Training in progress, epoch 4, checkpoint

checkpoint-2220/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1751a91e15485c2f6d57e2040398354fda8cbd997e63ebde3da4aff8a8f08c3c
+oid sha256:3fde9e031d986a8531c58c641f727b4fccd9dba34b250e11255973129990348d
 size 437958648

checkpoint-2220/trainer_state.json

@@ -8,1613 +8,1618 @@
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
-    {
-      "epoch": 0,
-      "eval_f1": 0.6562770562770562,
-      "eval_loss": 0.7476533651351929,
-      "eval_precision": 0.48976954555244456,
-      "eval_recall": 0.9943156974202011,
-      "eval_runtime": 16.4688,
-      "eval_samples_per_second": 283.749,
-      "eval_steps_per_second": 35.522,
-      "step": 0
-    },
     {
       "epoch": 0.018026137899954935,
-      "grad_norm": 14.20566463470459,
+      "grad_norm": 13.032843589782715,
       "learning_rate": 4.981949458483755e-05,
-      "loss": 0.6876,
+      "loss": 0.6633,
       "step": 10
     },
     {
       "epoch": 0.03605227579990987,
-      "grad_norm": 27.79998016357422,
+      "grad_norm": 35.63669204711914,
       "learning_rate": 4.963898916967509e-05,
-      "loss": 0.5988,
+      "loss": 0.5651,
       "step": 20
     },
     {
       "epoch": 0.054078413699864804,
-      "grad_norm": 39.63682556152344,
+      "grad_norm": 16.95796012878418,
       "learning_rate": 4.945848375451264e-05,
-      "loss": 0.5677,
+      "loss": 0.5359,
       "step": 30
     },
     {
       "epoch": 0.07210455159981974,
-      "grad_norm": 38.49867630004883,
+      "grad_norm": 14.045697212219238,
       "learning_rate": 4.927797833935018e-05,
-      "loss": 0.5922,
+      "loss": 0.5421,
       "step": 40
     },
     {
       "epoch": 0.09013068949977468,
-      "grad_norm": 52.28614044189453,
+      "grad_norm": 43.34794235229492,
       "learning_rate": 4.909747292418773e-05,
-      "loss": 0.5506,
+      "loss": 0.5222,
       "step": 50
     },
     {
       "epoch": 0.10815682739972961,
-      "grad_norm": 29.180490493774414,
+      "grad_norm": 32.69561767578125,
       "learning_rate": 4.891696750902527e-05,
-      "loss": 0.5173,
+      "loss": 0.4589,
       "step": 60
     },
     {
       "epoch": 0.12618296529968454,
-      "grad_norm": 33.55237579345703,
+      "grad_norm": 29.39051055908203,
       "learning_rate": 4.873646209386282e-05,
-      "loss": 0.4931,
+      "loss": 0.4858,
       "step": 70
     },
     {
       "epoch": 0.14420910319963948,
-      "grad_norm": 32.063228607177734,
+      "grad_norm": 30.52906608581543,
       "learning_rate": 4.855595667870036e-05,
-      "loss": 0.483,
+      "loss": 0.4418,
       "step": 80
     },
     {
       "epoch": 0.16223524109959442,
-      "grad_norm": 24.234926223754883,
+      "grad_norm": 15.103797912597656,
       "learning_rate": 4.837545126353791e-05,
-      "loss": 0.47,
+      "loss": 0.4713,
       "step": 90
     },
     {
       "epoch": 0.18026137899954936,
-      "grad_norm": 33.66775894165039,
+      "grad_norm": 40.57420349121094,
       "learning_rate": 4.819494584837546e-05,
-      "loss": 0.4564,
+      "loss": 0.468,
       "step": 100
     },
     {
       "epoch": 0.19828751689950427,
-      "grad_norm": 53.322574615478516,
+      "grad_norm": 37.67963409423828,
       "learning_rate": 4.8014440433213e-05,
-      "loss": 0.5083,
+      "loss": 0.492,
       "step": 110
     },
     {
       "epoch": 0.21631365479945922,
-      "grad_norm": 29.443470001220703,
+      "grad_norm": 30.945823669433594,
       "learning_rate": 4.783393501805055e-05,
-      "loss": 0.463,
+      "loss": 0.4633,
       "step": 120
     },
     {
       "epoch": 0.23433979269941416,
-      "grad_norm": 33.13877487182617,
+      "grad_norm": 40.709590911865234,
       "learning_rate": 4.765342960288809e-05,
-      "loss": 0.4626,
+      "loss": 0.4392,
       "step": 130
     },
     {
       "epoch": 0.25236593059936907,
-      "grad_norm": 25.509536743164062,
+      "grad_norm": 19.698150634765625,
       "learning_rate": 4.747292418772563e-05,
-      "loss": 0.5113,
+      "loss": 0.4707,
       "step": 140
     },
     {
       "epoch": 0.270392068499324,
-      "grad_norm": 21.015031814575195,
+      "grad_norm": 21.289947509765625,
       "learning_rate": 4.7292418772563177e-05,
-      "loss": 0.4611,
+      "loss": 0.4576,
       "step": 150
     },
     {
       "epoch": 0.28841820639927895,
-      "grad_norm": 28.079387664794922,
+      "grad_norm": 26.911935806274414,
       "learning_rate": 4.711191335740072e-05,
-      "loss": 0.4783,
+      "loss": 0.4583,
       "step": 160
     },
     {
       "epoch": 0.3064443442992339,
-      "grad_norm": 19.978599548339844,
+      "grad_norm": 19.308853149414062,
       "learning_rate": 4.693140794223827e-05,
-      "loss": 0.5026,
+      "loss": 0.4937,
       "step": 170
     },
     {
       "epoch": 0.32447048219918884,
-      "grad_norm": 29.316265106201172,
+      "grad_norm": 34.61503219604492,
       "learning_rate": 4.675090252707581e-05,
-      "loss": 0.4906,
+      "loss": 0.4836,
       "step": 180
     },
     {
       "epoch": 0.3424966200991438,
-      "grad_norm": 18.79282569885254,
+      "grad_norm": 16.551424026489258,
       "learning_rate": 4.657039711191336e-05,
-      "loss": 0.5016,
+      "loss": 0.4919,
       "step": 190
     },
     {
       "epoch": 0.3605227579990987,
-      "grad_norm": 18.66801643371582,
+      "grad_norm": 16.011377334594727,
       "learning_rate": 4.63898916967509e-05,
-      "loss": 0.4502,
+      "loss": 0.4518,
       "step": 200
     },
     {
       "epoch": 0.3785488958990536,
-      "grad_norm": 27.929332733154297,
+      "grad_norm": 22.71525764465332,
       "learning_rate": 4.620938628158845e-05,
-      "loss": 0.4302,
+      "loss": 0.4221,
       "step": 210
     },
     {
       "epoch": 0.39657503379900855,
-      "grad_norm": 20.541879653930664,
+      "grad_norm": 22.2515811920166,
       "learning_rate": 4.602888086642599e-05,
-      "loss": 0.4496,
+      "loss": 0.4286,
       "step": 220
     },
     {
       "epoch": 0.4146011716989635,
-      "grad_norm": 19.034873962402344,
+      "grad_norm": 18.233482360839844,
       "learning_rate": 4.584837545126354e-05,
-      "loss": 0.4541,
+      "loss": 0.4442,
       "step": 230
     },
     {
       "epoch": 0.43262730959891843,
-      "grad_norm": 32.077945709228516,
+      "grad_norm": 33.190948486328125,
       "learning_rate": 4.566787003610109e-05,
-      "loss": 0.4562,
+      "loss": 0.4584,
       "step": 240
     },
     {
       "epoch": 0.45065344749887337,
-      "grad_norm": 25.475997924804688,
+      "grad_norm": 29.945205688476562,
       "learning_rate": 4.548736462093863e-05,
-      "loss": 0.4456,
+      "loss": 0.4339,
       "step": 250
     },
     {
       "epoch": 0.4686795853988283,
-      "grad_norm": 17.132720947265625,
+      "grad_norm": 17.004419326782227,
       "learning_rate": 4.530685920577618e-05,
-      "loss": 0.4629,
+      "loss": 0.4504,
       "step": 260
     },
     {
       "epoch": 0.48670572329878325,
-      "grad_norm": 22.313941955566406,
+      "grad_norm": 23.36432647705078,
       "learning_rate": 4.5126353790613716e-05,
-      "loss": 0.4444,
+      "loss": 0.4405,
       "step": 270
     },
     {
       "epoch": 0.5047318611987381,
-      "grad_norm": 22.373477935791016,
+      "grad_norm": 29.156505584716797,
       "learning_rate": 4.494584837545127e-05,
-      "loss": 0.4404,
+      "loss": 0.4319,
       "step": 280
     },
     {
       "epoch": 0.5227579990986931,
-      "grad_norm": 16.72430992126465,
+      "grad_norm": 16.330259323120117,
       "learning_rate": 4.4765342960288806e-05,
-      "loss": 0.4611,
+      "loss": 0.453,
       "step": 290
     },
     {
       "epoch": 0.540784136998648,
-      "grad_norm": 27.037927627563477,
+      "grad_norm": 21.66246223449707,
       "learning_rate": 4.458483754512636e-05,
-      "loss": 0.4272,
+      "loss": 0.416,
       "step": 300
     },
     {
       "epoch": 0.558810274898603,
-      "grad_norm": 31.637985229492188,
+      "grad_norm": 22.177268981933594,
       "learning_rate": 4.44043321299639e-05,
-      "loss": 0.448,
+      "loss": 0.4456,
       "step": 310
     },
     {
       "epoch": 0.5768364127985579,
-      "grad_norm": 35.010459899902344,
+      "grad_norm": 38.450069427490234,
       "learning_rate": 4.422382671480145e-05,
-      "loss": 0.4215,
+      "loss": 0.4216,
       "step": 320
     },
     {
       "epoch": 0.5948625506985128,
-      "grad_norm": 26.25156021118164,
+      "grad_norm": 28.213655471801758,
       "learning_rate": 4.404332129963899e-05,
-      "loss": 0.4396,
+      "loss": 0.4392,
       "step": 330
     },
     {
       "epoch": 0.6128886885984678,
-      "grad_norm": 19.4619083404541,
+      "grad_norm": 19.354459762573242,
       "learning_rate": 4.386281588447654e-05,
-      "loss": 0.4378,
+      "loss": 0.4303,
       "step": 340
     },
     {
       "epoch": 0.6309148264984227,
-      "grad_norm": 19.16460609436035,
+      "grad_norm": 20.212663650512695,
       "learning_rate": 4.368231046931408e-05,
-      "loss": 0.4366,
+      "loss": 0.4426,
       "step": 350
     },
     {
       "epoch": 0.6489409643983777,
-      "grad_norm": 30.745384216308594,
+      "grad_norm": 34.96459197998047,
       "learning_rate": 4.350180505415163e-05,
-      "loss": 0.4062,
+      "loss": 0.4146,
       "step": 360
     },
     {
       "epoch": 0.6669671022983326,
-      "grad_norm": 16.873245239257812,
+      "grad_norm": 17.65025520324707,
       "learning_rate": 4.332129963898917e-05,
-      "loss": 0.439,
+      "loss": 0.4506,
       "step": 370
     },
     {
       "epoch": 0.6849932401982876,
-      "grad_norm": 25.771486282348633,
+      "grad_norm": 18.796554565429688,
       "learning_rate": 4.314079422382672e-05,
-      "loss": 0.4053,
+      "loss": 0.3765,
       "step": 380
     },
     {
       "epoch": 0.7030193780982424,
-      "grad_norm": 14.48513412475586,
+      "grad_norm": 21.690357208251953,
       "learning_rate": 4.296028880866426e-05,
-      "loss": 0.4164,
+      "loss": 0.4158,
       "step": 390
     },
     {
       "epoch": 0.7210455159981974,
-      "grad_norm": 34.993465423583984,
+      "grad_norm": 39.8782844543457,
       "learning_rate": 4.277978339350181e-05,
-      "loss": 0.4438,
+      "loss": 0.4602,
       "step": 400
     },
     {
       "epoch": 0.7390716538981523,
-      "grad_norm": 22.576608657836914,
+      "grad_norm": 26.3798885345459,
       "learning_rate": 4.259927797833935e-05,
-      "loss": 0.4518,
+      "loss": 0.4512,
       "step": 410
     },
     {
       "epoch": 0.7570977917981072,
-      "grad_norm": 19.088132858276367,
+      "grad_norm": 18.073617935180664,
       "learning_rate": 4.24187725631769e-05,
-      "loss": 0.4495,
+      "loss": 0.4347,
       "step": 420
     },
     {
       "epoch": 0.7751239296980622,
-      "grad_norm": 35.2998046875,
+      "grad_norm": 30.09633445739746,
       "learning_rate": 4.223826714801444e-05,
-      "loss": 0.3897,
+      "loss": 0.3858,
       "step": 430
     },
     {
       "epoch": 0.7931500675980171,
-      "grad_norm": 20.395572662353516,
+      "grad_norm": 17.02672576904297,
       "learning_rate": 4.205776173285199e-05,
-      "loss": 0.4818,
+      "loss": 0.4815,
       "step": 440
     },
     {
       "epoch": 0.8111762054979721,
-      "grad_norm": 21.174386978149414,
+      "grad_norm": 19.149808883666992,
       "learning_rate": 4.187725631768953e-05,
-      "loss": 0.4103,
+      "loss": 0.4196,
       "step": 450
     },
     {
       "epoch": 0.829202343397927,
-      "grad_norm": 32.97265625,
+      "grad_norm": 28.140705108642578,
       "learning_rate": 4.169675090252708e-05,
-      "loss": 0.4036,
+      "loss": 0.4085,
       "step": 460
     },
     {
       "epoch": 0.847228481297882,
-      "grad_norm": 23.78690528869629,
+      "grad_norm": 21.819814682006836,
       "learning_rate": 4.151624548736462e-05,
-      "loss": 0.4079,
+      "loss": 0.3986,
       "step": 470
     },
     {
       "epoch": 0.8652546191978369,
-      "grad_norm": 17.087247848510742,
+      "grad_norm": 15.050581932067871,
       "learning_rate": 4.1335740072202167e-05,
-      "loss": 0.4371,
+      "loss": 0.4267,
       "step": 480
     },
     {
       "epoch": 0.8832807570977917,
-      "grad_norm": 13.810160636901855,
+      "grad_norm": 14.929478645324707,
       "learning_rate": 4.115523465703972e-05,
-      "loss": 0.4249,
+      "loss": 0.4315,
       "step": 490
     },
     {
       "epoch": 0.9013068949977467,
-      "grad_norm": 20.197996139526367,
+      "grad_norm": 20.087432861328125,
       "learning_rate": 4.0974729241877256e-05,
-      "loss": 0.3991,
+      "loss": 0.3886,
       "step": 500
     },
     {
       "epoch": 0.9193330328977016,
-      "grad_norm": 33.1153564453125,
+      "grad_norm": 32.96128463745117,
       "learning_rate": 4.079422382671481e-05,
-      "loss": 0.3994,
+      "loss": 0.4046,
       "step": 510
     },
     {
       "epoch": 0.9373591707976566,
-      "grad_norm": 18.20081901550293,
+      "grad_norm": 19.673940658569336,
       "learning_rate": 4.0613718411552346e-05,
-      "loss": 0.4176,
+      "loss": 0.4127,
       "step": 520
     },
     {
       "epoch": 0.9553853086976115,
-      "grad_norm": 20.09193992614746,
+      "grad_norm": 16.12790298461914,
       "learning_rate": 4.043321299638989e-05,
-      "loss": 0.4297,
+      "loss": 0.4227,
       "step": 530
     },
     {
       "epoch": 0.9734114465975665,
-      "grad_norm": 15.939692497253418,
+      "grad_norm": 20.4881649017334,
       "learning_rate": 4.0252707581227436e-05,
-      "loss": 0.4309,
+      "loss": 0.4223,
       "step": 540
     },
     {
       "epoch": 0.9914375844975214,
-      "grad_norm": 24.811601638793945,
+      "grad_norm": 17.96915054321289,
       "learning_rate": 4.007220216606498e-05,
-      "loss": 0.3908,
+      "loss": 0.395,
       "step": 550
     },
     {
       "epoch": 1.0,
-      "eval_f1": 0.8367177786987153,
-      "eval_loss": 0.37827184796333313,
-      "eval_precision": 0.7951949586451359,
-      "eval_recall": 0.8828159160472234,
-      "eval_runtime": 15.9582,
-      "eval_samples_per_second": 292.827,
-      "eval_steps_per_second": 36.658,
+      "eval_f1": 0.8363481591828048,
+      "eval_fn": 322,
+      "eval_fp": 447,
+      "eval_loss": 0.372616171836853,
+      "eval_precision": 0.8146766169154229,
+      "eval_recall": 0.8592041976388282,
+      "eval_runtime": 12.0589,
+      "eval_samples_per_second": 387.513,
+      "eval_steps_per_second": 48.512,
+      "eval_tn": 1939,
+      "eval_tp": 1965,
       "step": 555
     },
     {
       "epoch": 1.0090130689499774,
-      "grad_norm": 16.960424423217773,
+      "grad_norm": 16.42650604248047,
       "learning_rate": 3.989169675090253e-05,
-      "loss": 0.3719,
+      "loss": 0.3725,
       "step": 560
     },
     {
       "epoch": 1.0270392068499323,
-      "grad_norm": 15.816173553466797,
+      "grad_norm": 14.740382194519043,
       "learning_rate": 3.971119133574007e-05,
-      "loss": 0.3238,
+      "loss": 0.3028,
       "step": 570
     },
     {
       "epoch": 1.0450653447498874,
-      "grad_norm": 19.375043869018555,
+      "grad_norm": 21.919588088989258,
       "learning_rate": 3.953068592057762e-05,
-      "loss": 0.3516,
+      "loss": 0.3371,
       "step": 580
     },
     {
       "epoch": 1.0630914826498423,
-      "grad_norm": 36.740943908691406,
+      "grad_norm": 23.157644271850586,
       "learning_rate": 3.935018050541516e-05,
-      "loss": 0.3329,
+      "loss": 0.3098,
       "step": 590
     },
     {
       "epoch": 1.0811176205497972,
-      "grad_norm": 23.23931884765625,
+      "grad_norm": 19.175317764282227,
       "learning_rate": 3.916967509025271e-05,
-      "loss": 0.345,
+      "loss": 0.3394,
       "step": 600
     },
     {
       "epoch": 1.099143758449752,
-      "grad_norm": 23.009553909301758,
+      "grad_norm": 24.65229034423828,
       "learning_rate": 3.898916967509025e-05,
-      "loss": 0.3054,
+      "loss": 0.3204,
       "step": 610
     },
     {
       "epoch": 1.117169896349707,
-      "grad_norm": 18.865903854370117,
+      "grad_norm": 20.54008674621582,
       "learning_rate": 3.88086642599278e-05,
-      "loss": 0.3236,
+      "loss": 0.2994,
       "step": 620
     },
     {
       "epoch": 1.135196034249662,
-      "grad_norm": 21.043642044067383,
+      "grad_norm": 24.291664123535156,
       "learning_rate": 3.862815884476535e-05,
-      "loss": 0.3092,
+      "loss": 0.3038,
       "step": 630
     },
     {
       "epoch": 1.153222172149617,
-      "grad_norm": 22.50505256652832,
+      "grad_norm": 21.287853240966797,
       "learning_rate": 3.844765342960289e-05,
-      "loss": 0.3274,
+      "loss": 0.3118,
       "step": 640
     },
     {
       "epoch": 1.1712483100495719,
-      "grad_norm": 20.4163818359375,
+      "grad_norm": 19.468006134033203,
       "learning_rate": 3.826714801444044e-05,
-      "loss": 0.2459,
+      "loss": 0.2484,
       "step": 650
     },
     {
       "epoch": 1.1892744479495267,
-      "grad_norm": 21.210535049438477,
+      "grad_norm": 26.11173439025879,
       "learning_rate": 3.8086642599277976e-05,
-      "loss": 0.3365,
+      "loss": 0.308,
       "step": 660
     },
     {
       "epoch": 1.2073005858494819,
-      "grad_norm": 21.084487915039062,
+      "grad_norm": 24.100196838378906,
       "learning_rate": 3.790613718411553e-05,
-      "loss": 0.3242,
+      "loss": 0.3032,
       "step": 670
     },
     {
       "epoch": 1.2253267237494367,
-      "grad_norm": 22.721454620361328,
+      "grad_norm": 23.543447494506836,
       "learning_rate": 3.7725631768953066e-05,
-      "loss": 0.2844,
+      "loss": 0.2704,
       "step": 680
     },
     {
       "epoch": 1.2433528616493916,
-      "grad_norm": 49.28988265991211,
+      "grad_norm": 34.42282485961914,
       "learning_rate": 3.754512635379062e-05,
-      "loss": 0.3247,
+      "loss": 0.3207,
       "step": 690
     },
     {
       "epoch": 1.2613789995493465,
-      "grad_norm": 48.35593795776367,
+      "grad_norm": 52.59242630004883,
       "learning_rate": 3.7364620938628155e-05,
-      "loss": 0.3217,
+      "loss": 0.303,
       "step": 700
     },
     {
       "epoch": 1.2794051374493014,
-      "grad_norm": 30.997228622436523,
+      "grad_norm": 28.380693435668945,
       "learning_rate": 3.718411552346571e-05,
-      "loss": 0.2854,
+      "loss": 0.2782,
       "step": 710
     },
     {
       "epoch": 1.2974312753492563,
-      "grad_norm": 35.57414627075195,
+      "grad_norm": 30.861343383789062,
       "learning_rate": 3.700361010830325e-05,
-      "loss": 0.3315,
+      "loss": 0.3218,
       "step": 720
     },
     {
       "epoch": 1.3154574132492114,
-      "grad_norm": 27.17759895324707,
+      "grad_norm": 24.913606643676758,
       "learning_rate": 3.68231046931408e-05,
-      "loss": 0.2965,
+      "loss": 0.2904,
       "step": 730
     },
     {
       "epoch": 1.3334835511491663,
-      "grad_norm": 31.296627044677734,
+      "grad_norm": 32.816200256347656,
       "learning_rate": 3.664259927797834e-05,
-      "loss": 0.3396,
+      "loss": 0.31,
       "step": 740
     },
     {
       "epoch": 1.3515096890491212,
-      "grad_norm": 18.639175415039062,
+      "grad_norm": 21.931764602661133,
       "learning_rate": 3.646209386281589e-05,
-      "loss": 0.3178,
+      "loss": 0.2986,
       "step": 750
     },
     {
       "epoch": 1.3695358269490763,
-      "grad_norm": 31.972665786743164,
+      "grad_norm": 18.701095581054688,
       "learning_rate": 3.628158844765343e-05,
-      "loss": 0.2927,
+      "loss": 0.2841,
       "step": 760
     },
     {
       "epoch": 1.3875619648490312,
-      "grad_norm": 14.907872200012207,
+      "grad_norm": 18.000774383544922,
       "learning_rate": 3.610108303249098e-05,
-      "loss": 0.3391,
+      "loss": 0.3354,
       "step": 770
     },
     {
       "epoch": 1.405588102748986,
-      "grad_norm": 33.58858871459961,
+      "grad_norm": 37.74651336669922,
       "learning_rate": 3.592057761732852e-05,
-      "loss": 0.3111,
+      "loss": 0.3101,
       "step": 780
     },
     {
       "epoch": 1.423614240648941,
-      "grad_norm": 17.019819259643555,
+      "grad_norm": 20.058916091918945,
       "learning_rate": 3.574007220216607e-05,
-      "loss": 0.3401,
+      "loss": 0.3324,
       "step": 790
     },
     {
       "epoch": 1.4416403785488958,
-      "grad_norm": 27.324296951293945,
+      "grad_norm": 25.672412872314453,
       "learning_rate": 3.555956678700361e-05,
-      "loss": 0.2917,
+      "loss": 0.2853,
       "step": 800
     },
     {
       "epoch": 1.4596665164488507,
-      "grad_norm": 30.170120239257812,
+      "grad_norm": 25.06855010986328,
       "learning_rate": 3.537906137184116e-05,
-      "loss": 0.2871,
+      "loss": 0.2913,
       "step": 810
     },
     {
       "epoch": 1.4776926543488058,
-      "grad_norm": 44.26614761352539,
+      "grad_norm": 28.99418067932129,
       "learning_rate": 3.51985559566787e-05,
-      "loss": 0.3307,
+      "loss": 0.3055,
       "step": 820
     },
     {
       "epoch": 1.4957187922487607,
-      "grad_norm": 28.978620529174805,
+      "grad_norm": 31.89482307434082,
       "learning_rate": 3.5018050541516247e-05,
-      "loss": 0.3207,
+      "loss": 0.3102,
       "step": 830
     },
     {
       "epoch": 1.5137449301487156,
-      "grad_norm": 22.961040496826172,
+      "grad_norm": 19.116968154907227,
       "learning_rate": 3.483754512635379e-05,
-      "loss": 0.3075,
+      "loss": 0.3079,
       "step": 840
     },
     {
       "epoch": 1.5317710680486707,
-      "grad_norm": 22.511327743530273,
+      "grad_norm": 17.446434020996094,
       "learning_rate": 3.4657039711191336e-05,
-      "loss": 0.3035,
+      "loss": 0.2972,
       "step": 850
     },
     {
       "epoch": 1.5497972059486256,
-      "grad_norm": 33.944190979003906,
+      "grad_norm": 24.87527847290039,
       "learning_rate": 3.447653429602888e-05,
-      "loss": 0.3243,
+      "loss": 0.3148,
       "step": 860
     },
     {
       "epoch": 1.5678233438485805,
-      "grad_norm": 24.72734832763672,
+      "grad_norm": 27.43445587158203,
       "learning_rate": 3.4296028880866426e-05,
-      "loss": 0.3403,
+      "loss": 0.3321,
       "step": 870
     },
     {
       "epoch": 1.5858494817485354,
-      "grad_norm": 31.211387634277344,
+      "grad_norm": 29.606460571289062,
       "learning_rate": 3.411552346570397e-05,
-      "loss": 0.3142,
+      "loss": 0.3132,
       "step": 880
     },
     {
       "epoch": 1.6038756196484902,
-      "grad_norm": 23.974918365478516,
+      "grad_norm": 26.478710174560547,
       "learning_rate": 3.3935018050541516e-05,
-      "loss": 0.3015,
+      "loss": 0.2844,
       "step": 890
     },
     {
       "epoch": 1.6219017575484451,
-      "grad_norm": 30.705829620361328,
+      "grad_norm": 39.11272048950195,
       "learning_rate": 3.375451263537907e-05,
-      "loss": 0.3529,
+      "loss": 0.3578,
       "step": 900
     },
     {
       "epoch": 1.6399278954484002,
-      "grad_norm": 20.697528839111328,
+      "grad_norm": 25.545093536376953,
       "learning_rate": 3.3574007220216606e-05,
-      "loss": 0.302,
+      "loss": 0.2977,
       "step": 910
     },
     {
       "epoch": 1.6579540333483551,
-      "grad_norm": 28.308101654052734,
+      "grad_norm": 32.68045425415039,
       "learning_rate": 3.339350180505416e-05,
-      "loss": 0.2735,
+      "loss": 0.2662,
       "step": 920
     },
     {
       "epoch": 1.67598017124831,
-      "grad_norm": 22.28900909423828,
+      "grad_norm": 21.527555465698242,
       "learning_rate": 3.3212996389891696e-05,
-      "loss": 0.3194,
+      "loss": 0.3121,
       "step": 930
     },
     {
       "epoch": 1.694006309148265,
-      "grad_norm": 29.04107666015625,
+      "grad_norm": 25.452423095703125,
       "learning_rate": 3.303249097472924e-05,
-      "loss": 0.3265,
+      "loss": 0.3243,
       "step": 940
     },
     {
       "epoch": 1.71203244704822,
-      "grad_norm": 40.08032989501953,
+      "grad_norm": 36.91832733154297,
       "learning_rate": 3.2851985559566786e-05,
-      "loss": 0.3331,
+      "loss": 0.3199,
       "step": 950
     },
     {
       "epoch": 1.7300585849481749,
-      "grad_norm": 29.40334129333496,
+      "grad_norm": 29.058032989501953,
       "learning_rate": 3.267148014440433e-05,
-      "loss": 0.2664,
+      "loss": 0.2745,
       "step": 960
     },
     {
       "epoch": 1.7480847228481298,
-      "grad_norm": 33.51088333129883,
+      "grad_norm": 34.39455032348633,
       "learning_rate": 3.249097472924188e-05,
-      "loss": 0.3031,
+      "loss": 0.294,
       "step": 970
     },
     {
       "epoch": 1.7661108607480847,
-      "grad_norm": 27.003469467163086,
+      "grad_norm": 32.51374053955078,
       "learning_rate": 3.231046931407942e-05,
-      "loss": 0.3195,
+      "loss": 0.2994,
       "step": 980
     },
     {
       "epoch": 1.7841369986480395,
-      "grad_norm": 26.585309982299805,
+      "grad_norm": 22.213062286376953,
       "learning_rate": 3.212996389891697e-05,
-      "loss": 0.3385,
+      "loss": 0.3296,
       "step": 990
     },
     {
       "epoch": 1.8021631365479944,
-      "grad_norm": 36.866294860839844,
+      "grad_norm": 26.40260887145996,
       "learning_rate": 3.194945848375451e-05,
-      "loss": 0.3029,
+      "loss": 0.2876,
       "step": 1000
     },
     {
       "epoch": 1.8201892744479495,
-      "grad_norm": 44.910362243652344,
+      "grad_norm": 26.672101974487305,
       "learning_rate": 3.176895306859206e-05,
-      "loss": 0.3239,
+      "loss": 0.3172,
       "step": 1010
     },
     {
       "epoch": 1.8382154123479044,
-      "grad_norm": 20.13945198059082,
+      "grad_norm": 20.305883407592773,
       "learning_rate": 3.15884476534296e-05,
-      "loss": 0.3071,
+      "loss": 0.3199,
       "step": 1020
     },
     {
       "epoch": 1.8562415502478595,
-      "grad_norm": 42.44215774536133,
+      "grad_norm": 33.50856018066406,
       "learning_rate": 3.140794223826715e-05,
-      "loss": 0.3246,
+      "loss": 0.2886,
       "step": 1030
     },
     {
       "epoch": 1.8742676881478144,
-      "grad_norm": 31.791410446166992,
+      "grad_norm": 38.9220085144043,
       "learning_rate": 3.12274368231047e-05,
-      "loss": 0.3273,
+      "loss": 0.3198,
       "step": 1040
     },
     {
       "epoch": 1.8922938260477693,
-      "grad_norm": 21.244722366333008,
+      "grad_norm": 31.592369079589844,
       "learning_rate": 3.104693140794224e-05,
-      "loss": 0.3093,
+      "loss": 0.3174,
       "step": 1050
     },
     {
       "epoch": 1.9103199639477242,
-      "grad_norm": 27.103349685668945,
+      "grad_norm": 23.867713928222656,
       "learning_rate": 3.086642599277979e-05,
-      "loss": 0.2694,
+      "loss": 0.2686,
       "step": 1060
     },
     {
       "epoch": 1.928346101847679,
-      "grad_norm": 35.73145294189453,
+      "grad_norm": 30.095354080200195,
       "learning_rate": 3.0685920577617325e-05,
-      "loss": 0.2934,
+      "loss": 0.2915,
       "step": 1070
     },
     {
       "epoch": 1.946372239747634,
-      "grad_norm": 19.97008514404297,
+      "grad_norm": 21.510950088500977,
       "learning_rate": 3.0505415162454877e-05,
-      "loss": 0.3493,
+      "loss": 0.3328,
       "step": 1080
     },
     {
       "epoch": 1.9643983776475888,
-      "grad_norm": 23.084754943847656,
+      "grad_norm": 25.863767623901367,
       "learning_rate": 3.032490974729242e-05,
-      "loss": 0.3243,
+      "loss": 0.3207,
       "step": 1090
     },
     {
       "epoch": 1.982424515547544,
-      "grad_norm": 29.515155792236328,
+      "grad_norm": 28.48052978515625,
       "learning_rate": 3.0144404332129967e-05,
-      "loss": 0.3235,
+      "loss": 0.3221,
       "step": 1100
     },
     {
       "epoch": 2.0,
-      "grad_norm": 24.17915916442871,
+      "grad_norm": 29.120187759399414,
       "learning_rate": 2.996389891696751e-05,
-      "loss": 0.2751,
+      "loss": 0.2667,
       "step": 1110
     },
     {
       "epoch": 2.0,
-      "eval_f1": 0.8428896708755951,
-      "eval_loss": 0.3940623104572296,
-      "eval_precision": 0.800314465408805,
-      "eval_recall": 0.890249234805422,
-      "eval_runtime": 12.3989,
-      "eval_samples_per_second": 376.889,
-      "eval_steps_per_second": 47.182,
+      "eval_f1": 0.8394543546694648,
+      "eval_fn": 287,
+      "eval_fp": 478,
+      "eval_loss": 0.3905850350856781,
+      "eval_precision": 0.8071025020177562,
+      "eval_recall": 0.8745080891998251,
+      "eval_runtime": 17.6994,
+      "eval_samples_per_second": 264.02,
+      "eval_steps_per_second": 33.052,
+      "eval_tn": 1908,
+      "eval_tp": 2000,
       "step": 1110
     },
     {
       "epoch": 2.018026137899955,
-      "grad_norm": 18.65660858154297,
+      "grad_norm": 17.400802612304688,
       "learning_rate": 2.9783393501805057e-05,
-      "loss": 0.1912,
+      "loss": 0.1682,
       "step": 1120
     },
     {
       "epoch": 2.0360522757999098,
-      "grad_norm": 15.253448486328125,
+      "grad_norm": 18.69659423828125,
       "learning_rate": 2.9602888086642598e-05,
-      "loss": 0.1425,
+      "loss": 0.1222,
       "step": 1130
     },
     {
       "epoch": 2.0540784136998647,
-      "grad_norm": 58.709293365478516,
+      "grad_norm": 50.107913970947266,
       "learning_rate": 2.9422382671480147e-05,
-      "loss": 0.1945,
+      "loss": 0.1668,
       "step": 1140
     },
     {
       "epoch": 2.0721045515998195,
-      "grad_norm": 45.032073974609375,
+      "grad_norm": 46.38024139404297,
       "learning_rate": 2.924187725631769e-05,
-      "loss": 0.1477,
+      "loss": 0.1498,
       "step": 1150
     },
     {
       "epoch": 2.090130689499775,
-      "grad_norm": 26.54606819152832,
+      "grad_norm": 25.969005584716797,
       "learning_rate": 2.906137184115524e-05,
-      "loss": 0.1731,
+      "loss": 0.159,
       "step": 1160
     },
     {
       "epoch": 2.1081568273997298,
-      "grad_norm": 24.949045181274414,
+      "grad_norm": 39.98933792114258,
       "learning_rate": 2.888086642599278e-05,
-      "loss": 0.1504,
+      "loss": 0.1309,
       "step": 1170
     },
     {
       "epoch": 2.1261829652996846,
-      "grad_norm": 71.09517669677734,
+      "grad_norm": 52.10431671142578,
       "learning_rate": 2.870036101083033e-05,
-      "loss": 0.149,
+      "loss": 0.1667,
       "step": 1180
     },
     {
       "epoch": 2.1442091031996395,
-      "grad_norm": 43.31296157836914,
+      "grad_norm": 40.3720703125,
       "learning_rate": 2.851985559566787e-05,
-      "loss": 0.1826,
+      "loss": 0.1893,
       "step": 1190
     },
     {
       "epoch": 2.1622352410995944,
-      "grad_norm": 23.145519256591797,
+      "grad_norm": 38.03019332885742,
       "learning_rate": 2.8339350180505413e-05,
-      "loss": 0.1716,
+      "loss": 0.1565,
       "step": 1200
     },
     {
       "epoch": 2.1802613789995493,
-      "grad_norm": 41.57735061645508,
+      "grad_norm": 45.700809478759766,
       "learning_rate": 2.815884476534296e-05,
-      "loss": 0.1784,
+      "loss": 0.1461,
       "step": 1210
     },
     {
       "epoch": 2.198287516899504,
-      "grad_norm": 51.243560791015625,
+      "grad_norm": 37.712379455566406,
       "learning_rate": 2.7978339350180506e-05,
-      "loss": 0.161,
+      "loss": 0.1561,
       "step": 1220
     },
     {
       "epoch": 2.216313654799459,
-      "grad_norm": 38.89030838012695,
+      "grad_norm": 33.16317367553711,
       "learning_rate": 2.779783393501805e-05,
-      "loss": 0.1688,
+      "loss": 0.1484,
       "step": 1230
     },
     {
       "epoch": 2.234339792699414,
-      "grad_norm": 43.14015197753906,
+      "grad_norm": 19.943294525146484,
       "learning_rate": 2.7617328519855596e-05,
-      "loss": 0.1916,
+      "loss": 0.1659,
       "step": 1240
     },
     {
       "epoch": 2.2523659305993693,
-      "grad_norm": 35.20539474487305,
+      "grad_norm": 20.473445892333984,
       "learning_rate": 2.7436823104693144e-05,
-      "loss": 0.1652,
+      "loss": 0.1515,
       "step": 1250
     },
     {
       "epoch": 2.270392068499324,
-      "grad_norm": 39.17034912109375,
+      "grad_norm": 38.28823471069336,
       "learning_rate": 2.7256317689530686e-05,
-      "loss": 0.1888,
+      "loss": 0.1856,
       "step": 1260
     },
     {
       "epoch": 2.288418206399279,
-      "grad_norm": 26.350337982177734,
+      "grad_norm": 17.913368225097656,
       "learning_rate": 2.7075812274368234e-05,
-      "loss": 0.1585,
+      "loss": 0.1326,
       "step": 1270
     },
     {
       "epoch": 2.306444344299234,
-      "grad_norm": 33.05204772949219,
+      "grad_norm": 20.06118392944336,
       "learning_rate": 2.6895306859205776e-05,
-      "loss": 0.1449,
+      "loss": 0.1426,
       "step": 1280
     },
     {
       "epoch": 2.324470482199189,
-      "grad_norm": 53.03055191040039,
+      "grad_norm": 30.24198341369629,
       "learning_rate": 2.6714801444043324e-05,
-      "loss": 0.1457,
+      "loss": 0.1237,
       "step": 1290
     },
     {
       "epoch": 2.3424966200991437,
-      "grad_norm": 13.711820602416992,
+      "grad_norm": 25.373851776123047,
       "learning_rate": 2.6534296028880866e-05,
-      "loss": 0.1451,
+      "loss": 0.1321,
       "step": 1300
     },
     {
       "epoch": 2.3605227579990986,
-      "grad_norm": 60.76357650756836,
+      "grad_norm": 61.367210388183594,
       "learning_rate": 2.6353790613718414e-05,
-      "loss": 0.1803,
+      "loss": 0.1628,
       "step": 1310
     },
     {
       "epoch": 2.3785488958990535,
-      "grad_norm": 47.53473663330078,
+      "grad_norm": 41.69776916503906,
       "learning_rate": 2.617328519855596e-05,
-      "loss": 0.1405,
+      "loss": 0.1245,
       "step": 1320
     },
     {
       "epoch": 2.3965750337990084,
-      "grad_norm": 46.01548385620117,
+      "grad_norm": 42.08649826049805,
       "learning_rate": 2.59927797833935e-05,
-      "loss": 0.1701,
+      "loss": 0.1612,
       "step": 1330
     },
     {
       "epoch": 2.4146011716989637,
-      "grad_norm": 36.47391128540039,
+      "grad_norm": 22.838258743286133,
       "learning_rate": 2.581227436823105e-05,
-      "loss": 0.1362,
+      "loss": 0.1373,
       "step": 1340
     },
     {
       "epoch": 2.4326273095989186,
-      "grad_norm": 69.90142059326172,
+      "grad_norm": 52.76067352294922,
       "learning_rate": 2.563176895306859e-05,
-      "loss": 0.143,
+      "loss": 0.1511,
       "step": 1350
     },
     {
       "epoch": 2.4506534474988735,
-      "grad_norm": 52.88728713989258,
+      "grad_norm": 50.871219635009766,
       "learning_rate": 2.545126353790614e-05,
-      "loss": 0.1776,
+      "loss": 0.1777,
       "step": 1360
     },
     {
       "epoch": 2.4686795853988284,
-      "grad_norm": 26.982271194458008,
+      "grad_norm": 19.779335021972656,
       "learning_rate": 2.527075812274368e-05,
-      "loss": 0.1922,
+      "loss": 0.1778,
       "step": 1370
     },
     {
       "epoch": 2.4867057232987833,
-      "grad_norm": 41.308414459228516,
+      "grad_norm": 40.787845611572266,
       "learning_rate": 2.509025270758123e-05,
-      "loss": 0.1699,
+      "loss": 0.1433,
       "step": 1380
     },
     {
       "epoch": 2.504731861198738,
-      "grad_norm": 68.30561828613281,
+      "grad_norm": 39.423404693603516,
       "learning_rate": 2.4909747292418774e-05,
-      "loss": 0.1503,
+      "loss": 0.1372,
       "step": 1390
     },
     {
       "epoch": 2.522757999098693,
-      "grad_norm": 32.563026428222656,
+      "grad_norm": 46.86770248413086,
       "learning_rate": 2.472924187725632e-05,
-      "loss": 0.2153,
+      "loss": 0.2135,
       "step": 1400
     },
     {
       "epoch": 2.540784136998648,
-      "grad_norm": 24.238431930541992,
+      "grad_norm": 17.858736038208008,
       "learning_rate": 2.4548736462093864e-05,
-      "loss": 0.1935,
+      "loss": 0.1631,
       "step": 1410
     },
     {
       "epoch": 2.558810274898603,
-      "grad_norm": 52.790916442871094,
+      "grad_norm": 36.45213317871094,
       "learning_rate": 2.436823104693141e-05,
-      "loss": 0.21,
+      "loss": 0.1753,
       "step": 1420
     },
     {
       "epoch": 2.576836412798558,
-      "grad_norm": 39.62615966796875,
+      "grad_norm": 51.208805084228516,
       "learning_rate": 2.4187725631768953e-05,
-      "loss": 0.1756,
+      "loss": 0.1722,
       "step": 1430
     },
     {
       "epoch": 2.5948625506985126,
-      "grad_norm": 28.220199584960938,
+      "grad_norm": 20.30636978149414,
       "learning_rate": 2.40072202166065e-05,
-      "loss": 0.1704,
+      "loss": 0.1516,
       "step": 1440
     },
     {
       "epoch": 2.612888688598468,
-      "grad_norm": 36.99165344238281,
+      "grad_norm": 57.81936264038086,
       "learning_rate": 2.3826714801444043e-05,
-      "loss": 0.1632,
+      "loss": 0.1488,
       "step": 1450
     },
     {
       "epoch": 2.630914826498423,
-      "grad_norm": 79.26138305664062,
+      "grad_norm": 33.060733795166016,
       "learning_rate": 2.3646209386281588e-05,
-      "loss": 0.1562,
+      "loss": 0.1625,
       "step": 1460
     },
     {
       "epoch": 2.6489409643983777,
-      "grad_norm": 47.28179931640625,
+      "grad_norm": 54.26327133178711,
       "learning_rate": 2.3465703971119137e-05,
-      "loss": 0.2061,
+      "loss": 0.1771,
       "step": 1470
     },
     {
       "epoch": 2.6669671022983326,
-      "grad_norm": 40.676551818847656,
+      "grad_norm": 25.87835693359375,
       "learning_rate": 2.328519855595668e-05,
-      "loss": 0.1955,
+      "loss": 0.191,
       "step": 1480
     },
     {
       "epoch": 2.6849932401982874,
-      "grad_norm": 29.851699829101562,
+      "grad_norm": 34.679718017578125,
       "learning_rate": 2.3104693140794227e-05,
-      "loss": 0.1447,
+      "loss": 0.138,
       "step": 1490
     },
     {
       "epoch": 2.7030193780982423,
-      "grad_norm": 46.22013473510742,
+      "grad_norm": 44.00432205200195,
       "learning_rate": 2.292418772563177e-05,
-      "loss": 0.1635,
+      "loss": 0.1652,
       "step": 1500
     },
     {
       "epoch": 2.721045515998197,
-      "grad_norm": 35.77067565917969,
+      "grad_norm": 11.452496528625488,
       "learning_rate": 2.2743682310469316e-05,
-      "loss": 0.1611,
+      "loss": 0.1457,
       "step": 1510
     },
     {
       "epoch": 2.7390716538981525,
-      "grad_norm": 44.474876403808594,
+      "grad_norm": 34.115055084228516,
       "learning_rate": 2.2563176895306858e-05,
-      "loss": 0.159,
+      "loss": 0.1607,
       "step": 1520
     },
     {
       "epoch": 2.757097791798107,
-      "grad_norm": 49.898887634277344,
+      "grad_norm": 37.13041305541992,
       "learning_rate": 2.2382671480144403e-05,
-      "loss": 0.1716,
+      "loss": 0.1519,
       "step": 1530
     },
     {
       "epoch": 2.7751239296980623,
-      "grad_norm": 54.783653259277344,
+      "grad_norm": 49.859092712402344,
       "learning_rate": 2.220216606498195e-05,
-      "loss": 0.1921,
+      "loss": 0.1842,
       "step": 1540
     },
     {
       "epoch": 2.793150067598017,
-      "grad_norm": 14.532942771911621,
+      "grad_norm": 20.27824592590332,
       "learning_rate": 2.2021660649819496e-05,
-      "loss": 0.1436,
+      "loss": 0.1352,
       "step": 1550
     },
     {
       "epoch": 2.811176205497972,
-      "grad_norm": 47.2420768737793,
+      "grad_norm": 54.31693649291992,
       "learning_rate": 2.184115523465704e-05,
-      "loss": 0.2074,
+      "loss": 0.2065,
       "step": 1560
     },
     {
       "epoch": 2.829202343397927,
-      "grad_norm": 39.21223449707031,
+      "grad_norm": 53.624446868896484,
       "learning_rate": 2.1660649819494586e-05,
-      "loss": 0.1871,
+      "loss": 0.167,
       "step": 1570
     },
     {
       "epoch": 2.847228481297882,
-      "grad_norm": 43.29368591308594,
+      "grad_norm": 22.161457061767578,
       "learning_rate": 2.148014440433213e-05,
-      "loss": 0.1474,
+      "loss": 0.1425,
       "step": 1580
     },
     {
       "epoch": 2.8652546191978367,
-      "grad_norm": 30.276391983032227,
+      "grad_norm": 27.219818115234375,
       "learning_rate": 2.1299638989169676e-05,
-      "loss": 0.2214,
+      "loss": 0.2153,
       "step": 1590
     },
     {
       "epoch": 2.8832807570977916,
-      "grad_norm": 38.97634506225586,
+      "grad_norm": 32.98075866699219,
       "learning_rate": 2.111913357400722e-05,
-      "loss": 0.1692,
+      "loss": 0.1473,
       "step": 1600
     },
     {
       "epoch": 2.901306894997747,
-      "grad_norm": 46.147945404052734,
+      "grad_norm": 71.80169677734375,
       "learning_rate": 2.0938628158844766e-05,
-      "loss": 0.1587,
+      "loss": 0.1333,
       "step": 1610
     },
     {
       "epoch": 2.9193330328977014,
-      "grad_norm": 57.70730209350586,
+      "grad_norm": 45.45034408569336,
       "learning_rate": 2.075812274368231e-05,
-      "loss": 0.1642,
+      "loss": 0.1525,
       "step": 1620
     },
     {
       "epoch": 2.9373591707976567,
-      "grad_norm": 48.976478576660156,
+      "grad_norm": 36.39427185058594,
       "learning_rate": 2.057761732851986e-05,
-      "loss": 0.1536,
+      "loss": 0.1315,
       "step": 1630
     },
     {
       "epoch": 2.9553853086976116,
-      "grad_norm": 49.945587158203125,
+      "grad_norm": 36.360191345214844,
       "learning_rate": 2.0397111913357404e-05,
-      "loss": 0.1612,
+      "loss": 0.1497,
       "step": 1640
     },
     {
       "epoch": 2.9734114465975665,
-      "grad_norm": 74.91495513916016,
+      "grad_norm": 45.537322998046875,
       "learning_rate": 2.0216606498194946e-05,
-      "loss": 0.1557,
+      "loss": 0.1357,
       "step": 1650
     },
     {
       "epoch": 2.9914375844975214,
-      "grad_norm": 30.24640464782715,
+      "grad_norm": 22.563854217529297,
       "learning_rate": 2.003610108303249e-05,
-      "loss": 0.1631,
+      "loss": 0.166,
       "step": 1660
     },
     {
       "epoch": 3.0,
-      "eval_f1": 0.8290598290598291,
-      "eval_loss": 0.5290467739105225,
-      "eval_precision": 0.8106978687839532,
-      "eval_recall": 0.8482728465238304,
-      "eval_runtime": 11.4074,
-      "eval_samples_per_second": 409.646,
-      "eval_steps_per_second": 51.282,
+      "eval_f1": 0.82996632996633,
+      "eval_fn": 315,
+      "eval_fp": 493,
+      "eval_loss": 0.5492772459983826,
+      "eval_precision": 0.8,
+      "eval_recall": 0.8622649759510276,
+      "eval_runtime": 12.9988,
+      "eval_samples_per_second": 359.495,
+      "eval_steps_per_second": 45.004,
+      "eval_tn": 1893,
+      "eval_tp": 1972,
       "step": 1665
     },
     {
       "epoch": 3.0090130689499777,
-      "grad_norm": 7.1513214111328125,
+      "grad_norm": 7.169532299041748,
       "learning_rate": 1.9855595667870036e-05,
-      "loss": 0.1199,
+      "loss": 0.1283,
       "step": 1670
     },
     {
       "epoch": 3.0270392068499326,
-      "grad_norm": 36.67481994628906,
+      "grad_norm": 53.082916259765625,
       "learning_rate": 1.967509025270758e-05,
-      "loss": 0.0607,
+      "loss": 0.0816,
       "step": 1680
     },
     {
       "epoch": 3.0450653447498874,
-      "grad_norm": 41.55238723754883,
+      "grad_norm": 47.888553619384766,
       "learning_rate": 1.9494584837545125e-05,
-      "loss": 0.0821,
+      "loss": 0.0742,
       "step": 1690
     },
     {
       "epoch": 3.0630914826498423,
-      "grad_norm": 37.779537200927734,
+      "grad_norm": 54.369449615478516,
       "learning_rate": 1.9314079422382674e-05,
-      "loss": 0.0624,
+      "loss": 0.0728,
       "step": 1700
     },
     {
       "epoch": 3.081117620549797,
-      "grad_norm": 13.272784233093262,
+      "grad_norm": 7.978189945220947,
       "learning_rate": 1.913357400722022e-05,
-      "loss": 0.0745,
+      "loss": 0.0702,
       "step": 1710
     },
     {
       "epoch": 3.099143758449752,
-      "grad_norm": 19.611099243164062,
+      "grad_norm": 30.30511474609375,
       "learning_rate": 1.8953068592057764e-05,
-      "loss": 0.0953,
+      "loss": 0.065,
       "step": 1720
     },
     {
       "epoch": 3.117169896349707,
-      "grad_norm": 26.23482894897461,
+      "grad_norm": 15.761763572692871,
       "learning_rate": 1.877256317689531e-05,
-      "loss": 0.0576,
+      "loss": 0.0573,
       "step": 1730
     },
     {
       "epoch": 3.135196034249662,
-      "grad_norm": 31.549488067626953,
+      "grad_norm": 47.82561492919922,
       "learning_rate": 1.8592057761732854e-05,
-      "loss": 0.1034,
+      "loss": 0.0842,
       "step": 1740
     },
     {
       "epoch": 3.1532221721496168,
-      "grad_norm": 44.37907409667969,
+      "grad_norm": 35.642642974853516,
       "learning_rate": 1.84115523465704e-05,
-      "loss": 0.0807,
+      "loss": 0.0745,
       "step": 1750
     },
     {
       "epoch": 3.171248310049572,
-      "grad_norm": 29.255582809448242,
+      "grad_norm": 27.68004608154297,
       "learning_rate": 1.8231046931407943e-05,
-      "loss": 0.08,
+      "loss": 0.0743,
       "step": 1760
     },
     {
       "epoch": 3.189274447949527,
-      "grad_norm": 11.359423637390137,
+      "grad_norm": 16.700593948364258,
       "learning_rate": 1.805054151624549e-05,
-      "loss": 0.073,
+      "loss": 0.061,
       "step": 1770
     },
     {
       "epoch": 3.207300585849482,
-      "grad_norm": 51.08259201049805,
+      "grad_norm": 8.846168518066406,
       "learning_rate": 1.7870036101083033e-05,
-      "loss": 0.0869,
+      "loss": 0.0581,
       "step": 1780
     },
     {
       "epoch": 3.2253267237494367,
-      "grad_norm": 28.016101837158203,
+      "grad_norm": 38.31391143798828,
       "learning_rate": 1.768953068592058e-05,
-      "loss": 0.1044,
+      "loss": 0.1093,
       "step": 1790
     },
     {
       "epoch": 3.2433528616493916,
-      "grad_norm": 81.0438003540039,
+      "grad_norm": 39.713565826416016,
       "learning_rate": 1.7509025270758123e-05,
-      "loss": 0.0818,
+      "loss": 0.0859,
       "step": 1800
     },
     {
       "epoch": 3.2613789995493465,
-      "grad_norm": 25.233104705810547,
+      "grad_norm": 51.820274353027344,
       "learning_rate": 1.7328519855595668e-05,
-      "loss": 0.0808,
+      "loss": 0.0768,
       "step": 1810
     },
     {
       "epoch": 3.2794051374493014,
-      "grad_norm": 127.22054290771484,
+      "grad_norm": 64.148681640625,
       "learning_rate": 1.7148014440433213e-05,
-      "loss": 0.0847,
+      "loss": 0.0688,
       "step": 1820
     },
     {
       "epoch": 3.2974312753492563,
-      "grad_norm": 37.59592819213867,
+      "grad_norm": 42.258243560791016,
       "learning_rate": 1.6967509025270758e-05,
-      "loss": 0.1276,
+      "loss": 0.0787,
       "step": 1830
     },
     {
       "epoch": 3.315457413249211,
-      "grad_norm": 33.60833740234375,
+      "grad_norm": 7.588295936584473,
       "learning_rate": 1.6787003610108303e-05,
-      "loss": 0.0718,
+      "loss": 0.0516,
       "step": 1840
     },
     {
       "epoch": 3.3334835511491665,
-      "grad_norm": 47.96291732788086,
+      "grad_norm": 61.41086196899414,
       "learning_rate": 1.6606498194945848e-05,
-      "loss": 0.0769,
+      "loss": 0.0759,
       "step": 1850
     },
     {
       "epoch": 3.3515096890491214,
-      "grad_norm": 1.6931202411651611,
+      "grad_norm": 2.3156187534332275,
       "learning_rate": 1.6425992779783393e-05,
-      "loss": 0.0694,
+      "loss": 0.0649,
       "step": 1860
     },
     {
       "epoch": 3.3695358269490763,
-      "grad_norm": 65.2172622680664,
+      "grad_norm": 37.482566833496094,
       "learning_rate": 1.624548736462094e-05,
-      "loss": 0.1009,
+      "loss": 0.0882,
       "step": 1870
     },
     {
       "epoch": 3.387561964849031,
-      "grad_norm": 11.240936279296875,
+      "grad_norm": 73.03086853027344,
       "learning_rate": 1.6064981949458486e-05,
-      "loss": 0.0749,
+      "loss": 0.0588,
       "step": 1880
     },
     {
       "epoch": 3.405588102748986,
-      "grad_norm": 24.050607681274414,
+      "grad_norm": 66.03443908691406,
       "learning_rate": 1.588447653429603e-05,
-      "loss": 0.0979,
+      "loss": 0.0786,
       "step": 1890
     },
     {
       "epoch": 3.423614240648941,
-      "grad_norm": 92.54638671875,
+      "grad_norm": 85.51991271972656,
       "learning_rate": 1.5703971119133576e-05,
-      "loss": 0.0968,
+      "loss": 0.1056,
       "step": 1900
     },
     {
       "epoch": 3.441640378548896,
-      "grad_norm": 10.564998626708984,
+      "grad_norm": 146.41830444335938,
       "learning_rate": 1.552346570397112e-05,
-      "loss": 0.0642,
+      "loss": 0.0603,
       "step": 1910
     },
     {
       "epoch": 3.4596665164488507,
-      "grad_norm": 63.50326919555664,
+      "grad_norm": 10.789298057556152,
       "learning_rate": 1.5342960288808663e-05,
-      "loss": 0.0941,
+      "loss": 0.081,
       "step": 1920
     },
     {
       "epoch": 3.4776926543488056,
-      "grad_norm": 74.58378601074219,
+      "grad_norm": 55.54345703125,
       "learning_rate": 1.516245487364621e-05,
-      "loss": 0.1205,
+      "loss": 0.0884,
       "step": 1930
     },
     {
       "epoch": 3.495718792248761,
-      "grad_norm": 9.73009967803955,
+      "grad_norm": 11.078765869140625,
       "learning_rate": 1.4981949458483754e-05,
-      "loss": 0.0826,
+      "loss": 0.0609,
       "step": 1940
     },
     {
       "epoch": 3.5137449301487154,
-      "grad_norm": 14.504748344421387,
+      "grad_norm": 67.22940063476562,
       "learning_rate": 1.4801444043321299e-05,
-      "loss": 0.0861,
+      "loss": 0.0923,
       "step": 1950
     },
     {
       "epoch": 3.5317710680486707,
-      "grad_norm": 3.4811439514160156,
+      "grad_norm": 21.463754653930664,
       "learning_rate": 1.4620938628158846e-05,
-      "loss": 0.0658,
+      "loss": 0.0784,
       "step": 1960
     },
     {
       "epoch": 3.5497972059486256,
-      "grad_norm": 60.301517486572266,
+      "grad_norm": 46.619197845458984,
       "learning_rate": 1.444043321299639e-05,
-      "loss": 0.074,
+      "loss": 0.0557,
       "step": 1970
     },
     {
       "epoch": 3.5678233438485805,
-      "grad_norm": 58.39499282836914,
+      "grad_norm": 38.9202766418457,
       "learning_rate": 1.4259927797833936e-05,
-      "loss": 0.1037,
+      "loss": 0.0773,
       "step": 1980
     },
     {
       "epoch": 3.5858494817485354,
-      "grad_norm": 28.403823852539062,
+      "grad_norm": 75.98242950439453,
       "learning_rate": 1.407942238267148e-05,
-      "loss": 0.0807,
+      "loss": 0.0704,
       "step": 1990
     },
     {
       "epoch": 3.6038756196484902,
-      "grad_norm": 47.89693832397461,
+      "grad_norm": 21.9030818939209,
       "learning_rate": 1.3898916967509026e-05,
-      "loss": 0.0818,
+      "loss": 0.0917,
       "step": 2000
     },
     {
       "epoch": 3.621901757548445,
-      "grad_norm": 43.21398162841797,
+      "grad_norm": 39.96767807006836,
       "learning_rate": 1.3718411552346572e-05,
-      "loss": 0.099,
+      "loss": 0.0596,
       "step": 2010
     },
     {
       "epoch": 3.6399278954484,
-      "grad_norm": 84.97303771972656,
+      "grad_norm": 34.662776947021484,
       "learning_rate": 1.3537906137184117e-05,
-      "loss": 0.0916,
+      "loss": 0.0943,
       "step": 2020
     },
     {
       "epoch": 3.6579540333483553,
-      "grad_norm": 62.820228576660156,
+      "grad_norm": 83.10199737548828,
       "learning_rate": 1.3357400722021662e-05,
-      "loss": 0.1033,
+      "loss": 0.0836,
       "step": 2030
     },
     {
       "epoch": 3.67598017124831,
-      "grad_norm": 41.46628952026367,
+      "grad_norm": 43.2503662109375,
       "learning_rate": 1.3176895306859207e-05,
-      "loss": 0.0775,
+      "loss": 0.0706,
       "step": 2040
     },
     {
       "epoch": 3.694006309148265,
-      "grad_norm": 9.337126731872559,
+      "grad_norm": 29.68882942199707,
       "learning_rate": 1.299638989169675e-05,
-      "loss": 0.083,
+      "loss": 0.0795,
       "step": 2050
     },
     {
       "epoch": 3.71203244704822,
-      "grad_norm": 86.4469985961914,
+      "grad_norm": 29.66282844543457,
       "learning_rate": 1.2815884476534295e-05,
-      "loss": 0.0702,
+      "loss": 0.0581,
       "step": 2060
     },
     {
       "epoch": 3.730058584948175,
-      "grad_norm": 44.70920944213867,
+      "grad_norm": 44.41033935546875,
       "learning_rate": 1.263537906137184e-05,
-      "loss": 0.0908,
+      "loss": 0.0801,
       "step": 2070
     },
     {
       "epoch": 3.7480847228481298,
-      "grad_norm": 83.01412963867188,
+      "grad_norm": 36.88494110107422,
       "learning_rate": 1.2454873646209387e-05,
-      "loss": 0.0831,
+      "loss": 0.0815,
       "step": 2080
     },
     {
       "epoch": 3.7661108607480847,
-      "grad_norm": 28.753332138061523,
+      "grad_norm": 49.79912185668945,
       "learning_rate": 1.2274368231046932e-05,
-      "loss": 0.1054,
+      "loss": 0.0758,
       "step": 2090
     },
     {
       "epoch": 3.7841369986480395,
-      "grad_norm": 39.22829055786133,
+      "grad_norm": 44.74293899536133,
       "learning_rate": 1.2093862815884477e-05,
-      "loss": 0.0976,
+      "loss": 0.0701,
       "step": 2100
     },
     {
       "epoch": 3.8021631365479944,
-      "grad_norm": 41.21415328979492,
+      "grad_norm": 19.233646392822266,
       "learning_rate": 1.1913357400722022e-05,
-      "loss": 0.0995,
+      "loss": 0.0836,
       "step": 2110
     },
     {
       "epoch": 3.8201892744479498,
-      "grad_norm": 25.80561065673828,
+      "grad_norm": 74.74971771240234,
       "learning_rate": 1.1732851985559568e-05,
-      "loss": 0.1185,
+      "loss": 0.1071,
       "step": 2120
     },
     {
       "epoch": 3.838215412347904,
-      "grad_norm": 23.873735427856445,
+      "grad_norm": 65.20303344726562,
       "learning_rate": 1.1552346570397113e-05,
-      "loss": 0.0618,
+      "loss": 0.0761,
       "step": 2130
     },
     {
       "epoch": 3.8562415502478595,
-      "grad_norm": 20.476320266723633,
+      "grad_norm": 67.05883026123047,
       "learning_rate": 1.1371841155234658e-05,
-      "loss": 0.1164,
+      "loss": 0.1228,
       "step": 2140
     },
     {
       "epoch": 3.8742676881478144,
-      "grad_norm": 75.58688354492188,
+      "grad_norm": 21.299774169921875,
       "learning_rate": 1.1191335740072201e-05,
-      "loss": 0.1166,
+      "loss": 0.1194,
       "step": 2150
     },
     {
       "epoch": 3.8922938260477693,
-      "grad_norm": 49.5202751159668,
+      "grad_norm": 12.536727905273438,
       "learning_rate": 1.1010830324909748e-05,
-      "loss": 0.0982,
+      "loss": 0.0878,
       "step": 2160
     },
     {
       "epoch": 3.910319963947724,
-      "grad_norm": 37.534568786621094,
+      "grad_norm": 75.28766632080078,
       "learning_rate": 1.0830324909747293e-05,
-      "loss": 0.0652,
+      "loss": 0.0555,
       "step": 2170
     },
     {
       "epoch": 3.928346101847679,
-      "grad_norm": 70.99764251708984,
+      "grad_norm": 30.614364624023438,
       "learning_rate": 1.0649819494584838e-05,
-      "loss": 0.0876,
+      "loss": 0.0766,
       "step": 2180
     },
     {
       "epoch": 3.946372239747634,
-      "grad_norm": 35.94860076904297,
+      "grad_norm": 62.048099517822266,
       "learning_rate": 1.0469314079422383e-05,
-      "loss": 0.0642,
+      "loss": 0.0459,
       "step": 2190
     },
     {
       "epoch": 3.964398377647589,
-      "grad_norm": 6.9634785652160645,
+      "grad_norm": 5.321977615356445,
       "learning_rate": 1.028880866425993e-05,
-      "loss": 0.1256,
+      "loss": 0.0751,
       "step": 2200
     },
     {
       "epoch": 3.982424515547544,
-      "grad_norm": 33.263797760009766,
+      "grad_norm": 109.93370819091797,
       "learning_rate": 1.0108303249097473e-05,
-      "loss": 0.0646,
+      "loss": 0.0572,
       "step": 2210
     },
     {
       "epoch": 4.0,
-      "grad_norm": 24.34364128112793,
+      "grad_norm": 51.198822021484375,
       "learning_rate": 9.927797833935018e-06,
-      "loss": 0.0782,
+      "loss": 0.0649,
       "step": 2220
     },
     {
       "epoch": 4.0,
-      "eval_f1": 0.827906976744186,
-      "eval_loss": 0.6959211826324463,
-      "eval_precision": 0.8014735980352026,
-      "eval_recall": 0.8561434193266287,
-      "eval_runtime": 10.9179,
-      "eval_samples_per_second": 428.013,
-      "eval_steps_per_second": 53.582,
+      "eval_f1": 0.8250753985351141,
+      "eval_fn": 372,
+      "eval_fp": 440,
+      "eval_loss": 0.7689136266708374,
+      "eval_precision": 0.8131634819532909,
+      "eval_recall": 0.8373414954088325,
+      "eval_runtime": 17.851,
+      "eval_samples_per_second": 261.777,
+      "eval_steps_per_second": 32.771,
+      "eval_tn": 1946,
+      "eval_tp": 1915,
       "step": 2220
     }
   ],

checkpoint-2220/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:92aaf17ba6b51dd0179acd28630a512a049c53024ba01863d7232c815d6018dd
+oid sha256:7e924750e5e12e6bb9372df333907b4d7f42d9dd2da839a2407f8f861ab40940
 size 5432