842650809897484,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 214460
-    },
-    {
-      "epoch": 3.484427547887118,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 214470
-    },
-    {
-      "epoch": 3.4845900147844877,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2164,
-      "step": 214480
-    },
-    {
-      "epoch": 3.4847524816818574,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 214490
-    },
-    {
-      "epoch": 3.484914948579227,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 214500
-    },
-    {
-      "epoch": 3.4850774154765967,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 214510
-    },
-    {
-      "epoch": 3.4852398823739663,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 214520
-    },
-    {
-      "epoch": 3.485402349271336,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2097,
-      "step": 214530
-    },
-    {
-      "epoch": 3.4855648161687056,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 214540
-    },
-    {
-      "epoch": 3.4857272830660753,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 214550
-    },
-    {
-      "epoch": 3.485889749963445,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2319,
-      "step": 214560
-    },
-    {
-      "epoch": 3.4860522168608146,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 214570
-    },
-    {
-      "epoch": 3.486214683758184,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2449,
-      "step": 214580
-    },
-    {
-      "epoch": 3.486377150655554,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 214590
-    },
-    {
-      "epoch": 3.4865396175529235,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2464,
-      "step": 214600
-    },
-    {
-      "epoch": 3.486702084450293,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2467,
-      "step": 214610
-    },
-    {
-      "epoch": 3.486864551347663,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2247,
-      "step": 214620
-    },
-    {
-      "epoch": 3.4870270182450325,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2171,
-      "step": 214630
-    },
-    {
-      "epoch": 3.487189485142402,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 214640
-    },
-    {
-      "epoch": 3.4873519520397718,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.232,
-      "step": 214650
-    },
-    {
-      "epoch": 3.4875144189371414,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2085,
-      "step": 214660
-    },
-    {
-      "epoch": 3.487676885834511,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 214670
-    },
-    {
-      "epoch": 3.4878393527318807,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 214680
-    },
-    {
-      "epoch": 3.4880018196292504,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2121,
-      "step": 214690
-    },
-    {
-      "epoch": 3.48816428652662,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2352,
-      "step": 214700
-    },
-    {
-      "epoch": 3.4883267534239897,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 214710
-    },
-    {
-      "epoch": 3.4884892203213593,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 214720
-    },
-    {
-      "epoch": 3.488651687218729,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 214730
-    },
-    {
-      "epoch": 3.488814154116099,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 214740
-    },
-    {
-      "epoch": 3.4889766210134683,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 214750
-    },
-    {
-      "epoch": 3.4891390879108384,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1737,
-      "step": 214760
-    },
-    {
-      "epoch": 3.4893015548082076,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 214770
-    },
-    {
-      "epoch": 3.4894640217055777,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2217,
-      "step": 214780
-    },
-    {
-      "epoch": 3.4896264886029473,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 214790
-    },
-    {
-      "epoch": 3.489788955500317,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 214800
-    },
-    {
-      "epoch": 3.4899514223976866,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1905,
-      "step": 214810
-    },
-    {
-      "epoch": 3.4901138892950563,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2195,
-      "step": 214820
-    },
-    {
-      "epoch": 3.490276356192426,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2,
-      "step": 214830
-    },
-    {
-      "epoch": 3.4904388230897956,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2011,
-      "step": 214840
-    },
-    {
-      "epoch": 3.4906012899871652,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 214850
-    },
-    {
-      "epoch": 3.490763756884535,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2137,
-      "step": 214860
-    },
-    {
-      "epoch": 3.4909262237819045,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2253,
-      "step": 214870
-    },
-    {
-      "epoch": 3.491088690679274,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1915,
-      "step": 214880
-    },
-    {
-      "epoch": 3.491251157576644,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 214890
-    },
-    {
-      "epoch": 3.4914136244740135,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 214900
-    },
-    {
-      "epoch": 3.491576091371383,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 214910
-    },
-    {
-      "epoch": 3.491738558268753,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2137,
-      "step": 214920
-    },
-    {
-      "epoch": 3.4919010251661224,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2264,
-      "step": 214930
-    },
-    {
-      "epoch": 3.492063492063492,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2129,
-      "step": 214940
-    },
-    {
-      "epoch": 3.4922259589608617,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 214950
-    },
-    {
-      "epoch": 3.4923884258582314,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1894,
-      "step": 214960
-    },
-    {
-      "epoch": 3.492550892755601,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2089,
-      "step": 214970
-    },
-    {
-      "epoch": 3.4927133596529707,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2102,
-      "step": 214980
-    },
-    {
-      "epoch": 3.4928758265503403,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 214990
-    },
-    {
-      "epoch": 3.49303829344771,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 215000
-    },
-    {
-      "epoch": 3.4932007603450796,
-      "grad_norm": 7.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 215010
-    },
-    {
-      "epoch": 3.4933632272424493,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2133,
-      "step": 215020
-    },
-    {
-      "epoch": 3.493525694139819,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 215030
-    },
-    {
-      "epoch": 3.4936881610371886,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 215040
-    },
-    {
-      "epoch": 3.4938506279345583,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2067,
-      "step": 215050
-    },
-    {
-      "epoch": 3.494013094831928,
-      "grad_norm": 9.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 215060
-    },
-    {
-      "epoch": 3.4941755617292976,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2309,
-      "step": 215070
-    },
-    {
-      "epoch": 3.494338028626667,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1952,
-      "step": 215080
-    },
-    {
-      "epoch": 3.494500495524037,
-      "grad_norm": 8.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2358,
-      "step": 215090
-    },
-    {
-      "epoch": 3.4946629624214065,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 215100
-    },
-    {
-      "epoch": 3.494825429318776,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 215110
-    },
-    {
-      "epoch": 3.494987896216146,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 215120
-    },
-    {
-      "epoch": 3.4951503631135155,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2086,
-      "step": 215130
-    },
-    {
-      "epoch": 3.495312830010885,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 215140
-    },
-    {
-      "epoch": 3.4954752969082548,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2189,
-      "step": 215150
-    },
-    {
-      "epoch": 3.4956377638056244,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 215160
-    },
-    {
-      "epoch": 3.4958002307029945,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 215170
-    },
-    {
-      "epoch": 3.4959626976003637,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 215180
-    },
-    {
-      "epoch": 3.496125164497734,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2222,
-      "step": 215190
-    },
-    {
-      "epoch": 3.496287631395103,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 215200
-    },
-    {
-      "epoch": 3.496450098292473,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 215210
-    },
-    {
-      "epoch": 3.4966125651898428,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 215220
-    },
-    {
-      "epoch": 3.4967750320872124,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 215230
-    },
-    {
-      "epoch": 3.496937498984582,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 215240
-    },
-    {
-      "epoch": 3.4970999658819517,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.193,
-      "step": 215250
-    },
-    {
-      "epoch": 3.4972624327793214,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 215260
-    },
-    {
-      "epoch": 3.497424899676691,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1865,
-      "step": 215270
-    },
-    {
-      "epoch": 3.4975873665740607,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 215280
-    },
-    {
-      "epoch": 3.4977498334714303,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 215290
-    },
-    {
-      "epoch": 3.4979123003688,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 215300
-    },
-    {
-      "epoch": 3.4980747672661696,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 215310
-    },
-    {
-      "epoch": 3.4982372341635393,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1845,
-      "step": 215320
-    },
-    {
-      "epoch": 3.498399701060909,
-      "grad_norm": 7.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 215330
-    },
-    {
-      "epoch": 3.4985621679582786,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 215340
-    },
-    {
-      "epoch": 3.498724634855648,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 215350
-    },
-    {
-      "epoch": 3.498887101753018,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 215360
-    },
-    {
-      "epoch": 3.4990495686503875,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 215370
-    },
-    {
-      "epoch": 3.499212035547757,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 215380
-    },
-    {
-      "epoch": 3.499374502445127,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 215390
-    },
-    {
-      "epoch": 3.4995369693424965,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 215400
-    },
-    {
-      "epoch": 3.499699436239866,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 215410
-    },
-    {
-      "epoch": 3.4998619031372358,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 215420
-    },
-    {
-      "epoch": 3.5000243700346054,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2255,
-      "step": 215430
-    },
-    {
-      "epoch": 3.500186836931975,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.221,
-      "step": 215440
-    },
-    {
-      "epoch": 3.5003493038293447,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2121,
-      "step": 215450
-    },
-    {
-      "epoch": 3.5005117707267144,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2244,
-      "step": 215460
-    },
-    {
-      "epoch": 3.500674237624084,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2307,
-      "step": 215470
-    },
-    {
-      "epoch": 3.5008367045214537,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 215480
-    },
-    {
-      "epoch": 3.5009991714188233,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 215490
-    },
-    {
-      "epoch": 3.501161638316193,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1944,
-      "step": 215500
-    },
-    {
-      "epoch": 3.5013241052135626,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 215510
-    },
-    {
-      "epoch": 3.5014865721109323,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 215520
-    },
-    {
-      "epoch": 3.501649039008302,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 215530
-    },
-    {
-      "epoch": 3.5018115059056716,
-      "grad_norm": 7.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 215540
-    },
-    {
-      "epoch": 3.5019739728030412,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 215550
-    },
-    {
-      "epoch": 3.502136439700411,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2136,
-      "step": 215560
-    },
-    {
-      "epoch": 3.5022989065977805,
-      "grad_norm": 7.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2367,
-      "step": 215570
-    },
-    {
-      "epoch": 3.5024613734951506,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2368,
-      "step": 215580
-    },
-    {
-      "epoch": 3.50262384039252,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1954,
-      "step": 215590
-    },
-    {
-      "epoch": 3.50278630728989,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2089,
-      "step": 215600
-    },
-    {
-      "epoch": 3.502948774187259,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2193,
-      "step": 215610
-    },
-    {
-      "epoch": 3.5031112410846292,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 215620
-    },
-    {
-      "epoch": 3.5032737079819984,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 215630
-    },
-    {
-      "epoch": 3.5034361748793685,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 215640
-    },
-    {
-      "epoch": 3.5035986417767377,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 215650
-    },
-    {
-      "epoch": 3.503761108674108,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1978,
-      "step": 215660
-    },
-    {
-      "epoch": 3.503923575571477,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 215670
-    },
-    {
-      "epoch": 3.504086042468847,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1854,
-      "step": 215680
-    },
-    {
-      "epoch": 3.504248509366217,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 215690
-    },
-    {
-      "epoch": 3.5044109762635864,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1887,
-      "step": 215700
-    },
-    {
-      "epoch": 3.504573443160956,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1667,
-      "step": 215710
-    },
-    {
-      "epoch": 3.5047359100583257,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1755,
-      "step": 215720
-    },
-    {
-      "epoch": 3.5048983769556954,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 215730
-    },
-    {
-      "epoch": 3.505060843853065,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.17,
-      "step": 215740
-    },
-    {
-      "epoch": 3.5052233107504347,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1976,
-      "step": 215750
-    },
-    {
-      "epoch": 3.5053857776478043,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1765,
-      "step": 215760
-    },
-    {
-      "epoch": 3.505548244545174,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 215770
-    },
-    {
-      "epoch": 3.5057107114425436,
-      "grad_norm": 8.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2406,
-      "step": 215780
-    },
-    {
-      "epoch": 3.5058731783399133,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2143,
-      "step": 215790
-    },
-    {
-      "epoch": 3.506035645237283,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 215800
-    },
-    {
-      "epoch": 3.5061981121346526,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 215810
-    },
-    {
-      "epoch": 3.5063605790320223,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 215820
-    },
-    {
-      "epoch": 3.506523045929392,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 215830
-    },
-    {
-      "epoch": 3.5066855128267616,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 215840
-    },
-    {
-      "epoch": 3.506847979724131,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2146,
-      "step": 215850
-    },
-    {
-      "epoch": 3.507010446621501,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 215860
-    },
-    {
-      "epoch": 3.5071729135188705,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1963,
-      "step": 215870
-    },
-    {
-      "epoch": 3.50733538041624,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 215880
-    },
-    {
-      "epoch": 3.50749784731361,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 215890
-    },
-    {
-      "epoch": 3.5076603142109795,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1772,
-      "step": 215900
-    },
-    {
-      "epoch": 3.507822781108349,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 215910
-    },
-    {
-      "epoch": 3.5079852480057188,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2336,
-      "step": 215920
-    },
-    {
-      "epoch": 3.5081477149030884,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 215930
-    },
-    {
-      "epoch": 3.508310181800458,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2139,
-      "step": 215940
-    },
-    {
-      "epoch": 3.5084726486978277,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 215950
-    },
-    {
-      "epoch": 3.5086351155951974,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1822,
-      "step": 215960
-    },
-    {
-      "epoch": 3.508797582492567,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1848,
-      "step": 215970
-    },
-    {
-      "epoch": 3.5089600493899367,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.163,
-      "step": 215980
-    },
-    {
-      "epoch": 3.5091225162873063,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1725,
-      "step": 215990
-    },
-    {
-      "epoch": 3.509284983184676,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1863,
-      "step": 216000
-    },
-    {
-      "epoch": 3.509447450082046,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.175,
-      "step": 216010
-    },
-    {
-      "epoch": 3.5096099169794153,
-      "grad_norm": 7.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 216020
-    },
-    {
-      "epoch": 3.5097723838767854,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 216030
-    },
-    {
-      "epoch": 3.5099348507741546,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 216040
-    },
-    {
-      "epoch": 3.5100973176715247,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 216050
-    },
-    {
-      "epoch": 3.510259784568894,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1713,
-      "step": 216060
-    },
-    {
-      "epoch": 3.510422251466264,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 216070
-    },
-    {
-      "epoch": 3.510584718363633,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1759,
-      "step": 216080
-    },
-    {
-      "epoch": 3.5107471852610033,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 216090
-    },
-    {
-      "epoch": 3.5109096521583725,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 216100
-    },
-    {
-      "epoch": 3.5110721190557426,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 216110
-    },
-    {
-      "epoch": 3.511234585953112,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 216120
-    },
-    {
-      "epoch": 3.511397052850482,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 216130
-    },
-    {
-      "epoch": 3.5115595197478515,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 216140
-    },
-    {
-      "epoch": 3.511721986645221,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 216150
-    },
-    {
-      "epoch": 3.511884453542591,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1952,
-      "step": 216160
-    },
-    {
-      "epoch": 3.5120469204399605,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1866,
-      "step": 216170
-    },
-    {
-      "epoch": 3.51220938733733,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1901,
-      "step": 216180
-    },
-    {
-      "epoch": 3.5123718542346998,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 216190
-    },
-    {
-      "epoch": 3.5125343211320694,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 216200
-    },
-    {
-      "epoch": 3.512696788029439,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.236,
-      "step": 216210
-    },
-    {
-      "epoch": 3.5128592549268087,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2216,
-      "step": 216220
-    },
-    {
-      "epoch": 3.5130217218241784,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 216230
-    },
-    {
-      "epoch": 3.513184188721548,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 216240
-    },
-    {
-      "epoch": 3.5133466556189177,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 216250
-    },
-    {
-      "epoch": 3.5135091225162873,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 216260
-    },
-    {
-      "epoch": 3.513671589413657,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 216270
-    },
-    {
-      "epoch": 3.5138340563110266,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 216280
-    },
-    {
-      "epoch": 3.5139965232083963,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2152,
-      "step": 216290
-    },
-    {
-      "epoch": 3.514158990105766,
-      "grad_norm": 7.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 216300
-    },
-    {
-      "epoch": 3.5143214570031356,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2154,
-      "step": 216310
-    },
-    {
-      "epoch": 3.5144839239005052,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.218,
-      "step": 216320
-    },
-    {
-      "epoch": 3.514646390797875,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 216330
-    },
-    {
-      "epoch": 3.5148088576952445,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 216340
-    },
-    {
-      "epoch": 3.514971324592614,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2166,
-      "step": 216350
-    },
-    {
-      "epoch": 3.515133791489984,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 216360
-    },
-    {
-      "epoch": 3.5152962583873535,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2139,
-      "step": 216370
-    },
-    {
-      "epoch": 3.515458725284723,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2251,
-      "step": 216380
-    },
-    {
-      "epoch": 3.515621192182093,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2082,
-      "step": 216390
-    },
-    {
-      "epoch": 3.5157836590794624,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2383,
-      "step": 216400
-    },
-    {
-      "epoch": 3.515946125976832,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 216410
-    },
-    {
-      "epoch": 3.5161085928742017,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2021,
-      "step": 216420
-    },
-    {
-      "epoch": 3.5162710597715714,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 216430
-    },
-    {
-      "epoch": 3.516433526668941,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 216440
-    },
-    {
-      "epoch": 3.5165959935663107,
-      "grad_norm": 7.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2021,
-      "step": 216450
-    },
-    {
-      "epoch": 3.516758460463681,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 216460
-    },
-    {
-      "epoch": 3.51692092736105,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 216470
-    },
-    {
-      "epoch": 3.51708339425842,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 216480
-    },
-    {
-      "epoch": 3.5172458611557893,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2145,
-      "step": 216490
-    },
-    {
-      "epoch": 3.5174083280531594,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2421,
-      "step": 216500
-    },
-    {
-      "epoch": 3.5175707949505286,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 216510
-    },
-    {
-      "epoch": 3.5177332618478987,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 216520
-    },
-    {
-      "epoch": 3.517895728745268,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2347,
-      "step": 216530
-    },
-    {
-      "epoch": 3.518058195642638,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 216540
-    },
-    {
-      "epoch": 3.518220662540007,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2361,
-      "step": 216550
-    },
-    {
-      "epoch": 3.5183831294373773,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2124,
-      "step": 216560
-    },
-    {
-      "epoch": 3.518545596334747,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 216570
-    },
-    {
-      "epoch": 3.5187080632321166,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 216580
-    },
-    {
-      "epoch": 3.5188705301294863,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 216590
-    },
-    {
-      "epoch": 3.519032997026856,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.191,
-      "step": 216600
-    },
-    {
-      "epoch": 3.5191954639242256,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 216610
-    },
-    {
-      "epoch": 3.519357930821595,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1771,
-      "step": 216620
-    },
-    {
-      "epoch": 3.519520397718965,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 216630
-    },
-    {
-      "epoch": 3.5196828646163345,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 216640
-    },
-    {
-      "epoch": 3.519845331513704,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1877,
-      "step": 216650
-    },
-    {
-      "epoch": 3.520007798411074,
-      "grad_norm": 7.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 216660
-    },
-    {
-      "epoch": 3.5201702653084435,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 216670
-    },
-    {
-      "epoch": 3.520332732205813,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1794,
-      "step": 216680
-    },
-    {
-      "epoch": 3.5204951991031828,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 216690
-    },
-    {
-      "epoch": 3.5206576660005524,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1763,
-      "step": 216700
-    },
-    {
-      "epoch": 3.520820132897922,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1813,
-      "step": 216710
-    },
-    {
-      "epoch": 3.5209825997952917,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1895,
-      "step": 216720
-    },
-    {
-      "epoch": 3.5211450666926614,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1769,
-      "step": 216730
-    },
-    {
-      "epoch": 3.521307533590031,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 216740
-    },
-    {
-      "epoch": 3.5214700004874007,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2129,
-      "step": 216750
-    },
-    {
-      "epoch": 3.5216324673847703,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 216760
-    },
-    {
-      "epoch": 3.52179493428214,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.184,
-      "step": 216770
-    },
-    {
-      "epoch": 3.5219574011795096,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 216780
-    },
-    {
-      "epoch": 3.5221198680768793,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 216790
-    },
-    {
-      "epoch": 3.522282334974249,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 216800
-    },
-    {
-      "epoch": 3.5224448018716186,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 216810
-    },
-    {
-      "epoch": 3.522607268768988,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 216820
-    },
-    {
-      "epoch": 3.522769735666358,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 216830
-    },
-    {
-      "epoch": 3.5229322025637275,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1852,
-      "step": 216840
-    },
-    {
-      "epoch": 3.523094669461097,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1838,
-      "step": 216850
-    },
-    {
-      "epoch": 3.523257136358467,
-      "grad_norm": 7.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2153,
-      "step": 216860
-    },
-    {
-      "epoch": 3.5234196032558365,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1842,
-      "step": 216870
-    },
-    {
-      "epoch": 3.523582070153206,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2109,
-      "step": 216880
-    },
-    {
-      "epoch": 3.523744537050576,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1734,
-      "step": 216890
-    },
-    {
-      "epoch": 3.5239070039479454,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1708,
-      "step": 216900
-    },
-    {
-      "epoch": 3.5240694708453155,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1745,
-      "step": 216910
-    },
-    {
-      "epoch": 3.5242319377426847,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1706,
-      "step": 216920
-    },
-    {
-      "epoch": 3.524394404640055,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 216930
-    },
-    {
-      "epoch": 3.524556871537424,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 216940
-    },
-    {
-      "epoch": 3.524719338434794,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 216950
-    },
-    {
-      "epoch": 3.5248818053321633,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 216960
-    },
-    {
-      "epoch": 3.5250442722295334,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 216970
-    },
-    {
-      "epoch": 3.5252067391269026,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 216980
-    },
-    {
-      "epoch": 3.5253692060242727,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 216990
-    },
-    {
-      "epoch": 3.5255316729216424,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 217000
-    },
-    {
-      "epoch": 3.525694139819012,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.192,
-      "step": 217010
-    },
-    {
-      "epoch": 3.5258566067163817,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1983,
-      "step": 217020
-    },
-    {
-      "epoch": 3.5260190736137513,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 217030
-    },
-    {
-      "epoch": 3.526181540511121,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2293,
-      "step": 217040
-    },
-    {
-      "epoch": 3.5263440074084906,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.216,
-      "step": 217050
-    },
-    {
-      "epoch": 3.5265064743058603,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 217060
-    },
-    {
-      "epoch": 3.52666894120323,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 217070
-    },
-    {
-      "epoch": 3.5268314081005996,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 217080
-    },
-    {
-      "epoch": 3.5269938749979692,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2353,
-      "step": 217090
-    },
-    {
-      "epoch": 3.527156341895339,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 217100
-    },
-    {
-      "epoch": 3.5273188087927085,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2211,
-      "step": 217110
-    },
-    {
-      "epoch": 3.527481275690078,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2107,
-      "step": 217120
-    },
-    {
-      "epoch": 3.527643742587448,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2112,
-      "step": 217130
-    },
-    {
-      "epoch": 3.5278062094848175,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 217140
-    },
-    {
-      "epoch": 3.527968676382187,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 217150
-    },
-    {
-      "epoch": 3.528131143279557,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 217160
-    },
-    {
-      "epoch": 3.5282936101769264,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2047,
-      "step": 217170
-    },
-    {
-      "epoch": 3.528456077074296,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 217180
-    },
-    {
-      "epoch": 3.5286185439716657,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 217190
-    },
-    {
-      "epoch": 3.5287810108690354,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 217200
-    },
-    {
-      "epoch": 3.528943477766405,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2258,
-      "step": 217210
-    },
-    {
-      "epoch": 3.5291059446637747,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2102,
-      "step": 217220
-    },
-    {
-      "epoch": 3.5292684115611443,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2153,
-      "step": 217230
-    },
-    {
-      "epoch": 3.529430878458514,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 217240
-    },
-    {
-      "epoch": 3.5295933453558836,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1849,
-      "step": 217250
-    },
-    {
-      "epoch": 3.5297558122532533,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 217260
-    },
-    {
-      "epoch": 3.529918279150623,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1709,
-      "step": 217270
-    },
-    {
-      "epoch": 3.5300807460479926,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 217280
-    },
-    {
-      "epoch": 3.5302432129453623,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 217290
-    },
-    {
-      "epoch": 3.530405679842732,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 217300
-    },
-    {
-      "epoch": 3.5305681467401016,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1808,
-      "step": 217310
-    },
-    {
-      "epoch": 3.530730613637471,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1812,
-      "step": 217320
-    },
-    {
-      "epoch": 3.530893080534841,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 217330
-    },
-    {
-      "epoch": 3.531055547432211,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1752,
-      "step": 217340
-    },
-    {
-      "epoch": 3.53121801432958,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 217350
-    },
-    {
-      "epoch": 3.5313804812269503,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 217360
-    },
-    {
-      "epoch": 3.5315429481243195,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 217370
-    },
-    {
-      "epoch": 3.5317054150216896,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1952,
-      "step": 217380
-    },
-    {
-      "epoch": 3.5318678819190588,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2302,
-      "step": 217390
-    },
-    {
-      "epoch": 3.532030348816429,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 217400
-    },
-    {
-      "epoch": 3.532192815713798,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 217410
-    },
-    {
-      "epoch": 3.532355282611168,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 217420
-    },
-    {
-      "epoch": 3.5325177495085374,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2085,
-      "step": 217430
-    },
-    {
-      "epoch": 3.5326802164059075,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 217440
-    },
-    {
-      "epoch": 3.532842683303277,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 217450
-    },
-    {
-      "epoch": 3.5330051502006468,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1776,
-      "step": 217460
-    },
-    {
-      "epoch": 3.5331676170980164,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2053,
-      "step": 217470
-    },
-    {
-      "epoch": 3.533330083995386,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1907,
-      "step": 217480
-    },
-    {
-      "epoch": 3.5334925508927557,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1927,
-      "step": 217490
-    },
-    {
-      "epoch": 3.5336550177901254,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2206,
-      "step": 217500
-    },
-    {
-      "epoch": 3.533817484687495,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1793,
-      "step": 217510
-    },
-    {
-      "epoch": 3.5339799515848647,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1889,
-      "step": 217520
-    },
-    {
-      "epoch": 3.5341424184822343,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 217530
-    },
-    {
-      "epoch": 3.534304885379604,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1837,
-      "step": 217540
-    },
-    {
-      "epoch": 3.5344673522769736,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1835,
-      "step": 217550
-    },
-    {
-      "epoch": 3.5346298191743433,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1899,
-      "step": 217560
-    },
-    {
-      "epoch": 3.534792286071713,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 217570
-    },
-    {
-      "epoch": 3.5349547529690826,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 217580
-    },
-    {
-      "epoch": 3.535117219866452,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1811,
-      "step": 217590
-    },
-    {
-      "epoch": 3.535279686763822,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 217600
-    },
-    {
-      "epoch": 3.5354421536611915,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 217610
-    },
-    {
-      "epoch": 3.535604620558561,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 217620
-    },
-    {
-      "epoch": 3.535767087455931,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2611,
-      "step": 217630
-    },
-    {
-      "epoch": 3.5359295543533005,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 217640
-    },
-    {
-      "epoch": 3.53609202125067,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 217650
-    },
-    {
-      "epoch": 3.5362544881480398,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 217660
-    },
-    {
-      "epoch": 3.5364169550454094,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.199,
-      "step": 217670
-    },
-    {
-      "epoch": 3.536579421942779,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 217680
-    },
-    {
-      "epoch": 3.5367418888401487,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.198,
-      "step": 217690
-    },
-    {
-      "epoch": 3.5369043557375184,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2114,
-      "step": 217700
-    },
-    {
-      "epoch": 3.537066822634888,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 217710
-    },
-    {
-      "epoch": 3.5372292895322577,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 217720
-    },
-    {
-      "epoch": 3.5373917564296273,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 217730
-    },
-    {
-      "epoch": 3.537554223326997,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2383,
-      "step": 217740
-    },
-    {
-      "epoch": 3.5377166902243666,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 217750
-    },
-    {
-      "epoch": 3.5378791571217363,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1811,
-      "step": 217760
-    },
-    {
-      "epoch": 3.5380416240191064,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1911,
-      "step": 217770
-    },
-    {
-      "epoch": 3.5382040909164756,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 217780
-    },
-    {
-      "epoch": 3.5383665578138457,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1867,
-      "step": 217790
-    },
-    {
-      "epoch": 3.538529024711215,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 217800
-    },
-    {
-      "epoch": 3.538691491608585,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1978,
-      "step": 217810
-    },
-    {
-      "epoch": 3.538853958505954,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2318,
-      "step": 217820
-    },
-    {
-      "epoch": 3.5390164254033243,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 217830
-    },
-    {
-      "epoch": 3.5391788923006935,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2221,
-      "step": 217840
-    },
-    {
-      "epoch": 3.5393413591980636,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.232,
-      "step": 217850
-    },
-    {
-      "epoch": 3.539503826095433,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.225,
-      "step": 217860
-    },
-    {
-      "epoch": 3.539666292992803,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2143,
-      "step": 217870
-    },
-    {
-      "epoch": 3.5398287598901725,
-      "grad_norm": 3.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 217880
-    },
-    {
-      "epoch": 3.539991226787542,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2033,
-      "step": 217890
-    },
-    {
-      "epoch": 3.540153693684912,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2175,
-      "step": 217900
-    },
-    {
-      "epoch": 3.5403161605822815,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 217910
-    },
-    {
-      "epoch": 3.540478627479651,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1858,
-      "step": 217920
-    },
-    {
-      "epoch": 3.540641094377021,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1707,
-      "step": 217930
-    },
-    {
-      "epoch": 3.5408035612743904,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 217940
-    },
-    {
-      "epoch": 3.54096602817176,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.174,
-      "step": 217950
-    },
-    {
-      "epoch": 3.5411284950691297,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 217960
-    },
-    {
-      "epoch": 3.5412909619664994,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 217970
-    },
-    {
-      "epoch": 3.541453428863869,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2389,
-      "step": 217980
-    },
-    {
-      "epoch": 3.5416158957612387,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 217990
-    },
-    {
-      "epoch": 3.5417783626586083,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2201,
-      "step": 218000
-    },
-    {
-      "epoch": 3.541940829555978,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 218010
-    },
-    {
-      "epoch": 3.5421032964533476,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 218020
-    },
-    {
-      "epoch": 3.5422657633507173,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 218030
-    },
-    {
-      "epoch": 3.542428230248087,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 218040
-    },
-    {
-      "epoch": 3.5425906971454566,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 218050
-    },
-    {
-      "epoch": 3.5427531640428263,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 218060
-    },
-    {
-      "epoch": 3.542915630940196,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 218070
-    },
-    {
-      "epoch": 3.5430780978375656,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2293,
-      "step": 218080
-    },
-    {
-      "epoch": 3.543240564734935,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 218090
-    },
-    {
-      "epoch": 3.543403031632305,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 218100
-    },
-    {
-      "epoch": 3.5435654985296745,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 218110
-    },
-    {
-      "epoch": 3.543727965427044,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.21,
-      "step": 218120
-    },
-    {
-      "epoch": 3.543890432324414,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2107,
-      "step": 218130
-    },
-    {
-      "epoch": 3.5440528992217835,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 218140
-    },
-    {
-      "epoch": 3.544215366119153,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2047,
-      "step": 218150
-    },
-    {
-      "epoch": 3.5443778330165228,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 218160
-    },
-    {
-      "epoch": 3.5445402999138924,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 218170
-    },
-    {
-      "epoch": 3.544702766811262,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.181,
-      "step": 218180
-    },
-    {
-      "epoch": 3.5448652337086317,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1908,
-      "step": 218190
-    },
-    {
-      "epoch": 3.5450277006060014,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 218200
-    },
-    {
-      "epoch": 3.545190167503371,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2189,
-      "step": 218210
-    },
-    {
-      "epoch": 3.545352634400741,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2085,
-      "step": 218220
-    },
-    {
-      "epoch": 3.5455151012981103,
-      "grad_norm": 7.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2008,
-      "step": 218230
-    },
-    {
-      "epoch": 3.5456775681954804,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 218240
-    },
-    {
-      "epoch": 3.5458400350928496,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 218250
-    },
-    {
-      "epoch": 3.5460025019902197,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 218260
-    },
-    {
-      "epoch": 3.546164968887589,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2234,
-      "step": 218270
-    },
-    {
-      "epoch": 3.546327435784959,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 218280
-    },
-    {
-      "epoch": 3.546489902682328,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2173,
-      "step": 218290
-    },
-    {
-      "epoch": 3.5466523695796983,
-      "grad_norm": 7.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 218300
-    },
-    {
-      "epoch": 3.5468148364770675,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 218310
-    },
-    {
-      "epoch": 3.5469773033744376,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 218320
-    },
-    {
-      "epoch": 3.5471397702718073,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2324,
-      "step": 218330
-    },
-    {
-      "epoch": 3.547302237169177,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 218340
-    },
-    {
-      "epoch": 3.5474647040665466,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 218350
-    },
-    {
-      "epoch": 3.547627170963916,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 218360
-    },
-    {
-      "epoch": 3.547789637861286,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 218370
-    },
-    {
-      "epoch": 3.5479521047586555,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 218380
-    },
-    {
-      "epoch": 3.548114571656025,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 218390
-    },
-    {
-      "epoch": 3.548277038553395,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2093,
-      "step": 218400
-    },
-    {
-      "epoch": 3.5484395054507645,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 218410
-    },
-    {
-      "epoch": 3.548601972348134,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1954,
-      "step": 218420
-    },
-    {
-      "epoch": 3.5487644392455038,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 218430
-    },
-    {
-      "epoch": 3.5489269061428734,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.235,
-      "step": 218440
-    },
-    {
-      "epoch": 3.549089373040243,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.213,
-      "step": 218450
-    },
-    {
-      "epoch": 3.5492518399376127,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2137,
-      "step": 218460
-    },
-    {
-      "epoch": 3.5494143068349824,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.215,
-      "step": 218470
-    },
-    {
-      "epoch": 3.549576773732352,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 218480
-    },
-    {
-      "epoch": 3.5497392406297217,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2292,
-      "step": 218490
-    },
-    {
-      "epoch": 3.5499017075270913,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2322,
-      "step": 218500
-    },
-    {
-      "epoch": 3.550064174424461,
-      "grad_norm": 3.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2446,
-      "step": 218510
-    },
-    {
-      "epoch": 3.5502266413218306,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 218520
-    },
-    {
-      "epoch": 3.5503891082192003,
-      "grad_norm": 3.921875,
-      "learning_rate": 5e-05,
-      "loss": 0.1768,
-      "step": 218530
-    },
-    {
-      "epoch": 3.55055157511657,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2044,
-      "step": 218540
-    },
-    {
-      "epoch": 3.5507140420139396,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1991,
-      "step": 218550
-    },
-    {
-      "epoch": 3.5508765089113092,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2272,
-      "step": 218560
-    },
-    {
-      "epoch": 3.551038975808679,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 218570
-    },
-    {
-      "epoch": 3.5512014427060485,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 218580
-    },
-    {
-      "epoch": 3.551363909603418,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 218590
-    },
-    {
-      "epoch": 3.551526376500788,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 218600
-    },
-    {
-      "epoch": 3.5516888433981575,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2308,
-      "step": 218610
-    },
-    {
-      "epoch": 3.551851310295527,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 218620
-    },
-    {
-      "epoch": 3.552013777192897,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2312,
-      "step": 218630
-    },
-    {
-      "epoch": 3.5521762440902664,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 218640
-    },
-    {
-      "epoch": 3.5523387109876365,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 218650
-    },
-    {
-      "epoch": 3.5525011778850057,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2233,
-      "step": 218660
-    },
-    {
-      "epoch": 3.552663644782376,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2266,
-      "step": 218670
-    },
-    {
-      "epoch": 3.552826111679745,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 218680
-    },
-    {
-      "epoch": 3.552988578577115,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2413,
-      "step": 218690
-    },
-    {
-      "epoch": 3.5531510454744843,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 218700
-    },
-    {
-      "epoch": 3.5533135123718544,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2257,
-      "step": 218710
-    },
-    {
-      "epoch": 3.5534759792692237,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2378,
-      "step": 218720
-    },
-    {
-      "epoch": 3.5536384461665937,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2153,
-      "step": 218730
-    },
-    {
-      "epoch": 3.553800913063963,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 218740
-    },
-    {
-      "epoch": 3.553963379961333,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.221,
-      "step": 218750
-    },
-    {
-      "epoch": 3.5541258468587027,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1978,
-      "step": 218760
-    },
-    {
-      "epoch": 3.5542883137560723,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 218770
-    },
-    {
-      "epoch": 3.554450780653442,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 218780
-    },
-    {
-      "epoch": 3.5546132475508116,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 218790
-    },
-    {
-      "epoch": 3.5547757144481813,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 218800
-    },
-    {
-      "epoch": 3.554938181345551,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.198,
-      "step": 218810
-    },
-    {
-      "epoch": 3.5551006482429206,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 218820
-    },
-    {
-      "epoch": 3.5552631151402903,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2231,
-      "step": 218830
-    },
-    {
-      "epoch": 3.55542558203766,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 218840
-    },
-    {
-      "epoch": 3.5555880489350296,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 218850
-    },
-    {
-      "epoch": 3.555750515832399,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2144,
-      "step": 218860
-    },
-    {
-      "epoch": 3.555912982729769,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.217,
-      "step": 218870
-    },
-    {
-      "epoch": 3.5560754496271385,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 218880
-    },
-    {
-      "epoch": 3.556237916524508,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 218890
-    },
-    {
-      "epoch": 3.556400383421878,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 218900
-    },
-    {
-      "epoch": 3.5565628503192475,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 218910
-    },
-    {
-      "epoch": 3.556725317216617,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 218920
-    },
-    {
-      "epoch": 3.5568877841139868,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2265,
-      "step": 218930
-    },
-    {
-      "epoch": 3.5570502510113564,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 218940
-    },
-    {
-      "epoch": 3.557212717908726,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1983,
-      "step": 218950
-    },
-    {
-      "epoch": 3.5573751848060957,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 218960
-    },
-    {
-      "epoch": 3.5575376517034654,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 218970
-    },
-    {
-      "epoch": 3.557700118600835,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 218980
-    },
-    {
-      "epoch": 3.5578625854982047,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 218990
-    },
-    {
-      "epoch": 3.5580250523955743,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1967,
-      "step": 219000
-    },
-    {
-      "epoch": 3.558187519292944,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 219010
-    },
-    {
-      "epoch": 3.5583499861903136,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 219020
-    },
-    {
-      "epoch": 3.5585124530876833,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 219030
-    },
-    {
-      "epoch": 3.558674919985053,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 219040
-    },
-    {
-      "epoch": 3.5588373868824226,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2004,
-      "step": 219050
-    },
-    {
-      "epoch": 3.558999853779792,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2175,
-      "step": 219060
-    },
-    {
-      "epoch": 3.559162320677162,
-      "grad_norm": 3.546875,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 219070
-    },
-    {
-      "epoch": 3.5593247875745315,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2115,
-      "step": 219080
-    },
-    {
-      "epoch": 3.559487254471901,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.197,
-      "step": 219090
-    },
-    {
-      "epoch": 3.5596497213692713,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 219100
-    },
-    {
-      "epoch": 3.5598121882666405,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 219110
-    },
-    {
-      "epoch": 3.5599746551640106,
-      "grad_norm": 7.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 219120
-    },
-    {
-      "epoch": 3.5601371220613798,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 219130
-    },
-    {
-      "epoch": 3.56029958895875,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 219140
-    },
-    {
-      "epoch": 3.560462055856119,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2174,
-      "step": 219150
-    },
-    {
-      "epoch": 3.560624522753489,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 219160
-    },
-    {
-      "epoch": 3.5607869896508584,
-      "grad_norm": 3.671875,
-      "learning_rate": 5e-05,
-      "loss": 0.2136,
-      "step": 219170
-    },
-    {
-      "epoch": 3.5609494565482285,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 219180
-    },
-    {
-      "epoch": 3.5611119234455977,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 219190
-    },
-    {
-      "epoch": 3.5612743903429678,
-      "grad_norm": 7.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 219200
-    },
-    {
-      "epoch": 3.5614368572403374,
-      "grad_norm": 7.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2159,
-      "step": 219210
-    },
-    {
-      "epoch": 3.561599324137707,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1855,
-      "step": 219220
-    },
-    {
-      "epoch": 3.5617617910350767,
-      "grad_norm": 7.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 219230
-    },
-    {
-      "epoch": 3.5619242579324464,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 219240
-    },
-    {
-      "epoch": 3.562086724829816,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 219250
-    },
-    {
-      "epoch": 3.5622491917271857,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1866,
-      "step": 219260
-    },
-    {
-      "epoch": 3.5624116586245553,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 219270
-    },
-    {
-      "epoch": 3.562574125521925,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1948,
-      "step": 219280
-    },
-    {
-      "epoch": 3.5627365924192946,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2155,
-      "step": 219290
-    },
-    {
-      "epoch": 3.5628990593166643,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 219300
-    },
-    {
-      "epoch": 3.563061526214034,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 219310
-    },
-    {
-      "epoch": 3.5632239931114036,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 219320
-    },
-    {
-      "epoch": 3.5633864600087732,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 219330
-    },
-    {
-      "epoch": 3.563548926906143,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 219340
-    },
-    {
-      "epoch": 3.5637113938035125,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2235,
-      "step": 219350
-    },
-    {
-      "epoch": 3.563873860700882,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2086,
-      "step": 219360
-    },
-    {
-      "epoch": 3.564036327598252,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 219370
-    },
-    {
-      "epoch": 3.5641987944956215,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2235,
-      "step": 219380
-    },
-    {
-      "epoch": 3.564361261392991,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 219390
-    },
-    {
-      "epoch": 3.564523728290361,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 219400
-    },
-    {
-      "epoch": 3.5646861951877304,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2186,
-      "step": 219410
-    },
-    {
-      "epoch": 3.5648486620851,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 219420
-    },
-    {
-      "epoch": 3.5650111289824697,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2132,
-      "step": 219430
-    },
-    {
-      "epoch": 3.5651735958798394,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 219440
-    },
-    {
-      "epoch": 3.565336062777209,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2097,
-      "step": 219450
-    },
-    {
-      "epoch": 3.5654985296745787,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1917,
-      "step": 219460
-    },
-    {
-      "epoch": 3.5656609965719483,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 219470
-    },
-    {
-      "epoch": 3.565823463469318,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1976,
-      "step": 219480
-    },
-    {
-      "epoch": 3.5659859303666877,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 219490
-    },
-    {
-      "epoch": 3.5661483972640573,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 219500
-    },
-    {
-      "epoch": 3.566310864161427,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 219510
-    },
-    {
-      "epoch": 3.5664733310587966,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1955,
-      "step": 219520
-    },
-    {
-      "epoch": 3.5666357979561667,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 219530
-    },
-    {
-      "epoch": 3.566798264853536,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 219540
-    },
-    {
-      "epoch": 3.566960731750906,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1888,
-      "step": 219550
-    },
-    {
-      "epoch": 3.567123198648275,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 219560
-    },
-    {
-      "epoch": 3.5672856655456453,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2179,
-      "step": 219570
-    },
-    {
-      "epoch": 3.5674481324430145,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 219580
-    },
-    {
-      "epoch": 3.5676105993403846,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 219590
-    },
-    {
-      "epoch": 3.567773066237754,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 219600
-    },
-    {
-      "epoch": 3.567935533135124,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 219610
-    },
-    {
-      "epoch": 3.568098000032493,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 219620
-    },
-    {
-      "epoch": 3.568260466929863,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 219630
-    },
-    {
-      "epoch": 3.568422933827233,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2115,
-      "step": 219640
-    },
-    {
-      "epoch": 3.5685854007246025,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2134,
-      "step": 219650
-    },
-    {
-      "epoch": 3.568747867621972,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 219660
-    },
-    {
-      "epoch": 3.568910334519342,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 219670
-    },
-    {
-      "epoch": 3.5690728014167115,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 219680
-    },
-    {
-      "epoch": 3.569235268314081,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1954,
-      "step": 219690
-    },
-    {
-      "epoch": 3.5693977352114508,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2265,
-      "step": 219700
-    },
-    {
-      "epoch": 3.5695602021088204,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 219710
-    },
-    {
-      "epoch": 3.56972266900619,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2147,
-      "step": 219720
-    },
-    {
-      "epoch": 3.5698851359035597,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 219730
-    },
-    {
-      "epoch": 3.5700476028009294,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.225,
-      "step": 219740
-    },
-    {
-      "epoch": 3.570210069698299,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 219750
-    },
-    {
-      "epoch": 3.5703725365956687,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 219760
-    },
-    {
-      "epoch": 3.5705350034930383,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 219770
-    },
-    {
-      "epoch": 3.570697470390408,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1809,
-      "step": 219780
-    },
-    {
-      "epoch": 3.5708599372877776,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1691,
-      "step": 219790
-    },
-    {
-      "epoch": 3.5710224041851473,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1672,
-      "step": 219800
-    },
-    {
-      "epoch": 3.571184871082517,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 219810
-    },
-    {
-      "epoch": 3.5713473379798866,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 219820
-    },
-    {
-      "epoch": 3.5715098048772562,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1887,
-      "step": 219830
-    },
-    {
-      "epoch": 3.571672271774626,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 219840
-    },
-    {
-      "epoch": 3.5718347386719955,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 219850
-    },
-    {
-      "epoch": 3.571997205569365,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 219860
-    },
-    {
-      "epoch": 3.572159672466735,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 219870
-    },
-    {
-      "epoch": 3.5723221393641045,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 219880
-    },
-    {
-      "epoch": 3.572484606261474,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2116,
-      "step": 219890
-    },
-    {
-      "epoch": 3.5726470731588438,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 219900
-    },
-    {
-      "epoch": 3.5728095400562134,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2381,
-      "step": 219910
-    },
-    {
-      "epoch": 3.572972006953583,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2223,
-      "step": 219920
-    },
-    {
-      "epoch": 3.5731344738509527,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1914,
-      "step": 219930
-    },
-    {
-      "epoch": 3.5732969407483224,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2338,
-      "step": 219940
-    },
-    {
-      "epoch": 3.573459407645692,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 219950
-    },
-    {
-      "epoch": 3.573621874543062,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2231,
-      "step": 219960
-    },
-    {
-      "epoch": 3.5737843414404313,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2343,
-      "step": 219970
-    },
-    {
-      "epoch": 3.5739468083378014,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2112,
-      "step": 219980
-    },
-    {
-      "epoch": 3.5741092752351706,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 219990
-    },
-    {
-      "epoch": 3.5742717421325407,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 220000
-    },
-    {
-      "epoch": 3.57443420902991,
-      "grad_norm": 3.984375,
-      "learning_rate": 5e-05,
-      "loss": 0.229,
-      "step": 220010
-    },
-    {
-      "epoch": 3.57459667592728,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 220020
-    },
-    {
-      "epoch": 3.5747591428246492,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 220030
-    },
-    {
-      "epoch": 3.5749216097220193,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2033,
-      "step": 220040
-    },
-    {
-      "epoch": 3.5750840766193885,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 220050
-    },
-    {
-      "epoch": 3.5752465435167586,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.221,
-      "step": 220060
-    },
-    {
-      "epoch": 3.5754090104141283,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2344,
-      "step": 220070
-    },
-    {
-      "epoch": 3.575571477311498,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2171,
-      "step": 220080
-    },
-    {
-      "epoch": 3.5757339442088676,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2233,
-      "step": 220090
-    },
-    {
-      "epoch": 3.5758964111062372,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2217,
-      "step": 220100
-    },
-    {
-      "epoch": 3.576058878003607,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2189,
-      "step": 220110
-    },
-    {
-      "epoch": 3.5762213449009765,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 220120
-    },
-    {
-      "epoch": 3.576383811798346,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 220130
-    },
-    {
-      "epoch": 3.576546278695716,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 220140
-    },
-    {
-      "epoch": 3.5767087455930855,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2356,
-      "step": 220150
-    },
-    {
-      "epoch": 3.576871212490455,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 220160
-    },
-    {
-      "epoch": 3.577033679387825,
-      "grad_norm": 3.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 220170
-    },
-    {
-      "epoch": 3.5771961462851944,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2294,
-      "step": 220180
-    },
-    {
-      "epoch": 3.577358613182564,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 220190
-    },
-    {
-      "epoch": 3.5775210800799337,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 220200
-    },
-    {
-      "epoch": 3.5776835469773034,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 220210
-    },
-    {
-      "epoch": 3.577846013874673,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1777,
-      "step": 220220
-    },
-    {
-      "epoch": 3.5780084807720427,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 220230
-    },
-    {
-      "epoch": 3.5781709476694123,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 220240
-    },
-    {
-      "epoch": 3.578333414566782,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 220250
-    },
-    {
-      "epoch": 3.5784958814641517,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2066,
-      "step": 220260
-    },
-    {
-      "epoch": 3.5786583483615213,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2295,
-      "step": 220270
-    },
-    {
-      "epoch": 3.578820815258891,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 220280
-    },
-    {
-      "epoch": 3.5789832821562606,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 220290
-    },
-    {
-      "epoch": 3.5791457490536303,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2114,
-      "step": 220300
-    },
-    {
-      "epoch": 3.579308215951,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2335,
-      "step": 220310
-    },
-    {
-      "epoch": 3.5794706828483696,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 220320
-    },
-    {
-      "epoch": 3.579633149745739,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 220330
-    },
-    {
-      "epoch": 3.579795616643109,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1968,
-      "step": 220340
-    },
-    {
-      "epoch": 3.5799580835404785,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.219,
-      "step": 220350
-    },
-    {
-      "epoch": 3.580120550437848,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 220360
-    },
-    {
-      "epoch": 3.580283017335218,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 220370
-    },
-    {
-      "epoch": 3.5804454842325875,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2297,
-      "step": 220380
-    },
-    {
-      "epoch": 3.580607951129957,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2318,
-      "step": 220390
-    },
-    {
-      "epoch": 3.5807704180273268,
-      "grad_norm": 3.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2332,
-      "step": 220400
-    },
-    {
-      "epoch": 3.580932884924697,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2321,
-      "step": 220410
-    },
-    {
-      "epoch": 3.581095351822066,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2327,
-      "step": 220420
-    },
-    {
-      "epoch": 3.581257818719436,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2443,
-      "step": 220430
-    },
-    {
-      "epoch": 3.5814202856168054,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2136,
-      "step": 220440
-    },
-    {
-      "epoch": 3.5815827525141755,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 220450
-    },
-    {
-      "epoch": 3.5817452194115447,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2076,
-      "step": 220460
-    },
-    {
-      "epoch": 3.5819076863089148,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.228,
-      "step": 220470
-    },
-    {
-      "epoch": 3.582070153206284,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 220480
-    },
-    {
-      "epoch": 3.582232620103654,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 220490
-    },
-    {
-      "epoch": 3.5823950870010233,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2145,
-      "step": 220500
-    },
-    {
-      "epoch": 3.5825575538983934,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2223,
-      "step": 220510
-    },
-    {
-      "epoch": 3.582720020795763,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2239,
-      "step": 220520
-    },
-    {
-      "epoch": 3.5828824876931327,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.252,
-      "step": 220530
-    },
-    {
-      "epoch": 3.5830449545905023,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2155,
-      "step": 220540
-    },
-    {
-      "epoch": 3.583207421487872,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2368,
-      "step": 220550
-    },
-    {
-      "epoch": 3.5833698883852416,
-      "grad_norm": 3.859375,
-      "learning_rate": 5e-05,
-      "loss": 0.2363,
-      "step": 220560
-    },
-    {
-      "epoch": 3.5835323552826113,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 220570
-    },
-    {
-      "epoch": 3.583694822179981,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 220580
-    },
-    {
-      "epoch": 3.5838572890773506,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 220590
-    },
-    {
-      "epoch": 3.5840197559747202,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 220600
-    },
-    {
-      "epoch": 3.58418222287209,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 220610
-    },
-    {
-      "epoch": 3.5843446897694595,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 220620
-    },
-    {
-      "epoch": 3.584507156666829,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 220630
-    },
-    {
-      "epoch": 3.584669623564199,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2236,
-      "step": 220640
-    },
-    {
-      "epoch": 3.5848320904615685,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.227,
-      "step": 220650
-    },
-    {
-      "epoch": 3.584994557358938,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 220660
-    },
-    {
-      "epoch": 3.585157024256308,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2175,
-      "step": 220670
-    },
-    {
-      "epoch": 3.5853194911536774,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2129,
-      "step": 220680
-    },
-    {
-      "epoch": 3.585481958051047,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 220690
-    },
-    {
-      "epoch": 3.5856444249484167,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 220700
-    },
-    {
-      "epoch": 3.5858068918457864,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2304,
-      "step": 220710
-    },
-    {
-      "epoch": 3.585969358743156,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 220720
-    },
-    {
-      "epoch": 3.5861318256405257,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 220730
-    },
-    {
-      "epoch": 3.5862942925378953,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.185,
-      "step": 220740
-    },
-    {
-      "epoch": 3.586456759435265,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 220750
-    },
-    {
-      "epoch": 3.5866192263326346,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.219,
-      "step": 220760
-    },
-    {
-      "epoch": 3.5867816932300043,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2089,
-      "step": 220770
-    },
-    {
-      "epoch": 3.586944160127374,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.223,
-      "step": 220780
-    },
-    {
-      "epoch": 3.5871066270247436,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 220790
-    },
-    {
-      "epoch": 3.5872690939221132,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2211,
-      "step": 220800
-    },
-    {
-      "epoch": 3.587431560819483,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 220810
-    },
-    {
-      "epoch": 3.5875940277168525,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2097,
-      "step": 220820
-    },
-    {
-      "epoch": 3.587756494614222,
-      "grad_norm": 7.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1896,
-      "step": 220830
-    },
-    {
-      "epoch": 3.5879189615115923,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 220840
-    },
-    {
-      "epoch": 3.5880814284089615,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 220850
-    },
-    {
-      "epoch": 3.5882438953063316,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 220860
-    },
-    {
-      "epoch": 3.588406362203701,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 220870
-    },
-    {
-      "epoch": 3.588568829101071,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 220880
-    },
-    {
-      "epoch": 3.58873129599844,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2001,
-      "step": 220890
-    },
-    {
-      "epoch": 3.58889376289581,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2046,
-      "step": 220900
-    },
-    {
-      "epoch": 3.5890562297931794,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 220910
-    },
-    {
-      "epoch": 3.5892186966905495,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2133,
-      "step": 220920
-    },
-    {
-      "epoch": 3.5893811635879187,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 220930
-    },
-    {
-      "epoch": 3.589543630485289,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 220940
-    },
-    {
-      "epoch": 3.5897060973826584,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 220950
-    },
-    {
-      "epoch": 3.589868564280028,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2312,
-      "step": 220960
-    },
-    {
-      "epoch": 3.5900310311773977,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 220970
-    },
-    {
-      "epoch": 3.5901934980747674,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 220980
-    },
-    {
-      "epoch": 3.590355964972137,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 220990
-    },
-    {
-      "epoch": 3.5905184318695067,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2164,
-      "step": 221000
-    },
-    {
-      "epoch": 3.5906808987668763,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 221010
-    },
-    {
-      "epoch": 3.590843365664246,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 221020
-    },
-    {
-      "epoch": 3.5910058325616157,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2283,
-      "step": 221030
-    },
-    {
-      "epoch": 3.5911682994589853,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2217,
-      "step": 221040
-    },
-    {
-      "epoch": 3.591330766356355,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2236,
-      "step": 221050
-    },
-    {
-      "epoch": 3.5914932332537246,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 221060
-    },
-    {
-      "epoch": 3.5916557001510943,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 221070
-    },
-    {
-      "epoch": 3.591818167048464,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2115,
-      "step": 221080
-    },
-    {
-      "epoch": 3.5919806339458336,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2431,
-      "step": 221090
-    },
-    {
-      "epoch": 3.592143100843203,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 221100
-    },
-    {
-      "epoch": 3.592305567740573,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 221110
-    },
-    {
-      "epoch": 3.5924680346379425,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 221120
-    },
-    {
-      "epoch": 3.592630501535312,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 221130
-    },
-    {
-      "epoch": 3.592792968432682,
-      "grad_norm": 7.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2203,
-      "step": 221140
-    },
-    {
-      "epoch": 3.5929554353300515,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 221150
-    },
-    {
-      "epoch": 3.593117902227421,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 221160
-    },
-    {
-      "epoch": 3.5932803691247908,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2126,
-      "step": 221170
-    },
-    {
-      "epoch": 3.5934428360221604,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 221180
-    },
-    {
-      "epoch": 3.59360530291953,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 221190
-    },
-    {
-      "epoch": 3.5937677698168997,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 221200
-    },
-    {
-      "epoch": 3.5939302367142694,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 221210
-    },
-    {
-      "epoch": 3.594092703611639,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 221220
-    },
-    {
-      "epoch": 3.5942551705090087,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 221230
-    },
-    {
-      "epoch": 3.5944176374063783,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1986,
-      "step": 221240
-    },
-    {
-      "epoch": 3.594580104303748,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1702,
-      "step": 221250
-    },
-    {
-      "epoch": 3.5947425712011176,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 221260
-    },
-    {
-      "epoch": 3.5949050380984873,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 221270
-    },
-    {
-      "epoch": 3.595067504995857,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 221280
-    },
-    {
-      "epoch": 3.595229971893227,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 221290
-    },
-    {
-      "epoch": 3.5953924387905962,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.184,
-      "step": 221300
-    },
-    {
-      "epoch": 3.5955549056879663,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 221310
-    },
-    {
-      "epoch": 3.5957173725853355,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2061,
-      "step": 221320
-    },
-    {
-      "epoch": 3.5958798394827056,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2118,
-      "step": 221330
-    },
-    {
-      "epoch": 3.596042306380075,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2264,
-      "step": 221340
-    },
-    {
-      "epoch": 3.596204773277445,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2153,
-      "step": 221350
-    },
-    {
-      "epoch": 3.596367240174814,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2255,
-      "step": 221360
-    },
-    {
-      "epoch": 3.5965297070721842,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 221370
-    },
-    {
-      "epoch": 3.5966921739695534,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1868,
-      "step": 221380
-    },
-    {
-      "epoch": 3.5968546408669235,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 221390
-    },
-    {
-      "epoch": 3.597017107764293,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1832,
-      "step": 221400
-    },
-    {
-      "epoch": 3.597179574661663,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1872,
-      "step": 221410
-    },
-    {
-      "epoch": 3.5973420415590325,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 221420
-    },
-    {
-      "epoch": 3.597504508456402,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1919,
-      "step": 221430
-    },
-    {
-      "epoch": 3.597666975353772,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 221440
-    },
-    {
-      "epoch": 3.5978294422511414,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 221450
-    },
-    {
-      "epoch": 3.597991909148511,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2224,
-      "step": 221460
-    },
-    {
-      "epoch": 3.5981543760458807,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 221470
-    },
-    {
-      "epoch": 3.5983168429432504,
-      "grad_norm": 7.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2356,
-      "step": 221480
-    },
-    {
-      "epoch": 3.59847930984062,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2268,
-      "step": 221490
-    },
-    {
-      "epoch": 3.5986417767379897,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 221500
-    },
-    {
-      "epoch": 3.5988042436353593,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 221510
-    },
-    {
-      "epoch": 3.598966710532729,
-      "grad_norm": 7.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 221520
-    },
-    {
-      "epoch": 3.5991291774300986,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2169,
-      "step": 221530
-    },
-    {
-      "epoch": 3.5992916443274683,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 221540
-    },
-    {
-      "epoch": 3.599454111224838,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 221550
-    },
-    {
-      "epoch": 3.5996165781222076,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 221560
-    },
-    {
-      "epoch": 3.5997790450195772,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 221570
-    },
-    {
-      "epoch": 3.599941511916947,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 221580
-    },
-    {
-      "epoch": 3.6001039788143165,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 221590
-    },
-    {
-      "epoch": 3.600266445711686,
-      "grad_norm": 7.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 221600
-    },
-    {
-      "epoch": 3.600428912609056,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 221610
-    },
-    {
-      "epoch": 3.6005913795064255,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2133,
-      "step": 221620
-    },
-    {
-      "epoch": 3.600753846403795,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 221630
-    },
-    {
-      "epoch": 3.600916313301165,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 221640
-    },
-    {
-      "epoch": 3.6010787801985344,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 221650
-    },
-    {
-      "epoch": 3.601241247095904,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1782,
-      "step": 221660
-    },
-    {
-      "epoch": 3.6014037139932737,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 221670
-    },
-    {
-      "epoch": 3.6015661808906434,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 221680
-    },
-    {
-      "epoch": 3.601728647788013,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 221690
-    },
-    {
-      "epoch": 3.6018911146853827,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 221700
-    },
-    {
-      "epoch": 3.6020535815827524,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.171,
-      "step": 221710
-    },
-    {
-      "epoch": 3.6022160484801224,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1763,
-      "step": 221720
-    },
-    {
-      "epoch": 3.6023785153774917,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1819,
-      "step": 221730
-    },
-    {
-      "epoch": 3.6025409822748617,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1823,
-      "step": 221740
-    },
-    {
-      "epoch": 3.602703449172231,
-      "grad_norm": 8.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 221750
-    },
-    {
-      "epoch": 3.602865916069601,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 221760
-    },
-    {
-      "epoch": 3.6030283829669703,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2222,
-      "step": 221770
-    },
-    {
-      "epoch": 3.6031908498643404,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.228,
-      "step": 221780
-    },
-    {
-      "epoch": 3.6033533167617096,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2328,
-      "step": 221790
-    },
-    {
-      "epoch": 3.6035157836590797,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2313,
-      "step": 221800
-    },
-    {
-      "epoch": 3.603678250556449,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 221810
-    },
-    {
-      "epoch": 3.603840717453819,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 221820
-    },
-    {
-      "epoch": 3.6040031843511886,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1821,
-      "step": 221830
-    },
-    {
-      "epoch": 3.6041656512485583,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1799,
-      "step": 221840
-    },
-    {
-      "epoch": 3.604328118145928,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1781,
-      "step": 221850
-    },
-    {
-      "epoch": 3.6044905850432976,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 221860
-    },
-    {
-      "epoch": 3.604653051940667,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 221870
-    },
-    {
-      "epoch": 3.604815518838037,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 221880
-    },
-    {
-      "epoch": 3.6049779857354065,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2044,
-      "step": 221890
-    },
-    {
-      "epoch": 3.605140452632776,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2247,
-      "step": 221900
-    },
-    {
-      "epoch": 3.605302919530146,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 221910
-    },
-    {
-      "epoch": 3.6054653864275155,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2281,
-      "step": 221920
-    },
-    {
-      "epoch": 3.605627853324885,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2309,
-      "step": 221930
-    },
-    {
-      "epoch": 3.6057903202222548,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2473,
-      "step": 221940
-    },
-    {
-      "epoch": 3.6059527871196244,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2242,
-      "step": 221950
-    },
-    {
-      "epoch": 3.606115254016994,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2359,
-      "step": 221960
-    },
-    {
-      "epoch": 3.6062777209143637,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 221970
-    },
-    {
-      "epoch": 3.6064401878117334,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 221980
-    },
-    {
-      "epoch": 3.606602654709103,
-      "grad_norm": 7.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.227,
-      "step": 221990
-    },
-    {
-      "epoch": 3.6067651216064727,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2236,
-      "step": 222000
-    },
-    {
-      "epoch": 3.6069275885038423,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2187,
-      "step": 222010
-    },
-    {
-      "epoch": 3.607090055401212,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2172,
-      "step": 222020
-    },
-    {
-      "epoch": 3.6072525222985816,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2084,
-      "step": 222030
-    },
-    {
-      "epoch": 3.6074149891959513,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 222040
-    },
-    {
-      "epoch": 3.607577456093321,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 222050
-    },
-    {
-      "epoch": 3.6077399229906906,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 222060
-    },
-    {
-      "epoch": 3.6079023898880602,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 222070
-    },
-    {
-      "epoch": 3.60806485678543,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2109,
-      "step": 222080
-    },
-    {
-      "epoch": 3.6082273236827995,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 222090
-    },
-    {
-      "epoch": 3.608389790580169,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.227,
-      "step": 222100
-    },
-    {
-      "epoch": 3.608552257477539,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 222110
-    },
-    {
-      "epoch": 3.6087147243749085,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.187,
-      "step": 222120
-    },
-    {
-      "epoch": 3.608877191272278,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 222130
-    },
-    {
-      "epoch": 3.609039658169648,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1863,
-      "step": 222140
-    },
-    {
-      "epoch": 3.6092021250670174,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2046,
-      "step": 222150
-    },
-    {
-      "epoch": 3.609364591964387,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 222160
-    },
-    {
-      "epoch": 3.609527058861757,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 222170
-    },
-    {
-      "epoch": 3.6096895257591264,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 222180
-    },
-    {
-      "epoch": 3.6098519926564965,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 222190
-    },
-    {
-      "epoch": 3.6100144595538657,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 222200
-    },
-    {
-      "epoch": 3.610176926451236,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2112,
-      "step": 222210
-    },
-    {
-      "epoch": 3.610339393348605,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 222220
-    },
-    {
-      "epoch": 3.610501860245975,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 222230
-    },
-    {
-      "epoch": 3.6106643271433443,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 222240
-    },
-    {
-      "epoch": 3.6108267940407144,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 222250
-    },
-    {
-      "epoch": 3.6109892609380836,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2255,
-      "step": 222260
-    },
-    {
-      "epoch": 3.6111517278354537,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2228,
-      "step": 222270
-    },
-    {
-      "epoch": 3.6113141947328233,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2192,
-      "step": 222280
-    },
-    {
-      "epoch": 3.611476661630193,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 222290
-    },
-    {
-      "epoch": 3.6116391285275626,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 222300
-    },
-    {
-      "epoch": 3.6118015954249323,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 222310
-    },
-    {
-      "epoch": 3.611964062322302,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.217,
-      "step": 222320
-    },
-    {
-      "epoch": 3.6121265292196716,
-      "grad_norm": 4.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.194,
-      "step": 222330
-    },
-    {
-      "epoch": 3.6122889961170412,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 222340
-    },
-    {
-      "epoch": 3.612451463014411,
-      "grad_norm": 7.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 222350
-    },
-    {
-      "epoch": 3.6126139299117805,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2121,
-      "step": 222360
-    },
-    {
-      "epoch": 3.61277639680915,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2412,
-      "step": 222370
-    },
-    {
-      "epoch": 3.61293886370652,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1904,
-      "step": 222380
-    },
-    {
-      "epoch": 3.6131013306038895,
-      "grad_norm": 4.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.207,
-      "step": 222390
-    },
-    {
-      "epoch": 3.613263797501259,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1961,
-      "step": 222400
-    },
-    {
-      "epoch": 3.613426264398629,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 222410
-    },
-    {
-      "epoch": 3.6135887312959984,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2281,
-      "step": 222420
-    },
-    {
-      "epoch": 3.613751198193368,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2239,
-      "step": 222430
-    },
-    {
-      "epoch": 3.6139136650907377,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 222440
-    },
-    {
-      "epoch": 3.6140761319881074,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2164,
-      "step": 222450
-    },
-    {
-      "epoch": 3.614238598885477,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2268,
-      "step": 222460
-    },
-    {
-      "epoch": 3.6144010657828467,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2169,
-      "step": 222470
-    },
-    {
-      "epoch": 3.6145635326802164,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 222480
-    },
-    {
-      "epoch": 3.614725999577586,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2267,
-      "step": 222490
-    },
-    {
-      "epoch": 3.6148884664749557,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1977,
-      "step": 222500
-    },
-    {
-      "epoch": 3.6150509333723253,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2132,
-      "step": 222510
-    },
-    {
-      "epoch": 3.615213400269695,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2195,
-      "step": 222520
-    },
-    {
-      "epoch": 3.6153758671670646,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 222530
-    },
-    {
-      "epoch": 3.6155383340644343,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2424,
-      "step": 222540
-    },
-    {
-      "epoch": 3.615700800961804,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2192,
-      "step": 222550
-    },
-    {
-      "epoch": 3.6158632678591736,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 222560
-    },
-    {
-      "epoch": 3.616025734756543,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1833,
-      "step": 222570
-    },
-    {
-      "epoch": 3.616188201653913,
-      "grad_norm": 7.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2142,
-      "step": 222580
-    },
-    {
-      "epoch": 3.6163506685512825,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 222590
-    },
-    {
-      "epoch": 3.6165131354486526,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 222600
-    },
-    {
-      "epoch": 3.616675602346022,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2189,
-      "step": 222610
-    },
-    {
-      "epoch": 3.616838069243392,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 222620
-    },
-    {
-      "epoch": 3.617000536140761,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 222630
-    },
-    {
-      "epoch": 3.617163003038131,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2102,
-      "step": 222640
-    },
-    {
-      "epoch": 3.6173254699355004,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2223,
-      "step": 222650
-    },
-    {
-      "epoch": 3.6174879368328705,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2187,
-      "step": 222660
-    },
-    {
-      "epoch": 3.6176504037302397,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2042,
-      "step": 222670
-    },
-    {
-      "epoch": 3.61781287062761,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2216,
-      "step": 222680
-    },
-    {
-      "epoch": 3.617975337524979,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 222690
-    },
-    {
-      "epoch": 3.618137804422349,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2116,
-      "step": 222700
-    },
-    {
-      "epoch": 3.6183002713197188,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 222710
-    },
-    {
-      "epoch": 3.6184627382170884,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 222720
-    },
-    {
-      "epoch": 3.618625205114458,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1988,
-      "step": 222730
-    },
-    {
-      "epoch": 3.6187876720118277,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1972,
-      "step": 222740
-    },
-    {
-      "epoch": 3.6189501389091974,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 222750
-    },
-    {
-      "epoch": 3.619112605806567,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 222760
-    },
-    {
-      "epoch": 3.6192750727039367,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1999,
-      "step": 222770
-    },
-    {
-      "epoch": 3.6194375396013063,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 222780
-    },
-    {
-      "epoch": 3.619600006498676,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 222790
-    },
-    {
-      "epoch": 3.6197624733960456,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2275,
-      "step": 222800
-    },
-    {
-      "epoch": 3.6199249402934153,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 222810
-    },
-    {
-      "epoch": 3.620087407190785,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2218,
-      "step": 222820
-    },
-    {
-      "epoch": 3.6202498740881546,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 222830
-    },
-    {
-      "epoch": 3.6204123409855242,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2066,
-      "step": 222840
-    },
-    {
-      "epoch": 3.620574807882894,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 222850
-    },
-    {
-      "epoch": 3.6207372747802635,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2089,
-      "step": 222860
-    },
-    {
-      "epoch": 3.620899741677633,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2002,
-      "step": 222870
-    },
-    {
-      "epoch": 3.621062208575003,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 222880
-    },
-    {
-      "epoch": 3.6212246754723725,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 222890
-    },
-    {
-      "epoch": 3.621387142369742,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 222900
-    },
-    {
-      "epoch": 3.621549609267112,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 222910
-    },
-    {
-      "epoch": 3.6217120761644814,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1831,
-      "step": 222920
-    },
-    {
-      "epoch": 3.621874543061851,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 222930
-    },
-    {
-      "epoch": 3.6220370099592207,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 222940
-    },
-    {
-      "epoch": 3.6221994768565904,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 222950
-    },
-    {
-      "epoch": 3.62236194375396,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2075,
-      "step": 222960
-    },
-    {
-      "epoch": 3.6225244106513297,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2076,
-      "step": 222970
-    },
-    {
-      "epoch": 3.6226868775486993,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2042,
-      "step": 222980
-    },
-    {
-      "epoch": 3.622849344446069,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2267,
-      "step": 222990
-    },
-    {
-      "epoch": 3.6230118113434386,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 223000
-    },
-    {
-      "epoch": 3.6231742782408083,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2275,
-      "step": 223010
-    },
-    {
-      "epoch": 3.623336745138178,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2222,
-      "step": 223020
-    },
-    {
-      "epoch": 3.623499212035548,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2387,
-      "step": 223030
-    },
-    {
-      "epoch": 3.6236616789329172,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2408,
-      "step": 223040
-    },
-    {
-      "epoch": 3.6238241458302873,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.255,
-      "step": 223050
-    },
-    {
-      "epoch": 3.6239866127276565,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2413,
-      "step": 223060
-    },
-    {
-      "epoch": 3.6241490796250266,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.224,
-      "step": 223070
-    },
-    {
-      "epoch": 3.624311546522396,
-      "grad_norm": 4.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2443,
-      "step": 223080
-    },
-    {
-      "epoch": 3.624474013419766,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2375,
-      "step": 223090
-    },
-    {
-      "epoch": 3.624636480317135,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2573,
-      "step": 223100
-    },
-    {
-      "epoch": 3.6247989472145052,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2454,
-      "step": 223110
-    },
-    {
-      "epoch": 3.6249614141118744,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2513,
-      "step": 223120
-    },
-    {
-      "epoch": 3.6251238810092445,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2414,
-      "step": 223130
-    },
-    {
-      "epoch": 3.625286347906614,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2177,
-      "step": 223140
-    },
-    {
-      "epoch": 3.625448814803984,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 223150
-    },
-    {
-      "epoch": 3.6256112817013535,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 223160
-    },
-    {
-      "epoch": 3.625773748598723,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1798,
-      "step": 223170
-    },
-    {
-      "epoch": 3.625936215496093,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2116,
-      "step": 223180
-    },
-    {
-      "epoch": 3.6260986823934624,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 223190
-    },
-    {
-      "epoch": 3.626261149290832,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2046,
-      "step": 223200
-    },
-    {
-      "epoch": 3.6264236161882017,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1989,
-      "step": 223210
-    },
-    {
-      "epoch": 3.6265860830855714,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 223220
-    },
-    {
-      "epoch": 3.626748549982941,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 223230
-    },
-    {
-      "epoch": 3.6269110168803107,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1923,
-      "step": 223240
-    },
-    {
-      "epoch": 3.6270734837776804,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2059,
-      "step": 223250
-    },
-    {
-      "epoch": 3.62723595067505,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 223260
-    },
-    {
-      "epoch": 3.6273984175724197,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 223270
-    },
-    {
-      "epoch": 3.6275608844697893,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 223280
-    },
-    {
-      "epoch": 3.627723351367159,
-      "grad_norm": 7.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2348,
-      "step": 223290
-    },
-    {
-      "epoch": 3.6278858182645286,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2233,
-      "step": 223300
-    },
-    {
-      "epoch": 3.6280482851618983,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 223310
-    },
-    {
-      "epoch": 3.628210752059268,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 223320
-    },
-    {
-      "epoch": 3.6283732189566376,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2265,
-      "step": 223330
-    },
-    {
-      "epoch": 3.628535685854007,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.223,
-      "step": 223340
-    },
-    {
-      "epoch": 3.628698152751377,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 223350
-    },
-    {
-      "epoch": 3.6288606196487465,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 223360
-    },
-    {
-      "epoch": 3.629023086546116,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 223370
-    },
-    {
-      "epoch": 3.629185553443486,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 223380
-    },
-    {
-      "epoch": 3.6293480203408555,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1876,
-      "step": 223390
-    },
-    {
-      "epoch": 3.629510487238225,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 223400
-    },
-    {
-      "epoch": 3.6296729541355948,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 223410
-    },
-    {
-      "epoch": 3.6298354210329644,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 223420
-    },
-    {
-      "epoch": 3.629997887930334,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 223430
-    },
-    {
-      "epoch": 3.6301603548277037,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.225,
-      "step": 223440
-    },
-    {
-      "epoch": 3.6303228217250734,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 223450
-    },
-    {
-      "epoch": 3.630485288622443,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 223460
-    },
-    {
-      "epoch": 3.6306477555198127,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 223470
-    },
-    {
-      "epoch": 3.6308102224171828,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2303,
-      "step": 223480
-    },
-    {
-      "epoch": 3.630972689314552,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 223490
-    },
-    {
-      "epoch": 3.631135156211922,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 223500
-    },
-    {
-      "epoch": 3.6312976231092913,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2144,
-      "step": 223510
-    },
-    {
-      "epoch": 3.6314600900066614,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 223520
-    },
-    {
-      "epoch": 3.6316225569040306,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2328,
-      "step": 223530
-    },
-    {
-      "epoch": 3.6317850238014007,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 223540
-    },
-    {
-      "epoch": 3.63194749069877,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 223550
-    },
-    {
-      "epoch": 3.63210995759614,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 223560
-    },
-    {
-      "epoch": 3.632272424493509,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2275,
-      "step": 223570
-    },
-    {
-      "epoch": 3.6324348913908793,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 223580
-    },
-    {
-      "epoch": 3.632597358288249,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2371,
-      "step": 223590
-    },
-    {
-      "epoch": 3.6327598251856186,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 223600
-    },
-    {
-      "epoch": 3.6329222920829882,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2085,
-      "step": 223610
-    },
-    {
-      "epoch": 3.633084758980358,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 223620
-    },
-    {
-      "epoch": 3.6332472258777275,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 223630
-    },
-    {
-      "epoch": 3.633409692775097,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 223640
-    },
-    {
-      "epoch": 3.633572159672467,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 223650
-    },
-    {
-      "epoch": 3.6337346265698365,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 223660
-    },
-    {
-      "epoch": 3.633897093467206,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2193,
-      "step": 223670
-    },
-    {
-      "epoch": 3.634059560364576,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 223680
-    },
-    {
-      "epoch": 3.6342220272619454,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2285,
-      "step": 223690
-    },
-    {
-      "epoch": 3.634384494159315,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 223700
-    },
-    {
-      "epoch": 3.6345469610566847,
-      "grad_norm": 3.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 223710
-    },
-    {
-      "epoch": 3.6347094279540544,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 223720
-    },
-    {
-      "epoch": 3.634871894851424,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2246,
-      "step": 223730
-    },
-    {
-      "epoch": 3.6350343617487937,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 223740
-    },
-    {
-      "epoch": 3.6351968286461633,
-      "grad_norm": 7.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2362,
-      "step": 223750
-    },
-    {
-      "epoch": 3.635359295543533,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2206,
-      "step": 223760
-    },
-    {
-      "epoch": 3.6355217624409026,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2298,
-      "step": 223770
-    },
-    {
-      "epoch": 3.6356842293382723,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2343,
-      "step": 223780
-    },
-    {
-      "epoch": 3.635846696235642,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2227,
-      "step": 223790
-    },
-    {
-      "epoch": 3.6360091631330116,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 223800
-    },
-    {
-      "epoch": 3.6361716300303812,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2119,
-      "step": 223810
-    },
-    {
-      "epoch": 3.636334096927751,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 223820
-    },
-    {
-      "epoch": 3.6364965638251205,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1956,
-      "step": 223830
-    },
-    {
-      "epoch": 3.63665903072249,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 223840
-    },
-    {
-      "epoch": 3.63682149761986,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 223850
-    },
-    {
-      "epoch": 3.6369839645172295,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2287,
-      "step": 223860
-    },
-    {
-      "epoch": 3.637146431414599,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 223870
-    },
-    {
-      "epoch": 3.637308898311969,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2239,
-      "step": 223880
-    },
-    {
-      "epoch": 3.6374713652093384,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2396,
-      "step": 223890
-    },
-    {
-      "epoch": 3.637633832106708,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2244,
-      "step": 223900
-    },
-    {
-      "epoch": 3.637796299004078,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 223910
-    },
-    {
-      "epoch": 3.6379587659014474,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 223920
-    },
-    {
-      "epoch": 3.6381212327988175,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 223930
-    },
-    {
-      "epoch": 3.6382836996961867,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2189,
-      "step": 223940
-    },
-    {
-      "epoch": 3.638446166593557,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2242,
-      "step": 223950
-    },
-    {
-      "epoch": 3.638608633490926,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2176,
-      "step": 223960
-    },
-    {
-      "epoch": 3.638771100388296,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 223970
-    },
-    {
-      "epoch": 3.6389335672856653,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.224,
-      "step": 223980
-    },
-    {
-      "epoch": 3.6390960341830354,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 223990
-    },
-    {
-      "epoch": 3.6392585010804046,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 224000
-    },
-    {
-      "epoch": 3.6394209679777747,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 224010
-    },
-    {
-      "epoch": 3.6395834348751444,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1991,
-      "step": 224020
-    },
-    {
-      "epoch": 3.639745901772514,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 224030
-    },
-    {
-      "epoch": 3.6399083686698837,
-      "grad_norm": 4.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 224040
-    },
-    {
-      "epoch": 3.6400708355672533,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2334,
-      "step": 224050
-    },
-    {
-      "epoch": 3.640233302464623,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2022,
-      "step": 224060
-    },
-    {
-      "epoch": 3.6403957693619926,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2257,
-      "step": 224070
-    },
-    {
-      "epoch": 3.6405582362593623,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2214,
-      "step": 224080
-    },
-    {
-      "epoch": 3.640720703156732,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2365,
-      "step": 224090
-    },
-    {
-      "epoch": 3.6408831700541016,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2267,
-      "step": 224100
-    },
-    {
-      "epoch": 3.641045636951471,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 224110
-    },
-    {
-      "epoch": 3.641208103848841,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 224120
-    },
-    {
-      "epoch": 3.6413705707462105,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 224130
-    },
-    {
-      "epoch": 3.64153303764358,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 224140
-    },
-    {
-      "epoch": 3.64169550454095,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 224150
-    },
-    {
-      "epoch": 3.6418579714383195,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 224160
-    },
-    {
-      "epoch": 3.642020438335689,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 224170
-    },
-    {
-      "epoch": 3.6421829052330588,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2272,
-      "step": 224180
-    },
-    {
-      "epoch": 3.6423453721304284,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2171,
-      "step": 224190
-    },
-    {
-      "epoch": 3.642507839027798,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2433,
-      "step": 224200
-    },
-    {
-      "epoch": 3.6426703059251677,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 224210
-    },
-    {
-      "epoch": 3.6428327728225374,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2207,
-      "step": 224220
-    },
-    {
-      "epoch": 3.642995239719907,
-      "grad_norm": 3.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2288,
-      "step": 224230
-    },
-    {
-      "epoch": 3.6431577066172767,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2341,
-      "step": 224240
-    },
-    {
-      "epoch": 3.6433201735146463,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2313,
-      "step": 224250
-    },
-    {
-      "epoch": 3.643482640412016,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 224260
-    },
-    {
-      "epoch": 3.6436451073093856,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1748,
-      "step": 224270
-    },
-    {
-      "epoch": 3.6438075742067553,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 224280
-    },
-    {
-      "epoch": 3.643970041104125,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 224290
-    },
-    {
-      "epoch": 3.6441325080014946,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.196,
-      "step": 224300
-    },
-    {
-      "epoch": 3.6442949748988642,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 224310
-    },
-    {
-      "epoch": 3.644457441796234,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1897,
-      "step": 224320
-    },
-    {
-      "epoch": 3.6446199086936035,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1773,
-      "step": 224330
-    },
-    {
-      "epoch": 3.644782375590973,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 224340
-    },
-    {
-      "epoch": 3.644944842488343,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2153,
-      "step": 224350
-    },
-    {
-      "epoch": 3.645107309385713,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 224360
-    },
-    {
-      "epoch": 3.645269776283082,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 224370
-    },
-    {
-      "epoch": 3.6454322431804522,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1851,
-      "step": 224380
-    },
-    {
-      "epoch": 3.6455947100778214,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1886,
-      "step": 224390
-    },
-    {
-      "epoch": 3.6457571769751915,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2515,
-      "step": 224400
-    },
-    {
-      "epoch": 3.6459196438725607,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 224410
-    },
-    {
-      "epoch": 3.646082110769931,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 224420
-    },
-    {
-      "epoch": 3.6462445776673,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2233,
-      "step": 224430
-    },
-    {
-      "epoch": 3.64640704456467,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2318,
-      "step": 224440
-    },
-    {
-      "epoch": 3.6465695114620393,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2266,
-      "step": 224450
-    },
-    {
-      "epoch": 3.6467319783594094,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 224460
-    },
-    {
-      "epoch": 3.646894445256779,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 224470
-    },
-    {
-      "epoch": 3.6470569121541487,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 224480
-    },
-    {
-      "epoch": 3.6472193790515184,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1922,
-      "step": 224490
-    },
-    {
-      "epoch": 3.647381845948888,
-      "grad_norm": 4.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 224500
-    },
-    {
-      "epoch": 3.6475443128462577,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 224510
-    },
-    {
-      "epoch": 3.6477067797436273,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 224520
-    },
-    {
-      "epoch": 3.647869246640997,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 224530
-    },
-    {
-      "epoch": 3.6480317135383666,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2201,
-      "step": 224540
-    },
-    {
-      "epoch": 3.6481941804357363,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 224550
-    },
-    {
-      "epoch": 3.648356647333106,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 224560
-    },
-    {
-      "epoch": 3.6485191142304756,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 224570
-    },
-    {
-      "epoch": 3.6486815811278452,
-      "grad_norm": 7.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2255,
-      "step": 224580
-    },
-    {
-      "epoch": 3.648844048025215,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2239,
-      "step": 224590
-    },
-    {
-      "epoch": 3.6490065149225845,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 224600
-    },
-    {
-      "epoch": 3.649168981819954,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 224610
-    },
-    {
-      "epoch": 3.649331448717324,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 224620
-    },
-    {
-      "epoch": 3.6494939156146935,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 224630
-    },
-    {
-      "epoch": 3.649656382512063,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1987,
-      "step": 224640
-    },
-    {
-      "epoch": 3.649818849409433,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 224650
-    },
-    {
-      "epoch": 3.6499813163068024,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 224660
-    },
-    {
-      "epoch": 3.650143783204172,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2146,
-      "step": 224670
-    },
-    {
-      "epoch": 3.6503062501015417,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 224680
-    },
-    {
-      "epoch": 3.6504687169989114,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2105,
-      "step": 224690
-    },
-    {
-      "epoch": 3.650631183896281,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 224700
-    },
-    {
-      "epoch": 3.6507936507936507,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1949,
-      "step": 224710
-    },
-    {
-      "epoch": 3.6509561176910204,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 224720
-    },
-    {
-      "epoch": 3.65111858458839,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2239,
-      "step": 224730
-    },
-    {
-      "epoch": 3.6512810514857597,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2191,
-      "step": 224740
-    },
-    {
-      "epoch": 3.6514435183831293,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2374,
-      "step": 224750
-    },
-    {
-      "epoch": 3.651605985280499,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 224760
-    },
-    {
-      "epoch": 3.6517684521778686,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2072,
-      "step": 224770
-    },
-    {
-      "epoch": 3.6519309190752383,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 224780
-    },
-    {
-      "epoch": 3.6520933859726084,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2122,
-      "step": 224790
-    },
-    {
-      "epoch": 3.6522558528699776,
-      "grad_norm": 6.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.223,
-      "step": 224800
-    },
-    {
-      "epoch": 3.6524183197673477,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 224810
-    },
-    {
-      "epoch": 3.652580786664717,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 224820
-    },
-    {
-      "epoch": 3.652743253562087,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 224830
-    },
-    {
-      "epoch": 3.652905720459456,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1964,
-      "step": 224840
-    },
-    {
-      "epoch": 3.6530681873568263,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 224850
-    },
-    {
-      "epoch": 3.6532306542541955,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 224860
-    },
-    {
-      "epoch": 3.6533931211515656,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 224870
-    },
-    {
-      "epoch": 3.6535555880489348,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1909,
-      "step": 224880
-    },
-    {
-      "epoch": 3.653718054946305,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 224890
-    },
-    {
-      "epoch": 3.6538805218436745,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 224900
-    },
-    {
-      "epoch": 3.654042988741044,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 224910
-    },
-    {
-      "epoch": 3.654205455638414,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 224920
-    },
-    {
-      "epoch": 3.6543679225357835,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2283,
-      "step": 224930
-    },
-    {
-      "epoch": 3.654530389433153,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1864,
-      "step": 224940
-    },
-    {
-      "epoch": 3.6546928563305228,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1788,
-      "step": 224950
-    },
-    {
-      "epoch": 3.6548553232278924,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 224960
-    },
-    {
-      "epoch": 3.655017790125262,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1993,
-      "step": 224970
-    },
-    {
-      "epoch": 3.6551802570226317,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 224980
-    },
-    {
-      "epoch": 3.6553427239200014,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 224990
-    },
-    {
-      "epoch": 3.655505190817371,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2236,
-      "step": 225000
-    },
-    {
-      "epoch": 3.6556676577147407,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 225010
-    },
-    {
-      "epoch": 3.6558301246121103,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2246,
-      "step": 225020
-    },
-    {
-      "epoch": 3.65599259150948,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2224,
-      "step": 225030
-    },
-    {
-      "epoch": 3.6561550584068496,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.1984,
-      "step": 225040
-    },
-    {
-      "epoch": 3.6563175253042193,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2084,
-      "step": 225050
-    },
-    {
-      "epoch": 3.656479992201589,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 225060
-    },
-    {
-      "epoch": 3.6566424590989586,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2019,
-      "step": 225070
-    },
-    {
-      "epoch": 3.6568049259963282,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1913,
-      "step": 225080
-    },
-    {
-      "epoch": 3.656967392893698,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 225090
-    },
-    {
-      "epoch": 3.6571298597910675,
-      "grad_norm": 7.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 225100
-    },
-    {
-      "epoch": 3.657292326688437,
-      "grad_norm": 4.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1787,
-      "step": 225110
-    },
-    {
-      "epoch": 3.657454793585807,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 225120
-    },
-    {
-      "epoch": 3.6576172604831765,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 225130
-    },
-    {
-      "epoch": 3.657779727380546,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 225140
-    },
-    {
-      "epoch": 3.657942194277916,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2176,
-      "step": 225150
-    },
-    {
-      "epoch": 3.6581046611752854,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.237,
-      "step": 225160
-    },
-    {
-      "epoch": 3.658267128072655,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2285,
-      "step": 225170
-    },
-    {
-      "epoch": 3.6584295949700247,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2234,
-      "step": 225180
-    },
-    {
-      "epoch": 3.6585920618673944,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2355,
-      "step": 225190
-    },
-    {
-      "epoch": 3.658754528764764,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2131,
-      "step": 225200
-    },
-    {
-      "epoch": 3.6589169956621337,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 225210
-    },
-    {
-      "epoch": 3.6590794625595033,
-      "grad_norm": 7.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 225220
-    },
-    {
-      "epoch": 3.659241929456873,
-      "grad_norm": 4.125,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 225230
-    },
-    {
-      "epoch": 3.659404396354243,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2481,
-      "step": 225240
-    },
-    {
-      "epoch": 3.6595668632516123,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 225250
-    },
-    {
-      "epoch": 3.6597293301489824,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 225260
-    },
-    {
-      "epoch": 3.6598917970463516,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 225270
-    },
-    {
-      "epoch": 3.6600542639437217,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 225280
-    },
-    {
-      "epoch": 3.660216730841091,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 225290
-    },
-    {
-      "epoch": 3.660379197738461,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2227,
-      "step": 225300
-    },
-    {
-      "epoch": 3.66054166463583,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 225310
-    },
-    {
-      "epoch": 3.6607041315332003,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2004,
-      "step": 225320
-    },
-    {
-      "epoch": 3.6608665984305695,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2028,
-      "step": 225330
-    },
-    {
-      "epoch": 3.6610290653279396,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 225340
-    },
-    {
-      "epoch": 3.6611915322253092,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2087,
-      "step": 225350
-    },
-    {
-      "epoch": 3.661353999122679,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2209,
-      "step": 225360
-    },
-    {
-      "epoch": 3.6615164660200485,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2205,
-      "step": 225370
-    },
-    {
-      "epoch": 3.661678932917418,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2155,
-      "step": 225380
-    },
-    {
-      "epoch": 3.661841399814788,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 225390
-    },
-    {
-      "epoch": 3.6620038667121575,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 225400
-    },
-    {
-      "epoch": 3.662166333609527,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1996,
-      "step": 225410
-    },
-    {
-      "epoch": 3.662328800506897,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2146,
-      "step": 225420
-    },
-    {
-      "epoch": 3.6624912674042664,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 225430
-    },
-    {
-      "epoch": 3.662653734301636,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2275,
-      "step": 225440
-    },
-    {
-      "epoch": 3.6628162011990058,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 225450
-    },
-    {
-      "epoch": 3.6629786680963754,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 225460
-    },
-    {
-      "epoch": 3.663141134993745,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2063,
-      "step": 225470
-    },
-    {
-      "epoch": 3.6633036018911147,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.19,
-      "step": 225480
-    },
-    {
-      "epoch": 3.6634660687884844,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 225490
-    },
-    {
-      "epoch": 3.663628535685854,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 225500
-    },
-    {
-      "epoch": 3.6637910025832237,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2177,
-      "step": 225510
-    },
-    {
-      "epoch": 3.6639534694805933,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2195,
-      "step": 225520
-    },
-    {
-      "epoch": 3.664115936377963,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2221,
-      "step": 225530
-    },
-    {
-      "epoch": 3.6642784032753326,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2383,
-      "step": 225540
-    },
-    {
-      "epoch": 3.6644408701727023,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2251,
-      "step": 225550
-    },
-    {
-      "epoch": 3.664603337070072,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 225560
-    },
-    {
-      "epoch": 3.6647658039674416,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 225570
-    },
-    {
-      "epoch": 3.664928270864811,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 225580
-    },
-    {
-      "epoch": 3.665090737762181,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 225590
-    },
-    {
-      "epoch": 3.6652532046595505,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1938,
-      "step": 225600
-    },
-    {
-      "epoch": 3.66541567155692,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 225610
-    },
-    {
-      "epoch": 3.66557813845429,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1863,
-      "step": 225620
-    },
-    {
-      "epoch": 3.6657406053516595,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.217,
-      "step": 225630
-    },
-    {
-      "epoch": 3.665903072249029,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1975,
-      "step": 225640
-    },
-    {
-      "epoch": 3.6660655391463988,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 225650
-    },
-    {
-      "epoch": 3.6662280060437684,
-      "grad_norm": 7.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 225660
-    },
-    {
-      "epoch": 3.6663904729411385,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1875,
-      "step": 225670
-    },
-    {
-      "epoch": 3.6665529398385077,
-      "grad_norm": 3.453125,
-      "learning_rate": 5e-05,
-      "loss": 0.1877,
-      "step": 225680
-    },
-    {
-      "epoch": 3.666715406735878,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 225690
-    },
-    {
-      "epoch": 3.666877873633247,
-      "grad_norm": 7.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2004,
-      "step": 225700
-    },
-    {
-      "epoch": 3.667040340530617,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1747,
-      "step": 225710
-    },
-    {
-      "epoch": 3.6672028074279863,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2005,
-      "step": 225720
-    },
-    {
-      "epoch": 3.6673652743253564,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1892,
-      "step": 225730
-    },
-    {
-      "epoch": 3.6675277412227256,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2121,
-      "step": 225740
-    },
-    {
-      "epoch": 3.6676902081200957,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 225750
-    },
-    {
-      "epoch": 3.667852675017465,
-      "grad_norm": 3.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1878,
-      "step": 225760
-    },
-    {
-      "epoch": 3.668015141914835,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2223,
-      "step": 225770
-    },
-    {
-      "epoch": 3.6681776088122047,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 225780
-    },
-    {
-      "epoch": 3.6683400757095743,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2275,
-      "step": 225790
-    },
-    {
-      "epoch": 3.668502542606944,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2393,
-      "step": 225800
-    },
-    {
-      "epoch": 3.6686650095043136,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2316,
-      "step": 225810
-    },
-    {
-      "epoch": 3.6688274764016833,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2286,
-      "step": 225820
-    },
-    {
-      "epoch": 3.668989943299053,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 225830
-    },
-    {
-      "epoch": 3.6691524101964226,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2171,
-      "step": 225840
-    },
-    {
-      "epoch": 3.6693148770937922,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 225850
-    },
-    {
-      "epoch": 3.669477343991162,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1976,
-      "step": 225860
-    },
-    {
-      "epoch": 3.6696398108885315,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2284,
-      "step": 225870
-    },
-    {
-      "epoch": 3.669802277785901,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 225880
-    },
-    {
-      "epoch": 3.669964744683271,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2132,
-      "step": 225890
-    },
-    {
-      "epoch": 3.6701272115806405,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2052,
-      "step": 225900
-    },
-    {
-      "epoch": 3.67028967847801,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2132,
-      "step": 225910
-    },
-    {
-      "epoch": 3.67045214537538,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 225920
-    },
-    {
-      "epoch": 3.6706146122727494,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 225930
-    },
-    {
-      "epoch": 3.670777079170119,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 225940
-    },
-    {
-      "epoch": 3.6709395460674887,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 225950
-    },
-    {
-      "epoch": 3.6711020129648584,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2264,
-      "step": 225960
-    },
-    {
-      "epoch": 3.671264479862228,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 225970
-    },
-    {
-      "epoch": 3.6714269467595977,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 225980
-    },
-    {
-      "epoch": 3.6715894136569673,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2331,
-      "step": 225990
-    },
-    {
-      "epoch": 3.671751880554337,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 226000
-    },
-    {
-      "epoch": 3.6719143474517066,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.187,
-      "step": 226010
-    },
-    {
-      "epoch": 3.6720768143490763,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1844,
-      "step": 226020
-    },
-    {
-      "epoch": 3.672239281246446,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2235,
-      "step": 226030
-    },
-    {
-      "epoch": 3.6724017481438156,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1921,
-      "step": 226040
-    },
-    {
-      "epoch": 3.6725642150411852,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2209,
-      "step": 226050
-    },
-    {
-      "epoch": 3.672726681938555,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2204,
-      "step": 226060
-    },
-    {
-      "epoch": 3.6728891488359245,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2159,
-      "step": 226070
-    },
-    {
-      "epoch": 3.673051615733294,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 226080
-    },
-    {
-      "epoch": 3.673214082630664,
-      "grad_norm": 8.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2404,
-      "step": 226090
-    },
-    {
-      "epoch": 3.6733765495280335,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2198,
-      "step": 226100
-    },
-    {
-      "epoch": 3.673539016425403,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.232,
-      "step": 226110
-    },
-    {
-      "epoch": 3.6737014833227732,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2109,
-      "step": 226120
-    },
-    {
-      "epoch": 3.6738639502201424,
-      "grad_norm": 6.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2296,
-      "step": 226130
-    },
-    {
-      "epoch": 3.6740264171175125,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.232,
-      "step": 226140
-    },
-    {
-      "epoch": 3.6741888840148818,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2322,
-      "step": 226150
-    },
-    {
-      "epoch": 3.674351350912252,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2263,
-      "step": 226160
-    },
-    {
-      "epoch": 3.674513817809621,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.231,
-      "step": 226170
-    },
-    {
-      "epoch": 3.674676284706991,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2464,
-      "step": 226180
-    },
-    {
-      "epoch": 3.6748387516043604,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2388,
-      "step": 226190
-    },
-    {
-      "epoch": 3.6750012185017304,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1916,
-      "step": 226200
-    },
-    {
-      "epoch": 3.6751636853991,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2386,
-      "step": 226210
-    },
-    {
-      "epoch": 3.6753261522964698,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2296,
-      "step": 226220
-    },
-    {
-      "epoch": 3.6754886191938394,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2274,
-      "step": 226230
-    },
-    {
-      "epoch": 3.675651086091209,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2172,
-      "step": 226240
-    },
-    {
-      "epoch": 3.6758135529885787,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2152,
-      "step": 226250
-    },
-    {
-      "epoch": 3.6759760198859484,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2155,
-      "step": 226260
-    },
-    {
-      "epoch": 3.676138486783318,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2306,
-      "step": 226270
-    },
-    {
-      "epoch": 3.6763009536806877,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2269,
-      "step": 226280
-    },
-    {
-      "epoch": 3.6764634205780573,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 226290
-    },
-    {
-      "epoch": 3.676625887475427,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2354,
-      "step": 226300
-    },
-    {
-      "epoch": 3.6767883543727966,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 226310
-    },
-    {
-      "epoch": 3.6769508212701663,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 226320
-    },
-    {
-      "epoch": 3.677113288167536,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2205,
-      "step": 226330
-    },
-    {
-      "epoch": 3.6772757550649056,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1953,
-      "step": 226340
-    },
-    {
-      "epoch": 3.677438221962275,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 226350
-    },
-    {
-      "epoch": 3.677600688859645,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 226360
-    },
-    {
-      "epoch": 3.6777631557570145,
-      "grad_norm": 6.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 226370
-    },
-    {
-      "epoch": 3.677925622654384,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2068,
-      "step": 226380
-    },
-    {
-      "epoch": 3.678088089551754,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2515,
-      "step": 226390
-    },
-    {
-      "epoch": 3.6782505564491235,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2287,
-      "step": 226400
-    },
-    {
-      "epoch": 3.678413023346493,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.25,
-      "step": 226410
-    },
-    {
-      "epoch": 3.6785754902438628,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2116,
-      "step": 226420
-    },
-    {
-      "epoch": 3.6787379571412324,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2265,
-      "step": 226430
-    },
-    {
-      "epoch": 3.678900424038602,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2328,
-      "step": 226440
-    },
-    {
-      "epoch": 3.6790628909359717,
-      "grad_norm": 7.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 226450
-    },
-    {
-      "epoch": 3.6792253578333414,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2081,
-      "step": 226460
-    },
-    {
-      "epoch": 3.679387824730711,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 226470
-    },
-    {
-      "epoch": 3.6795502916280807,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2155,
-      "step": 226480
-    },
-    {
-      "epoch": 3.6797127585254503,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2152,
-      "step": 226490
-    },
-    {
-      "epoch": 3.67987522542282,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1789,
-      "step": 226500
-    },
-    {
-      "epoch": 3.6800376923201896,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 226510
-    },
-    {
-      "epoch": 3.6802001592175593,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2142,
-      "step": 226520
-    },
-    {
-      "epoch": 3.680362626114929,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 226530
-    },
-    {
-      "epoch": 3.6805250930122986,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 226540
-    },
-    {
-      "epoch": 3.6806875599096687,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 226550
-    },
-    {
-      "epoch": 3.680850026807038,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2136,
-      "step": 226560
-    },
-    {
-      "epoch": 3.681012493704408,
-      "grad_norm": 4.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2124,
-      "step": 226570
-    },
-    {
-      "epoch": 3.681174960601777,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 226580
-    },
-    {
-      "epoch": 3.6813374274991473,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2299,
-      "step": 226590
-    },
-    {
-      "epoch": 3.6814998943965165,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 226600
-    },
-    {
-      "epoch": 3.6816623612938866,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2327,
-      "step": 226610
-    },
-    {
-      "epoch": 3.681824828191256,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 226620
-    },
-    {
-      "epoch": 3.681987295088626,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 226630
-    },
-    {
-      "epoch": 3.682149761985995,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2316,
-      "step": 226640
-    },
-    {
-      "epoch": 3.682312228883365,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2216,
-      "step": 226650
-    },
-    {
-      "epoch": 3.682474695780735,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1951,
-      "step": 226660
-    },
-    {
-      "epoch": 3.6826371626781045,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 226670
-    },
-    {
-      "epoch": 3.682799629575474,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1873,
-      "step": 226680
-    },
-    {
-      "epoch": 3.682962096472844,
-      "grad_norm": 3.625,
-      "learning_rate": 5e-05,
-      "loss": 0.187,
-      "step": 226690
-    },
-    {
-      "epoch": 3.6831245633702134,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1881,
-      "step": 226700
-    },
-    {
-      "epoch": 3.683287030267583,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2016,
-      "step": 226710
-    },
-    {
-      "epoch": 3.6834494971649527,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.203,
-      "step": 226720
-    },
-    {
-      "epoch": 3.6836119640623224,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2008,
-      "step": 226730
-    },
-    {
-      "epoch": 3.683774430959692,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2123,
-      "step": 226740
-    },
-    {
-      "epoch": 3.6839368978570617,
-      "grad_norm": 7.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2109,
-      "step": 226750
-    },
-    {
-      "epoch": 3.6840993647544313,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 226760
-    },
-    {
-      "epoch": 3.684261831651801,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1976,
-      "step": 226770
-    },
-    {
-      "epoch": 3.6844242985491706,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2058,
-      "step": 226780
-    },
-    {
-      "epoch": 3.6845867654465403,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 226790
-    },
-    {
-      "epoch": 3.68474923234391,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 226800
-    },
-    {
-      "epoch": 3.6849116992412796,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2176,
-      "step": 226810
-    },
-    {
-      "epoch": 3.6850741661386492,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2203,
-      "step": 226820
-    },
-    {
-      "epoch": 3.685236633036019,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2229,
-      "step": 226830
-    },
-    {
-      "epoch": 3.6853990999333885,
-      "grad_norm": 7.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2261,
-      "step": 226840
-    },
-    {
-      "epoch": 3.685561566830758,
-      "grad_norm": 4.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 226850
-    },
-    {
-      "epoch": 3.685724033728128,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2283,
-      "step": 226860
-    },
-    {
-      "epoch": 3.6858865006254975,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 226870
-    },
-    {
-      "epoch": 3.686048967522867,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2,
-      "step": 226880
-    },
-    {
-      "epoch": 3.686211434420237,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2308,
-      "step": 226890
-    },
-    {
-      "epoch": 3.6863739013176064,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2337,
-      "step": 226900
-    },
-    {
-      "epoch": 3.686536368214976,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2285,
-      "step": 226910
-    },
-    {
-      "epoch": 3.6866988351123458,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.23,
-      "step": 226920
-    },
-    {
-      "epoch": 3.6868613020097154,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2403,
-      "step": 226930
-    },
-    {
-      "epoch": 3.687023768907085,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 226940
-    },
-    {
-      "epoch": 3.6871862358044547,
-      "grad_norm": 7.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2359,
-      "step": 226950
-    },
-    {
-      "epoch": 3.6873487027018244,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2288,
-      "step": 226960
-    },
-    {
-      "epoch": 3.687511169599194,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2374,
-      "step": 226970
-    },
-    {
-      "epoch": 3.687673636496564,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2311,
-      "step": 226980
-    },
-    {
-      "epoch": 3.6878361033939333,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 226990
-    },
-    {
-      "epoch": 3.6879985702913034,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2102,
-      "step": 227000
-    },
-    {
-      "epoch": 3.6881610371886726,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2062,
-      "step": 227010
-    },
-    {
-      "epoch": 3.6883235040860427,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1954,
-      "step": 227020
-    },
-    {
-      "epoch": 3.688485970983412,
-      "grad_norm": 7.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2435,
-      "step": 227030
-    },
-    {
-      "epoch": 3.688648437880782,
-      "grad_norm": 8.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2128,
-      "step": 227040
-    },
-    {
-      "epoch": 3.688810904778151,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1791,
-      "step": 227050
-    },
-    {
-      "epoch": 3.6889733716755213,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 227060
-    },
-    {
-      "epoch": 3.6891358385728905,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 227070
-    },
-    {
-      "epoch": 3.6892983054702606,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 227080
-    },
-    {
-      "epoch": 3.6894607723676303,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 227090
-    },
-    {
-      "epoch": 3.689623239265,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 227100
-    },
-    {
-      "epoch": 3.6897857061623696,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2098,
-      "step": 227110
-    },
-    {
-      "epoch": 3.689948173059739,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2173,
-      "step": 227120
-    },
-    {
-      "epoch": 3.690110639957109,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 227130
-    },
-    {
-      "epoch": 3.6902731068544785,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 227140
-    },
-    {
-      "epoch": 3.690435573751848,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2142,
-      "step": 227150
-    },
-    {
-      "epoch": 3.690598040649218,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 227160
-    },
-    {
-      "epoch": 3.6907605075465875,
-      "grad_norm": 8.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2384,
-      "step": 227170
-    },
-    {
-      "epoch": 3.690922974443957,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2305,
-      "step": 227180
-    },
-    {
-      "epoch": 3.6910854413413268,
-      "grad_norm": 7.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2194,
-      "step": 227190
-    },
-    {
-      "epoch": 3.6912479082386964,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2137,
-      "step": 227200
-    },
-    {
-      "epoch": 3.691410375136066,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 227210
-    },
-    {
-      "epoch": 3.6915728420334357,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 227220
-    },
-    {
-      "epoch": 3.6917353089308054,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2234,
-      "step": 227230
-    },
-    {
-      "epoch": 3.691897775828175,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 227240
-    },
-    {
-      "epoch": 3.6920602427255447,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 227250
-    },
-    {
-      "epoch": 3.6922227096229143,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 227260
-    },
-    {
-      "epoch": 3.692385176520284,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2157,
-      "step": 227270
-    },
-    {
-      "epoch": 3.6925476434176536,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 227280
-    },
-    {
-      "epoch": 3.6927101103150233,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2187,
-      "step": 227290
-    },
-    {
-      "epoch": 3.692872577212393,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2012,
-      "step": 227300
-    },
-    {
-      "epoch": 3.6930350441097626,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2131,
-      "step": 227310
-    },
-    {
-      "epoch": 3.6931975110071322,
-      "grad_norm": 7.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2335,
-      "step": 227320
-    },
-    {
-      "epoch": 3.693359977904502,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1883,
-      "step": 227330
-    },
-    {
-      "epoch": 3.6935224448018715,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1857,
-      "step": 227340
-    },
-    {
-      "epoch": 3.693684911699241,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1748,
-      "step": 227350
-    },
-    {
-      "epoch": 3.693847378596611,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.1662,
-      "step": 227360
-    },
-    {
-      "epoch": 3.6940098454939805,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 227370
-    },
-    {
-      "epoch": 3.69417231239135,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1931,
-      "step": 227380
-    },
-    {
-      "epoch": 3.69433477928872,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 227390
-    },
-    {
-      "epoch": 3.6944972461860894,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2015,
-      "step": 227400
-    },
-    {
-      "epoch": 3.694659713083459,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 227410
-    },
-    {
-      "epoch": 3.6948221799808287,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 227420
-    },
-    {
-      "epoch": 3.694984646878199,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1792,
-      "step": 227430
-    },
-    {
-      "epoch": 3.695147113775568,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1751,
-      "step": 227440
-    },
-    {
-      "epoch": 3.695309580672938,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 227450
-    },
-    {
-      "epoch": 3.6954720475703073,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1985,
-      "step": 227460
-    },
-    {
-      "epoch": 3.6956345144676774,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 227470
-    },
-    {
-      "epoch": 3.6957969813650466,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2282,
-      "step": 227480
-    },
-    {
-      "epoch": 3.6959594482624167,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 227490
-    },
-    {
-      "epoch": 3.696121915159786,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2282,
-      "step": 227500
-    },
-    {
-      "epoch": 3.696284382057156,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 227510
-    },
-    {
-      "epoch": 3.6964468489545252,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2212,
-      "step": 227520
-    },
-    {
-      "epoch": 3.6966093158518953,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2154,
-      "step": 227530
-    },
-    {
-      "epoch": 3.696771782749265,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 227540
-    },
-    {
-      "epoch": 3.6969342496466346,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 227550
-    },
-    {
-      "epoch": 3.6970967165440043,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2338,
-      "step": 227560
-    },
-    {
-      "epoch": 3.697259183441374,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2258,
-      "step": 227570
-    },
-    {
-      "epoch": 3.6974216503387436,
-      "grad_norm": 4.0,
-      "learning_rate": 5e-05,
-      "loss": 0.223,
-      "step": 227580
-    },
-    {
-      "epoch": 3.6975841172361132,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 227590
-    },
-    {
-      "epoch": 3.697746584133483,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 227600
-    },
-    {
-      "epoch": 3.6979090510308525,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 227610
-    },
-    {
-      "epoch": 3.698071517928222,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2201,
-      "step": 227620
-    },
-    {
-      "epoch": 3.698233984825592,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 227630
-    },
-    {
-      "epoch": 3.6983964517229615,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1965,
-      "step": 227640
-    },
-    {
-      "epoch": 3.698558918620331,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 227650
-    },
-    {
-      "epoch": 3.698721385517701,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2009,
-      "step": 227660
-    },
-    {
-      "epoch": 3.6988838524150705,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2137,
-      "step": 227670
-    },
-    {
-      "epoch": 3.69904631931244,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2467,
-      "step": 227680
-    },
-    {
-      "epoch": 3.6992087862098098,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2152,
-      "step": 227690
-    },
-    {
-      "epoch": 3.6993712531071794,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 227700
-    },
-    {
-      "epoch": 3.699533720004549,
-      "grad_norm": 8.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.195,
-      "step": 227710
-    },
-    {
-      "epoch": 3.6996961869019187,
-      "grad_norm": 7.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2443,
-      "step": 227720
-    },
-    {
-      "epoch": 3.6998586537992884,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1941,
-      "step": 227730
-    },
-    {
-      "epoch": 3.700021120696658,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 227740
-    },
-    {
-      "epoch": 3.7001835875940277,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2256,
-      "step": 227750
-    },
-    {
-      "epoch": 3.7003460544913973,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1847,
-      "step": 227760
-    },
-    {
-      "epoch": 3.700508521388767,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2076,
-      "step": 227770
-    },
-    {
-      "epoch": 3.7006709882861366,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2006,
-      "step": 227780
-    },
-    {
-      "epoch": 3.7008334551835063,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1846,
-      "step": 227790
-    },
-    {
-      "epoch": 3.700995922080876,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1998,
-      "step": 227800
-    },
-    {
-      "epoch": 3.7011583889782456,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 227810
-    },
-    {
-      "epoch": 3.701320855875615,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2422,
-      "step": 227820
-    },
-    {
-      "epoch": 3.701483322772985,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.1918,
-      "step": 227830
-    },
-    {
-      "epoch": 3.7016457896703545,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2379,
-      "step": 227840
-    },
-    {
-      "epoch": 3.701808256567724,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2211,
-      "step": 227850
-    },
-    {
-      "epoch": 3.7019707234650943,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 227860
-    },
-    {
-      "epoch": 3.7021331903624635,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.201,
-      "step": 227870
-    },
-    {
-      "epoch": 3.7022956572598336,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2201,
-      "step": 227880
-    },
-    {
-      "epoch": 3.7024581241572028,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 227890
-    },
-    {
-      "epoch": 3.702620591054573,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2288,
-      "step": 227900
-    },
-    {
-      "epoch": 3.702783057951942,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2338,
-      "step": 227910
-    },
-    {
-      "epoch": 3.702945524849312,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2215,
-      "step": 227920
-    },
-    {
-      "epoch": 3.7031079917466814,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2273,
-      "step": 227930
-    },
-    {
-      "epoch": 3.7032704586440515,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.21,
-      "step": 227940
-    },
-    {
-      "epoch": 3.7034329255414207,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2157,
-      "step": 227950
-    },
-    {
-      "epoch": 3.7035953924387908,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1946,
-      "step": 227960
-    },
-    {
-      "epoch": 3.7037578593361604,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.1994,
-      "step": 227970
-    },
-    {
-      "epoch": 3.70392032623353,
-      "grad_norm": 6.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.219,
-      "step": 227980
-    },
-    {
-      "epoch": 3.7040827931308997,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2166,
-      "step": 227990
-    },
-    {
-      "epoch": 3.7042452600282694,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.1995,
-      "step": 228000
-    },
-    {
-      "epoch": 3.704407726925639,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2059,
-      "step": 228010
-    },
-    {
-      "epoch": 3.7045701938230087,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2017,
-      "step": 228020
-    },
-    {
-      "epoch": 3.7047326607203783,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2096,
-      "step": 228030
-    },
-    {
-      "epoch": 3.704895127617748,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 228040
-    },
-    {
-      "epoch": 3.7050575945151176,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 228050
-    },
-    {
-      "epoch": 3.7052200614124873,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1805,
-      "step": 228060
-    },
-    {
-      "epoch": 3.705382528309857,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1882,
-      "step": 228070
-    },
-    {
-      "epoch": 3.7055449952072266,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 228080
-    },
-    {
-      "epoch": 3.7057074621045962,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 228090
-    },
-    {
-      "epoch": 3.705869929001966,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2169,
-      "step": 228100
-    },
-    {
-      "epoch": 3.7060323958993355,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2069,
-      "step": 228110
-    },
-    {
-      "epoch": 3.706194862796705,
-      "grad_norm": 7.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 228120
-    },
-    {
-      "epoch": 3.706357329694075,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2446,
-      "step": 228130
-    },
-    {
-      "epoch": 3.7065197965914445,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 228140
-    },
-    {
-      "epoch": 3.706682263488814,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.232,
-      "step": 228150
-    },
-    {
-      "epoch": 3.706844730386184,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2203,
-      "step": 228160
-    },
-    {
-      "epoch": 3.7070071972835534,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 228170
-    },
-    {
-      "epoch": 3.707169664180923,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2349,
-      "step": 228180
-    },
-    {
-      "epoch": 3.7073321310782927,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2112,
-      "step": 228190
-    },
-    {
-      "epoch": 3.7074945979756624,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2374,
-      "step": 228200
-    },
-    {
-      "epoch": 3.707657064873032,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2057,
-      "step": 228210
-    },
-    {
-      "epoch": 3.7078195317704017,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1958,
-      "step": 228220
-    },
-    {
-      "epoch": 3.7079819986677713,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.1974,
-      "step": 228230
-    },
-    {
-      "epoch": 3.708144465565141,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 228240
-    },
-    {
-      "epoch": 3.7083069324625106,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 228250
-    },
-    {
-      "epoch": 3.7084693993598803,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2351,
-      "step": 228260
-    },
-    {
-      "epoch": 3.70863186625725,
-      "grad_norm": 3.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2117,
-      "step": 228270
-    },
-    {
-      "epoch": 3.7087943331546196,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 228280
-    },
-    {
-      "epoch": 3.7089568000519892,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 228290
-    },
-    {
-      "epoch": 3.709119266949359,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 228300
-    },
-    {
-      "epoch": 3.709281733846729,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 228310
-    },
-    {
-      "epoch": 3.709444200744098,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 228320
-    },
-    {
-      "epoch": 3.7096066676414683,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 228330
-    },
-    {
-      "epoch": 3.7097691345388375,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2361,
-      "step": 228340
-    },
-    {
-      "epoch": 3.7099316014362076,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2049,
-      "step": 228350
-    },
-    {
-      "epoch": 3.710094068333577,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 228360
-    },
-    {
-      "epoch": 3.710256535230947,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2095,
-      "step": 228370
-    },
-    {
-      "epoch": 3.710419002128316,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2306,
-      "step": 228380
-    },
-    {
-      "epoch": 3.710581469025686,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.224,
-      "step": 228390
-    },
-    {
-      "epoch": 3.7107439359230554,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 228400
-    },
-    {
-      "epoch": 3.7109064028204255,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2106,
-      "step": 228410
-    },
-    {
-      "epoch": 3.711068869717795,
-      "grad_norm": 7.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2385,
-      "step": 228420
-    },
-    {
-      "epoch": 3.711231336615165,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2344,
-      "step": 228430
-    },
-    {
-      "epoch": 3.7113938035125345,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2342,
-      "step": 228440
-    },
-    {
-      "epoch": 3.711556270409904,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 228450
-    },
-    {
-      "epoch": 3.7117187373072738,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 228460
-    },
-    {
-      "epoch": 3.7118812042046434,
-      "grad_norm": 7.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2204,
-      "step": 228470
-    },
-    {
-      "epoch": 3.712043671102013,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1933,
-      "step": 228480
-    },
-    {
-      "epoch": 3.7122061379993827,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1971,
-      "step": 228490
-    },
-    {
-      "epoch": 3.7123686048967524,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 228500
-    },
-    {
-      "epoch": 3.712531071794122,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 228510
-    },
-    {
-      "epoch": 3.7126935386914917,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.1996,
-      "step": 228520
-    },
-    {
-      "epoch": 3.7128560055888613,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 228530
-    },
-    {
-      "epoch": 3.713018472486231,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2226,
-      "step": 228540
-    },
-    {
-      "epoch": 3.7131809393836006,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 228550
-    },
-    {
-      "epoch": 3.7133434062809703,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.215,
-      "step": 228560
-    },
-    {
-      "epoch": 3.71350587317834,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 228570
-    },
-    {
-      "epoch": 3.7136683400757096,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2407,
-      "step": 228580
-    },
-    {
-      "epoch": 3.713830806973079,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.233,
-      "step": 228590
-    },
-    {
-      "epoch": 3.713993273870449,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 228600
-    },
-    {
-      "epoch": 3.7141557407678185,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2337,
-      "step": 228610
-    },
-    {
-      "epoch": 3.714318207665188,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2294,
-      "step": 228620
-    },
-    {
-      "epoch": 3.714480674562558,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.208,
-      "step": 228630
-    },
-    {
-      "epoch": 3.7146431414599275,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2226,
-      "step": 228640
-    },
-    {
-      "epoch": 3.714805608357297,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2382,
-      "step": 228650
-    },
-    {
-      "epoch": 3.7149680752546668,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2299,
-      "step": 228660
-    },
-    {
-      "epoch": 3.7151305421520364,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2295,
-      "step": 228670
-    },
-    {
-      "epoch": 3.715293009049406,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2146,
-      "step": 228680
-    },
-    {
-      "epoch": 3.7154554759467757,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2368,
-      "step": 228690
-    },
-    {
-      "epoch": 3.7156179428441454,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2269,
-      "step": 228700
-    },
-    {
-      "epoch": 3.715780409741515,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.1947,
-      "step": 228710
-    },
-    {
-      "epoch": 3.7159428766388847,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 228720
-    },
-    {
-      "epoch": 3.7161053435362543,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 228730
-    },
-    {
-      "epoch": 3.7162678104336244,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2237,
-      "step": 228740
-    },
-    {
-      "epoch": 3.7164302773309936,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2254,
-      "step": 228750
-    },
-    {
-      "epoch": 3.7165927442283637,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2026,
-      "step": 228760
-    },
-    {
-      "epoch": 3.716755211125733,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2227,
-      "step": 228770
-    },
-    {
-      "epoch": 3.716917678023103,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2256,
-      "step": 228780
-    },
-    {
-      "epoch": 3.7170801449204722,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2351,
-      "step": 228790
-    },
-    {
-      "epoch": 3.7172426118178423,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2275,
-      "step": 228800
-    },
-    {
-      "epoch": 3.7174050787152115,
-      "grad_norm": 7.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 228810
-    },
-    {
-      "epoch": 3.7175675456125816,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2309,
-      "step": 228820
-    },
-    {
-      "epoch": 3.717730012509951,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2498,
-      "step": 228830
-    },
-    {
-      "epoch": 3.717892479407321,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2056,
-      "step": 228840
-    },
-    {
-      "epoch": 3.7180549463046906,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2345,
-      "step": 228850
-    },
-    {
-      "epoch": 3.7182174132020602,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2336,
-      "step": 228860
-    },
-    {
-      "epoch": 3.71837988009943,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2094,
-      "step": 228870
-    },
-    {
-      "epoch": 3.7185423469967995,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2172,
-      "step": 228880
-    },
-    {
-      "epoch": 3.718704813894169,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2335,
-      "step": 228890
-    },
-    {
-      "epoch": 3.718867280791539,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1957,
-      "step": 228900
-    },
-    {
-      "epoch": 3.7190297476889085,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2051,
-      "step": 228910
-    },
-    {
-      "epoch": 3.719192214586278,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2325,
-      "step": 228920
-    },
-    {
-      "epoch": 3.719354681483648,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2036,
-      "step": 228930
-    },
-    {
-      "epoch": 3.7195171483810174,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2264,
-      "step": 228940
-    },
-    {
-      "epoch": 3.719679615278387,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2226,
-      "step": 228950
-    },
-    {
-      "epoch": 3.7198420821757567,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2292,
-      "step": 228960
-    },
-    {
-      "epoch": 3.7200045490731264,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 228970
-    },
-    {
-      "epoch": 3.720167015970496,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2091,
-      "step": 228980
-    },
-    {
-      "epoch": 3.7203294828678657,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2,
-      "step": 228990
-    },
-    {
-      "epoch": 3.7204919497652353,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.1889,
-      "step": 229000
-    },
-    {
-      "epoch": 3.720654416662605,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1903,
-      "step": 229010
-    },
-    {
-      "epoch": 3.7208168835599746,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.1997,
-      "step": 229020
-    },
-    {
-      "epoch": 3.7209793504573443,
-      "grad_norm": 6.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 229030
-    },
-    {
-      "epoch": 3.721141817354714,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1874,
-      "step": 229040
-    },
-    {
-      "epoch": 3.7213042842520836,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2035,
-      "step": 229050
-    },
-    {
-      "epoch": 3.7214667511494532,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.242,
-      "step": 229060
-    },
-    {
-      "epoch": 3.721629218046823,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.216,
-      "step": 229070
-    },
-    {
-      "epoch": 3.7217916849441925,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2029,
-      "step": 229080
-    },
-    {
-      "epoch": 3.721954151841562,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2152,
-      "step": 229090
-    },
-    {
-      "epoch": 3.722116618738932,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2302,
-      "step": 229100
-    },
-    {
-      "epoch": 3.7222790856363015,
-      "grad_norm": 7.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 229110
-    },
-    {
-      "epoch": 3.722441552533671,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1856,
-      "step": 229120
-    },
-    {
-      "epoch": 3.722604019431041,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 229130
-    },
-    {
-      "epoch": 3.7227664863284105,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.204,
-      "step": 229140
-    },
-    {
-      "epoch": 3.72292895322578,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2059,
-      "step": 229150
-    },
-    {
-      "epoch": 3.7230914201231498,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 229160
-    },
-    {
-      "epoch": 3.7232538870205194,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2338,
-      "step": 229170
-    },
-    {
-      "epoch": 3.723416353917889,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1992,
-      "step": 229180
-    },
-    {
-      "epoch": 3.723578820815259,
-      "grad_norm": 6.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2191,
-      "step": 229190
-    },
-    {
-      "epoch": 3.7237412877126284,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2073,
-      "step": 229200
-    },
-    {
-      "epoch": 3.7239037546099985,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 229210
-    },
-    {
-      "epoch": 3.7240662215073677,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2279,
-      "step": 229220
-    },
-    {
-      "epoch": 3.7242286884047378,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2399,
-      "step": 229230
-    },
-    {
-      "epoch": 3.724391155302107,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2419,
-      "step": 229240
-    },
-    {
-      "epoch": 3.724553622199477,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 229250
-    },
-    {
-      "epoch": 3.7247160890968463,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2048,
-      "step": 229260
-    },
-    {
-      "epoch": 3.7248785559942164,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1877,
-      "step": 229270
-    },
-    {
-      "epoch": 3.7250410228915856,
-      "grad_norm": 7.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2196,
-      "step": 229280
-    },
-    {
-      "epoch": 3.7252034897889557,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2111,
-      "step": 229290
-    },
-    {
-      "epoch": 3.7253659566863253,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 229300
-    },
-    {
-      "epoch": 3.725528423583695,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 229310
-    },
-    {
-      "epoch": 3.7256908904810646,
-      "grad_norm": 7.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2124,
-      "step": 229320
-    },
-    {
-      "epoch": 3.7258533573784343,
-      "grad_norm": 7.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2053,
-      "step": 229330
-    },
-    {
-      "epoch": 3.726015824275804,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 229340
-    },
-    {
-      "epoch": 3.7261782911731736,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 229350
-    },
-    {
-      "epoch": 3.726340758070543,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2113,
-      "step": 229360
-    },
-    {
-      "epoch": 3.726503224967913,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2039,
-      "step": 229370
-    },
-    {
-      "epoch": 3.7266656918652825,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1924,
-      "step": 229380
-    },
-    {
-      "epoch": 3.726828158762652,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.189,
-      "step": 229390
-    },
-    {
-      "epoch": 3.726990625660022,
-      "grad_norm": 6.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2192,
-      "step": 229400
-    },
-    {
-      "epoch": 3.7271530925573915,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2598,
-      "step": 229410
-    },
-    {
-      "epoch": 3.727315559454761,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2211,
-      "step": 229420
-    },
-    {
-      "epoch": 3.7274780263521308,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 229430
-    },
-    {
-      "epoch": 3.7276404932495004,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.209,
-      "step": 229440
-    },
-    {
-      "epoch": 3.72780296014687,
-      "grad_norm": 7.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2319,
-      "step": 229450
-    },
-    {
-      "epoch": 3.7279654270442397,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 229460
-    },
-    {
-      "epoch": 3.7281278939416094,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2312,
-      "step": 229470
-    },
-    {
-      "epoch": 3.728290360838979,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2347,
-      "step": 229480
-    },
-    {
-      "epoch": 3.7284528277363487,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2554,
-      "step": 229490
-    },
-    {
-      "epoch": 3.7286152946337183,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2298,
-      "step": 229500
-    },
-    {
-      "epoch": 3.728777761531088,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2268,
-      "step": 229510
-    },
-    {
-      "epoch": 3.7289402284284576,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2149,
-      "step": 229520
-    },
-    {
-      "epoch": 3.7291026953258273,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 229530
-    },
-    {
-      "epoch": 3.729265162223197,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2269,
-      "step": 229540
-    },
-    {
-      "epoch": 3.7294276291205666,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2426,
-      "step": 229550
-    },
-    {
-      "epoch": 3.7295900960179362,
-      "grad_norm": 3.953125,
-      "learning_rate": 5e-05,
-      "loss": 0.2272,
-      "step": 229560
-    },
-    {
-      "epoch": 3.729752562915306,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 229570
-    },
-    {
-      "epoch": 3.7299150298126755,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1982,
-      "step": 229580
-    },
-    {
-      "epoch": 3.730077496710045,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1879,
-      "step": 229590
-    },
-    {
-      "epoch": 3.730239963607415,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2199,
-      "step": 229600
-    },
-    {
-      "epoch": 3.7304024305047845,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2213,
-      "step": 229610
-    },
-    {
-      "epoch": 3.7305648974021546,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2251,
-      "step": 229620
-    },
-    {
-      "epoch": 3.730727364299524,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2327,
-      "step": 229630
-    },
-    {
-      "epoch": 3.730889831196894,
-      "grad_norm": 4.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2086,
-      "step": 229640
-    },
-    {
-      "epoch": 3.731052298094263,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2353,
-      "step": 229650
-    },
-    {
-      "epoch": 3.731214764991633,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2185,
-      "step": 229660
-    },
-    {
-      "epoch": 3.7313772318890024,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.214,
-      "step": 229670
-    },
-    {
-      "epoch": 3.7315396987863725,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2684,
-      "step": 229680
-    },
-    {
-      "epoch": 3.7317021656837417,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2501,
-      "step": 229690
-    },
-    {
-      "epoch": 3.731864632581112,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2333,
-      "step": 229700
-    },
-    {
-      "epoch": 3.732027099478481,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2379,
-      "step": 229710
-    },
-    {
-      "epoch": 3.732189566375851,
-      "grad_norm": 4.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2233,
-      "step": 229720
-    },
-    {
-      "epoch": 3.7323520332732207,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2422,
-      "step": 229730
-    },
-    {
-      "epoch": 3.7325145001705904,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 229740
-    },
-    {
-      "epoch": 3.73267696706796,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2333,
-      "step": 229750
-    },
-    {
-      "epoch": 3.7328394339653297,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2037,
-      "step": 229760
-    },
-    {
-      "epoch": 3.7330019008626993,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1981,
-      "step": 229770
-    },
-    {
-      "epoch": 3.733164367760069,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2344,
-      "step": 229780
-    },
-    {
-      "epoch": 3.7333268346574386,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2503,
-      "step": 229790
-    },
-    {
-      "epoch": 3.7334893015548083,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2274,
-      "step": 229800
-    },
-    {
-      "epoch": 3.733651768452178,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2067,
-      "step": 229810
-    },
-    {
-      "epoch": 3.7338142353495476,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2254,
-      "step": 229820
-    },
-    {
-      "epoch": 3.7339767022469172,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.236,
-      "step": 229830
-    },
-    {
-      "epoch": 3.734139169144287,
-      "grad_norm": 4.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2104,
-      "step": 229840
-    },
-    {
-      "epoch": 3.7343016360416565,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2288,
-      "step": 229850
-    },
-    {
-      "epoch": 3.734464102939026,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 229860
-    },
-    {
-      "epoch": 3.734626569836396,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2055,
-      "step": 229870
-    },
-    {
-      "epoch": 3.7347890367337655,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.215,
-      "step": 229880
-    },
-    {
-      "epoch": 3.734951503631135,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 229890
-    },
-    {
-      "epoch": 3.735113970528505,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 229900
-    },
-    {
-      "epoch": 3.7352764374258745,
-      "grad_norm": 6.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2007,
-      "step": 229910
-    },
-    {
-      "epoch": 3.735438904323244,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 229920
-    },
-    {
-      "epoch": 3.7356013712206138,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2153,
-      "step": 229930
-    },
-    {
-      "epoch": 3.7357638381179834,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2124,
-      "step": 229940
-    },
-    {
-      "epoch": 3.735926305015353,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.206,
-      "step": 229950
-    },
-    {
-      "epoch": 3.7360887719127227,
-      "grad_norm": 7.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2271,
-      "step": 229960
-    },
-    {
-      "epoch": 3.7362512388100924,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1861,
-      "step": 229970
-    },
-    {
-      "epoch": 3.736413705707462,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2178,
-      "step": 229980
-    },
-    {
-      "epoch": 3.7365761726048317,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2368,
-      "step": 229990
-    },
-    {
-      "epoch": 3.7367386395022013,
-      "grad_norm": 5.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2233,
-      "step": 230000
-    },
-    {
-      "epoch": 3.736901106399571,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2018,
-      "step": 230010
-    },
-    {
-      "epoch": 3.7370635732969406,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2027,
-      "step": 230020
-    },
-    {
-      "epoch": 3.7372260401943103,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.234,
-      "step": 230030
-    },
-    {
-      "epoch": 3.73738850709168,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2299,
-      "step": 230040
-    },
-    {
-      "epoch": 3.73755097398905,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 230050
-    },
-    {
-      "epoch": 3.737713440886419,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 230060
-    },
-    {
-      "epoch": 3.7378759077837893,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 230070
-    },
-    {
-      "epoch": 3.7380383746811585,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.1942,
-      "step": 230080
-    },
-    {
-      "epoch": 3.7382008415785286,
-      "grad_norm": 8.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2034,
-      "step": 230090
-    },
-    {
-      "epoch": 3.738363308475898,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2296,
-      "step": 230100
-    },
-    {
-      "epoch": 3.738525775373268,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 230110
-    },
-    {
-      "epoch": 3.738688242270637,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.1976,
-      "step": 230120
-    },
-    {
-      "epoch": 3.738850709168007,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1862,
-      "step": 230130
-    },
-    {
-      "epoch": 3.7390131760653764,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.1969,
-      "step": 230140
-    },
-    {
-      "epoch": 3.7391756429627465,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 230150
-    },
-    {
-      "epoch": 3.739338109860116,
-      "grad_norm": 7.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2293,
-      "step": 230160
-    },
-    {
-      "epoch": 3.739500576757486,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.248,
-      "step": 230170
-    },
-    {
-      "epoch": 3.7396630436548555,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2206,
-      "step": 230180
-    },
-    {
-      "epoch": 3.739825510552225,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2399,
-      "step": 230190
-    },
-    {
-      "epoch": 3.7399879774495948,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 230200
-    },
-    {
-      "epoch": 3.7401504443469644,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2415,
-      "step": 230210
-    },
-    {
-      "epoch": 3.740312911244334,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2494,
-      "step": 230220
-    },
-    {
-      "epoch": 3.7404753781417037,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2417,
-      "step": 230230
-    },
-    {
-      "epoch": 3.7406378450390734,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2245,
-      "step": 230240
-    },
-    {
-      "epoch": 3.740800311936443,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2162,
-      "step": 230250
-    },
-    {
-      "epoch": 3.7409627788338127,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 230260
-    },
-    {
-      "epoch": 3.7411252457311823,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2394,
-      "step": 230270
-    },
-    {
-      "epoch": 3.741287712628552,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2193,
-      "step": 230280
-    },
-    {
-      "epoch": 3.7414501795259216,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.1891,
-      "step": 230290
-    },
-    {
-      "epoch": 3.7416126464232913,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2196,
-      "step": 230300
-    },
-    {
-      "epoch": 3.741775113320661,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2163,
-      "step": 230310
-    },
-    {
-      "epoch": 3.7419375802180306,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2291,
-      "step": 230320
-    },
-    {
-      "epoch": 3.7421000471154002,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2315,
-      "step": 230330
-    },
-    {
-      "epoch": 3.74226251401277,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2454,
-      "step": 230340
-    },
-    {
-      "epoch": 3.7424249809101395,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2205,
-      "step": 230350
-    },
-    {
-      "epoch": 3.742587447807509,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2349,
-      "step": 230360
-    },
-    {
-      "epoch": 3.742749914704879,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 230370
-    },
-    {
-      "epoch": 3.7429123816022485,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2266,
-      "step": 230380
-    },
-    {
-      "epoch": 3.743074848499618,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.236,
-      "step": 230390
-    },
-    {
-      "epoch": 3.743237315396988,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2033,
-      "step": 230400
-    },
-    {
-      "epoch": 3.7433997822943574,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2046,
-      "step": 230410
-    },
-    {
-      "epoch": 3.743562249191727,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2083,
-      "step": 230420
-    },
-    {
-      "epoch": 3.7437247160890967,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2151,
-      "step": 230430
-    },
-    {
-      "epoch": 3.7438871829864664,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2327,
-      "step": 230440
-    },
-    {
-      "epoch": 3.744049649883836,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2023,
-      "step": 230450
-    },
-    {
-      "epoch": 3.7442121167812057,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2058,
-      "step": 230460
-    },
-    {
-      "epoch": 3.7443745836785753,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2074,
-      "step": 230470
-    },
-    {
-      "epoch": 3.744537050575945,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1647,
-      "step": 230480
-    },
-    {
-      "epoch": 3.7446995174733146,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1945,
-      "step": 230490
-    },
-    {
-      "epoch": 3.7448619843706847,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2184,
-      "step": 230500
-    },
-    {
-      "epoch": 3.745024451268054,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2134,
-      "step": 230510
-    },
-    {
-      "epoch": 3.745186918165424,
-      "grad_norm": 4.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 230520
-    },
-    {
-      "epoch": 3.7453493850627932,
-      "grad_norm": 4.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.1912,
-      "step": 230530
-    },
-    {
-      "epoch": 3.7455118519601633,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1934,
-      "step": 230540
-    },
-    {
-      "epoch": 3.7456743188575325,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1928,
-      "step": 230550
-    },
-    {
-      "epoch": 3.7458367857549026,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2065,
-      "step": 230560
-    },
-    {
-      "epoch": 3.745999252652272,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2099,
-      "step": 230570
-    },
-    {
-      "epoch": 3.746161719549642,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.216,
-      "step": 230580
-    },
-    {
-      "epoch": 3.746324186447011,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 230590
-    },
-    {
-      "epoch": 3.7464866533443812,
-      "grad_norm": 6.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2135,
-      "step": 230600
-    },
-    {
-      "epoch": 3.746649120241751,
-      "grad_norm": 7.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2346,
-      "step": 230610
-    },
-    {
-      "epoch": 3.7468115871391205,
-      "grad_norm": 6.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2241,
-      "step": 230620
-    },
-    {
-      "epoch": 3.74697405403649,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2332,
-      "step": 230630
-    },
-    {
-      "epoch": 3.74713652093386,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2328,
-      "step": 230640
-    },
-    {
-      "epoch": 3.7472989878312295,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2543,
-      "step": 230650
-    },
-    {
-      "epoch": 3.747461454728599,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2546,
-      "step": 230660
-    },
-    {
-      "epoch": 3.747623921625969,
-      "grad_norm": 5.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2157,
-      "step": 230670
-    },
-    {
-      "epoch": 3.7477863885233385,
-      "grad_norm": 7.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2289,
-      "step": 230680
-    },
-    {
-      "epoch": 3.747948855420708,
-      "grad_norm": 6.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2165,
-      "step": 230690
-    },
-    {
-      "epoch": 3.7481113223180778,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2318,
-      "step": 230700
-    },
-    {
-      "epoch": 3.7482737892154474,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.231,
-      "step": 230710
-    },
-    {
-      "epoch": 3.748436256112817,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.231,
-      "step": 230720
-    },
-    {
-      "epoch": 3.7485987230101867,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2269,
-      "step": 230730
-    },
-    {
-      "epoch": 3.7487611899075564,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2167,
-      "step": 230740
-    },
-    {
-      "epoch": 3.748923656804926,
-      "grad_norm": 7.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2226,
-      "step": 230750
-    },
-    {
-      "epoch": 3.7490861237022957,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2021,
-      "step": 230760
-    },
-    {
-      "epoch": 3.7492485905996653,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.1935,
-      "step": 230770
-    },
-    {
-      "epoch": 3.749411057497035,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2127,
-      "step": 230780
-    },
-    {
-      "epoch": 3.7495735243944046,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 230790
-    },
-    {
-      "epoch": 3.7497359912917743,
-      "grad_norm": 4.5,
-      "learning_rate": 5e-05,
-      "loss": 0.1884,
-      "step": 230800
-    },
-    {
-      "epoch": 3.749898458189144,
-      "grad_norm": 8.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2045,
-      "step": 230810
-    },
-    {
-      "epoch": 3.7500609250865136,
-      "grad_norm": 6.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 230820
-    },
-    {
-      "epoch": 3.750223391983883,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2428,
-      "step": 230830
-    },
-    {
-      "epoch": 3.750385858881253,
-      "grad_norm": 6.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2197,
-      "step": 230840
-    },
-    {
-      "epoch": 3.7505483257786225,
-      "grad_norm": 4.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2327,
-      "step": 230850
-    },
-    {
-      "epoch": 3.750710792675992,
-      "grad_norm": 5.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2351,
-      "step": 230860
-    },
-    {
-      "epoch": 3.750873259573362,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2383,
-      "step": 230870
-    },
-    {
-      "epoch": 3.7510357264707315,
-      "grad_norm": 6.0625,
-      "learning_rate": 5e-05,
-      "loss": 0.2238,
-      "step": 230880
-    },
-    {
-      "epoch": 3.751198193368101,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2509,
-      "step": 230890
-    },
-    {
-      "epoch": 3.7513606602654708,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.218,
-      "step": 230900
-    },
-    {
-      "epoch": 3.7515231271628404,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2461,
-      "step": 230910
-    },
-    {
-      "epoch": 3.75168559406021,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2252,
-      "step": 230920
-    },
-    {
-      "epoch": 3.75184806095758,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2221,
-      "step": 230930
-    },
-    {
-      "epoch": 3.7520105278549494,
-      "grad_norm": 7.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2219,
-      "step": 230940
-    },
-    {
-      "epoch": 3.7521729947523195,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2353,
-      "step": 230950
-    },
-    {
-      "epoch": 3.7523354616496887,
-      "grad_norm": 4.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2458,
-      "step": 230960
-    },
-    {
-      "epoch": 3.7524979285470588,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2339,
-      "step": 230970
-    },
-    {
-      "epoch": 3.752660395444428,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2307,
-      "step": 230980
-    },
-    {
-      "epoch": 3.752822862341798,
-      "grad_norm": 5.4375,
-      "learning_rate": 5e-05,
-      "loss": 0.2242,
-      "step": 230990
-    },
-    {
-      "epoch": 3.7529853292391673,
-      "grad_norm": 5.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2101,
-      "step": 231000
-    },
-    {
-      "epoch": 3.7531477961365374,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2077,
-      "step": 231010
-    },
-    {
-      "epoch": 3.7533102630339066,
-      "grad_norm": 5.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.2389,
-      "step": 231020
-    },
-    {
-      "epoch": 3.7534727299312767,
-      "grad_norm": 4.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2392,
-      "step": 231030
-    },
-    {
-      "epoch": 3.7536351968286463,
-      "grad_norm": 6.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2403,
-      "step": 231040
-    },
-    {
-      "epoch": 3.753797663726016,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2378,
-      "step": 231050
-    },
-    {
-      "epoch": 3.7539601306233856,
-      "grad_norm": 4.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2447,
-      "step": 231060
-    },
-    {
-      "epoch": 3.7541225975207553,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2209,
-      "step": 231070
-    },
-    {
-      "epoch": 3.754285064418125,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2359,
-      "step": 231080
-    },
-    {
-      "epoch": 3.7544475313154946,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.1962,
-      "step": 231090
-    },
-    {
-      "epoch": 3.7546099982128642,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2204,
-      "step": 231100
-    },
-    {
-      "epoch": 3.754772465110234,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2143,
-      "step": 231110
-    },
-    {
-      "epoch": 3.7549349320076035,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.1926,
-      "step": 231120
-    },
-    {
-      "epoch": 3.755097398904973,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.1929,
-      "step": 231130
-    },
-    {
-      "epoch": 3.755259865802343,
-      "grad_norm": 4.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.2138,
-      "step": 231140
-    },
-    {
-      "epoch": 3.7554223326997125,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.1979,
-      "step": 231150
-    },
-    {
-      "epoch": 3.755584799597082,
-      "grad_norm": 5.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2088,
-      "step": 231160
-    },
-    {
-      "epoch": 3.755747266494452,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2218,
-      "step": 231170
-    },
-    {
-      "epoch": 3.7559097333918214,
-      "grad_norm": 7.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2291,
-      "step": 231180
-    },
-    {
-      "epoch": 3.756072200289191,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2222,
-      "step": 231190
-    },
-    {
-      "epoch": 3.7562346671865607,
-      "grad_norm": 6.71875,
-      "learning_rate": 5e-05,
-      "loss": 0.2202,
-      "step": 231200
-    },
-    {
-      "epoch": 3.7563971340839304,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2307,
-      "step": 231210
-    },
-    {
-      "epoch": 3.7565596009813,
-      "grad_norm": 5.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2054,
-      "step": 231220
-    },
-    {
-      "epoch": 3.7567220678786697,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2115,
-      "step": 231230
-    },
-    {
-      "epoch": 3.7568845347760393,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2332,
-      "step": 231240
-    },
-    {
-      "epoch": 3.757047001673409,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2079,
-      "step": 231250
-    },
-    {
-      "epoch": 3.7572094685707786,
-      "grad_norm": 5.5625,
-      "learning_rate": 5e-05,
-      "loss": 0.2181,
-      "step": 231260
-    },
-    {
-      "epoch": 3.7573719354681483,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2285,
-      "step": 231270
-    },
-    {
-      "epoch": 3.757534402365518,
-      "grad_norm": 6.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 231280
-    },
-    {
-      "epoch": 3.7576968692628876,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2242,
-      "step": 231290
-    },
-    {
-      "epoch": 3.7578593361602572,
-      "grad_norm": 5.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2129,
-      "step": 231300
-    },
-    {
-      "epoch": 3.758021803057627,
-      "grad_norm": 5.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2182,
-      "step": 231310
-    },
-    {
-      "epoch": 3.7581842699549965,
-      "grad_norm": 7.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2434,
-      "step": 231320
-    },
-    {
-      "epoch": 3.758346736852366,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2293,
-      "step": 231330
-    },
-    {
-      "epoch": 3.758509203749736,
-      "grad_norm": 5.3125,
-      "learning_rate": 5e-05,
-      "loss": 0.2164,
-      "step": 231340
-    },
-    {
-      "epoch": 3.7586716706471055,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2043,
-      "step": 231350
-    },
-    {
-      "epoch": 3.758834137544475,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 231360
-    },
-    {
-      "epoch": 3.758996604441845,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 231370
-    },
-    {
-      "epoch": 3.759159071339215,
-      "grad_norm": 7.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.2321,
-      "step": 231380
-    },
-    {
-      "epoch": 3.759321538236584,
-      "grad_norm": 5.375,
-      "learning_rate": 5e-05,
-      "loss": 0.22,
-      "step": 231390
-    },
-    {
-      "epoch": 3.759484005133954,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2158,
-      "step": 231400
-    },
-    {
-      "epoch": 3.7596464720313234,
-      "grad_norm": 4.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2024,
-      "step": 231410
-    },
-    {
-      "epoch": 3.7598089389286935,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.1937,
-      "step": 231420
-    },
-    {
-      "epoch": 3.7599714058260627,
-      "grad_norm": 6.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2204,
-      "step": 231430
-    },
-    {
-      "epoch": 3.760133872723433,
-      "grad_norm": 5.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2071,
-      "step": 231440
-    },
-    {
-      "epoch": 3.760296339620802,
-      "grad_norm": 6.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2186,
-      "step": 231450
-    },
-    {
-      "epoch": 3.760458806518172,
-      "grad_norm": 5.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2033,
-      "step": 231460
-    },
-    {
-      "epoch": 3.7606212734155413,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 231470
-    },
-    {
-      "epoch": 3.7607837403129114,
-      "grad_norm": 4.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2003,
-      "step": 231480
-    },
-    {
-      "epoch": 3.760946207210281,
-      "grad_norm": 5.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2041,
-      "step": 231490
-    },
-    {
-      "epoch": 3.7611086741076507,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2102,
-      "step": 231500
-    },
-    {
-      "epoch": 3.7612711410050204,
-      "grad_norm": 6.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2287,
-      "step": 231510
-    },
-    {
-      "epoch": 3.76143360790239,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2205,
-      "step": 231520
-    },
-    {
-      "epoch": 3.7615960747997597,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2042,
-      "step": 231530
-    },
-    {
-      "epoch": 3.7617585416971293,
-      "grad_norm": 6.5,
-      "learning_rate": 5e-05,
-      "loss": 0.2225,
-      "step": 231540
-    },
-    {
-      "epoch": 3.761921008594499,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2266,
-      "step": 231550
-    },
-    {
-      "epoch": 3.7620834754918686,
-      "grad_norm": 4.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2038,
-      "step": 231560
-    },
-    {
-      "epoch": 3.7622459423892383,
-      "grad_norm": 4.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2287,
-      "step": 231570
-    },
-    {
-      "epoch": 3.762408409286608,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.213,
-      "step": 231580
-    },
-    {
-      "epoch": 3.7625708761839776,
-      "grad_norm": 4.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.205,
-      "step": 231590
-    },
-    {
-      "epoch": 3.762733343081347,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2166,
-      "step": 231600
-    },
-    {
-      "epoch": 3.762895809978717,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2288,
-      "step": 231610
-    },
-    {
-      "epoch": 3.7630582768760865,
-      "grad_norm": 6.65625,
-      "learning_rate": 5e-05,
-      "loss": 0.2292,
-      "step": 231620
-    },
-    {
-      "epoch": 3.763220743773456,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2078,
-      "step": 231630
-    },
-    {
-      "epoch": 3.763383210670826,
-      "grad_norm": 6.25,
-      "learning_rate": 5e-05,
-      "loss": 0.2025,
-      "step": 231640
-    },
-    {
-      "epoch": 3.7635456775681955,
-      "grad_norm": 4.75,
-      "learning_rate": 5e-05,
-      "loss": 0.2031,
-      "step": 231650
-    },
-    {
-      "epoch": 3.763708144465565,
-      "grad_norm": 6.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.2269,
-      "step": 231660
-    },
-    {
-      "epoch": 3.7638706113629348,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2102,
-      "step": 231670
-    },
-    {
-      "epoch": 3.7640330782603044,
-      "grad_norm": 5.0,
-      "learning_rate": 5e-05,
-      "loss": 0.2103,
-      "step": 231680
-    },
-    {
-      "epoch": 3.764195545157674,
-      "grad_norm": 5.8125,
-      "learning_rate": 5e-05,
-      "loss": 0.1973,
-      "step": 231690
-    },
-    {
-      "epoch": 3.7643580120550437,
-      "grad_norm": 6.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2235,
-      "step": 231700
-    },
-    {
-      "epoch": 3.7645204789524134,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.1818,
-      "step": 231710
-    },
-    {
-      "epoch": 3.764682945849783,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.1806,
-      "step": 231720
-    },
-    {
-      "epoch": 3.7648454127471527,
-      "grad_norm": 6.03125,
-      "learning_rate": 5e-05,
-      "loss": 0.1925,
-      "step": 231730
-    },
-    {
-      "epoch": 3.7650078796445223,
-      "grad_norm": 5.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.184,
-      "step": 231740
-    },
-    {
-      "epoch": 3.765170346541892,
-      "grad_norm": 5.625,
-      "learning_rate": 5e-05,
-      "loss": 0.2092,
-      "step": 231750
-    },
-    {
-      "epoch": 3.7653328134392616,
-      "grad_norm": 5.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2148,
-      "step": 231760
-    },
-    {
-      "epoch": 3.7654952803366313,
-      "grad_norm": 5.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2125,
-      "step": 231770
-    },
-    {
-      "epoch": 3.765657747234001,
-      "grad_norm": 5.90625,
-      "learning_rate": 5e-05,
-      "loss": 0.2313,
-      "step": 231780
-    },
-    {
-      "epoch": 3.7658202141313706,
-      "grad_norm": 6.375,
-      "learning_rate": 5e-05,
-      "loss": 0.2256,
-      "step": 231790
-    },
-    {
-      "epoch": 3.7659826810287402,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2108,
-      "step": 231800
-    },
-    {
-      "epoch": 3.7661451479261103,
-      "grad_norm": 6.75,
-      "learning_rate": 5e-05,
-      "loss": 0.225,
-      "step": 231810
-    },
-    {
-      "epoch": 3.7663076148234795,
-      "grad_norm": 4.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2216,
-      "step": 231820
-    },
-    {
-      "epoch": 3.7664700817208496,
-      "grad_norm": 5.15625,
-      "learning_rate": 5e-05,
-      "loss": 0.1932,
-      "step": 231830
-    },
-    {
-      "epoch": 3.766632548618219,
-      "grad_norm": 7.28125,
-      "learning_rate": 5e-05,
-      "loss": 0.2168,
-      "step": 231840
-    },
-    {
-      "epoch": 3.766795015515589,
-      "grad_norm": 4.53125,
-      "learning_rate": 5e-05,
-      "loss": 0.2013,
-      "step": 231850
-    },
-    {
-      "epoch": 3.766957482412958,
-      "grad_norm": 5.09375,
-      "learning_rate": 5e-05,
-      "loss": 0.2183,
-      "step": 231860
-    },
-    {
-      "epoch": 3.7671199493103282,
-      "grad_norm": 5.6875,
-      "learning_rate": 5e-05,
-      "loss": 0.2367,
-      "step": 231870
-    },
-    {
-      "epoch": 3.7672824162076974,
-      "grad_norm": 5.78125,
-      "learning_rate": 5e-05,
-      "loss": 0.222,
-      "step": 231880
-    },
-    {
-      "epoch": 3.7674448831050675,
-      "grad_norm": 5.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2014,
-      "step": 231890
-    },
-    {
-      "epoch": 3.7676073500024367,
-      "grad_norm": 6.84375,
-      "learning_rate": 5e-05,
-      "loss": 0.2514,
-      "step": 231900
-    },
-    {
-      "epoch": 3.767769816899807,
-      "grad_norm": 5.40625,
-      "learning_rate": 5e-05,
-      "loss": 0.2283,
-      "step": 231910
-    },
-    {
-      "epoch": 3.7679322837971765,
-      "grad_norm": 5.46875,
-      "learning_rate": 5e-05,
-      "loss": 0.2469,
-      "step": 231920
-    },
-    {
-      "epoch": 3.768094750694546,
-      "grad_norm": 4.96875,
-      "learning_rate": 5e-05,
-      "loss": 0.2222,
-      "step": 231930
-    },
-    {
-      "epoch": 3.768257217591916,
-      "grad_norm": 4.59375,
-      "learning_rate": 5e-05,
-      "loss": 0.2159,
-      "step": 231940
-    },
-    {
-      "epoch": 3.7684196844892854,
-      "grad_norm": 6.125,
-      "learning_rate": 5e-05,
-      "loss": 0.2306,
-      "step": 231950
-    },
-    {
-      "epoch": 3.768582151386655,
-      "grad_norm": 6.34375,
-      "learning_rate": 5e-05,
-      "loss": 0.2161,
-      "step": 231960
-    },
-    {
-      "epoch": 3.7687446182840247,
-      "grad_norm": 4.875,
-      "learning_rate": 5e-05,
-      "loss": 0.2212,
-      "step": 231970
-    },
-    {
-      "epoch": 3.7689070851813944,
-      "grad_norm": 5.9375,
-      "learning_rate": 5e-05,
-      "loss": 0.212,
-      "step": 231980
-    },
-    {
-      "epoch": 3.769069552078764,
-      "grad_norm": 5.21875,
-      "learning_rate": 5e-05,
-      "loss": 0.2226,
-      "step": 231990
-    },
-    {
-      "epoch": 3.7692320189761337,
-      "grad_norm": 6.1875,
-      "learning_rate": 5e-05,
-      "loss": 0.2139,
-      "step": 232000
     }
   ],
   "logging_steps": 10,
@@ -162427,7 +100827,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.2257290287220933e+19,
+  "total_flos": 2.885752155158778e+19,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null