diff --git "a/tmp-checkpoint-5335/trainer_state.json" "b/tmp-checkpoint-5335/trainer_state.json"
--- "a/tmp-checkpoint-5335/trainer_state.json"
+++ "b/tmp-checkpoint-5335/trainer_state.json"
@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.3931359052658081,
-  "best_model_checkpoint": "MarkKisker/RoBERTa-base-RottenTomatoes_v2\\checkpoint-2134",
+  "best_metric": 0.6987214088439941,
+  "best_model_checkpoint": "MarkKisker/RoBERTa-base-RottenTomatoes_v2\\checkpoint-1067",
   "epoch": 5.0,
   "eval_steps": 500,
   "global_step": 5335,
@@ -10,3793 +10,3793 @@
   "log_history": [
     {
       "epoch": 0.01,
-      "grad_norm": 3.744861602783203,
+      "grad_norm": 0.18819734454154968,
       "learning_rate": 1.0000000000000002e-06,
-      "loss": 0.7101,
+      "loss": 0.2519,
       "step": 10
     },
     {
       "epoch": 0.02,
-      "grad_norm": 4.64638090133667,
+      "grad_norm": 0.9499709010124207,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 0.7158,
+      "loss": 0.2018,
       "step": 20
     },
     {
       "epoch": 0.03,
-      "grad_norm": 1.5597444772720337,
+      "grad_norm": 20.164424896240234,
       "learning_rate": 3e-06,
-      "loss": 0.6905,
+      "loss": 0.472,
       "step": 30
     },
     {
       "epoch": 0.04,
-      "grad_norm": 2.208892345428467,
+      "grad_norm": 1.5840502977371216,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 0.6901,
+      "loss": 0.1679,
       "step": 40
     },
     {
       "epoch": 0.05,
-      "grad_norm": 4.314438819885254,
+      "grad_norm": 79.08979034423828,
       "learning_rate": 5e-06,
-      "loss": 0.6825,
+      "loss": 0.3838,
       "step": 50
     },
     {
       "epoch": 0.06,
-      "grad_norm": 3.1201136112213135,
+      "grad_norm": 3.6526267528533936,
       "learning_rate": 6e-06,
-      "loss": 0.7093,
+      "loss": 0.3122,
       "step": 60
     },
     {
       "epoch": 0.07,
-      "grad_norm": 1.3445158004760742,
+      "grad_norm": 88.9712142944336,
       "learning_rate": 7.000000000000001e-06,
-      "loss": 0.691,
+      "loss": 0.2529,
       "step": 70
     },
     {
       "epoch": 0.07,
-      "grad_norm": 2.153444766998291,
+      "grad_norm": 1.6937979459762573,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.7013,
+      "loss": 0.1963,
       "step": 80
     },
     {
       "epoch": 0.08,
-      "grad_norm": 1.514472246170044,
+      "grad_norm": 0.20718297362327576,
       "learning_rate": 9e-06,
-      "loss": 0.6843,
+      "loss": 0.3032,
       "step": 90
     },
     {
       "epoch": 0.09,
-      "grad_norm": 7.173806190490723,
+      "grad_norm": 0.26690974831581116,
       "learning_rate": 1e-05,
-      "loss": 0.6641,
+      "loss": 0.2491,
       "step": 100
     },
     {
       "epoch": 0.1,
-      "grad_norm": 11.494062423706055,
+      "grad_norm": 60.143089294433594,
       "learning_rate": 1.1000000000000001e-05,
-      "loss": 0.5903,
+      "loss": 0.2035,
       "step": 110
     },
     {
       "epoch": 0.11,
-      "grad_norm": 28.968042373657227,
+      "grad_norm": 81.73893737792969,
       "learning_rate": 1.2e-05,
-      "loss": 0.4994,
+      "loss": 0.2426,
       "step": 120
     },
     {
       "epoch": 0.12,
-      "grad_norm": 44.565860748291016,
+      "grad_norm": 0.23848050832748413,
       "learning_rate": 1.3000000000000001e-05,
-      "loss": 0.4509,
+      "loss": 0.1964,
       "step": 130
     },
     {
       "epoch": 0.13,
-      "grad_norm": 28.26981544494629,
+      "grad_norm": 150.04852294921875,
       "learning_rate": 1.4000000000000001e-05,
-      "loss": 0.4284,
+      "loss": 0.3048,
       "step": 140
     },
     {
       "epoch": 0.14,
-      "grad_norm": 8.198760032653809,
+      "grad_norm": 0.18502970039844513,
       "learning_rate": 1.5e-05,
-      "loss": 0.4876,
+      "loss": 0.0782,
       "step": 150
     },
     {
       "epoch": 0.15,
-      "grad_norm": 50.528987884521484,
+      "grad_norm": 4.491607666015625,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 0.4672,
+      "loss": 0.2463,
       "step": 160
     },
     {
       "epoch": 0.16,
-      "grad_norm": 16.226484298706055,
+      "grad_norm": 0.17159606516361237,
       "learning_rate": 1.7000000000000003e-05,
-      "loss": 0.6434,
+      "loss": 0.1454,
       "step": 170
     },
     {
       "epoch": 0.17,
-      "grad_norm": 14.364855766296387,
+      "grad_norm": 8.306988716125488,
       "learning_rate": 1.8e-05,
-      "loss": 0.5164,
+      "loss": 0.1481,
       "step": 180
     },
     {
       "epoch": 0.18,
-      "grad_norm": 7.223922252655029,
+      "grad_norm": 68.59893798828125,
       "learning_rate": 1.9e-05,
-      "loss": 0.4885,
+      "loss": 0.0853,
       "step": 190
     },
     {
       "epoch": 0.19,
-      "grad_norm": 8.672370910644531,
+      "grad_norm": 0.32962626218795776,
       "learning_rate": 2e-05,
-      "loss": 0.4994,
+      "loss": 0.2061,
       "step": 200
     },
     {
       "epoch": 0.2,
-      "grad_norm": 52.54035949707031,
+      "grad_norm": 0.44911274313926697,
       "learning_rate": 2.1e-05,
-      "loss": 0.462,
+      "loss": 0.2424,
       "step": 210
     },
     {
       "epoch": 0.21,
-      "grad_norm": 0.855399489402771,
+      "grad_norm": 0.1533774435520172,
       "learning_rate": 2.2000000000000003e-05,
-      "loss": 0.217,
+      "loss": 0.0086,
       "step": 220
     },
     {
       "epoch": 0.22,
-      "grad_norm": 66.37751007080078,
+      "grad_norm": 0.11695315688848495,
       "learning_rate": 2.3000000000000003e-05,
-      "loss": 0.6573,
+      "loss": 0.1169,
       "step": 230
     },
     {
       "epoch": 0.22,
-      "grad_norm": 6.453437805175781,
+      "grad_norm": 0.11531256884336472,
       "learning_rate": 2.4e-05,
-      "loss": 0.5052,
+      "loss": 0.0582,
       "step": 240
     },
     {
       "epoch": 0.23,
-      "grad_norm": 7.456045627593994,
+      "grad_norm": 0.08368314802646637,
       "learning_rate": 2.5e-05,
-      "loss": 0.3942,
+      "loss": 0.1063,
       "step": 250
     },
     {
       "epoch": 0.24,
-      "grad_norm": 16.87609100341797,
+      "grad_norm": 0.08130346238613129,
       "learning_rate": 2.6000000000000002e-05,
-      "loss": 0.4712,
+      "loss": 0.1673,
       "step": 260
     },
     {
       "epoch": 0.25,
-      "grad_norm": 22.84710693359375,
+      "grad_norm": 0.07953976094722748,
       "learning_rate": 2.7000000000000002e-05,
-      "loss": 0.4618,
+      "loss": 0.1073,
       "step": 270
     },
     {
       "epoch": 0.26,
-      "grad_norm": 1.8799303770065308,
+      "grad_norm": 0.08237360417842865,
       "learning_rate": 2.8000000000000003e-05,
-      "loss": 0.59,
+      "loss": 0.1379,
       "step": 280
     },
     {
       "epoch": 0.27,
-      "grad_norm": 80.71158599853516,
+      "grad_norm": 0.09733965992927551,
       "learning_rate": 2.9e-05,
-      "loss": 0.392,
+      "loss": 0.1136,
       "step": 290
     },
     {
       "epoch": 0.28,
-      "grad_norm": 2.51127290725708,
+      "grad_norm": 0.9005267024040222,
       "learning_rate": 3e-05,
-      "loss": 0.4768,
+      "loss": 0.2757,
       "step": 300
     },
     {
       "epoch": 0.29,
-      "grad_norm": 4.434652805328369,
+      "grad_norm": 216.63461303710938,
       "learning_rate": 3.1e-05,
-      "loss": 0.3732,
+      "loss": 0.1274,
       "step": 310
     },
     {
       "epoch": 0.3,
-      "grad_norm": 12.18094253540039,
+      "grad_norm": 0.08846700191497803,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 0.4689,
+      "loss": 0.1549,
       "step": 320
     },
     {
       "epoch": 0.31,
-      "grad_norm": 22.681018829345703,
+      "grad_norm": 0.10107693076133728,
       "learning_rate": 3.3e-05,
-      "loss": 0.4172,
+      "loss": 0.0722,
       "step": 330
     },
     {
       "epoch": 0.32,
-      "grad_norm": 12.466619491577148,
+      "grad_norm": 0.11054098606109619,
       "learning_rate": 3.4000000000000007e-05,
-      "loss": 0.4661,
+      "loss": 0.0776,
       "step": 340
     },
     {
       "epoch": 0.33,
-      "grad_norm": 7.371677398681641,
+      "grad_norm": 0.1684148609638214,
       "learning_rate": 3.5e-05,
-      "loss": 0.3976,
+      "loss": 0.1409,
       "step": 350
     },
     {
       "epoch": 0.34,
-      "grad_norm": 4.3304595947265625,
+      "grad_norm": 13.547722816467285,
       "learning_rate": 3.6e-05,
-      "loss": 0.6109,
+      "loss": 0.1295,
       "step": 360
     },
     {
       "epoch": 0.35,
-      "grad_norm": 165.4158172607422,
+      "grad_norm": 0.09100760519504547,
       "learning_rate": 3.7e-05,
-      "loss": 0.3678,
+      "loss": 0.1349,
       "step": 370
     },
     {
       "epoch": 0.36,
-      "grad_norm": 4.8646016120910645,
+      "grad_norm": 0.11295680701732635,
       "learning_rate": 3.8e-05,
-      "loss": 0.6374,
+      "loss": 0.2007,
       "step": 380
     },
     {
       "epoch": 0.37,
-      "grad_norm": 1.7399684190750122,
+      "grad_norm": 0.28185102343559265,
       "learning_rate": 3.9000000000000006e-05,
-      "loss": 0.4887,
+      "loss": 0.0049,
       "step": 390
     },
     {
       "epoch": 0.37,
-      "grad_norm": 16.256572723388672,
+      "grad_norm": 44.035606384277344,
       "learning_rate": 4e-05,
-      "loss": 0.5702,
+      "loss": 0.1354,
       "step": 400
     },
     {
       "epoch": 0.38,
-      "grad_norm": 45.110679626464844,
+      "grad_norm": 88.09428405761719,
       "learning_rate": 4.1e-05,
-      "loss": 0.5779,
+      "loss": 0.2693,
       "step": 410
     },
     {
       "epoch": 0.39,
-      "grad_norm": 10.500941276550293,
+      "grad_norm": 0.13857559859752655,
       "learning_rate": 4.2e-05,
-      "loss": 0.6049,
+      "loss": 0.309,
       "step": 420
     },
     {
       "epoch": 0.4,
-      "grad_norm": 46.1656494140625,
+      "grad_norm": 0.13726216554641724,
       "learning_rate": 4.3e-05,
-      "loss": 0.5598,
+      "loss": 0.2309,
       "step": 430
     },
     {
       "epoch": 0.41,
-      "grad_norm": 11.01528549194336,
+      "grad_norm": 0.35133376717567444,
       "learning_rate": 4.4000000000000006e-05,
-      "loss": 0.4441,
+      "loss": 0.3066,
       "step": 440
     },
     {
       "epoch": 0.42,
-      "grad_norm": 11.193294525146484,
+      "grad_norm": 2.9189860820770264,
       "learning_rate": 4.5e-05,
-      "loss": 0.5585,
+      "loss": 0.2049,
       "step": 450
     },
     {
       "epoch": 0.43,
-      "grad_norm": 18.2698917388916,
+      "grad_norm": 8.981446266174316,
       "learning_rate": 4.600000000000001e-05,
-      "loss": 0.4241,
+      "loss": 0.1587,
       "step": 460
     },
     {
       "epoch": 0.44,
-      "grad_norm": 166.50296020507812,
+      "grad_norm": 1.4933860301971436,
       "learning_rate": 4.7e-05,
-      "loss": 0.3354,
+      "loss": 0.1674,
       "step": 470
     },
     {
       "epoch": 0.45,
-      "grad_norm": 128.95260620117188,
+      "grad_norm": 0.10708652436733246,
       "learning_rate": 4.8e-05,
-      "loss": 0.5357,
+      "loss": 0.0669,
       "step": 480
     },
     {
       "epoch": 0.46,
-      "grad_norm": 5.128959655761719,
+      "grad_norm": 27.164520263671875,
       "learning_rate": 4.9e-05,
-      "loss": 0.4072,
+      "loss": 0.1222,
       "step": 490
     },
     {
       "epoch": 0.47,
-      "grad_norm": 3.268726348876953,
+      "grad_norm": 0.1280289888381958,
       "learning_rate": 5e-05,
-      "loss": 0.474,
+      "loss": 0.2507,
       "step": 500
     },
     {
       "epoch": 0.48,
-      "grad_norm": 37.61716842651367,
+      "grad_norm": 0.14575959742069244,
       "learning_rate": 4.9896587383660806e-05,
-      "loss": 0.6837,
+      "loss": 0.2695,
       "step": 510
     },
     {
       "epoch": 0.49,
-      "grad_norm": 5.7842302322387695,
+      "grad_norm": 1.213277816772461,
       "learning_rate": 4.9793174767321616e-05,
-      "loss": 0.4518,
+      "loss": 0.1962,
       "step": 520
     },
     {
       "epoch": 0.5,
-      "grad_norm": 17.824024200439453,
+      "grad_norm": 110.40786743164062,
       "learning_rate": 4.968976215098242e-05,
-      "loss": 0.5325,
+      "loss": 0.1626,
       "step": 530
     },
     {
       "epoch": 0.51,
-      "grad_norm": 13.727099418640137,
+      "grad_norm": 3.0125222206115723,
       "learning_rate": 4.958634953464323e-05,
-      "loss": 0.4236,
+      "loss": 0.3283,
       "step": 540
     },
     {
       "epoch": 0.52,
-      "grad_norm": 11.013121604919434,
+      "grad_norm": 2.7797441482543945,
       "learning_rate": 4.948293691830403e-05,
-      "loss": 0.3925,
+      "loss": 0.259,
       "step": 550
     },
     {
       "epoch": 0.52,
-      "grad_norm": 10.758121490478516,
+      "grad_norm": 0.21556439995765686,
       "learning_rate": 4.937952430196484e-05,
-      "loss": 0.4354,
+      "loss": 0.1826,
       "step": 560
     },
     {
       "epoch": 0.53,
-      "grad_norm": 63.49281692504883,
+      "grad_norm": 4.356193542480469,
       "learning_rate": 4.9276111685625646e-05,
-      "loss": 0.456,
+      "loss": 0.1792,
       "step": 570
     },
     {
       "epoch": 0.54,
-      "grad_norm": 14.087475776672363,
+      "grad_norm": 10.7008056640625,
       "learning_rate": 4.9172699069286456e-05,
-      "loss": 0.514,
+      "loss": 0.6648,
       "step": 580
     },
     {
       "epoch": 0.55,
-      "grad_norm": 2.470353841781616,
+      "grad_norm": 5.572720050811768,
       "learning_rate": 4.906928645294726e-05,
-      "loss": 0.5102,
+      "loss": 0.5023,
       "step": 590
     },
     {
       "epoch": 0.56,
-      "grad_norm": 43.60289001464844,
+      "grad_norm": 41.6730842590332,
       "learning_rate": 4.896587383660807e-05,
-      "loss": 0.4577,
+      "loss": 0.2741,
       "step": 600
     },
     {
       "epoch": 0.57,
-      "grad_norm": 13.848435401916504,
+      "grad_norm": 39.446434020996094,
       "learning_rate": 4.886246122026887e-05,
-      "loss": 0.3769,
+      "loss": 0.1501,
       "step": 610
     },
     {
       "epoch": 0.58,
-      "grad_norm": 85.09272003173828,
+      "grad_norm": 84.94667053222656,
       "learning_rate": 4.8759048603929683e-05,
-      "loss": 0.4745,
+      "loss": 0.2907,
       "step": 620
     },
     {
       "epoch": 0.59,
-      "grad_norm": 2.0452070236206055,
+      "grad_norm": 1.8101928234100342,
       "learning_rate": 4.865563598759049e-05,
-      "loss": 0.3618,
+      "loss": 0.2514,
       "step": 630
     },
     {
       "epoch": 0.6,
-      "grad_norm": 5.641032695770264,
+      "grad_norm": 102.29499053955078,
       "learning_rate": 4.855222337125129e-05,
-      "loss": 0.5703,
+      "loss": 0.5467,
       "step": 640
     },
     {
       "epoch": 0.61,
-      "grad_norm": 19.266847610473633,
+      "grad_norm": 2.9631879329681396,
       "learning_rate": 4.84488107549121e-05,
-      "loss": 0.5826,
+      "loss": 0.2997,
       "step": 650
     },
     {
       "epoch": 0.62,
-      "grad_norm": 73.6524658203125,
+      "grad_norm": 0.1405537873506546,
       "learning_rate": 4.8345398138572904e-05,
-      "loss": 0.5455,
+      "loss": 0.2319,
       "step": 660
     },
     {
       "epoch": 0.63,
-      "grad_norm": 1.602294683456421,
+      "grad_norm": 0.13494689762592316,
       "learning_rate": 4.8241985522233714e-05,
-      "loss": 0.5497,
+      "loss": 0.4836,
       "step": 670
     },
     {
       "epoch": 0.64,
-      "grad_norm": 22.117345809936523,
+      "grad_norm": 3.714601755142212,
       "learning_rate": 4.813857290589452e-05,
-      "loss": 0.4378,
+      "loss": 0.3312,
       "step": 680
     },
     {
       "epoch": 0.65,
-      "grad_norm": 110.3412094116211,
+      "grad_norm": 0.274013876914978,
       "learning_rate": 4.803516028955533e-05,
-      "loss": 0.4393,
+      "loss": 0.1327,
       "step": 690
     },
     {
       "epoch": 0.66,
-      "grad_norm": 3.794416904449463,
+      "grad_norm": 0.6073850989341736,
       "learning_rate": 4.793174767321613e-05,
-      "loss": 0.5334,
+      "loss": 0.1932,
       "step": 700
     },
     {
       "epoch": 0.67,
-      "grad_norm": 21.734609603881836,
+      "grad_norm": 0.08534292876720428,
       "learning_rate": 4.782833505687694e-05,
-      "loss": 0.5833,
+      "loss": 0.0918,
       "step": 710
     },
     {
       "epoch": 0.67,
-      "grad_norm": 15.466120719909668,
+      "grad_norm": 0.1106233149766922,
       "learning_rate": 4.772492244053775e-05,
-      "loss": 0.6138,
+      "loss": 0.0718,
       "step": 720
     },
     {
       "epoch": 0.68,
-      "grad_norm": 15.851093292236328,
+      "grad_norm": 0.14169247448444366,
       "learning_rate": 4.7621509824198554e-05,
-      "loss": 0.474,
+      "loss": 0.2155,
       "step": 730
     },
     {
       "epoch": 0.69,
-      "grad_norm": 11.810195922851562,
+      "grad_norm": 5.554646015167236,
       "learning_rate": 4.7518097207859365e-05,
-      "loss": 0.5048,
+      "loss": 0.2144,
       "step": 740
     },
     {
       "epoch": 0.7,
-      "grad_norm": 7.658618450164795,
+      "grad_norm": 0.28406473994255066,
       "learning_rate": 4.741468459152017e-05,
-      "loss": 0.3721,
+      "loss": 0.2004,
       "step": 750
     },
     {
       "epoch": 0.71,
-      "grad_norm": 5.516314506530762,
+      "grad_norm": 0.2573547959327698,
       "learning_rate": 4.731127197518098e-05,
-      "loss": 0.4057,
+      "loss": 0.1411,
       "step": 760
     },
     {
       "epoch": 0.72,
-      "grad_norm": 29.502197265625,
+      "grad_norm": 0.1661107838153839,
       "learning_rate": 4.720785935884178e-05,
-      "loss": 0.4228,
+      "loss": 0.1108,
       "step": 770
     },
     {
       "epoch": 0.73,
-      "grad_norm": 5.364101409912109,
+      "grad_norm": 0.1617022007703781,
       "learning_rate": 4.710444674250259e-05,
-      "loss": 0.6065,
+      "loss": 0.1855,
       "step": 780
     },
     {
       "epoch": 0.74,
-      "grad_norm": 3.6527419090270996,
+      "grad_norm": 0.2379370480775833,
       "learning_rate": 4.7001034126163395e-05,
-      "loss": 0.4702,
+      "loss": 0.145,
       "step": 790
     },
     {
       "epoch": 0.75,
-      "grad_norm": 234.80238342285156,
+      "grad_norm": 0.22972096502780914,
       "learning_rate": 4.6897621509824205e-05,
-      "loss": 0.4136,
+      "loss": 0.1885,
       "step": 800
     },
     {
       "epoch": 0.76,
-      "grad_norm": 3.4541990756988525,
+      "grad_norm": 0.21119792759418488,
       "learning_rate": 4.679420889348501e-05,
-      "loss": 0.4633,
+      "loss": 0.1153,
       "step": 810
     },
     {
       "epoch": 0.77,
-      "grad_norm": 10.891709327697754,
+      "grad_norm": 0.5091124773025513,
       "learning_rate": 4.669079627714581e-05,
-      "loss": 0.4773,
+      "loss": 0.2018,
       "step": 820
     },
     {
       "epoch": 0.78,
-      "grad_norm": 2.062467575073242,
+      "grad_norm": 0.09518251568078995,
       "learning_rate": 4.658738366080662e-05,
-      "loss": 0.4229,
+      "loss": 0.0668,
       "step": 830
     },
     {
       "epoch": 0.79,
-      "grad_norm": 1.5813032388687134,
+      "grad_norm": 0.08498405665159225,
       "learning_rate": 4.6483971044467425e-05,
-      "loss": 0.4163,
+      "loss": 0.1655,
       "step": 840
     },
     {
       "epoch": 0.8,
-      "grad_norm": 2.950200080871582,
+      "grad_norm": 0.11825589835643768,
       "learning_rate": 4.6380558428128236e-05,
-      "loss": 0.4312,
+      "loss": 0.3369,
       "step": 850
     },
     {
       "epoch": 0.81,
-      "grad_norm": 15.027694702148438,
+      "grad_norm": 0.19106203317642212,
       "learning_rate": 4.627714581178904e-05,
-      "loss": 0.4475,
+      "loss": 0.3149,
       "step": 860
     },
     {
       "epoch": 0.82,
-      "grad_norm": 6.54260778427124,
+      "grad_norm": 0.15981145203113556,
       "learning_rate": 4.617373319544985e-05,
-      "loss": 0.4871,
+      "loss": 0.1569,
       "step": 870
     },
     {
       "epoch": 0.82,
-      "grad_norm": 40.894622802734375,
+      "grad_norm": 1.3525559902191162,
       "learning_rate": 4.607032057911065e-05,
-      "loss": 0.7644,
+      "loss": 0.0151,
       "step": 880
     },
     {
       "epoch": 0.83,
-      "grad_norm": 40.53754425048828,
+      "grad_norm": 0.14920440316200256,
       "learning_rate": 4.596690796277146e-05,
-      "loss": 0.6225,
+      "loss": 0.3345,
       "step": 890
     },
     {
       "epoch": 0.84,
-      "grad_norm": 2.118955373764038,
+      "grad_norm": 0.42381948232650757,
       "learning_rate": 4.5863495346432266e-05,
-      "loss": 0.38,
+      "loss": 0.3653,
       "step": 900
     },
     {
       "epoch": 0.85,
-      "grad_norm": 4.862216472625732,
+      "grad_norm": 2.133568048477173,
       "learning_rate": 4.5760082730093076e-05,
-      "loss": 0.4339,
+      "loss": 0.2117,
       "step": 910
     },
     {
       "epoch": 0.86,
-      "grad_norm": 32.316802978515625,
+      "grad_norm": 0.3465505540370941,
       "learning_rate": 4.565667011375388e-05,
-      "loss": 0.6216,
+      "loss": 0.1258,
       "step": 920
     },
     {
       "epoch": 0.87,
-      "grad_norm": 8.157034873962402,
+      "grad_norm": 6.085052967071533,
       "learning_rate": 4.555325749741469e-05,
-      "loss": 0.3952,
+      "loss": 0.0828,
       "step": 930
     },
     {
       "epoch": 0.88,
-      "grad_norm": 34.675384521484375,
+      "grad_norm": 43.05193328857422,
       "learning_rate": 4.544984488107549e-05,
-      "loss": 0.4752,
+      "loss": 0.1615,
       "step": 940
     },
     {
       "epoch": 0.89,
-      "grad_norm": 1.4986870288848877,
+      "grad_norm": 0.3297998607158661,
       "learning_rate": 4.5346432264736296e-05,
-      "loss": 0.5754,
+      "loss": 0.2717,
       "step": 950
     },
     {
       "epoch": 0.9,
-      "grad_norm": 5.362887859344482,
+      "grad_norm": 0.6388803720474243,
       "learning_rate": 4.5243019648397106e-05,
-      "loss": 0.4518,
+      "loss": 0.3202,
       "step": 960
     },
     {
       "epoch": 0.91,
-      "grad_norm": 8.203104972839355,
+      "grad_norm": 0.17811286449432373,
       "learning_rate": 4.513960703205791e-05,
-      "loss": 0.4244,
+      "loss": 0.0243,
       "step": 970
     },
     {
       "epoch": 0.92,
-      "grad_norm": 146.74620056152344,
+      "grad_norm": 0.07790106534957886,
       "learning_rate": 4.503619441571872e-05,
-      "loss": 0.4484,
+      "loss": 0.1374,
       "step": 980
     },
     {
       "epoch": 0.93,
-      "grad_norm": 38.64397048950195,
+      "grad_norm": 4.508092403411865,
       "learning_rate": 4.493278179937952e-05,
-      "loss": 0.5375,
+      "loss": 0.458,
       "step": 990
     },
     {
       "epoch": 0.94,
-      "grad_norm": 9.168529510498047,
+      "grad_norm": 0.2888680100440979,
       "learning_rate": 4.4829369183040333e-05,
-      "loss": 0.4305,
+      "loss": 0.315,
       "step": 1000
     },
     {
       "epoch": 0.95,
-      "grad_norm": 11.166984558105469,
+      "grad_norm": 0.36161041259765625,
       "learning_rate": 4.472595656670114e-05,
-      "loss": 0.5018,
+      "loss": 0.0882,
       "step": 1010
     },
     {
       "epoch": 0.96,
-      "grad_norm": 2.078636646270752,
+      "grad_norm": 0.2502145767211914,
       "learning_rate": 4.462254395036195e-05,
-      "loss": 0.5365,
+      "loss": 0.063,
       "step": 1020
     },
     {
       "epoch": 0.97,
-      "grad_norm": 2.5594077110290527,
+      "grad_norm": 0.17374250292778015,
       "learning_rate": 4.451913133402275e-05,
-      "loss": 0.2867,
+      "loss": 0.1916,
       "step": 1030
     },
     {
       "epoch": 0.97,
-      "grad_norm": 40.97541046142578,
+      "grad_norm": 2.548640251159668,
       "learning_rate": 4.441571871768356e-05,
-      "loss": 0.5615,
+      "loss": 0.3626,
       "step": 1040
     },
     {
       "epoch": 0.98,
-      "grad_norm": 5.333548545837402,
+      "grad_norm": 0.27965047955513,
       "learning_rate": 4.4312306101344364e-05,
-      "loss": 0.371,
+      "loss": 0.2225,
       "step": 1050
     },
     {
       "epoch": 0.99,
-      "grad_norm": 29.254602432250977,
+      "grad_norm": 4.920745372772217,
       "learning_rate": 4.420889348500517e-05,
-      "loss": 0.7758,
+      "loss": 0.5524,
       "step": 1060
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8180112570356473,
-      "eval_f1": 0.831304347826087,
-      "eval_loss": 0.4565158784389496,
-      "eval_precision": 0.7734627831715211,
-      "eval_recall": 0.8984962406015038,
-      "eval_runtime": 1.5762,
-      "eval_samples_per_second": 338.164,
-      "eval_steps_per_second": 42.508,
+      "eval_accuracy": 0.7936210131332082,
+      "eval_f1": 0.8166666666666667,
+      "eval_loss": 0.6987214088439941,
+      "eval_precision": 0.7335329341317365,
+      "eval_recall": 0.9210526315789473,
+      "eval_runtime": 1.4067,
+      "eval_samples_per_second": 378.902,
+      "eval_steps_per_second": 47.629,
       "step": 1067
     },
     {
       "epoch": 1.0,
-      "grad_norm": 4.774388313293457,
+      "grad_norm": 5.866567611694336,
       "learning_rate": 4.410548086866598e-05,
-      "loss": 0.4325,
+      "loss": 0.3374,
       "step": 1070
     },
     {
       "epoch": 1.01,
-      "grad_norm": 25.7803897857666,
+      "grad_norm": 0.8940852284431458,
       "learning_rate": 4.400206825232678e-05,
-      "loss": 0.3303,
+      "loss": 0.3523,
       "step": 1080
     },
     {
       "epoch": 1.02,
-      "grad_norm": 4.480411529541016,
+      "grad_norm": 0.5038263201713562,
       "learning_rate": 4.389865563598759e-05,
-      "loss": 0.425,
+      "loss": 0.2949,
       "step": 1090
     },
     {
       "epoch": 1.03,
-      "grad_norm": 2.4002604484558105,
+      "grad_norm": 4.7829179763793945,
       "learning_rate": 4.3795243019648394e-05,
-      "loss": 0.386,
+      "loss": 0.246,
       "step": 1100
     },
     {
       "epoch": 1.04,
-      "grad_norm": 61.24464416503906,
+      "grad_norm": 2.0902047157287598,
       "learning_rate": 4.369183040330921e-05,
-      "loss": 0.277,
+      "loss": 0.3073,
       "step": 1110
     },
     {
       "epoch": 1.05,
-      "grad_norm": 25.068614959716797,
+      "grad_norm": 1.7576603889465332,
       "learning_rate": 4.3588417786970015e-05,
-      "loss": 0.354,
+      "loss": 0.4624,
       "step": 1120
     },
     {
       "epoch": 1.06,
-      "grad_norm": 1.0058590173721313,
+      "grad_norm": 1.3686978816986084,
       "learning_rate": 4.3485005170630825e-05,
-      "loss": 0.3038,
+      "loss": 0.2415,
       "step": 1130
     },
     {
       "epoch": 1.07,
-      "grad_norm": 408.1586608886719,
+      "grad_norm": 0.6715884804725647,
       "learning_rate": 4.338159255429163e-05,
-      "loss": 0.5157,
+      "loss": 0.2274,
       "step": 1140
     },
     {
       "epoch": 1.08,
-      "grad_norm": 6.401839256286621,
+      "grad_norm": 2.798624038696289,
       "learning_rate": 4.327817993795243e-05,
-      "loss": 0.4993,
+      "loss": 0.2798,
       "step": 1150
     },
     {
       "epoch": 1.09,
-      "grad_norm": 6.1133317947387695,
+      "grad_norm": 0.3907501697540283,
       "learning_rate": 4.317476732161324e-05,
-      "loss": 0.4571,
+      "loss": 0.1533,
       "step": 1160
     },
     {
       "epoch": 1.1,
-      "grad_norm": 2.3999035358428955,
+      "grad_norm": 0.4938846230506897,
       "learning_rate": 4.3071354705274045e-05,
-      "loss": 0.4561,
+      "loss": 0.284,
       "step": 1170
     },
     {
       "epoch": 1.11,
-      "grad_norm": 12.716032028198242,
+      "grad_norm": 1.830908179283142,
       "learning_rate": 4.2967942088934855e-05,
-      "loss": 0.2751,
+      "loss": 0.1958,
       "step": 1180
     },
     {
       "epoch": 1.12,
-      "grad_norm": 36.668941497802734,
+      "grad_norm": 0.467227965593338,
       "learning_rate": 4.286452947259566e-05,
-      "loss": 0.328,
+      "loss": 0.1826,
       "step": 1190
     },
     {
       "epoch": 1.12,
-      "grad_norm": 8.840494155883789,
+      "grad_norm": 1.6289818286895752,
       "learning_rate": 4.276111685625647e-05,
-      "loss": 0.4708,
+      "loss": 0.2573,
       "step": 1200
     },
     {
       "epoch": 1.13,
-      "grad_norm": 2.9924654960632324,
+      "grad_norm": 0.42744210362434387,
       "learning_rate": 4.265770423991727e-05,
-      "loss": 0.2318,
+      "loss": 0.0986,
       "step": 1210
     },
     {
       "epoch": 1.14,
-      "grad_norm": 0.4895036816596985,
+      "grad_norm": 0.48458239436149597,
       "learning_rate": 4.255429162357808e-05,
-      "loss": 0.3037,
+      "loss": 0.2624,
       "step": 1220
     },
     {
       "epoch": 1.15,
-      "grad_norm": 44.88699722290039,
+      "grad_norm": 0.6485515832901001,
       "learning_rate": 4.2450879007238886e-05,
-      "loss": 0.3585,
+      "loss": 0.317,
       "step": 1230
     },
     {
       "epoch": 1.16,
-      "grad_norm": 5.701717376708984,
+      "grad_norm": 3.006849527359009,
       "learning_rate": 4.2347466390899696e-05,
-      "loss": 0.5978,
+      "loss": 0.2392,
       "step": 1240
     },
     {
       "epoch": 1.17,
-      "grad_norm": 1.0451136827468872,
+      "grad_norm": 1.1750074625015259,
       "learning_rate": 4.22440537745605e-05,
-      "loss": 0.8103,
+      "loss": 0.2915,
       "step": 1250
     },
     {
       "epoch": 1.18,
-      "grad_norm": 2.085301399230957,
+      "grad_norm": 0.35948407649993896,
       "learning_rate": 4.21406411582213e-05,
-      "loss": 0.3038,
+      "loss": 0.1648,
       "step": 1260
     },
     {
       "epoch": 1.19,
-      "grad_norm": 66.31008911132812,
+      "grad_norm": 0.35395777225494385,
       "learning_rate": 4.203722854188211e-05,
-      "loss": 0.5508,
+      "loss": 0.1433,
       "step": 1270
     },
     {
       "epoch": 1.2,
-      "grad_norm": 19.349102020263672,
+      "grad_norm": 0.526642382144928,
       "learning_rate": 4.1933815925542916e-05,
-      "loss": 0.2065,
+      "loss": 0.1706,
       "step": 1280
     },
     {
       "epoch": 1.21,
-      "grad_norm": 2.037997007369995,
+      "grad_norm": 2.7297651767730713,
       "learning_rate": 4.1830403309203726e-05,
-      "loss": 0.4177,
+      "loss": 0.2143,
       "step": 1290
     },
     {
       "epoch": 1.22,
-      "grad_norm": 0.6965357065200806,
+      "grad_norm": 0.2764374911785126,
       "learning_rate": 4.172699069286453e-05,
-      "loss": 0.3867,
+      "loss": 0.2931,
       "step": 1300
     },
     {
       "epoch": 1.23,
-      "grad_norm": 0.722122848033905,
+      "grad_norm": 2.820064067840576,
       "learning_rate": 4.162357807652534e-05,
-      "loss": 0.4882,
+      "loss": 0.3882,
       "step": 1310
     },
     {
       "epoch": 1.24,
-      "grad_norm": 9.377044677734375,
+      "grad_norm": 3.801469326019287,
       "learning_rate": 4.152016546018614e-05,
-      "loss": 0.4661,
+      "loss": 0.2653,
       "step": 1320
     },
     {
       "epoch": 1.25,
-      "grad_norm": 0.40828177332878113,
+      "grad_norm": 0.3793887495994568,
       "learning_rate": 4.141675284384695e-05,
-      "loss": 0.4306,
+      "loss": 0.2912,
       "step": 1330
     },
     {
       "epoch": 1.26,
-      "grad_norm": 4.25218391418457,
+      "grad_norm": 1.7214317321777344,
       "learning_rate": 4.1313340227507756e-05,
-      "loss": 0.4858,
+      "loss": 0.3062,
       "step": 1340
     },
     {
       "epoch": 1.27,
-      "grad_norm": 3.3388795852661133,
+      "grad_norm": 0.3773093819618225,
       "learning_rate": 4.1209927611168567e-05,
-      "loss": 0.5039,
+      "loss": 0.1882,
       "step": 1350
     },
     {
       "epoch": 1.27,
-      "grad_norm": 6.391818523406982,
+      "grad_norm": 4.08633279800415,
       "learning_rate": 4.110651499482937e-05,
-      "loss": 0.4893,
+      "loss": 0.3492,
       "step": 1360
     },
     {
       "epoch": 1.28,
-      "grad_norm": 27.01968765258789,
+      "grad_norm": 0.6732409596443176,
       "learning_rate": 4.100310237849018e-05,
-      "loss": 0.4502,
+      "loss": 0.2416,
       "step": 1370
     },
     {
       "epoch": 1.29,
-      "grad_norm": 29.210132598876953,
+      "grad_norm": 2.674741268157959,
       "learning_rate": 4.0899689762150983e-05,
-      "loss": 0.3657,
+      "loss": 0.2195,
       "step": 1380
     },
     {
       "epoch": 1.3,
-      "grad_norm": 54.26066970825195,
+      "grad_norm": 2.5985941886901855,
       "learning_rate": 4.079627714581179e-05,
-      "loss": 0.2958,
+      "loss": 0.2772,
       "step": 1390
     },
     {
       "epoch": 1.31,
-      "grad_norm": 36.26032257080078,
+      "grad_norm": 2.2892799377441406,
       "learning_rate": 4.06928645294726e-05,
-      "loss": 0.2768,
+      "loss": 0.3021,
       "step": 1400
     },
     {
       "epoch": 1.32,
-      "grad_norm": 1.0523189306259155,
+      "grad_norm": 1.6137515306472778,
       "learning_rate": 4.05894519131334e-05,
-      "loss": 0.3634,
+      "loss": 0.4407,
       "step": 1410
     },
     {
       "epoch": 1.33,
-      "grad_norm": 191.5191192626953,
+      "grad_norm": 1.6823959350585938,
       "learning_rate": 4.048603929679421e-05,
-      "loss": 0.3442,
+      "loss": 0.1717,
       "step": 1420
     },
     {
       "epoch": 1.34,
-      "grad_norm": 0.6843079924583435,
+      "grad_norm": 0.7000110745429993,
       "learning_rate": 4.0382626680455014e-05,
-      "loss": 0.4025,
+      "loss": 0.4317,
       "step": 1430
     },
     {
       "epoch": 1.35,
-      "grad_norm": 0.2217019647359848,
+      "grad_norm": 14.843025207519531,
       "learning_rate": 4.0279214064115824e-05,
-      "loss": 0.2445,
+      "loss": 0.5217,
       "step": 1440
     },
     {
       "epoch": 1.36,
-      "grad_norm": 7.919774055480957,
+      "grad_norm": 1.8638509511947632,
       "learning_rate": 4.017580144777663e-05,
-      "loss": 0.548,
+      "loss": 0.6968,
       "step": 1450
     },
     {
       "epoch": 1.37,
-      "grad_norm": 1.322029709815979,
+      "grad_norm": 4.495730876922607,
       "learning_rate": 4.007238883143744e-05,
-      "loss": 0.3518,
+      "loss": 0.3835,
       "step": 1460
     },
     {
       "epoch": 1.38,
-      "grad_norm": 0.4105088412761688,
+      "grad_norm": 80.59029388427734,
       "learning_rate": 3.996897621509824e-05,
-      "loss": 0.3859,
+      "loss": 0.3564,
       "step": 1470
     },
     {
       "epoch": 1.39,
-      "grad_norm": 0.31217288970947266,
+      "grad_norm": 0.4503944218158722,
       "learning_rate": 3.986556359875905e-05,
-      "loss": 0.3287,
+      "loss": 0.2868,
       "step": 1480
     },
     {
       "epoch": 1.4,
-      "grad_norm": 39.67789077758789,
+      "grad_norm": 1.2642120122909546,
       "learning_rate": 3.9762150982419854e-05,
-      "loss": 0.4649,
+      "loss": 0.272,
       "step": 1490
     },
     {
       "epoch": 1.41,
-      "grad_norm": 17.59617042541504,
+      "grad_norm": 5.5861005783081055,
       "learning_rate": 3.965873836608066e-05,
-      "loss": 0.4667,
+      "loss": 0.3955,
       "step": 1500
     },
     {
       "epoch": 1.42,
-      "grad_norm": 14.470635414123535,
+      "grad_norm": 64.48252868652344,
       "learning_rate": 3.9555325749741475e-05,
-      "loss": 0.4061,
+      "loss": 0.3061,
       "step": 1510
     },
     {
       "epoch": 1.42,
-      "grad_norm": 62.437618255615234,
+      "grad_norm": 8.142120361328125,
       "learning_rate": 3.945191313340228e-05,
-      "loss": 0.466,
+      "loss": 0.2618,
       "step": 1520
     },
     {
       "epoch": 1.43,
-      "grad_norm": 23.588979721069336,
+      "grad_norm": 3.7314465045928955,
       "learning_rate": 3.934850051706309e-05,
-      "loss": 0.3333,
+      "loss": 0.2106,
       "step": 1530
     },
     {
       "epoch": 1.44,
-      "grad_norm": 3.491382598876953,
+      "grad_norm": 1.7706559896469116,
       "learning_rate": 3.924508790072389e-05,
-      "loss": 0.2536,
+      "loss": 0.4735,
       "step": 1540
     },
     {
       "epoch": 1.45,
-      "grad_norm": 0.3138933479785919,
+      "grad_norm": 0.5831993222236633,
       "learning_rate": 3.91416752843847e-05,
-      "loss": 0.3297,
+      "loss": 0.474,
       "step": 1550
     },
     {
       "epoch": 1.46,
-      "grad_norm": 4.9050211906433105,
+      "grad_norm": 1.258558750152588,
       "learning_rate": 3.9038262668045505e-05,
-      "loss": 0.3643,
+      "loss": 0.4618,
       "step": 1560
     },
     {
       "epoch": 1.47,
-      "grad_norm": 13.311896324157715,
+      "grad_norm": 1.3511230945587158,
       "learning_rate": 3.8934850051706315e-05,
-      "loss": 0.2224,
+      "loss": 0.3205,
       "step": 1570
     },
     {
       "epoch": 1.48,
-      "grad_norm": 136.0790252685547,
+      "grad_norm": 1.4303970336914062,
       "learning_rate": 3.883143743536712e-05,
-      "loss": 0.5125,
+      "loss": 0.2386,
       "step": 1580
     },
     {
       "epoch": 1.49,
-      "grad_norm": 1.9536207914352417,
+      "grad_norm": 2.5400567054748535,
       "learning_rate": 3.872802481902792e-05,
-      "loss": 0.3984,
+      "loss": 0.3399,
       "step": 1590
     },
     {
       "epoch": 1.5,
-      "grad_norm": 15.781469345092773,
+      "grad_norm": 1.0604122877120972,
       "learning_rate": 3.862461220268873e-05,
-      "loss": 0.4071,
+      "loss": 0.2023,
       "step": 1600
     },
     {
       "epoch": 1.51,
-      "grad_norm": 4.010928153991699,
+      "grad_norm": 5.520498275756836,
       "learning_rate": 3.8521199586349535e-05,
-      "loss": 0.3088,
+      "loss": 0.2842,
       "step": 1610
     },
     {
       "epoch": 1.52,
-      "grad_norm": 15.730233192443848,
+      "grad_norm": 0.4855865240097046,
       "learning_rate": 3.8417786970010346e-05,
-      "loss": 0.4905,
+      "loss": 0.2712,
       "step": 1620
     },
     {
       "epoch": 1.53,
-      "grad_norm": 18.30491065979004,
+      "grad_norm": 0.35705462098121643,
       "learning_rate": 3.831437435367115e-05,
-      "loss": 0.366,
+      "loss": 0.2409,
       "step": 1630
     },
     {
       "epoch": 1.54,
-      "grad_norm": 0.39613863825798035,
+      "grad_norm": 0.26277977228164673,
       "learning_rate": 3.821096173733196e-05,
-      "loss": 0.3205,
+      "loss": 0.2584,
       "step": 1640
     },
     {
       "epoch": 1.55,
-      "grad_norm": 51.20091247558594,
+      "grad_norm": 3.6166861057281494,
       "learning_rate": 3.810754912099276e-05,
-      "loss": 0.5628,
+      "loss": 0.4416,
       "step": 1650
     },
     {
       "epoch": 1.56,
-      "grad_norm": 17.219799041748047,
+      "grad_norm": 2.64589524269104,
       "learning_rate": 3.800413650465357e-05,
-      "loss": 0.6222,
+      "loss": 0.3003,
       "step": 1660
     },
     {
       "epoch": 1.57,
-      "grad_norm": 41.373634338378906,
+      "grad_norm": 62.66883087158203,
       "learning_rate": 3.7900723888314376e-05,
-      "loss": 0.293,
+      "loss": 0.2302,
       "step": 1670
     },
     {
       "epoch": 1.57,
-      "grad_norm": 77.48905181884766,
+      "grad_norm": 1.1689012050628662,
       "learning_rate": 3.7797311271975186e-05,
-      "loss": 0.3165,
+      "loss": 0.5299,
       "step": 1680
     },
     {
       "epoch": 1.58,
-      "grad_norm": 94.227783203125,
+      "grad_norm": 0.7352681756019592,
       "learning_rate": 3.769389865563599e-05,
-      "loss": 0.455,
+      "loss": 0.1651,
       "step": 1690
     },
     {
       "epoch": 1.59,
-      "grad_norm": 21.350830078125,
+      "grad_norm": 0.66200852394104,
       "learning_rate": 3.759048603929679e-05,
-      "loss": 0.4369,
+      "loss": 0.2608,
       "step": 1700
     },
     {
       "epoch": 1.6,
-      "grad_norm": 17.99802589416504,
+      "grad_norm": 13.166357040405273,
       "learning_rate": 3.74870734229576e-05,
-      "loss": 0.4513,
+      "loss": 0.3542,
       "step": 1710
     },
     {
       "epoch": 1.61,
-      "grad_norm": 2.999746322631836,
+      "grad_norm": 1.8984485864639282,
       "learning_rate": 3.7383660806618406e-05,
-      "loss": 0.3328,
+      "loss": 0.3457,
       "step": 1720
     },
     {
       "epoch": 1.62,
-      "grad_norm": 1.245676040649414,
+      "grad_norm": 1.0245816707611084,
       "learning_rate": 3.7280248190279217e-05,
-      "loss": 0.5112,
+      "loss": 0.3729,
       "step": 1730
     },
     {
       "epoch": 1.63,
-      "grad_norm": 0.8955879211425781,
+      "grad_norm": 2.527099370956421,
       "learning_rate": 3.717683557394002e-05,
-      "loss": 0.2696,
+      "loss": 0.2096,
       "step": 1740
     },
     {
       "epoch": 1.64,
-      "grad_norm": 14.041428565979004,
+      "grad_norm": 0.7097912430763245,
       "learning_rate": 3.707342295760083e-05,
-      "loss": 0.3887,
+      "loss": 0.241,
       "step": 1750
     },
     {
       "epoch": 1.65,
-      "grad_norm": 10.70749568939209,
+      "grad_norm": 0.5669069886207581,
       "learning_rate": 3.6970010341261633e-05,
-      "loss": 0.5087,
+      "loss": 0.2875,
       "step": 1760
     },
     {
       "epoch": 1.66,
-      "grad_norm": 26.841726303100586,
+      "grad_norm": 2.8181681632995605,
       "learning_rate": 3.6866597724922444e-05,
-      "loss": 0.5083,
+      "loss": 0.3424,
       "step": 1770
     },
     {
       "epoch": 1.67,
-      "grad_norm": 9.196104049682617,
+      "grad_norm": 2.679267406463623,
       "learning_rate": 3.676318510858325e-05,
-      "loss": 0.3208,
+      "loss": 0.2514,
       "step": 1780
     },
     {
       "epoch": 1.68,
-      "grad_norm": 15.969416618347168,
+      "grad_norm": 5.605787754058838,
       "learning_rate": 3.665977249224406e-05,
-      "loss": 0.3257,
+      "loss": 0.4296,
       "step": 1790
     },
     {
       "epoch": 1.69,
-      "grad_norm": 27.856632232666016,
+      "grad_norm": 1.4018586874008179,
       "learning_rate": 3.655635987590486e-05,
-      "loss": 0.3663,
+      "loss": 0.2786,
       "step": 1800
     },
     {
       "epoch": 1.7,
-      "grad_norm": 1.3687829971313477,
+      "grad_norm": 0.8706427216529846,
       "learning_rate": 3.645294725956567e-05,
-      "loss": 0.3494,
+      "loss": 0.2925,
       "step": 1810
     },
     {
       "epoch": 1.71,
-      "grad_norm": 24.28046989440918,
+      "grad_norm": 0.6901386976242065,
       "learning_rate": 3.6349534643226474e-05,
-      "loss": 0.4308,
+      "loss": 0.1585,
       "step": 1820
     },
     {
       "epoch": 1.72,
-      "grad_norm": 0.7270779013633728,
+      "grad_norm": 2.785902500152588,
       "learning_rate": 3.624612202688728e-05,
-      "loss": 0.3686,
+      "loss": 0.2355,
       "step": 1830
     },
     {
       "epoch": 1.72,
-      "grad_norm": 0.5241263508796692,
+      "grad_norm": 2.0460734367370605,
       "learning_rate": 3.614270941054809e-05,
-      "loss": 0.3533,
+      "loss": 0.1592,
       "step": 1840
     },
     {
       "epoch": 1.73,
-      "grad_norm": 120.1414794921875,
+      "grad_norm": 1.6502751111984253,
       "learning_rate": 3.603929679420889e-05,
-      "loss": 0.1999,
+      "loss": 0.3338,
       "step": 1850
     },
     {
       "epoch": 1.74,
-      "grad_norm": 26.968687057495117,
+      "grad_norm": 1.587488055229187,
       "learning_rate": 3.59358841778697e-05,
-      "loss": 0.3365,
+      "loss": 0.4915,
       "step": 1860
     },
     {
       "epoch": 1.75,
-      "grad_norm": 31.189796447753906,
+      "grad_norm": 0.7224433422088623,
       "learning_rate": 3.5832471561530504e-05,
-      "loss": 0.5752,
+      "loss": 0.6087,
       "step": 1870
     },
     {
       "epoch": 1.76,
-      "grad_norm": 3.3193278312683105,
+      "grad_norm": 0.8321720957756042,
       "learning_rate": 3.5729058945191315e-05,
-      "loss": 0.4337,
+      "loss": 0.2709,
       "step": 1880
     },
     {
       "epoch": 1.77,
-      "grad_norm": 17.27393913269043,
+      "grad_norm": 0.8136805891990662,
       "learning_rate": 3.562564632885212e-05,
-      "loss": 0.4251,
+      "loss": 0.1631,
       "step": 1890
     },
     {
       "epoch": 1.78,
-      "grad_norm": 44.38993835449219,
+      "grad_norm": 18.07825469970703,
       "learning_rate": 3.552223371251293e-05,
-      "loss": 0.3813,
+      "loss": 0.2628,
       "step": 1900
     },
     {
       "epoch": 1.79,
-      "grad_norm": 14.854008674621582,
+      "grad_norm": 0.4381904602050781,
       "learning_rate": 3.541882109617373e-05,
-      "loss": 0.349,
+      "loss": 0.0763,
       "step": 1910
     },
     {
       "epoch": 1.8,
-      "grad_norm": 6.35029935836792,
+      "grad_norm": 0.4933386743068695,
       "learning_rate": 3.531540847983454e-05,
-      "loss": 0.3621,
+      "loss": 0.1871,
       "step": 1920
     },
     {
       "epoch": 1.81,
-      "grad_norm": 38.28048324584961,
+      "grad_norm": 0.3448088765144348,
       "learning_rate": 3.521199586349535e-05,
-      "loss": 0.3143,
+      "loss": 0.277,
       "step": 1930
     },
     {
       "epoch": 1.82,
-      "grad_norm": 3.274022340774536,
+      "grad_norm": 0.33174896240234375,
       "learning_rate": 3.5108583247156155e-05,
-      "loss": 0.3147,
+      "loss": 0.0749,
       "step": 1940
     },
     {
       "epoch": 1.83,
-      "grad_norm": 24.270750045776367,
+      "grad_norm": 0.7721774578094482,
       "learning_rate": 3.5005170630816965e-05,
-      "loss": 0.486,
+      "loss": 0.1818,
       "step": 1950
     },
     {
       "epoch": 1.84,
-      "grad_norm": 26.76897621154785,
+      "grad_norm": 0.26144295930862427,
       "learning_rate": 3.490175801447777e-05,
-      "loss": 0.3353,
+      "loss": 0.1525,
       "step": 1960
     },
     {
       "epoch": 1.85,
-      "grad_norm": 28.255657196044922,
+      "grad_norm": 0.190896674990654,
       "learning_rate": 3.479834539813858e-05,
-      "loss": 0.3074,
+      "loss": 0.0797,
       "step": 1970
     },
     {
       "epoch": 1.86,
-      "grad_norm": 14.969733238220215,
+      "grad_norm": 3.312373638153076,
       "learning_rate": 3.469493278179938e-05,
-      "loss": 0.3648,
+      "loss": 0.2365,
       "step": 1980
     },
     {
       "epoch": 1.87,
-      "grad_norm": 13.978160858154297,
+      "grad_norm": 0.1972302347421646,
       "learning_rate": 3.459152016546019e-05,
-      "loss": 0.3108,
+      "loss": 0.1235,
       "step": 1990
     },
     {
       "epoch": 1.87,
-      "grad_norm": 4.878652095794678,
+      "grad_norm": 3.559814214706421,
       "learning_rate": 3.4488107549120996e-05,
-      "loss": 0.3236,
+      "loss": 0.2742,
       "step": 2000
     },
     {
       "epoch": 1.88,
-      "grad_norm": 9.4730806350708,
+      "grad_norm": 1.8290061950683594,
       "learning_rate": 3.4384694932781806e-05,
-      "loss": 0.3036,
+      "loss": 0.1612,
       "step": 2010
     },
     {
       "epoch": 1.89,
-      "grad_norm": 71.49247741699219,
+      "grad_norm": 0.30610841512680054,
       "learning_rate": 3.428128231644261e-05,
-      "loss": 0.3228,
+      "loss": 0.1185,
       "step": 2020
     },
     {
       "epoch": 1.9,
-      "grad_norm": 11.899968147277832,
+      "grad_norm": 0.32188501954078674,
       "learning_rate": 3.417786970010341e-05,
-      "loss": 0.422,
+      "loss": 0.1858,
       "step": 2030
     },
     {
       "epoch": 1.91,
-      "grad_norm": 10.609174728393555,
+      "grad_norm": 0.20377053320407867,
       "learning_rate": 3.407445708376422e-05,
-      "loss": 0.2226,
+      "loss": 0.1092,
       "step": 2040
     },
     {
       "epoch": 1.92,
-      "grad_norm": 0.9701952338218689,
+      "grad_norm": 0.182037815451622,
       "learning_rate": 3.3971044467425026e-05,
-      "loss": 0.3561,
+      "loss": 0.1015,
       "step": 2050
     },
     {
       "epoch": 1.93,
-      "grad_norm": 16.32933235168457,
+      "grad_norm": 25.248992919921875,
       "learning_rate": 3.3867631851085836e-05,
-      "loss": 0.3124,
+      "loss": 0.2925,
       "step": 2060
     },
     {
       "epoch": 1.94,
-      "grad_norm": 1.6818139553070068,
+      "grad_norm": 0.8001343607902527,
       "learning_rate": 3.376421923474664e-05,
-      "loss": 0.274,
+      "loss": 0.4151,
       "step": 2070
     },
     {
       "epoch": 1.95,
-      "grad_norm": 2.9835000038146973,
+      "grad_norm": 1.3534541130065918,
       "learning_rate": 3.366080661840745e-05,
-      "loss": 0.3897,
+      "loss": 0.4384,
       "step": 2080
     },
     {
       "epoch": 1.96,
-      "grad_norm": 0.44764024019241333,
+      "grad_norm": 0.3796235918998718,
       "learning_rate": 3.355739400206825e-05,
-      "loss": 0.3541,
+      "loss": 0.1112,
       "step": 2090
     },
     {
       "epoch": 1.97,
-      "grad_norm": 34.775390625,
+      "grad_norm": 0.280908465385437,
       "learning_rate": 3.345398138572906e-05,
-      "loss": 0.3058,
+      "loss": 0.1345,
       "step": 2100
     },
     {
       "epoch": 1.98,
-      "grad_norm": 10.107741355895996,
+      "grad_norm": 0.2541828155517578,
       "learning_rate": 3.3350568769389867e-05,
-      "loss": 0.2805,
+      "loss": 0.1643,
       "step": 2110
     },
     {
       "epoch": 1.99,
-      "grad_norm": 9.582823753356934,
+      "grad_norm": 0.25121456384658813,
       "learning_rate": 3.324715615305068e-05,
-      "loss": 0.3837,
+      "loss": 0.064,
       "step": 2120
     },
     {
       "epoch": 2.0,
-      "grad_norm": 86.71916198730469,
+      "grad_norm": 0.19068244099617004,
       "learning_rate": 3.314374353671148e-05,
-      "loss": 0.4249,
+      "loss": 0.0118,
       "step": 2130
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8555347091932458,
-      "eval_f1": 0.8560747663551402,
-      "eval_loss": 0.3931359052658081,
-      "eval_precision": 0.8513011152416357,
-      "eval_recall": 0.8609022556390977,
-      "eval_runtime": 1.5356,
-      "eval_samples_per_second": 347.106,
-      "eval_steps_per_second": 43.632,
+      "eval_accuracy": 0.851782363977486,
+      "eval_f1": 0.8500948766603416,
+      "eval_loss": 0.7103251218795776,
+      "eval_precision": 0.8582375478927203,
+      "eval_recall": 0.8421052631578947,
+      "eval_runtime": 1.4454,
+      "eval_samples_per_second": 368.759,
+      "eval_steps_per_second": 46.354,
       "step": 2134
     },
     {
       "epoch": 2.01,
-      "grad_norm": 2.3712680339813232,
+      "grad_norm": 0.4352416694164276,
       "learning_rate": 3.3040330920372283e-05,
-      "loss": 0.1953,
+      "loss": 0.1945,
       "step": 2140
     },
     {
       "epoch": 2.01,
-      "grad_norm": 28.71424102783203,
+      "grad_norm": 3.1081056594848633,
       "learning_rate": 3.2936918304033094e-05,
-      "loss": 0.1712,
+      "loss": 0.192,
       "step": 2150
     },
     {
       "epoch": 2.02,
-      "grad_norm": 0.18505160510540009,
+      "grad_norm": 0.20173217356204987,
       "learning_rate": 3.28335056876939e-05,
-      "loss": 0.3299,
+      "loss": 0.1651,
       "step": 2160
     },
     {
       "epoch": 2.03,
-      "grad_norm": 74.10963439941406,
+      "grad_norm": 3.0507891178131104,
       "learning_rate": 3.273009307135471e-05,
-      "loss": 0.2948,
+      "loss": 0.1877,
       "step": 2170
     },
     {
       "epoch": 2.04,
-      "grad_norm": 1.3842068910598755,
+      "grad_norm": 0.13503152132034302,
       "learning_rate": 3.262668045501551e-05,
-      "loss": 0.3453,
+      "loss": 0.2316,
       "step": 2180
     },
     {
       "epoch": 2.05,
-      "grad_norm": 10.536425590515137,
+      "grad_norm": 0.2812797725200653,
       "learning_rate": 3.252326783867632e-05,
-      "loss": 0.23,
+      "loss": 0.3679,
       "step": 2190
     },
     {
       "epoch": 2.06,
-      "grad_norm": 28.119539260864258,
+      "grad_norm": 0.2773813307285309,
       "learning_rate": 3.2419855222337124e-05,
-      "loss": 0.2037,
+      "loss": 0.0743,
       "step": 2200
     },
     {
       "epoch": 2.07,
-      "grad_norm": 0.34018927812576294,
+      "grad_norm": 0.19849738478660583,
       "learning_rate": 3.2316442605997934e-05,
-      "loss": 0.1796,
+      "loss": 0.1193,
       "step": 2210
     },
     {
       "epoch": 2.08,
-      "grad_norm": 2.8183701038360596,
+      "grad_norm": 5.642420768737793,
       "learning_rate": 3.221302998965874e-05,
-      "loss": 0.5003,
+      "loss": 0.2526,
       "step": 2220
     },
     {
       "epoch": 2.09,
-      "grad_norm": 0.478957861661911,
+      "grad_norm": 0.19763706624507904,
       "learning_rate": 3.210961737331955e-05,
-      "loss": 0.14,
+      "loss": 0.1146,
       "step": 2230
     },
     {
       "epoch": 2.1,
-      "grad_norm": 2.9625442028045654,
+      "grad_norm": 0.19127729535102844,
       "learning_rate": 3.200620475698035e-05,
-      "loss": 0.3229,
+      "loss": 0.1787,
       "step": 2240
     },
     {
       "epoch": 2.11,
-      "grad_norm": 0.5814473032951355,
+      "grad_norm": 0.23325544595718384,
       "learning_rate": 3.190279214064116e-05,
-      "loss": 0.1369,
+      "loss": 0.2123,
       "step": 2250
     },
     {
       "epoch": 2.12,
-      "grad_norm": 1.5362516641616821,
+      "grad_norm": 3.2584617137908936,
       "learning_rate": 3.1799379524301965e-05,
-      "loss": 0.1266,
+      "loss": 0.1797,
       "step": 2260
     },
     {
       "epoch": 2.13,
-      "grad_norm": 30.91729736328125,
+      "grad_norm": 66.83531951904297,
       "learning_rate": 3.169596690796277e-05,
-      "loss": 0.4261,
+      "loss": 0.315,
       "step": 2270
     },
     {
       "epoch": 2.14,
-      "grad_norm": 6.7020673751831055,
+      "grad_norm": 13.469511032104492,
       "learning_rate": 3.159255429162358e-05,
-      "loss": 0.3651,
+      "loss": 0.3296,
       "step": 2280
     },
     {
       "epoch": 2.15,
-      "grad_norm": 3.024595022201538,
+      "grad_norm": 1.0547977685928345,
       "learning_rate": 3.148914167528438e-05,
-      "loss": 0.2789,
+      "loss": 0.2146,
       "step": 2290
     },
     {
       "epoch": 2.16,
-      "grad_norm": 387.9389953613281,
+      "grad_norm": 32.664710998535156,
       "learning_rate": 3.138572905894519e-05,
-      "loss": 0.4301,
+      "loss": 0.2088,
       "step": 2300
     },
     {
       "epoch": 2.16,
-      "grad_norm": 0.2992786467075348,
+      "grad_norm": 1.26264226436615,
       "learning_rate": 3.1282316442605995e-05,
-      "loss": 0.276,
+      "loss": 0.1234,
       "step": 2310
     },
     {
       "epoch": 2.17,
-      "grad_norm": 4.288240432739258,
+      "grad_norm": 0.5536983609199524,
       "learning_rate": 3.117890382626681e-05,
-      "loss": 0.0722,
+      "loss": 0.1124,
       "step": 2320
     },
     {
       "epoch": 2.18,
-      "grad_norm": 0.20439238846302032,
+      "grad_norm": 34.93430709838867,
       "learning_rate": 3.1075491209927615e-05,
-      "loss": 0.1687,
+      "loss": 0.1095,
       "step": 2330
     },
     {
       "epoch": 2.19,
-      "grad_norm": 0.17426259815692902,
+      "grad_norm": 0.1729714721441269,
       "learning_rate": 3.097207859358842e-05,
-      "loss": 0.3444,
+      "loss": 0.2375,
       "step": 2340
     },
     {
       "epoch": 2.2,
-      "grad_norm": 2.802089214324951,
+      "grad_norm": 0.18604367971420288,
       "learning_rate": 3.086866597724923e-05,
-      "loss": 0.4715,
+      "loss": 0.0587,
       "step": 2350
     },
     {
       "epoch": 2.21,
-      "grad_norm": 0.27714866399765015,
+      "grad_norm": 0.1283494383096695,
       "learning_rate": 3.076525336091003e-05,
-      "loss": 0.2581,
+      "loss": 0.1216,
       "step": 2360
     },
     {
       "epoch": 2.22,
-      "grad_norm": 9.906682014465332,
+      "grad_norm": 78.32118225097656,
       "learning_rate": 3.066184074457084e-05,
-      "loss": 0.1077,
+      "loss": 0.2107,
       "step": 2370
     },
     {
       "epoch": 2.23,
-      "grad_norm": 17.114538192749023,
+      "grad_norm": 2.105651378631592,
       "learning_rate": 3.0558428128231646e-05,
-      "loss": 0.3108,
+      "loss": 0.3923,
       "step": 2380
     },
     {
       "epoch": 2.24,
-      "grad_norm": 5.7600178718566895,
+      "grad_norm": 0.33363184332847595,
       "learning_rate": 3.0455015511892452e-05,
-      "loss": 0.3055,
+      "loss": 0.213,
       "step": 2390
     },
     {
       "epoch": 2.25,
-      "grad_norm": 0.660796582698822,
+      "grad_norm": 0.2842453122138977,
       "learning_rate": 3.035160289555326e-05,
-      "loss": 0.1998,
+      "loss": 0.051,
       "step": 2400
     },
     {
       "epoch": 2.26,
-      "grad_norm": 0.19979248940944672,
+      "grad_norm": 0.2736751139163971,
       "learning_rate": 3.0248190279214066e-05,
-      "loss": 0.3523,
+      "loss": 0.3842,
       "step": 2410
     },
     {
       "epoch": 2.27,
-      "grad_norm": 0.6000668406486511,
+      "grad_norm": 0.2922070622444153,
       "learning_rate": 3.0144777662874873e-05,
-      "loss": 0.1009,
+      "loss": 0.2233,
       "step": 2420
     },
     {
       "epoch": 2.28,
-      "grad_norm": 0.15889067947864532,
+      "grad_norm": 0.28500527143478394,
       "learning_rate": 3.004136504653568e-05,
-      "loss": 0.0861,
+      "loss": 0.0617,
       "step": 2430
     },
     {
       "epoch": 2.29,
-      "grad_norm": 0.32534798979759216,
+      "grad_norm": 7.861285209655762,
       "learning_rate": 2.9937952430196486e-05,
-      "loss": 0.1119,
+      "loss": 0.2619,
       "step": 2440
     },
     {
       "epoch": 2.3,
-      "grad_norm": 6.747263431549072,
+      "grad_norm": 0.8819804787635803,
       "learning_rate": 2.9834539813857293e-05,
-      "loss": 0.2019,
+      "loss": 0.1612,
       "step": 2450
     },
     {
       "epoch": 2.31,
-      "grad_norm": 0.0872272402048111,
+      "grad_norm": 0.41944587230682373,
       "learning_rate": 2.97311271975181e-05,
-      "loss": 0.3013,
+      "loss": 0.1717,
       "step": 2460
     },
     {
       "epoch": 2.31,
-      "grad_norm": 0.418996125459671,
+      "grad_norm": 0.20882445573806763,
       "learning_rate": 2.9627714581178906e-05,
-      "loss": 0.3679,
+      "loss": 0.3839,
       "step": 2470
     },
     {
       "epoch": 2.32,
-      "grad_norm": 34.88379669189453,
+      "grad_norm": 3.0659005641937256,
       "learning_rate": 2.9524301964839713e-05,
-      "loss": 0.3748,
+      "loss": 0.2136,
       "step": 2480
     },
     {
       "epoch": 2.33,
-      "grad_norm": 0.8229822516441345,
+      "grad_norm": 0.2248445302248001,
       "learning_rate": 2.942088934850052e-05,
-      "loss": 0.3046,
+      "loss": 0.1435,
       "step": 2490
     },
     {
       "epoch": 2.34,
-      "grad_norm": 1.254721999168396,
+      "grad_norm": 0.26371681690216064,
       "learning_rate": 2.9317476732161327e-05,
-      "loss": 0.4162,
+      "loss": 0.1847,
       "step": 2500
     },
     {
       "epoch": 2.35,
-      "grad_norm": 0.805443286895752,
+      "grad_norm": 554.4179077148438,
       "learning_rate": 2.921406411582213e-05,
-      "loss": 0.2995,
+      "loss": 0.3933,
       "step": 2510
     },
     {
       "epoch": 2.36,
-      "grad_norm": 0.21225698292255402,
+      "grad_norm": 0.27695804834365845,
       "learning_rate": 2.9110651499482937e-05,
-      "loss": 0.2081,
+      "loss": 0.1195,
       "step": 2520
     },
     {
       "epoch": 2.37,
-      "grad_norm": 1.794771671295166,
+      "grad_norm": 0.5627778172492981,
       "learning_rate": 2.9007238883143744e-05,
-      "loss": 0.2845,
+      "loss": 0.2075,
       "step": 2530
     },
     {
       "epoch": 2.38,
-      "grad_norm": 0.45042118430137634,
+      "grad_norm": 0.26693472266197205,
       "learning_rate": 2.890382626680455e-05,
-      "loss": 0.1312,
+      "loss": 0.1704,
       "step": 2540
     },
     {
       "epoch": 2.39,
-      "grad_norm": 2.9366912841796875,
+      "grad_norm": 2.9638285636901855,
       "learning_rate": 2.8800413650465357e-05,
-      "loss": 0.2553,
+      "loss": 0.3962,
       "step": 2550
     },
     {
       "epoch": 2.4,
-      "grad_norm": 0.3621287941932678,
+      "grad_norm": 0.25458648800849915,
       "learning_rate": 2.8697001034126164e-05,
-      "loss": 0.419,
+      "loss": 0.1056,
       "step": 2560
     },
     {
       "epoch": 2.41,
-      "grad_norm": 0.33347466588020325,
+      "grad_norm": 0.3186055123806,
       "learning_rate": 2.859358841778697e-05,
-      "loss": 0.1815,
+      "loss": 0.2689,
       "step": 2570
     },
     {
       "epoch": 2.42,
-      "grad_norm": 2.7859878540039062,
+      "grad_norm": 0.272903710603714,
       "learning_rate": 2.8490175801447777e-05,
-      "loss": 0.2801,
+      "loss": 0.0668,
       "step": 2580
     },
     {
       "epoch": 2.43,
-      "grad_norm": 0.30300045013427734,
+      "grad_norm": 2.9451920986175537,
       "learning_rate": 2.8386763185108584e-05,
-      "loss": 0.1433,
+      "loss": 0.0815,
       "step": 2590
     },
     {
       "epoch": 2.44,
-      "grad_norm": 98.14810943603516,
+      "grad_norm": 193.8448486328125,
       "learning_rate": 2.828335056876939e-05,
-      "loss": 0.2292,
+      "loss": 0.2416,
       "step": 2600
     },
     {
       "epoch": 2.45,
-      "grad_norm": 4.406097412109375,
+      "grad_norm": 0.24242287874221802,
       "learning_rate": 2.8179937952430198e-05,
-      "loss": 0.1548,
+      "loss": 0.2184,
       "step": 2610
     },
     {
       "epoch": 2.46,
-      "grad_norm": 5.7536234855651855,
+      "grad_norm": 0.1883140653371811,
       "learning_rate": 2.8076525336091004e-05,
-      "loss": 0.1126,
+      "loss": 0.06,
       "step": 2620
     },
     {
       "epoch": 2.46,
-      "grad_norm": 36.3065299987793,
+      "grad_norm": 152.6375274658203,
       "learning_rate": 2.7973112719751808e-05,
-      "loss": 0.3183,
+      "loss": 0.2696,
       "step": 2630
     },
     {
       "epoch": 2.47,
-      "grad_norm": 34.32221603393555,
+      "grad_norm": 52.7486686706543,
       "learning_rate": 2.7869700103412615e-05,
-      "loss": 0.342,
+      "loss": 0.2919,
       "step": 2640
     },
     {
       "epoch": 2.48,
-      "grad_norm": 0.14024418592453003,
+      "grad_norm": 0.2107856720685959,
       "learning_rate": 2.776628748707342e-05,
-      "loss": 0.3198,
+      "loss": 0.2631,
       "step": 2650
     },
     {
       "epoch": 2.49,
-      "grad_norm": 0.32867252826690674,
+      "grad_norm": 0.21232011914253235,
       "learning_rate": 2.7662874870734228e-05,
-      "loss": 0.34,
+      "loss": 0.1414,
       "step": 2660
     },
     {
       "epoch": 2.5,
-      "grad_norm": 4.173491954803467,
+      "grad_norm": 0.27856922149658203,
       "learning_rate": 2.7559462254395035e-05,
-      "loss": 0.2842,
+      "loss": 0.0647,
       "step": 2670
     },
     {
       "epoch": 2.51,
-      "grad_norm": 167.7888641357422,
+      "grad_norm": 18.96074676513672,
       "learning_rate": 2.745604963805584e-05,
-      "loss": 0.2955,
+      "loss": 0.2706,
       "step": 2680
     },
     {
       "epoch": 2.52,
-      "grad_norm": 0.7819741368293762,
+      "grad_norm": 0.1531578004360199,
       "learning_rate": 2.735263702171665e-05,
-      "loss": 0.0981,
+      "loss": 0.0589,
       "step": 2690
     },
     {
       "epoch": 2.53,
-      "grad_norm": 3.086498498916626,
+      "grad_norm": 0.13904044032096863,
       "learning_rate": 2.7249224405377455e-05,
-      "loss": 0.2673,
+      "loss": 0.1598,
       "step": 2700
     },
     {
       "epoch": 2.54,
-      "grad_norm": 0.086996890604496,
+      "grad_norm": 0.5579696893692017,
       "learning_rate": 2.7145811789038262e-05,
-      "loss": 0.2067,
+      "loss": 0.0233,
       "step": 2710
     },
     {
       "epoch": 2.55,
-      "grad_norm": 0.09881442785263062,
+      "grad_norm": 3.028477430343628,
       "learning_rate": 2.7042399172699072e-05,
-      "loss": 0.365,
+      "loss": 0.2927,
       "step": 2720
     },
     {
       "epoch": 2.56,
-      "grad_norm": 43.90235137939453,
+      "grad_norm": 0.15513668954372406,
       "learning_rate": 2.693898655635988e-05,
-      "loss": 0.1421,
+      "loss": 0.0464,
       "step": 2730
     },
     {
       "epoch": 2.57,
-      "grad_norm": 0.06362267583608627,
+      "grad_norm": 0.11886773258447647,
       "learning_rate": 2.6835573940020685e-05,
-      "loss": 0.3048,
+      "loss": 0.0602,
       "step": 2740
     },
     {
       "epoch": 2.58,
-      "grad_norm": 0.08752346783876419,
+      "grad_norm": 0.1136392205953598,
       "learning_rate": 2.6732161323681492e-05,
-      "loss": 0.0187,
+      "loss": 0.007,
       "step": 2750
     },
     {
       "epoch": 2.59,
-      "grad_norm": 0.5300313234329224,
+      "grad_norm": 0.10946289449930191,
       "learning_rate": 2.66287487073423e-05,
-      "loss": 0.1348,
+      "loss": 0.2097,
       "step": 2760
     },
     {
       "epoch": 2.6,
-      "grad_norm": 0.04888344556093216,
+      "grad_norm": 0.10061541199684143,
       "learning_rate": 2.6525336091003106e-05,
-      "loss": 0.1468,
+      "loss": 0.1469,
       "step": 2770
     },
     {
       "epoch": 2.61,
-      "grad_norm": 13.05820369720459,
+      "grad_norm": 6.262234687805176,
       "learning_rate": 2.6421923474663913e-05,
-      "loss": 0.4363,
+      "loss": 0.2607,
       "step": 2780
     },
     {
       "epoch": 2.61,
-      "grad_norm": 3.226611852645874,
+      "grad_norm": 40.82796096801758,
       "learning_rate": 2.631851085832472e-05,
-      "loss": 0.2585,
+      "loss": 0.3085,
       "step": 2790
     },
     {
       "epoch": 2.62,
-      "grad_norm": 127.92499542236328,
+      "grad_norm": 0.13032160699367523,
       "learning_rate": 2.6215098241985526e-05,
-      "loss": 0.2791,
+      "loss": 0.0238,
       "step": 2800
     },
     {
       "epoch": 2.63,
-      "grad_norm": 4.131041526794434,
+      "grad_norm": 5.553137302398682,
       "learning_rate": 2.6111685625646333e-05,
-      "loss": 0.3342,
+      "loss": 0.0747,
       "step": 2810
     },
     {
       "epoch": 2.64,
-      "grad_norm": 9.45926570892334,
+      "grad_norm": 3.196852684020996,
       "learning_rate": 2.600827300930714e-05,
-      "loss": 0.4095,
+      "loss": 0.3373,
       "step": 2820
     },
     {
       "epoch": 2.65,
-      "grad_norm": 0.3252735435962677,
+      "grad_norm": 0.13691984117031097,
       "learning_rate": 2.5904860392967943e-05,
-      "loss": 0.1566,
+      "loss": 0.0726,
       "step": 2830
     },
     {
       "epoch": 2.66,
-      "grad_norm": 17.8929386138916,
+      "grad_norm": 0.11953440308570862,
       "learning_rate": 2.580144777662875e-05,
-      "loss": 0.2914,
+      "loss": 0.0064,
       "step": 2840
     },
     {
       "epoch": 2.67,
-      "grad_norm": 0.4428044557571411,
+      "grad_norm": 0.11957331746816635,
       "learning_rate": 2.5698035160289556e-05,
-      "loss": 0.1937,
+      "loss": 0.1053,
       "step": 2850
     },
     {
       "epoch": 2.68,
-      "grad_norm": 0.196681946516037,
+      "grad_norm": 54.33175277709961,
       "learning_rate": 2.5594622543950363e-05,
-      "loss": 0.1334,
+      "loss": 0.3292,
       "step": 2860
     },
     {
       "epoch": 2.69,
-      "grad_norm": 14.465700149536133,
+      "grad_norm": 4.716727256774902,
       "learning_rate": 2.549120992761117e-05,
-      "loss": 0.4305,
+      "loss": 0.3259,
       "step": 2870
     },
     {
       "epoch": 2.7,
-      "grad_norm": 0.25671684741973877,
+      "grad_norm": 3.1103293895721436,
       "learning_rate": 2.5387797311271977e-05,
-      "loss": 0.3097,
+      "loss": 0.2598,
       "step": 2880
     },
     {
       "epoch": 2.71,
-      "grad_norm": 8.047564506530762,
+      "grad_norm": 3.2426257133483887,
       "learning_rate": 2.5284384694932783e-05,
-      "loss": 0.483,
+      "loss": 0.2396,
       "step": 2890
     },
     {
       "epoch": 2.72,
-      "grad_norm": 12.367530822753906,
+      "grad_norm": 28.899869918823242,
       "learning_rate": 2.518097207859359e-05,
-      "loss": 0.1829,
+      "loss": 0.064,
       "step": 2900
     },
     {
       "epoch": 2.73,
-      "grad_norm": 0.40190449357032776,
+      "grad_norm": 0.18505991995334625,
       "learning_rate": 2.5077559462254397e-05,
-      "loss": 0.2601,
+      "loss": 0.1202,
       "step": 2910
     },
     {
       "epoch": 2.74,
-      "grad_norm": 0.18794836103916168,
+      "grad_norm": 0.1994311809539795,
       "learning_rate": 2.4974146845915204e-05,
-      "loss": 0.1182,
+      "loss": 0.302,
       "step": 2920
     },
     {
       "epoch": 2.75,
-      "grad_norm": 4.731940269470215,
+      "grad_norm": 0.22046104073524475,
       "learning_rate": 2.487073422957601e-05,
-      "loss": 0.4801,
+      "loss": 0.2079,
       "step": 2930
     },
     {
       "epoch": 2.76,
-      "grad_norm": 72.07913208007812,
+      "grad_norm": 2.033285617828369,
       "learning_rate": 2.4767321613236817e-05,
-      "loss": 0.1826,
+      "loss": 0.2222,
       "step": 2940
     },
     {
       "epoch": 2.76,
-      "grad_norm": 0.34144583344459534,
+      "grad_norm": 0.3028343915939331,
       "learning_rate": 2.466390899689762e-05,
-      "loss": 0.2458,
+      "loss": 0.2839,
       "step": 2950
     },
     {
       "epoch": 2.77,
-      "grad_norm": 2.847762107849121,
+      "grad_norm": 0.3970238268375397,
       "learning_rate": 2.4560496380558427e-05,
-      "loss": 0.2508,
+      "loss": 0.111,
       "step": 2960
     },
     {
       "epoch": 2.78,
-      "grad_norm": 1.0531044006347656,
+      "grad_norm": 38.20304489135742,
       "learning_rate": 2.4457083764219234e-05,
-      "loss": 0.3881,
+      "loss": 0.1972,
       "step": 2970
     },
     {
       "epoch": 2.79,
-      "grad_norm": 0.8281335830688477,
+      "grad_norm": 0.2983644902706146,
       "learning_rate": 2.435367114788004e-05,
-      "loss": 0.221,
+      "loss": 0.1076,
       "step": 2980
     },
     {
       "epoch": 2.8,
-      "grad_norm": 12.635905265808105,
+      "grad_norm": 0.24695682525634766,
       "learning_rate": 2.4250258531540848e-05,
-      "loss": 0.2008,
+      "loss": 0.0882,
       "step": 2990
     },
     {
       "epoch": 2.81,
-      "grad_norm": 0.37540680170059204,
+      "grad_norm": 0.22463589906692505,
       "learning_rate": 2.4146845915201654e-05,
-      "loss": 0.3256,
+      "loss": 0.2107,
       "step": 3000
     },
     {
       "epoch": 2.82,
-      "grad_norm": 0.306692510843277,
+      "grad_norm": 41.61905288696289,
       "learning_rate": 2.404343329886246e-05,
-      "loss": 0.2346,
+      "loss": 0.274,
       "step": 3010
     },
     {
       "epoch": 2.83,
-      "grad_norm": 0.23924803733825684,
+      "grad_norm": 0.194807231426239,
       "learning_rate": 2.394002068252327e-05,
-      "loss": 0.2269,
+      "loss": 0.1362,
       "step": 3020
     },
     {
       "epoch": 2.84,
-      "grad_norm": 43.663082122802734,
+      "grad_norm": 0.16440044343471527,
       "learning_rate": 2.3836608066184078e-05,
-      "loss": 0.2109,
+      "loss": 0.099,
       "step": 3030
     },
     {
       "epoch": 2.85,
-      "grad_norm": 3.9084434509277344,
+      "grad_norm": 0.43695831298828125,
       "learning_rate": 2.3733195449844885e-05,
-      "loss": 0.4289,
+      "loss": 0.148,
       "step": 3040
     },
     {
       "epoch": 2.86,
-      "grad_norm": 1.4723049402236938,
+      "grad_norm": 0.16196587681770325,
       "learning_rate": 2.3629782833505688e-05,
-      "loss": 0.2405,
+      "loss": 0.3657,
       "step": 3050
     },
     {
       "epoch": 2.87,
-      "grad_norm": 3.2489137649536133,
+      "grad_norm": 0.13062936067581177,
       "learning_rate": 2.3526370217166495e-05,
-      "loss": 0.2141,
+      "loss": 0.1996,
       "step": 3060
     },
     {
       "epoch": 2.88,
-      "grad_norm": 10.305293083190918,
+      "grad_norm": 17.27379608154297,
       "learning_rate": 2.3422957600827302e-05,
-      "loss": 0.4549,
+      "loss": 0.2486,
       "step": 3070
     },
     {
       "epoch": 2.89,
-      "grad_norm": 18.369117736816406,
+      "grad_norm": 2.8673007488250732,
       "learning_rate": 2.331954498448811e-05,
-      "loss": 0.3059,
+      "loss": 0.1317,
       "step": 3080
     },
     {
       "epoch": 2.9,
-      "grad_norm": 24.559249877929688,
+      "grad_norm": 156.50790405273438,
       "learning_rate": 2.3216132368148915e-05,
-      "loss": 0.1978,
+      "loss": 0.1618,
       "step": 3090
     },
     {
       "epoch": 2.91,
-      "grad_norm": 0.5222854018211365,
+      "grad_norm": 0.14409740269184113,
       "learning_rate": 2.3112719751809722e-05,
-      "loss": 0.1677,
+      "loss": 0.2678,
       "step": 3100
     },
     {
       "epoch": 2.91,
-      "grad_norm": 7.490653038024902,
+      "grad_norm": 55.17076873779297,
       "learning_rate": 2.300930713547053e-05,
-      "loss": 0.3634,
+      "loss": 0.2296,
       "step": 3110
     },
     {
       "epoch": 2.92,
-      "grad_norm": 0.39714524149894714,
+      "grad_norm": 0.35058945417404175,
       "learning_rate": 2.2905894519131335e-05,
-      "loss": 0.2277,
+      "loss": 0.1965,
       "step": 3120
     },
     {
       "epoch": 2.93,
-      "grad_norm": 0.16488385200500488,
+      "grad_norm": 0.17419961094856262,
       "learning_rate": 2.2802481902792142e-05,
-      "loss": 0.1301,
+      "loss": 0.2563,
       "step": 3130
     },
     {
       "epoch": 2.94,
-      "grad_norm": 3.3054399490356445,
+      "grad_norm": 0.13027483224868774,
       "learning_rate": 2.269906928645295e-05,
-      "loss": 0.3157,
+      "loss": 0.3372,
       "step": 3140
     },
     {
       "epoch": 2.95,
-      "grad_norm": 0.1860102266073227,
+      "grad_norm": 0.11983130872249603,
       "learning_rate": 2.2595656670113756e-05,
-      "loss": 0.1737,
+      "loss": 0.0509,
       "step": 3150
     },
     {
       "epoch": 2.96,
-      "grad_norm": 6.283505916595459,
+      "grad_norm": 2.41009783744812,
       "learning_rate": 2.2492244053774563e-05,
-      "loss": 0.1935,
+      "loss": 0.0769,
       "step": 3160
     },
     {
       "epoch": 2.97,
-      "grad_norm": 0.3177931010723114,
+      "grad_norm": 0.11595846712589264,
       "learning_rate": 2.2388831437435366e-05,
-      "loss": 0.2413,
+      "loss": 0.1942,
       "step": 3170
     },
     {
       "epoch": 2.98,
-      "grad_norm": 9.307920455932617,
+      "grad_norm": 136.9188690185547,
       "learning_rate": 2.2285418821096173e-05,
-      "loss": 0.2362,
+      "loss": 0.0619,
       "step": 3180
     },
     {
       "epoch": 2.99,
-      "grad_norm": 20.813262939453125,
+      "grad_norm": 10.190946578979492,
       "learning_rate": 2.218200620475698e-05,
-      "loss": 0.2896,
+      "loss": 0.196,
       "step": 3190
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.17595571279525757,
+      "grad_norm": 0.11687775701284409,
       "learning_rate": 2.2078593588417786e-05,
-      "loss": 0.1038,
+      "loss": 0.1802,
       "step": 3200
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8667917448405253,
-      "eval_f1": 0.864244741873805,
-      "eval_loss": 0.5857056975364685,
-      "eval_precision": 0.8793774319066148,
-      "eval_recall": 0.849624060150376,
-      "eval_runtime": 1.5831,
-      "eval_samples_per_second": 336.679,
-      "eval_steps_per_second": 42.322,
+      "eval_accuracy": 0.8574108818011257,
+      "eval_f1": 0.8608058608058609,
+      "eval_loss": 0.7037833333015442,
+      "eval_precision": 0.8392857142857143,
+      "eval_recall": 0.8834586466165414,
+      "eval_runtime": 1.4332,
+      "eval_samples_per_second": 371.895,
+      "eval_steps_per_second": 46.749,
       "step": 3201
     },
     {
       "epoch": 3.01,
-      "grad_norm": 0.1377851516008377,
+      "grad_norm": 0.2782648205757141,
       "learning_rate": 2.1975180972078593e-05,
-      "loss": 0.0617,
+      "loss": 0.0062,
       "step": 3210
     },
     {
       "epoch": 3.02,
-      "grad_norm": 8.743398666381836,
+      "grad_norm": 32.99894714355469,
       "learning_rate": 2.1871768355739403e-05,
-      "loss": 0.2428,
+      "loss": 0.1245,
       "step": 3220
     },
     {
       "epoch": 3.03,
-      "grad_norm": 1.2507226467132568,
+      "grad_norm": 0.12525002658367157,
       "learning_rate": 2.176835573940021e-05,
-      "loss": 0.187,
+      "loss": 0.19,
       "step": 3230
     },
     {
       "epoch": 3.04,
-      "grad_norm": 0.17167934775352478,
+      "grad_norm": 0.10308614373207092,
       "learning_rate": 2.1664943123061017e-05,
-      "loss": 0.1137,
+      "loss": 0.0759,
       "step": 3240
     },
     {
       "epoch": 3.05,
-      "grad_norm": 5.110489845275879,
+      "grad_norm": 3.754702568054199,
       "learning_rate": 2.1561530506721823e-05,
-      "loss": 0.0637,
+      "loss": 0.0751,
       "step": 3250
     },
     {
       "epoch": 3.06,
-      "grad_norm": 0.11725780367851257,
+      "grad_norm": 0.09151796996593475,
       "learning_rate": 2.145811789038263e-05,
-      "loss": 0.1933,
+      "loss": 0.0074,
       "step": 3260
     },
     {
       "epoch": 3.06,
-      "grad_norm": 0.11883195489645004,
+      "grad_norm": 311.2441711425781,
       "learning_rate": 2.1354705274043433e-05,
-      "loss": 0.1123,
+      "loss": 0.1058,
       "step": 3270
     },
     {
       "epoch": 3.07,
-      "grad_norm": 7.5306315422058105,
+      "grad_norm": 0.10257299989461899,
       "learning_rate": 2.125129265770424e-05,
-      "loss": 0.1364,
+      "loss": 0.0054,
       "step": 3280
     },
     {
       "epoch": 3.08,
-      "grad_norm": 0.7140617370605469,
+      "grad_norm": 0.6756200194358826,
       "learning_rate": 2.1147880041365047e-05,
-      "loss": 0.148,
+      "loss": 0.0657,
       "step": 3290
     },
     {
       "epoch": 3.09,
-      "grad_norm": 0.9037841558456421,
+      "grad_norm": 0.131881982088089,
       "learning_rate": 2.1044467425025854e-05,
-      "loss": 0.0663,
+      "loss": 0.0735,
       "step": 3300
     },
     {
       "epoch": 3.1,
-      "grad_norm": 0.05070972070097923,
+      "grad_norm": 0.0882406160235405,
       "learning_rate": 2.094105480868666e-05,
-      "loss": 0.0595,
+      "loss": 0.0732,
       "step": 3310
     },
     {
       "epoch": 3.11,
-      "grad_norm": 0.06952597945928574,
+      "grad_norm": 0.06303223967552185,
       "learning_rate": 2.0837642192347467e-05,
-      "loss": 0.0802,
+      "loss": 0.0034,
       "step": 3320
     },
     {
       "epoch": 3.12,
-      "grad_norm": 0.09682538360357285,
+      "grad_norm": 0.06847697496414185,
       "learning_rate": 2.0734229576008274e-05,
-      "loss": 0.0756,
+      "loss": 0.1137,
       "step": 3330
     },
     {
       "epoch": 3.13,
-      "grad_norm": 7.944037914276123,
+      "grad_norm": 0.05840703472495079,
       "learning_rate": 2.063081695966908e-05,
-      "loss": 0.1648,
+      "loss": 0.003,
       "step": 3340
     },
     {
       "epoch": 3.14,
-      "grad_norm": 0.049520719796419144,
+      "grad_norm": 4.701480388641357,
       "learning_rate": 2.0527404343329888e-05,
-      "loss": 0.2814,
+      "loss": 0.057,
       "step": 3350
     },
     {
       "epoch": 3.15,
-      "grad_norm": 0.12401607632637024,
+      "grad_norm": 0.06599827110767365,
       "learning_rate": 2.0423991726990694e-05,
-      "loss": 0.1665,
+      "loss": 0.1948,
       "step": 3360
     },
     {
       "epoch": 3.16,
-      "grad_norm": 3.3132777214050293,
+      "grad_norm": 0.07265453785657883,
       "learning_rate": 2.03205791106515e-05,
-      "loss": 0.2163,
+      "loss": 0.2271,
       "step": 3370
     },
     {
       "epoch": 3.17,
-      "grad_norm": 0.06092935800552368,
+      "grad_norm": 0.07454103976488113,
       "learning_rate": 2.0217166494312308e-05,
-      "loss": 0.055,
+      "loss": 0.0787,
       "step": 3380
     },
     {
       "epoch": 3.18,
-      "grad_norm": 0.09410246461629868,
+      "grad_norm": 0.06254041194915771,
       "learning_rate": 2.011375387797311e-05,
-      "loss": 0.0564,
+      "loss": 0.0035,
       "step": 3390
     },
     {
       "epoch": 3.19,
-      "grad_norm": 0.09851285070180893,
+      "grad_norm": 0.07290501892566681,
       "learning_rate": 2.0010341261633918e-05,
-      "loss": 0.0996,
+      "loss": 0.0705,
       "step": 3400
     },
     {
       "epoch": 3.2,
-      "grad_norm": 0.15248961746692657,
+      "grad_norm": 0.07300185412168503,
       "learning_rate": 1.9906928645294725e-05,
-      "loss": 0.202,
+      "loss": 0.2761,
       "step": 3410
     },
     {
       "epoch": 3.21,
-      "grad_norm": 0.06347668915987015,
+      "grad_norm": 0.09265441447496414,
       "learning_rate": 1.980351602895553e-05,
-      "loss": 0.0841,
+      "loss": 0.2529,
       "step": 3420
     },
     {
       "epoch": 3.21,
-      "grad_norm": 193.50746154785156,
+      "grad_norm": 0.3778844475746155,
       "learning_rate": 1.970010341261634e-05,
-      "loss": 0.4587,
+      "loss": 0.1832,
       "step": 3430
     },
     {
       "epoch": 3.22,
-      "grad_norm": 0.1516660451889038,
+      "grad_norm": 2.613478899002075,
       "learning_rate": 1.959669079627715e-05,
-      "loss": 0.0655,
+      "loss": 0.1361,
       "step": 3440
     },
     {
       "epoch": 3.23,
-      "grad_norm": 0.42711296677589417,
+      "grad_norm": 0.08724066615104675,
       "learning_rate": 1.9493278179937955e-05,
-      "loss": 0.1857,
+      "loss": 0.2884,
       "step": 3450
     },
     {
       "epoch": 3.24,
-      "grad_norm": 0.11023885011672974,
+      "grad_norm": 0.09438508003950119,
       "learning_rate": 1.9389865563598762e-05,
-      "loss": 0.0828,
+      "loss": 0.1839,
       "step": 3460
     },
     {
       "epoch": 3.25,
-      "grad_norm": 0.061248745769262314,
+      "grad_norm": 0.09784245491027832,
       "learning_rate": 1.928645294725957e-05,
-      "loss": 0.1167,
+      "loss": 0.173,
       "step": 3470
     },
     {
       "epoch": 3.26,
-      "grad_norm": 0.06203962489962578,
+      "grad_norm": 0.08967234194278717,
       "learning_rate": 1.9183040330920375e-05,
-      "loss": 0.1089,
+      "loss": 0.1881,
       "step": 3480
     },
     {
       "epoch": 3.27,
-      "grad_norm": 19.638856887817383,
+      "grad_norm": 0.09921535849571228,
       "learning_rate": 1.907962771458118e-05,
-      "loss": 0.1296,
+      "loss": 0.1038,
       "step": 3490
     },
     {
       "epoch": 3.28,
-      "grad_norm": 0.05891411006450653,
+      "grad_norm": 0.08819156140089035,
       "learning_rate": 1.8976215098241985e-05,
-      "loss": 0.0137,
+      "loss": 0.0053,
       "step": 3500
     },
     {
       "epoch": 3.29,
-      "grad_norm": 0.10509899258613586,
+      "grad_norm": 74.00690460205078,
       "learning_rate": 1.8872802481902792e-05,
-      "loss": 0.0542,
+      "loss": 0.014,
       "step": 3510
     },
     {
       "epoch": 3.3,
-      "grad_norm": 3.27602481842041,
+      "grad_norm": 2.492738723754883,
       "learning_rate": 1.87693898655636e-05,
-      "loss": 0.1655,
+      "loss": 0.1475,
       "step": 3520
     },
     {
       "epoch": 3.31,
-      "grad_norm": 0.08927629142999649,
+      "grad_norm": 0.08610210567712784,
       "learning_rate": 1.8665977249224406e-05,
-      "loss": 0.0737,
+      "loss": 0.0365,
       "step": 3530
     },
     {
       "epoch": 3.32,
-      "grad_norm": 0.2776730954647064,
+      "grad_norm": 0.45213979482650757,
       "learning_rate": 1.8562564632885213e-05,
-      "loss": 0.1569,
+      "loss": 0.3655,
       "step": 3540
     },
     {
       "epoch": 3.33,
-      "grad_norm": 3.598266124725342,
+      "grad_norm": 4.162100791931152,
       "learning_rate": 1.845915201654602e-05,
-      "loss": 0.3042,
+      "loss": 0.2041,
       "step": 3550
     },
     {
       "epoch": 3.34,
-      "grad_norm": 3.0984160900115967,
+      "grad_norm": 2.3223049640655518,
       "learning_rate": 1.8355739400206826e-05,
-      "loss": 0.1836,
+      "loss": 0.0742,
       "step": 3560
     },
     {
       "epoch": 3.35,
-      "grad_norm": 0.09482892602682114,
+      "grad_norm": 0.09031692892313004,
       "learning_rate": 1.8252326783867633e-05,
-      "loss": 0.0684,
+      "loss": 0.0773,
       "step": 3570
     },
     {
       "epoch": 3.36,
-      "grad_norm": 4.153615474700928,
+      "grad_norm": 0.09234865009784698,
       "learning_rate": 1.814891416752844e-05,
-      "loss": 0.188,
+      "loss": 0.196,
       "step": 3580
     },
     {
       "epoch": 3.36,
-      "grad_norm": 241.65493774414062,
+      "grad_norm": 0.1055426225066185,
       "learning_rate": 1.8045501551189246e-05,
-      "loss": 0.1741,
+      "loss": 0.1085,
       "step": 3590
     },
     {
       "epoch": 3.37,
-      "grad_norm": 0.13454625010490417,
+      "grad_norm": 0.07118582725524902,
       "learning_rate": 1.7942088934850053e-05,
-      "loss": 0.0606,
+      "loss": 0.0076,
       "step": 3600
     },
     {
       "epoch": 3.38,
-      "grad_norm": 0.12659063935279846,
+      "grad_norm": 1.206527829170227,
       "learning_rate": 1.7838676318510856e-05,
-      "loss": 0.0743,
+      "loss": 0.0053,
       "step": 3610
     },
     {
       "epoch": 3.39,
-      "grad_norm": 0.12111490219831467,
+      "grad_norm": 446.3911437988281,
       "learning_rate": 1.7735263702171663e-05,
-      "loss": 0.0042,
+      "loss": 0.0745,
       "step": 3620
     },
     {
       "epoch": 3.4,
-      "grad_norm": 3.1387064456939697,
+      "grad_norm": 2.56962513923645,
       "learning_rate": 1.7631851085832473e-05,
-      "loss": 0.3181,
+      "loss": 0.2187,
       "step": 3630
     },
     {
       "epoch": 3.41,
-      "grad_norm": 66.55416870117188,
+      "grad_norm": 0.36162999272346497,
       "learning_rate": 1.752843846949328e-05,
-      "loss": 0.1883,
+      "loss": 0.0797,
       "step": 3640
     },
     {
       "epoch": 3.42,
-      "grad_norm": 9.684922218322754,
+      "grad_norm": 0.1231626346707344,
       "learning_rate": 1.7425025853154087e-05,
-      "loss": 0.129,
+      "loss": 0.0706,
       "step": 3650
     },
     {
       "epoch": 3.43,
-      "grad_norm": 0.1191384568810463,
+      "grad_norm": 0.07382514327764511,
       "learning_rate": 1.7321613236814894e-05,
-      "loss": 0.1684,
+      "loss": 0.2472,
       "step": 3660
     },
     {
       "epoch": 3.44,
-      "grad_norm": 0.09780813008546829,
+      "grad_norm": 0.08766525983810425,
       "learning_rate": 1.72182006204757e-05,
-      "loss": 0.0052,
+      "loss": 0.0043,
       "step": 3670
     },
     {
       "epoch": 3.45,
-      "grad_norm": 3.4690449237823486,
+      "grad_norm": 0.07739522308111191,
       "learning_rate": 1.7114788004136507e-05,
-      "loss": 0.2599,
+      "loss": 0.0544,
       "step": 3680
     },
     {
       "epoch": 3.46,
-      "grad_norm": 0.1335831582546234,
+      "grad_norm": 0.07944560796022415,
       "learning_rate": 1.7011375387797314e-05,
-      "loss": 0.2327,
+      "loss": 0.0655,
       "step": 3690
     },
     {
       "epoch": 3.47,
-      "grad_norm": 0.1467386782169342,
+      "grad_norm": 9.697904586791992,
       "learning_rate": 1.690796277145812e-05,
-      "loss": 0.1308,
+      "loss": 0.1178,
       "step": 3700
     },
     {
       "epoch": 3.48,
-      "grad_norm": 373.8140869140625,
+      "grad_norm": 2.5675041675567627,
       "learning_rate": 1.6804550155118924e-05,
-      "loss": 0.2381,
+      "loss": 0.1142,
       "step": 3710
     },
     {
       "epoch": 3.49,
-      "grad_norm": 0.08402853459119797,
+      "grad_norm": 0.14886552095413208,
       "learning_rate": 1.670113753877973e-05,
-      "loss": 0.0955,
+      "loss": 0.0074,
       "step": 3720
     },
     {
       "epoch": 3.5,
-      "grad_norm": 171.42713928222656,
+      "grad_norm": 106.72728729248047,
       "learning_rate": 1.6597724922440538e-05,
-      "loss": 0.127,
+      "loss": 0.355,
       "step": 3730
     },
     {
       "epoch": 3.51,
-      "grad_norm": 0.11677587032318115,
+      "grad_norm": 0.06539061665534973,
       "learning_rate": 1.6494312306101344e-05,
-      "loss": 0.0415,
+      "loss": 0.0476,
       "step": 3740
     },
     {
       "epoch": 3.51,
-      "grad_norm": 0.0933157354593277,
+      "grad_norm": 70.10369873046875,
       "learning_rate": 1.639089968976215e-05,
-      "loss": 0.1434,
+      "loss": 0.0446,
       "step": 3750
     },
     {
       "epoch": 3.52,
-      "grad_norm": 0.06422335654497147,
+      "grad_norm": 0.06210116297006607,
       "learning_rate": 1.6287487073422958e-05,
-      "loss": 0.2715,
+      "loss": 0.0033,
       "step": 3760
     },
     {
       "epoch": 3.53,
-      "grad_norm": 0.08810001611709595,
+      "grad_norm": 0.05276573449373245,
       "learning_rate": 1.6184074457083765e-05,
-      "loss": 0.077,
+      "loss": 0.003,
       "step": 3770
     },
     {
       "epoch": 3.54,
-      "grad_norm": 0.08342567831277847,
+      "grad_norm": 0.053160134702920914,
       "learning_rate": 1.608066184074457e-05,
-      "loss": 0.2652,
+      "loss": 0.1561,
       "step": 3780
     },
     {
       "epoch": 3.55,
-      "grad_norm": 0.10808378458023071,
+      "grad_norm": 0.07925640046596527,
       "learning_rate": 1.5977249224405378e-05,
-      "loss": 0.0818,
+      "loss": 0.0036,
       "step": 3790
     },
     {
       "epoch": 3.56,
-      "grad_norm": 0.06099862605333328,
+      "grad_norm": 0.053619351238012314,
       "learning_rate": 1.5873836608066185e-05,
-      "loss": 0.195,
+      "loss": 0.0907,
       "step": 3800
     },
     {
       "epoch": 3.57,
-      "grad_norm": 17.37858772277832,
+      "grad_norm": 0.6370459198951721,
       "learning_rate": 1.577042399172699e-05,
-      "loss": 0.2626,
+      "loss": 0.0026,
       "step": 3810
     },
     {
       "epoch": 3.58,
-      "grad_norm": 0.15915565192699432,
+      "grad_norm": 0.062730573117733,
       "learning_rate": 1.56670113753878e-05,
-      "loss": 0.0859,
+      "loss": 0.191,
       "step": 3820
     },
     {
       "epoch": 3.59,
-      "grad_norm": 13.893326759338379,
+      "grad_norm": 0.043475110083818436,
       "learning_rate": 1.5563598759048605e-05,
-      "loss": 0.1901,
+      "loss": 0.1617,
       "step": 3830
     },
     {
       "epoch": 3.6,
-      "grad_norm": 0.19214822351932526,
+      "grad_norm": 240.55931091308594,
       "learning_rate": 1.5460186142709412e-05,
-      "loss": 0.1584,
+      "loss": 0.0346,
       "step": 3840
     },
     {
       "epoch": 3.61,
-      "grad_norm": 0.07823023200035095,
+      "grad_norm": 0.06091690808534622,
       "learning_rate": 1.535677352637022e-05,
-      "loss": 0.1354,
+      "loss": 0.2554,
       "step": 3850
     },
     {
       "epoch": 3.62,
-      "grad_norm": 0.2098669558763504,
+      "grad_norm": 0.8543986678123474,
       "learning_rate": 1.5253360910031025e-05,
-      "loss": 0.2055,
+      "loss": 0.153,
       "step": 3860
     },
     {
       "epoch": 3.63,
-      "grad_norm": 3.7674732208251953,
+      "grad_norm": 11.088409423828125,
       "learning_rate": 1.5149948293691832e-05,
-      "loss": 0.1174,
+      "loss": 0.0748,
       "step": 3870
     },
     {
       "epoch": 3.64,
-      "grad_norm": 0.2199217975139618,
+      "grad_norm": 0.06965257227420807,
       "learning_rate": 1.5046535677352639e-05,
-      "loss": 0.2572,
+      "loss": 0.2284,
       "step": 3880
     },
     {
       "epoch": 3.65,
-      "grad_norm": 13.79918098449707,
+      "grad_norm": 116.66575622558594,
       "learning_rate": 1.4943123061013446e-05,
-      "loss": 0.1485,
+      "loss": 0.2113,
       "step": 3890
     },
     {
       "epoch": 3.66,
-      "grad_norm": 57.75919723510742,
+      "grad_norm": 1.2777670621871948,
       "learning_rate": 1.483971044467425e-05,
-      "loss": 0.2868,
+      "loss": 0.2787,
       "step": 3900
     },
     {
       "epoch": 3.66,
-      "grad_norm": 0.26549792289733887,
+      "grad_norm": 0.09148576855659485,
       "learning_rate": 1.4736297828335057e-05,
-      "loss": 0.1935,
+      "loss": 0.0842,
       "step": 3910
     },
     {
       "epoch": 3.67,
-      "grad_norm": 0.11525550484657288,
+      "grad_norm": 0.07752866297960281,
       "learning_rate": 1.4632885211995864e-05,
-      "loss": 0.1079,
+      "loss": 0.004,
       "step": 3920
     },
     {
       "epoch": 3.68,
-      "grad_norm": 0.11162678152322769,
+      "grad_norm": 3.558208703994751,
       "learning_rate": 1.4529472595656671e-05,
-      "loss": 0.2091,
+      "loss": 0.1208,
       "step": 3930
     },
     {
       "epoch": 3.69,
-      "grad_norm": 0.18426716327667236,
+      "grad_norm": 0.09885440021753311,
       "learning_rate": 1.4426059979317478e-05,
-      "loss": 0.188,
+      "loss": 0.1466,
       "step": 3940
     },
     {
       "epoch": 3.7,
-      "grad_norm": 0.4274304509162903,
+      "grad_norm": 3.113973617553711,
       "learning_rate": 1.4322647362978284e-05,
-      "loss": 0.272,
+      "loss": 0.2648,
       "step": 3950
     },
     {
       "epoch": 3.71,
-      "grad_norm": 0.33479902148246765,
+      "grad_norm": 1.2651716470718384,
       "learning_rate": 1.421923474663909e-05,
-      "loss": 0.0622,
+      "loss": 0.1344,
       "step": 3960
     },
     {
       "epoch": 3.72,
-      "grad_norm": 0.14429700374603271,
+      "grad_norm": 0.12687820196151733,
       "learning_rate": 1.4115822130299896e-05,
-      "loss": 0.3126,
+      "loss": 0.1268,
       "step": 3970
     },
     {
       "epoch": 3.73,
-      "grad_norm": 17.044048309326172,
+      "grad_norm": 3.607964038848877,
       "learning_rate": 1.4012409513960703e-05,
-      "loss": 0.1594,
+      "loss": 0.068,
       "step": 3980
     },
     {
       "epoch": 3.74,
-      "grad_norm": 0.23532900214195251,
+      "grad_norm": 0.4581145644187927,
       "learning_rate": 1.390899689762151e-05,
-      "loss": 0.1205,
+      "loss": 0.0703,
       "step": 3990
     },
     {
       "epoch": 3.75,
-      "grad_norm": 0.15815971791744232,
+      "grad_norm": 0.23489171266555786,
       "learning_rate": 1.3805584281282317e-05,
-      "loss": 0.1607,
+      "loss": 0.0055,
       "step": 4000
     },
     {
       "epoch": 3.76,
-      "grad_norm": 0.1729772984981537,
+      "grad_norm": 0.10075139254331589,
       "learning_rate": 1.3702171664943123e-05,
-      "loss": 0.0638,
+      "loss": 0.0048,
       "step": 4010
     },
     {
       "epoch": 3.77,
-      "grad_norm": 10.028935432434082,
+      "grad_norm": 0.10037043690681458,
       "learning_rate": 1.3598759048603928e-05,
-      "loss": 0.1221,
+      "loss": 0.0683,
       "step": 4020
     },
     {
       "epoch": 3.78,
-      "grad_norm": 0.1198321133852005,
+      "grad_norm": 0.08124475181102753,
       "learning_rate": 1.3495346432264739e-05,
-      "loss": 0.2659,
+      "loss": 0.004,
       "step": 4030
     },
     {
       "epoch": 3.79,
-      "grad_norm": 0.061706289649009705,
+      "grad_norm": 0.07579688727855682,
       "learning_rate": 1.3391933815925545e-05,
-      "loss": 0.0927,
+      "loss": 0.013,
       "step": 4040
     },
     {
       "epoch": 3.8,
-      "grad_norm": 18.801136016845703,
+      "grad_norm": 0.07704436779022217,
       "learning_rate": 1.3288521199586352e-05,
-      "loss": 0.3013,
+      "loss": 0.2172,
       "step": 4050
     },
     {
       "epoch": 3.81,
-      "grad_norm": 3.1690914630889893,
+      "grad_norm": 39.54595947265625,
       "learning_rate": 1.3185108583247157e-05,
-      "loss": 0.1682,
+      "loss": 0.1718,
       "step": 4060
     },
     {
       "epoch": 3.81,
-      "grad_norm": 1.2492997646331787,
+      "grad_norm": 0.06458831578493118,
       "learning_rate": 1.3081695966907964e-05,
-      "loss": 0.1623,
+      "loss": 0.1191,
       "step": 4070
     },
     {
       "epoch": 3.82,
-      "grad_norm": 0.12981970608234406,
+      "grad_norm": 0.06719919294118881,
       "learning_rate": 1.297828335056877e-05,
-      "loss": 0.091,
+      "loss": 0.2055,
       "step": 4080
     },
     {
       "epoch": 3.83,
-      "grad_norm": 1.0257341861724854,
+      "grad_norm": 462.3251953125,
       "learning_rate": 1.2874870734229577e-05,
-      "loss": 0.3142,
+      "loss": 0.2145,
       "step": 4090
     },
     {
       "epoch": 3.84,
-      "grad_norm": 0.25799766182899475,
+      "grad_norm": 6.475925922393799,
       "learning_rate": 1.2771458117890384e-05,
-      "loss": 0.1454,
+      "loss": 0.1391,
       "step": 4100
     },
     {
       "epoch": 3.85,
-      "grad_norm": 40.982479095458984,
+      "grad_norm": 2.5063371658325195,
       "learning_rate": 1.2668045501551191e-05,
-      "loss": 0.1361,
+      "loss": 0.1575,
       "step": 4110
     },
     {
       "epoch": 3.86,
-      "grad_norm": 0.11332831531763077,
+      "grad_norm": 0.09441405534744263,
       "learning_rate": 1.2564632885211996e-05,
-      "loss": 0.0518,
+      "loss": 0.0038,
       "step": 4120
     },
     {
       "epoch": 3.87,
-      "grad_norm": 0.04680272936820984,
+      "grad_norm": 0.06849459558725357,
       "learning_rate": 1.2461220268872803e-05,
-      "loss": 0.1327,
+      "loss": 0.0034,
       "step": 4130
     },
     {
       "epoch": 3.88,
-      "grad_norm": 0.40334421396255493,
+      "grad_norm": 0.056417569518089294,
       "learning_rate": 1.235780765253361e-05,
-      "loss": 0.2707,
+      "loss": 0.0033,
       "step": 4140
     },
     {
       "epoch": 3.89,
-      "grad_norm": 0.09748423099517822,
+      "grad_norm": 0.06035558879375458,
       "learning_rate": 1.2254395036194416e-05,
-      "loss": 0.081,
+      "loss": 0.0027,
       "step": 4150
     },
     {
       "epoch": 3.9,
-      "grad_norm": 0.09192584455013275,
+      "grad_norm": 2.3210813999176025,
       "learning_rate": 1.2150982419855223e-05,
-      "loss": 0.2279,
+      "loss": 0.0739,
       "step": 4160
     },
     {
       "epoch": 3.91,
-      "grad_norm": 0.05917090177536011,
+      "grad_norm": 0.052967045456171036,
       "learning_rate": 1.204756980351603e-05,
-      "loss": 0.0764,
+      "loss": 0.0245,
       "step": 4170
     },
     {
       "epoch": 3.92,
-      "grad_norm": 10.152327537536621,
+      "grad_norm": 0.05248899757862091,
       "learning_rate": 1.1944157187176836e-05,
-      "loss": 0.1237,
+      "loss": 0.0029,
       "step": 4180
     },
     {
       "epoch": 3.93,
-      "grad_norm": 0.07578109949827194,
+      "grad_norm": 0.05719129368662834,
       "learning_rate": 1.1840744570837643e-05,
-      "loss": 0.136,
+      "loss": 0.0805,
       "step": 4190
     },
     {
       "epoch": 3.94,
-      "grad_norm": 0.13006837666034698,
+      "grad_norm": 2.66552734375,
       "learning_rate": 1.173733195449845e-05,
-      "loss": 0.0878,
+      "loss": 0.1218,
       "step": 4200
     },
     {
       "epoch": 3.95,
-      "grad_norm": 0.2893955111503601,
+      "grad_norm": 24.393230438232422,
       "learning_rate": 1.1633919338159257e-05,
-      "loss": 0.1322,
+      "loss": 0.1188,
       "step": 4210
     },
     {
       "epoch": 3.96,
-      "grad_norm": 0.04363380745053291,
+      "grad_norm": 0.0423579104244709,
       "learning_rate": 1.1530506721820064e-05,
-      "loss": 0.0566,
+      "loss": 0.0032,
       "step": 4220
     },
     {
       "epoch": 3.96,
-      "grad_norm": 17.78626823425293,
+      "grad_norm": 17.768678665161133,
       "learning_rate": 1.1427094105480869e-05,
-      "loss": 0.1348,
+      "loss": 0.0915,
       "step": 4230
     },
     {
       "epoch": 3.97,
-      "grad_norm": 0.08856278657913208,
+      "grad_norm": 0.04556003212928772,
       "learning_rate": 1.1323681489141675e-05,
-      "loss": 0.0527,
+      "loss": 0.0023,
       "step": 4240
     },
     {
       "epoch": 3.98,
-      "grad_norm": 3.122804880142212,
+      "grad_norm": 2.7925045490264893,
       "learning_rate": 1.1220268872802482e-05,
-      "loss": 0.1003,
+      "loss": 0.0756,
       "step": 4250
     },
     {
       "epoch": 3.99,
-      "grad_norm": 0.032136429101228714,
+      "grad_norm": 0.05207202211022377,
       "learning_rate": 1.1116856256463289e-05,
-      "loss": 0.1428,
+      "loss": 0.0026,
       "step": 4260
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8799249530956847,
-      "eval_f1": 0.8801498127340824,
-      "eval_loss": 0.6459251046180725,
-      "eval_precision": 0.8768656716417911,
-      "eval_recall": 0.8834586466165414,
-      "eval_runtime": 1.6082,
-      "eval_samples_per_second": 331.423,
-      "eval_steps_per_second": 41.661,
+      "eval_accuracy": 0.8555347091932458,
+      "eval_f1": 0.8566108007448789,
+      "eval_loss": 0.7934179306030273,
+      "eval_precision": 0.8487084870848709,
+      "eval_recall": 0.8646616541353384,
+      "eval_runtime": 1.4626,
+      "eval_samples_per_second": 364.425,
+      "eval_steps_per_second": 45.809,
       "step": 4268
     },
     {
       "epoch": 4.0,
-      "grad_norm": 12.53495979309082,
+      "grad_norm": 5.6651482582092285,
       "learning_rate": 1.1013443640124096e-05,
-      "loss": 0.2006,
+      "loss": 0.2031,
       "step": 4270
     },
     {
       "epoch": 4.01,
-      "grad_norm": 0.36368077993392944,
+      "grad_norm": 0.05581251159310341,
       "learning_rate": 1.0910031023784902e-05,
-      "loss": 0.0682,
+      "loss": 0.0802,
       "step": 4280
     },
     {
       "epoch": 4.02,
-      "grad_norm": 0.036783114075660706,
+      "grad_norm": 0.06616955250501633,
       "learning_rate": 1.0806618407445709e-05,
-      "loss": 0.0578,
+      "loss": 0.1276,
       "step": 4290
     },
     {
       "epoch": 4.03,
-      "grad_norm": 0.032453957945108414,
+      "grad_norm": 0.0667022317647934,
       "learning_rate": 1.0703205791106516e-05,
-      "loss": 0.0511,
+      "loss": 0.0744,
       "step": 4300
     },
     {
       "epoch": 4.04,
-      "grad_norm": 0.0314195491373539,
+      "grad_norm": 0.0499153807759285,
       "learning_rate": 1.0599793174767323e-05,
-      "loss": 0.0016,
+      "loss": 0.0026,
       "step": 4310
     },
     {
       "epoch": 4.05,
-      "grad_norm": 0.34390100836753845,
+      "grad_norm": 2.668574810028076,
       "learning_rate": 1.049638055842813e-05,
-      "loss": 0.063,
+      "loss": 0.0706,
       "step": 4320
     },
     {
       "epoch": 4.06,
-      "grad_norm": 0.08080201596021652,
+      "grad_norm": 0.07278812676668167,
       "learning_rate": 1.0392967942088936e-05,
-      "loss": 0.0625,
+      "loss": 0.0751,
       "step": 4330
     },
     {
       "epoch": 4.07,
-      "grad_norm": 0.07401977479457855,
+      "grad_norm": 0.062381431460380554,
       "learning_rate": 1.0289555325749741e-05,
-      "loss": 0.1792,
+      "loss": 0.0459,
       "step": 4340
     },
     {
       "epoch": 4.08,
-      "grad_norm": 0.02694222889840603,
+      "grad_norm": 0.07199248671531677,
       "learning_rate": 1.0186142709410548e-05,
-      "loss": 0.064,
+      "loss": 0.0763,
       "step": 4350
     },
     {
       "epoch": 4.09,
-      "grad_norm": 5.257876396179199,
+      "grad_norm": 0.04358450695872307,
       "learning_rate": 1.0082730093071355e-05,
-      "loss": 0.094,
+      "loss": 0.0025,
       "step": 4360
     },
     {
       "epoch": 4.1,
-      "grad_norm": 0.13338564336299896,
+      "grad_norm": 1.5874137878417969,
       "learning_rate": 9.979317476732161e-06,
-      "loss": 0.0827,
+      "loss": 0.0025,
       "step": 4370
     },
     {
       "epoch": 4.1,
-      "grad_norm": 0.04957311972975731,
+      "grad_norm": 0.05725986883044243,
       "learning_rate": 9.875904860392968e-06,
-      "loss": 0.224,
+      "loss": 0.0853,
       "step": 4380
     },
     {
       "epoch": 4.11,
-      "grad_norm": 0.3861539363861084,
+      "grad_norm": 2.704402446746826,
       "learning_rate": 9.772492244053775e-06,
-      "loss": 0.0612,
+      "loss": 0.1137,
       "step": 4390
     },
     {
       "epoch": 4.12,
-      "grad_norm": 0.1312892585992813,
+      "grad_norm": 0.05556065961718559,
       "learning_rate": 9.669079627714582e-06,
-      "loss": 0.1828,
+      "loss": 0.0309,
       "step": 4400
     },
     {
       "epoch": 4.13,
-      "grad_norm": 0.11486046761274338,
+      "grad_norm": 0.041595447808504105,
       "learning_rate": 9.565667011375389e-06,
-      "loss": 0.0034,
+      "loss": 0.0025,
       "step": 4410
     },
     {
       "epoch": 4.14,
-      "grad_norm": 0.09139719605445862,
+      "grad_norm": 0.033831994980573654,
       "learning_rate": 9.462254395036195e-06,
-      "loss": 0.079,
+      "loss": 0.002,
       "step": 4420
     },
     {
       "epoch": 4.15,
-      "grad_norm": 0.033867135643959045,
+      "grad_norm": 0.046929676085710526,
       "learning_rate": 9.358841778697002e-06,
-      "loss": 0.0016,
+      "loss": 0.0023,
       "step": 4430
     },
     {
       "epoch": 4.16,
-      "grad_norm": 0.30136638879776,
+      "grad_norm": 0.04889731854200363,
       "learning_rate": 9.255429162357807e-06,
-      "loss": 0.0799,
+      "loss": 0.1443,
       "step": 4440
     },
     {
       "epoch": 4.17,
-      "grad_norm": 0.03786735609173775,
+      "grad_norm": 0.04792765527963638,
       "learning_rate": 9.152016546018614e-06,
-      "loss": 0.2652,
+      "loss": 0.084,
       "step": 4450
     },
     {
       "epoch": 4.18,
-      "grad_norm": 0.047059349715709686,
+      "grad_norm": 0.05229383334517479,
       "learning_rate": 9.04860392967942e-06,
-      "loss": 0.0014,
+      "loss": 0.0081,
       "step": 4460
     },
     {
       "epoch": 4.19,
-      "grad_norm": 0.054901909083127975,
+      "grad_norm": 0.0474727489054203,
       "learning_rate": 8.945191313340227e-06,
-      "loss": 0.0509,
+      "loss": 0.0022,
       "step": 4470
     },
     {
       "epoch": 4.2,
-      "grad_norm": 227.820556640625,
+      "grad_norm": 0.041655492037534714,
       "learning_rate": 8.841778697001036e-06,
-      "loss": 0.0222,
+      "loss": 0.0021,
       "step": 4480
     },
     {
       "epoch": 4.21,
-      "grad_norm": 0.043635886162519455,
+      "grad_norm": 0.06119935214519501,
       "learning_rate": 8.738366080661841e-06,
-      "loss": 0.0011,
+      "loss": 0.0021,
       "step": 4490
     },
     {
       "epoch": 4.22,
-      "grad_norm": 0.0816226452589035,
+      "grad_norm": 0.05464434251189232,
       "learning_rate": 8.634953464322648e-06,
-      "loss": 0.0873,
+      "loss": 0.0717,
       "step": 4500
     },
     {
       "epoch": 4.23,
-      "grad_norm": 0.07973457872867584,
+      "grad_norm": 0.057566653937101364,
       "learning_rate": 8.531540847983454e-06,
-      "loss": 0.0016,
+      "loss": 0.017,
       "step": 4510
     },
     {
       "epoch": 4.24,
-      "grad_norm": 0.03563067317008972,
+      "grad_norm": 0.04644325375556946,
       "learning_rate": 8.428128231644261e-06,
-      "loss": 0.0758,
+      "loss": 0.0019,
       "step": 4520
     },
     {
       "epoch": 4.25,
-      "grad_norm": 0.09928788989782333,
+      "grad_norm": 0.046763256192207336,
       "learning_rate": 8.324715615305068e-06,
-      "loss": 0.0595,
+      "loss": 0.0019,
       "step": 4530
     },
     {
       "epoch": 4.25,
-      "grad_norm": 0.027137843891978264,
+      "grad_norm": 0.030382856726646423,
       "learning_rate": 8.221302998965875e-06,
-      "loss": 0.0263,
+      "loss": 0.0053,
       "step": 4540
     },
     {
       "epoch": 4.26,
-      "grad_norm": 0.021461045369505882,
+      "grad_norm": 0.03069426119327545,
       "learning_rate": 8.11789038262668e-06,
-      "loss": 0.0635,
+      "loss": 0.0097,
       "step": 4550
     },
     {
       "epoch": 4.27,
-      "grad_norm": 0.026392286643385887,
+      "grad_norm": 0.03841962292790413,
       "learning_rate": 8.014477766287486e-06,
-      "loss": 0.2159,
+      "loss": 0.1735,
       "step": 4560
     },
     {
       "epoch": 4.28,
-      "grad_norm": 15.914426803588867,
+      "grad_norm": 0.05885951593518257,
       "learning_rate": 7.911065149948293e-06,
-      "loss": 0.0597,
+      "loss": 0.0048,
       "step": 4570
     },
     {
       "epoch": 4.29,
-      "grad_norm": 0.028651505708694458,
+      "grad_norm": 0.032764993607997894,
       "learning_rate": 7.807652533609102e-06,
-      "loss": 0.0014,
+      "loss": 0.0016,
       "step": 4580
     },
     {
       "epoch": 4.3,
-      "grad_norm": 0.07976741343736649,
+      "grad_norm": 0.033063892275094986,
       "learning_rate": 7.704239917269908e-06,
-      "loss": 0.0019,
+      "loss": 0.0015,
       "step": 4590
     },
     {
       "epoch": 4.31,
-      "grad_norm": 0.7245219945907593,
+      "grad_norm": 121.3236083984375,
       "learning_rate": 7.600827300930714e-06,
-      "loss": 0.1196,
+      "loss": 0.1054,
       "step": 4600
     },
     {
       "epoch": 4.32,
-      "grad_norm": 3.421379327774048,
+      "grad_norm": 4.730100631713867,
       "learning_rate": 7.497414684591521e-06,
-      "loss": 0.1707,
+      "loss": 0.0787,
       "step": 4610
     },
     {
       "epoch": 4.33,
-      "grad_norm": 54.59723663330078,
+      "grad_norm": 69.30009460449219,
       "learning_rate": 7.394002068252327e-06,
-      "loss": 0.1257,
+      "loss": 0.0041,
       "step": 4620
     },
     {
       "epoch": 4.34,
-      "grad_norm": 0.02159339189529419,
+      "grad_norm": 0.03414592146873474,
       "learning_rate": 7.290589451913134e-06,
-      "loss": 0.0011,
+      "loss": 0.0349,
       "step": 4630
     },
     {
       "epoch": 4.35,
-      "grad_norm": 0.02801990881562233,
+      "grad_norm": 0.06496775895357132,
       "learning_rate": 7.1871768355739405e-06,
-      "loss": 0.0013,
+      "loss": 0.0017,
       "step": 4640
     },
     {
       "epoch": 4.36,
-      "grad_norm": 0.12835495173931122,
+      "grad_norm": 0.03325049951672554,
       "learning_rate": 7.0837642192347465e-06,
-      "loss": 0.1343,
+      "loss": 0.0014,
       "step": 4650
     },
     {
       "epoch": 4.37,
-      "grad_norm": 0.06059630960226059,
+      "grad_norm": 0.026971573010087013,
       "learning_rate": 6.980351602895553e-06,
-      "loss": 0.0029,
+      "loss": 0.0014,
       "step": 4660
     },
     {
       "epoch": 4.38,
-      "grad_norm": 0.021486638113856316,
+      "grad_norm": 0.028269225731492043,
       "learning_rate": 6.87693898655636e-06,
-      "loss": 0.0872,
+      "loss": 0.0918,
       "step": 4670
     },
     {
       "epoch": 4.39,
-      "grad_norm": 0.024658987298607826,
+      "grad_norm": 0.026689806953072548,
       "learning_rate": 6.773526370217166e-06,
-      "loss": 0.0013,
+      "loss": 0.0909,
       "step": 4680
     },
     {
       "epoch": 4.4,
-      "grad_norm": 0.020257245749235153,
+      "grad_norm": 0.03431650996208191,
       "learning_rate": 6.670113753877974e-06,
-      "loss": 0.0012,
+      "loss": 0.0015,
       "step": 4690
     },
     {
       "epoch": 4.4,
-      "grad_norm": 0.022628970444202423,
+      "grad_norm": 0.031223468482494354,
       "learning_rate": 6.56670113753878e-06,
-      "loss": 0.0879,
+      "loss": 0.0015,
       "step": 4700
     },
     {
       "epoch": 4.41,
-      "grad_norm": 0.17840439081192017,
+      "grad_norm": 0.03299291059374809,
       "learning_rate": 6.463288521199587e-06,
-      "loss": 0.1423,
+      "loss": 0.0013,
       "step": 4710
     },
     {
       "epoch": 4.42,
-      "grad_norm": 248.1060333251953,
+      "grad_norm": 132.70225524902344,
       "learning_rate": 6.359875904860394e-06,
-      "loss": 0.1144,
+      "loss": 0.2171,
       "step": 4720
     },
     {
       "epoch": 4.43,
-      "grad_norm": 3.23333740234375,
+      "grad_norm": 2.8790526390075684,
       "learning_rate": 6.2564632885212e-06,
-      "loss": 0.0923,
+      "loss": 0.0792,
       "step": 4730
     },
     {
       "epoch": 4.44,
-      "grad_norm": 0.022127114236354828,
+      "grad_norm": 0.03240809589624405,
       "learning_rate": 6.153050672182006e-06,
-      "loss": 0.1237,
+      "loss": 0.0014,
       "step": 4740
     },
     {
       "epoch": 4.45,
-      "grad_norm": 0.03299199789762497,
+      "grad_norm": 0.03706195205450058,
       "learning_rate": 6.049638055842813e-06,
-      "loss": 0.0742,
+      "loss": 0.1544,
       "step": 4750
     },
     {
       "epoch": 4.46,
-      "grad_norm": 0.041477978229522705,
+      "grad_norm": 446.900634765625,
       "learning_rate": 5.94622543950362e-06,
-      "loss": 0.0012,
+      "loss": 0.0677,
       "step": 4760
     },
     {
       "epoch": 4.47,
-      "grad_norm": 0.28576919436454773,
+      "grad_norm": 3.300240993499756,
       "learning_rate": 5.842812823164427e-06,
-      "loss": 0.0737,
+      "loss": 0.2351,
       "step": 4770
     },
     {
       "epoch": 4.48,
-      "grad_norm": 0.03505215421319008,
+      "grad_norm": 0.03419492393732071,
       "learning_rate": 5.739400206825233e-06,
-      "loss": 0.168,
+      "loss": 0.0015,
       "step": 4780
     },
     {
       "epoch": 4.49,
-      "grad_norm": 0.023467887192964554,
+      "grad_norm": 0.03599884361028671,
       "learning_rate": 5.635987590486039e-06,
-      "loss": 0.0011,
+      "loss": 0.0019,
       "step": 4790
     },
     {
       "epoch": 4.5,
-      "grad_norm": 0.0256861113011837,
+      "grad_norm": 0.028266843408346176,
       "learning_rate": 5.532574974146846e-06,
-      "loss": 0.1276,
+      "loss": 0.0016,
       "step": 4800
     },
     {
       "epoch": 4.51,
-      "grad_norm": 89.85589599609375,
+      "grad_norm": 0.11148073524236679,
       "learning_rate": 5.429162357807653e-06,
-      "loss": 0.0627,
+      "loss": 0.0027,
       "step": 4810
     },
     {
       "epoch": 4.52,
-      "grad_norm": 0.019946351647377014,
+      "grad_norm": 0.03269073739647865,
       "learning_rate": 5.32574974146846e-06,
-      "loss": 0.0012,
+      "loss": 0.0315,
       "step": 4820
     },
     {
       "epoch": 4.53,
-      "grad_norm": 0.1361633539199829,
+      "grad_norm": 0.03238886967301369,
       "learning_rate": 5.222337125129266e-06,
-      "loss": 0.0013,
+      "loss": 0.0449,
       "step": 4830
     },
     {
       "epoch": 4.54,
-      "grad_norm": 0.06002867594361305,
+      "grad_norm": 0.0361521914601326,
       "learning_rate": 5.118924508790072e-06,
       "loss": 0.0014,
       "step": 4840
     },
     {
       "epoch": 4.55,
-      "grad_norm": 0.02299017459154129,
+      "grad_norm": 0.029334863647818565,
       "learning_rate": 5.015511892450879e-06,
-      "loss": 0.0895,
+      "loss": 0.1483,
       "step": 4850
     },
     {
       "epoch": 4.55,
-      "grad_norm": 0.01912473328411579,
+      "grad_norm": 0.03407660499215126,
       "learning_rate": 4.912099276111686e-06,
-      "loss": 0.0163,
+      "loss": 0.0015,
       "step": 4860
     },
     {
       "epoch": 4.56,
-      "grad_norm": 0.020831523463129997,
+      "grad_norm": 0.031936611980199814,
       "learning_rate": 4.8086866597724926e-06,
-      "loss": 0.001,
+      "loss": 0.0043,
       "step": 4870
     },
     {
       "epoch": 4.57,
-      "grad_norm": 0.017449194565415382,
+      "grad_norm": 0.024707814678549767,
       "learning_rate": 4.705274043433299e-06,
-      "loss": 0.0637,
+      "loss": 0.0309,
       "step": 4880
     },
     {
       "epoch": 4.58,
-      "grad_norm": 0.14713767170906067,
+      "grad_norm": 0.0379851795732975,
       "learning_rate": 4.601861427094105e-06,
-      "loss": 0.0035,
+      "loss": 0.0016,
       "step": 4890
     },
     {
       "epoch": 4.59,
-      "grad_norm": 3.178408145904541,
+      "grad_norm": 0.031074857339262962,
       "learning_rate": 4.498448810754912e-06,
-      "loss": 0.091,
+      "loss": 0.0021,
       "step": 4900
     },
     {
       "epoch": 4.6,
-      "grad_norm": 0.0492083765566349,
+      "grad_norm": 0.03219422325491905,
       "learning_rate": 4.39503619441572e-06,
-      "loss": 0.24,
+      "loss": 0.078,
       "step": 4910
     },
     {
       "epoch": 4.61,
-      "grad_norm": 0.01996445097029209,
+      "grad_norm": 0.02124185487627983,
       "learning_rate": 4.2916235780765255e-06,
-      "loss": 0.1126,
+      "loss": 0.0909,
       "step": 4920
     },
     {
       "epoch": 4.62,
-      "grad_norm": 0.021161267533898354,
+      "grad_norm": 0.028623027727007866,
       "learning_rate": 4.188210961737332e-06,
-      "loss": 0.0011,
+      "loss": 0.0022,
       "step": 4930
     },
     {
       "epoch": 4.63,
-      "grad_norm": 0.019828295335173607,
+      "grad_norm": 0.0264423917979002,
       "learning_rate": 4.084798345398139e-06,
-      "loss": 0.0025,
+      "loss": 0.0019,
       "step": 4940
     },
     {
       "epoch": 4.64,
-      "grad_norm": 0.021803775802254677,
+      "grad_norm": 0.033193424344062805,
       "learning_rate": 3.981385729058945e-06,
-      "loss": 0.1434,
+      "loss": 0.1499,
       "step": 4950
     },
     {
       "epoch": 4.65,
-      "grad_norm": 0.13735263049602509,
+      "grad_norm": 0.03795929625630379,
       "learning_rate": 3.8779731127197525e-06,
-      "loss": 0.1412,
+      "loss": 0.0799,
       "step": 4960
     },
     {
       "epoch": 4.66,
-      "grad_norm": 0.24444641172885895,
+      "grad_norm": 0.03771360218524933,
       "learning_rate": 3.774560496380559e-06,
-      "loss": 0.0046,
+      "loss": 0.0015,
       "step": 4970
     },
     {
       "epoch": 4.67,
-      "grad_norm": 0.023776857182383537,
+      "grad_norm": 0.03688567876815796,
       "learning_rate": 3.671147880041365e-06,
-      "loss": 0.1191,
+      "loss": 0.0028,
       "step": 4980
     },
     {
       "epoch": 4.68,
-      "grad_norm": 0.024539895355701447,
+      "grad_norm": 0.02699306234717369,
       "learning_rate": 3.5677352637021715e-06,
-      "loss": 0.1884,
+      "loss": 0.082,
       "step": 4990
     },
     {
       "epoch": 4.69,
-      "grad_norm": 0.01984153315424919,
+      "grad_norm": 0.03286897763609886,
       "learning_rate": 3.4643226473629783e-06,
-      "loss": 0.1128,
+      "loss": 0.0013,
       "step": 5000
     },
     {
       "epoch": 4.7,
-      "grad_norm": 2.897346258163452,
+      "grad_norm": 0.03281855955719948,
       "learning_rate": 3.3609100310237855e-06,
-      "loss": 0.0782,
+      "loss": 0.0016,
       "step": 5010
     },
     {
       "epoch": 4.7,
-      "grad_norm": 0.04367586970329285,
+      "grad_norm": 0.028866132721304893,
       "learning_rate": 3.257497414684592e-06,
-      "loss": 0.1729,
+      "loss": 0.2369,
       "step": 5020
     },
     {
       "epoch": 4.71,
-      "grad_norm": 0.022409643977880478,
+      "grad_norm": 0.031186798587441444,
       "learning_rate": 3.154084798345398e-06,
-      "loss": 0.0013,
+      "loss": 0.0014,
       "step": 5030
     },
     {
       "epoch": 4.72,
-      "grad_norm": 0.02230711840093136,
+      "grad_norm": 0.02863004244863987,
       "learning_rate": 3.050672182006205e-06,
-      "loss": 0.0605,
+      "loss": 0.0018,
       "step": 5040
     },
     {
       "epoch": 4.73,
-      "grad_norm": 0.08053728193044662,
+      "grad_norm": 0.030251089483499527,
       "learning_rate": 2.9472595656670117e-06,
-      "loss": 0.0581,
+      "loss": 0.0866,
       "step": 5050
     },
     {
       "epoch": 4.74,
-      "grad_norm": 0.382304847240448,
+      "grad_norm": 0.11614389717578888,
       "learning_rate": 2.843846949327818e-06,
-      "loss": 0.0396,
+      "loss": 0.1602,
       "step": 5060
     },
     {
       "epoch": 4.75,
-      "grad_norm": 0.0988391637802124,
+      "grad_norm": 0.02916896715760231,
       "learning_rate": 2.7404343329886247e-06,
-      "loss": 0.0291,
+      "loss": 0.002,
       "step": 5070
     },
     {
       "epoch": 4.76,
-      "grad_norm": 0.025470012798905373,
+      "grad_norm": 0.03606222942471504,
       "learning_rate": 2.6370217166494315e-06,
-      "loss": 0.0012,
+      "loss": 0.0014,
       "step": 5080
     },
     {
       "epoch": 4.77,
-      "grad_norm": 0.01994398981332779,
+      "grad_norm": 0.03456837683916092,
       "learning_rate": 2.533609100310238e-06,
-      "loss": 0.0021,
+      "loss": 0.0015,
       "step": 5090
     },
     {
       "epoch": 4.78,
-      "grad_norm": 0.062303327023983,
+      "grad_norm": 0.02772245742380619,
       "learning_rate": 2.4301964839710446e-06,
-      "loss": 0.1794,
+      "loss": 0.0022,
       "step": 5100
     },
     {
       "epoch": 4.79,
-      "grad_norm": 0.02839079312980175,
+      "grad_norm": 0.0416284054517746,
       "learning_rate": 2.3267838676318514e-06,
-      "loss": 0.0827,
+      "loss": 0.0014,
       "step": 5110
     },
     {
       "epoch": 4.8,
-      "grad_norm": 0.02093564346432686,
+      "grad_norm": 0.028300166130065918,
       "learning_rate": 2.2233712512926577e-06,
-      "loss": 0.1439,
+      "loss": 0.002,
       "step": 5120
     },
     {
       "epoch": 4.81,
-      "grad_norm": 0.024724125862121582,
+      "grad_norm": 0.03322126343846321,
       "learning_rate": 2.1199586349534644e-06,
-      "loss": 0.2626,
+      "loss": 0.2574,
       "step": 5130
     },
     {
       "epoch": 4.82,
-      "grad_norm": 0.025263186544179916,
+      "grad_norm": 0.025889359414577484,
       "learning_rate": 2.0165460186142708e-06,
-      "loss": 0.0011,
+      "loss": 0.0014,
       "step": 5140
     },
     {
       "epoch": 4.83,
-      "grad_norm": 0.026100702583789825,
+      "grad_norm": 0.037663087248802185,
       "learning_rate": 1.913133402275078e-06,
-      "loss": 0.0012,
+      "loss": 0.079,
       "step": 5150
     },
     {
       "epoch": 4.84,
-      "grad_norm": 0.42885711789131165,
+      "grad_norm": 0.024740027263760567,
       "learning_rate": 1.8097207859358843e-06,
-      "loss": 0.0014,
+      "loss": 0.102,
       "step": 5160
     },
     {
       "epoch": 4.85,
-      "grad_norm": 0.09349822252988815,
+      "grad_norm": 0.029313886538147926,
       "learning_rate": 1.7063081695966906e-06,
-      "loss": 0.0014,
+      "loss": 0.0966,
       "step": 5170
     },
     {
       "epoch": 4.85,
-      "grad_norm": 0.020912567153573036,
+      "grad_norm": 0.02809065394103527,
       "learning_rate": 1.6028955532574976e-06,
-      "loss": 0.2671,
+      "loss": 0.2161,
       "step": 5180
     },
     {
       "epoch": 4.86,
-      "grad_norm": 0.0201337318867445,
+      "grad_norm": 0.02252083458006382,
       "learning_rate": 1.499482936918304e-06,
-      "loss": 0.0653,
+      "loss": 0.0016,
       "step": 5190
     },
     {
       "epoch": 4.87,
-      "grad_norm": 0.12457222491502762,
+      "grad_norm": 0.030461903661489487,
       "learning_rate": 1.3960703205791107e-06,
-      "loss": 0.0855,
+      "loss": 0.1729,
       "step": 5200
     },
     {
       "epoch": 4.88,
-      "grad_norm": 0.06253860145807266,
+      "grad_norm": 0.032880086451768875,
       "learning_rate": 1.2926577042399172e-06,
-      "loss": 0.0576,
+      "loss": 0.0815,
       "step": 5210
     },
     {
       "epoch": 4.89,
-      "grad_norm": 0.033292561769485474,
+      "grad_norm": 0.033105768263339996,
       "learning_rate": 1.189245087900724e-06,
-      "loss": 0.1763,
+      "loss": 0.0186,
       "step": 5220
     },
     {
       "epoch": 4.9,
-      "grad_norm": 5.004137992858887,
+      "grad_norm": 0.03529038280248642,
       "learning_rate": 1.0858324715615305e-06,
-      "loss": 0.0863,
+      "loss": 0.1081,
       "step": 5230
     },
     {
       "epoch": 4.91,
-      "grad_norm": 10.605862617492676,
+      "grad_norm": 3.352219343185425,
       "learning_rate": 9.82419855222337e-07,
-      "loss": 0.1183,
+      "loss": 0.0897,
       "step": 5240
     },
     {
       "epoch": 4.92,
-      "grad_norm": 0.02466098591685295,
+      "grad_norm": 0.035365719348192215,
       "learning_rate": 8.790072388831437e-07,
-      "loss": 0.0013,
+      "loss": 0.0021,
       "step": 5250
     },
     {
       "epoch": 4.93,
-      "grad_norm": 47.83183288574219,
+      "grad_norm": 6.3308587074279785,
       "learning_rate": 7.755946225439504e-07,
-      "loss": 0.1172,
+      "loss": 0.092,
       "step": 5260
     },
     {
       "epoch": 4.94,
-      "grad_norm": 0.05490906536579132,
+      "grad_norm": 0.03557182475924492,
       "learning_rate": 6.72182006204757e-07,
-      "loss": 0.0862,
+      "loss": 0.0022,
       "step": 5270
     },
     {
       "epoch": 4.95,
-      "grad_norm": 0.09379950165748596,
+      "grad_norm": 0.05220606178045273,
       "learning_rate": 5.687693898655636e-07,
-      "loss": 0.0013,
+      "loss": 0.0743,
       "step": 5280
     },
     {
       "epoch": 4.96,
-      "grad_norm": 0.023985665291547775,
+      "grad_norm": 248.59735107421875,
       "learning_rate": 4.6535677352637023e-07,
-      "loss": 0.0016,
+      "loss": 0.0558,
       "step": 5290
     },
     {
       "epoch": 4.97,
-      "grad_norm": 0.18879929184913635,
+      "grad_norm": 0.029809845611453056,
       "learning_rate": 3.619441571871769e-07,
-      "loss": 0.0607,
+      "loss": 0.0015,
       "step": 5300
     },
     {
       "epoch": 4.98,
-      "grad_norm": 0.043404027819633484,
+      "grad_norm": 0.030335601419210434,
       "learning_rate": 2.585315408479835e-07,
-      "loss": 0.0933,
+      "loss": 0.0736,
       "step": 5310
     },
     {
       "epoch": 4.99,
-      "grad_norm": 10.754716873168945,
+      "grad_norm": 0.039527297019958496,
       "learning_rate": 1.5511892450879008e-07,
-      "loss": 0.0017,
+      "loss": 0.137,
       "step": 5320
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.027110379189252853,
+      "grad_norm": 0.023937204852700233,
       "learning_rate": 5.1706308169596694e-08,
-      "loss": 0.0889,
+      "loss": 0.0819,
       "step": 5330
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8761726078799249,
-      "eval_f1": 0.8735632183908046,
-      "eval_loss": 0.7370651364326477,
-      "eval_precision": 0.890625,
-      "eval_recall": 0.8571428571428571,
-      "eval_runtime": 1.5139,
-      "eval_samples_per_second": 352.061,
-      "eval_steps_per_second": 44.255,
+      "eval_accuracy": 0.8667917448405253,
+      "eval_f1": 0.8672897196261682,
+      "eval_loss": 0.7846159934997559,
+      "eval_precision": 0.862453531598513,
+      "eval_recall": 0.8721804511278195,
+      "eval_runtime": 1.4269,
+      "eval_samples_per_second": 373.542,
+      "eval_steps_per_second": 46.956,
       "step": 5335
     }
   ],