End of training

all_results.json

@@ -1,34 +1,34 @@
 {
     "epoch": 10.0,
-    "eval_CAUSATOR_f1": 0.8,
+    "eval_CAUSATOR_f1": 0.8444444444444446,
     "eval_CAUSATOR_number": 91,
-    "eval_CAUSATOR_precision": 0.7676767676767676,
+    "eval_CAUSATOR_precision": 0.8539325842696629,
     "eval_CAUSATOR_recall": 0.8351648351648352,
-    "eval_EXPIRIENCER_f1": 0.9359605911330049,
-    "eval_EXPIRIENCER_number": 98,
-    "eval_EXPIRIENCER_precision": 0.9047619047619048,
-    "eval_EXPIRIENCER_recall": 0.9693877551020408,
-    "eval_INSTRUMENT_f1": 0.7499999999999999,
-    "eval_INSTRUMENT_number": 6,
-    "eval_INSTRUMENT_precision": 0.6,
+    "eval_EXPIRIENCER_f1": 0.949367088607595,
+    "eval_EXPIRIENCER_number": 77,
+    "eval_EXPIRIENCER_precision": 0.9259259259259259,
+    "eval_EXPIRIENCER_recall": 0.974025974025974,
+    "eval_INSTRUMENT_f1": 0.5454545454545454,
+    "eval_INSTRUMENT_number": 3,
+    "eval_INSTRUMENT_precision": 0.375,
     "eval_INSTRUMENT_recall": 1.0,
     "eval_OTHER_f1": 0.0,
     "eval_OTHER_number": 1,
     "eval_OTHER_precision": 0.0,
     "eval_OTHER_recall": 0.0,
-    "eval_PREDICATE_f1": 0.9477611940298507,
-    "eval_PREDICATE_number": 129,
-    "eval_PREDICATE_precision": 0.9136690647482014,
-    "eval_PREDICATE_recall": 0.9844961240310077,
-    "eval_loss": 0.15370479226112366,
-    "eval_overall_accuracy": 0.9660852713178295,
-    "eval_overall_f1": 0.8967551622418879,
-    "eval_overall_precision": 0.8611898016997167,
-    "eval_overall_recall": 0.9353846153846154,
-    "eval_runtime": 1.9588,
+    "eval_PREDICATE_f1": 0.961904761904762,
+    "eval_PREDICATE_number": 102,
+    "eval_PREDICATE_precision": 0.9351851851851852,
+    "eval_PREDICATE_recall": 0.9901960784313726,
+    "eval_loss": 0.17230771481990814,
+    "eval_overall_accuracy": 0.966686496133254,
+    "eval_overall_f1": 0.9107142857142856,
+    "eval_overall_precision": 0.8916083916083916,
+    "eval_overall_recall": 0.9306569343065694,
+    "eval_runtime": 1.8028,
     "eval_samples": 99,
-    "eval_samples_per_second": 50.54,
-    "eval_steps_per_second": 6.637,
+    "eval_samples_per_second": 54.914,
+    "eval_steps_per_second": 7.211,
     "eval_\u0414\u0420\u0423\u0413\u041e\u0415_f1": 0.0,
     "eval_\u0414\u0420\u0423\u0413\u041e\u0415_number": 1,
     "eval_\u0414\u0420\u0423\u0413\u041e\u0415_precision": 0.0,
@@ -53,9 +53,9 @@
     "eval_\u044d\u043a\u0441\u043f\u0435\u0440\u0438\u0435\u043d\u0446\u0435\u0440_number": 3,
     "eval_\u044d\u043a\u0441\u043f\u0435\u0440\u0438\u0435\u043d\u0446\u0435\u0440_precision": 0.0,
     "eval_\u044d\u043a\u0441\u043f\u0435\u0440\u0438\u0435\u043d\u0446\u0435\u0440_recall": 0.0,
-    "train_loss": 0.14471116384956986,
-    "train_runtime": 174.2063,
+    "train_loss": 0.13590615599782074,
+    "train_runtime": 186.3702,
     "train_samples": 886,
-    "train_samples_per_second": 50.859,
-    "train_steps_per_second": 3.215
+    "train_samples_per_second": 47.54,
+    "train_steps_per_second": 3.005
 }

eval_results.json

@@ -1,32 +1,32 @@
 {
     "epoch": 10.0,
-    "eval_CAUSATOR_f1": 0.8,
+    "eval_CAUSATOR_f1": 0.8444444444444446,
     "eval_CAUSATOR_number": 91,
-    "eval_CAUSATOR_precision": 0.7676767676767676,
+    "eval_CAUSATOR_precision": 0.8539325842696629,
     "eval_CAUSATOR_recall": 0.8351648351648352,
-    "eval_EXPIRIENCER_f1": 0.9359605911330049,
-    "eval_EXPIRIENCER_number": 98,
-    "eval_EXPIRIENCER_precision": 0.9047619047619048,
-    "eval_EXPIRIENCER_recall": 0.9693877551020408,
-    "eval_INSTRUMENT_f1": 0.7499999999999999,
-    "eval_INSTRUMENT_number": 6,
-    "eval_INSTRUMENT_precision": 0.6,
+    "eval_EXPIRIENCER_f1": 0.949367088607595,
+    "eval_EXPIRIENCER_number": 77,
+    "eval_EXPIRIENCER_precision": 0.9259259259259259,
+    "eval_EXPIRIENCER_recall": 0.974025974025974,
+    "eval_INSTRUMENT_f1": 0.5454545454545454,
+    "eval_INSTRUMENT_number": 3,
+    "eval_INSTRUMENT_precision": 0.375,
     "eval_INSTRUMENT_recall": 1.0,
     "eval_OTHER_f1": 0.0,
     "eval_OTHER_number": 1,
     "eval_OTHER_precision": 0.0,
     "eval_OTHER_recall": 0.0,
-    "eval_PREDICATE_f1": 0.9477611940298507,
-    "eval_PREDICATE_number": 129,
-    "eval_PREDICATE_precision": 0.9136690647482014,
-    "eval_PREDICATE_recall": 0.9844961240310077,
-    "eval_loss": 0.15370479226112366,
-    "eval_overall_accuracy": 0.9660852713178295,
-    "eval_overall_f1": 0.8967551622418879,
-    "eval_overall_precision": 0.8611898016997167,
-    "eval_overall_recall": 0.9353846153846154,
-    "eval_runtime": 1.9588,
+    "eval_PREDICATE_f1": 0.961904761904762,
+    "eval_PREDICATE_number": 102,
+    "eval_PREDICATE_precision": 0.9351851851851852,
+    "eval_PREDICATE_recall": 0.9901960784313726,
+    "eval_loss": 0.17230771481990814,
+    "eval_overall_accuracy": 0.966686496133254,
+    "eval_overall_f1": 0.9107142857142856,
+    "eval_overall_precision": 0.8916083916083916,
+    "eval_overall_recall": 0.9306569343065694,
+    "eval_runtime": 1.8028,
     "eval_samples": 99,
-    "eval_samples_per_second": 50.54,
-    "eval_steps_per_second": 6.637
+    "eval_samples_per_second": 54.914,
+    "eval_steps_per_second": 7.211
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:79281edcc0c9c393bbb9682545941dc09ff0978f2ed4385766707aff3105b4b1
+oid sha256:b9a96c0d7d6046fd4b744e9af0faccc1c00f24b2c5f1d3c91908e5ab6f1c3757
 size 710971889

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
-    "train_loss": 0.14471116384956986,
-    "train_runtime": 174.2063,
+    "train_loss": 0.13590615599782074,
+    "train_runtime": 186.3702,
     "train_samples": 886,
-    "train_samples_per_second": 50.859,
-    "train_steps_per_second": 3.215
+    "train_samples_per_second": 47.54,
+    "train_steps_per_second": 3.005
 }

trainer_state.json

@@ -10,827 +10,827 @@
     {
       "epoch": 0.02,
       "learning_rate": 1.4705882352941177e-06,
-      "loss": 1.9403,
+      "loss": 1.9542,
       "step": 1
     },
     {
       "epoch": 0.09,
       "learning_rate": 7.3529411764705884e-06,
-      "loss": 1.8862,
+      "loss": 1.898,
       "step": 5
     },
     {
       "epoch": 0.18,
       "learning_rate": 1.4705882352941177e-05,
-      "loss": 1.7434,
+      "loss": 1.7464,
       "step": 10
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.2058823529411766e-05,
-      "loss": 1.3329,
+      "loss": 1.3278,
       "step": 15
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.9411764705882354e-05,
-      "loss": 0.9204,
+      "loss": 0.8686,
       "step": 20
     },
     {
       "epoch": 0.45,
       "learning_rate": 3.6764705882352945e-05,
-      "loss": 1.1761,
+      "loss": 1.1514,
       "step": 25
     },
     {
       "epoch": 0.54,
       "learning_rate": 4.411764705882353e-05,
-      "loss": 0.6574,
+      "loss": 0.6278,
       "step": 30
     },
     {
       "epoch": 0.62,
       "learning_rate": 4.9999554100506465e-05,
-      "loss": 0.8733,
+      "loss": 0.8248,
       "step": 35
     },
     {
       "epoch": 0.71,
       "learning_rate": 4.998394928831034e-05,
-      "loss": 0.5708,
+      "loss": 0.514,
       "step": 40
     },
     {
       "epoch": 0.8,
       "learning_rate": 4.9946065404995485e-05,
-      "loss": 0.4218,
+      "loss": 0.3524,
       "step": 45
     },
     {
       "epoch": 0.89,
       "learning_rate": 4.988593623296038e-05,
-      "loss": 0.4296,
+      "loss": 0.3661,
       "step": 50
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.980361539151416e-05,
-      "loss": 0.3043,
+      "loss": 0.2552,
       "step": 55
     },
     {
       "epoch": 1.0,
-      "eval_CAUSATOR_f1": 0.6826347305389222,
+      "eval_CAUSATOR_f1": 0.7625000000000001,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.75,
-      "eval_CAUSATOR_recall": 0.6263736263736264,
-      "eval_EXPIRIENCER_f1": 0.821782178217822,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.7980769230769231,
-      "eval_EXPIRIENCER_recall": 0.8469387755102041,
+      "eval_CAUSATOR_precision": 0.8840579710144928,
+      "eval_CAUSATOR_recall": 0.6703296703296703,
+      "eval_EXPIRIENCER_f1": 0.8366013071895425,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.8421052631578947,
+      "eval_EXPIRIENCER_recall": 0.8311688311688312,
       "eval_INSTRUMENT_f1": 0.0,
-      "eval_INSTRUMENT_number": 6,
+      "eval_INSTRUMENT_number": 3,
       "eval_INSTRUMENT_precision": 0.0,
       "eval_INSTRUMENT_recall": 0.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9191176470588236,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.8741258741258742,
-      "eval_PREDICATE_recall": 0.9689922480620154,
-      "eval_loss": 0.35376405715942383,
-      "eval_overall_accuracy": 0.9142441860465116,
-      "eval_overall_f1": 0.8179012345679012,
-      "eval_overall_precision": 0.8204334365325078,
-      "eval_overall_recall": 0.8153846153846154,
-      "eval_runtime": 0.5833,
-      "eval_samples_per_second": 169.727,
-      "eval_steps_per_second": 22.287,
+      "eval_PREDICATE_f1": 0.9523809523809523,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9259259259259259,
+      "eval_PREDICATE_recall": 0.9803921568627451,
+      "eval_loss": 0.3471083343029022,
+      "eval_overall_accuracy": 0.9202855443188578,
+      "eval_overall_f1": 0.8538899430740038,
+      "eval_overall_precision": 0.8893280632411067,
+      "eval_overall_recall": 0.8211678832116789,
+      "eval_runtime": 0.5374,
+      "eval_samples_per_second": 184.214,
+      "eval_steps_per_second": 24.19,
       "step": 56
     },
     {
       "epoch": 1.07,
       "learning_rate": 4.969917628906234e-05,
-      "loss": 0.2351,
+      "loss": 0.22,
       "step": 60
     },
     {
       "epoch": 1.16,
       "learning_rate": 4.9572712057646e-05,
-      "loss": 0.2885,
+      "loss": 0.2635,
       "step": 65
     },
     {
       "epoch": 1.25,
       "learning_rate": 4.9424335469892594e-05,
-      "loss": 0.2677,
+      "loss": 0.2427,
       "step": 70
     },
     {
       "epoch": 1.34,
       "learning_rate": 4.925417883845266e-05,
-      "loss": 0.3071,
+      "loss": 0.2734,
       "step": 75
     },
     {
       "epoch": 1.43,
       "learning_rate": 4.9062393898011904e-05,
-      "loss": 0.2608,
+      "loss": 0.2406,
       "step": 80
     },
     {
       "epoch": 1.52,
       "learning_rate": 4.884915166998405e-05,
-      "loss": 0.2315,
+      "loss": 0.2305,
       "step": 85
     },
     {
       "epoch": 1.61,
       "learning_rate": 4.8614642310004974e-05,
-      "loss": 0.2369,
+      "loss": 0.1872,
       "step": 90
     },
     {
       "epoch": 1.7,
       "learning_rate": 4.835907493836421e-05,
-      "loss": 0.1395,
+      "loss": 0.1041,
       "step": 95
     },
     {
       "epoch": 1.79,
       "learning_rate": 4.808267745352502e-05,
-      "loss": 0.2028,
+      "loss": 0.1932,
       "step": 100
     },
     {
       "epoch": 1.88,
       "learning_rate": 4.778569632889919e-05,
-      "loss": 0.2415,
+      "loss": 0.1829,
       "step": 105
     },
     {
       "epoch": 1.96,
       "learning_rate": 4.746839639305808e-05,
-      "loss": 0.2664,
+      "loss": 0.2385,
       "step": 110
     },
     {
       "epoch": 2.0,
-      "eval_CAUSATOR_f1": 0.787878787878788,
+      "eval_CAUSATOR_f1": 0.8402366863905325,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.8783783783783784,
-      "eval_CAUSATOR_recall": 0.7142857142857143,
-      "eval_EXPIRIENCER_f1": 0.9128205128205128,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9175257731958762,
-      "eval_EXPIRIENCER_recall": 0.9081632653061225,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
-      "eval_INSTRUMENT_recall": 1.0,
+      "eval_CAUSATOR_precision": 0.9102564102564102,
+      "eval_CAUSATOR_recall": 0.7802197802197802,
+      "eval_EXPIRIENCER_f1": 0.9554140127388535,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9375,
+      "eval_EXPIRIENCER_recall": 0.974025974025974,
+      "eval_INSTRUMENT_f1": 0.4,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.2857142857142857,
+      "eval_INSTRUMENT_recall": 0.6666666666666666,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9541984732824427,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9398496240601504,
-      "eval_PREDICATE_recall": 0.9689922480620154,
-      "eval_loss": 0.19607584178447723,
-      "eval_overall_accuracy": 0.939922480620155,
-      "eval_overall_f1": 0.8920187793427231,
-      "eval_overall_precision": 0.9076433121019108,
-      "eval_overall_recall": 0.8769230769230769,
-      "eval_runtime": 0.5651,
-      "eval_samples_per_second": 175.19,
-      "eval_steps_per_second": 23.005,
+      "eval_PREDICATE_f1": 0.9611650485436893,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9519230769230769,
+      "eval_PREDICATE_recall": 0.9705882352941176,
+      "eval_loss": 0.1607949137687683,
+      "eval_overall_accuracy": 0.9553837001784652,
+      "eval_overall_f1": 0.9097605893186002,
+      "eval_overall_precision": 0.9182156133828996,
+      "eval_overall_recall": 0.9014598540145985,
+      "eval_runtime": 0.5406,
+      "eval_samples_per_second": 183.129,
+      "eval_steps_per_second": 24.047,
       "step": 112
     },
     {
       "epoch": 2.05,
       "learning_rate": 4.713106059357554e-05,
-      "loss": 0.1366,
+      "loss": 0.1171,
       "step": 115
     },
     {
       "epoch": 2.14,
       "learning_rate": 4.6773989744713596e-05,
-      "loss": 0.1735,
+      "loss": 0.1598,
       "step": 120
     },
     {
       "epoch": 2.23,
       "learning_rate": 4.639750225917573e-05,
-      "loss": 0.0874,
+      "loss": 0.0799,
       "step": 125
     },
     {
       "epoch": 2.32,
       "learning_rate": 4.600193386416697e-05,
-      "loss": 0.114,
+      "loss": 0.0935,
       "step": 130
     },
     {
       "epoch": 2.41,
       "learning_rate": 4.558763730201413e-05,
-      "loss": 0.1326,
+      "loss": 0.1022,
       "step": 135
     },
     {
       "epoch": 2.5,
       "learning_rate": 4.515498201561296e-05,
-      "loss": 0.1554,
+      "loss": 0.1408,
       "step": 140
     },
     {
       "epoch": 2.59,
       "learning_rate": 4.470435381898287e-05,
-      "loss": 0.1238,
+      "loss": 0.1019,
       "step": 145
     },
     {
       "epoch": 2.68,
       "learning_rate": 4.423615455322293e-05,
-      "loss": 0.0882,
+      "loss": 0.0569,
       "step": 150
     },
     {
       "epoch": 2.77,
       "learning_rate": 4.375080172817591e-05,
-      "loss": 0.0908,
+      "loss": 0.0964,
       "step": 155
     },
     {
       "epoch": 2.86,
       "learning_rate": 4.324872815012005e-05,
-      "loss": 0.1086,
+      "loss": 0.0827,
       "step": 160
     },
     {
       "epoch": 2.95,
       "learning_rate": 4.273038153582034e-05,
-      "loss": 0.0373,
+      "loss": 0.0367,
       "step": 165
     },
     {
       "epoch": 3.0,
-      "eval_CAUSATOR_f1": 0.8409090909090909,
+      "eval_CAUSATOR_f1": 0.8439306358381503,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.8705882352941177,
-      "eval_CAUSATOR_recall": 0.8131868131868132,
-      "eval_EXPIRIENCER_f1": 0.9452736318407962,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9223300970873787,
-      "eval_EXPIRIENCER_recall": 0.9693877551020408,
-      "eval_INSTRUMENT_f1": 0.7142857142857143,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.625,
-      "eval_INSTRUMENT_recall": 0.8333333333333334,
+      "eval_CAUSATOR_precision": 0.8902439024390244,
+      "eval_CAUSATOR_recall": 0.8021978021978022,
+      "eval_EXPIRIENCER_f1": 0.9554140127388535,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9375,
+      "eval_EXPIRIENCER_recall": 0.974025974025974,
+      "eval_INSTRUMENT_f1": 0.6,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.42857142857142855,
+      "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9584905660377359,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9338235294117647,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.12746019661426544,
-      "eval_overall_accuracy": 0.9641472868217055,
-      "eval_overall_f1": 0.9162861491628614,
-      "eval_overall_precision": 0.9066265060240963,
-      "eval_overall_recall": 0.9261538461538461,
-      "eval_runtime": 1.9297,
-      "eval_samples_per_second": 51.302,
-      "eval_steps_per_second": 6.737,
+      "eval_PREDICATE_f1": 0.975609756097561,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.970873786407767,
+      "eval_PREDICATE_recall": 0.9803921568627451,
+      "eval_loss": 0.13108175992965698,
+      "eval_overall_accuracy": 0.9672813801308745,
+      "eval_overall_f1": 0.9194139194139195,
+      "eval_overall_precision": 0.9227941176470589,
+      "eval_overall_recall": 0.916058394160584,
+      "eval_runtime": 1.832,
+      "eval_samples_per_second": 54.038,
+      "eval_steps_per_second": 7.096,
       "step": 168
     },
     {
       "epoch": 3.04,
       "learning_rate": 4.21962241132837e-05,
-      "loss": 0.0976,
+      "loss": 0.0686,
       "step": 170
     },
     {
       "epoch": 3.12,
       "learning_rate": 4.164673220957387e-05,
-      "loss": 0.0896,
+      "loss": 0.0729,
       "step": 175
     },
     {
       "epoch": 3.21,
       "learning_rate": 4.108239582605374e-05,
-      "loss": 0.0494,
+      "loss": 0.05,
       "step": 180
     },
     {
       "epoch": 3.3,
       "learning_rate": 4.050371820143384e-05,
-      "loss": 0.0447,
+      "loss": 0.052,
       "step": 185
     },
     {
       "epoch": 3.39,
       "learning_rate": 3.991121536301653e-05,
-      "loss": 0.0834,
+      "loss": 0.0798,
       "step": 190
     },
     {
       "epoch": 3.48,
       "learning_rate": 3.9305415666536236e-05,
-      "loss": 0.0545,
+      "loss": 0.0468,
       "step": 195
     },
     {
       "epoch": 3.57,
       "learning_rate": 3.868685932500595e-05,
-      "loss": 0.0579,
+      "loss": 0.0489,
       "step": 200
     },
     {
       "epoch": 3.66,
       "learning_rate": 3.80560979269902e-05,
-      "loss": 0.0586,
+      "loss": 0.0377,
       "step": 205
     },
     {
       "epoch": 3.75,
       "learning_rate": 3.7413693944733995e-05,
-      "loss": 0.0674,
+      "loss": 0.0648,
       "step": 210
     },
     {
       "epoch": 3.84,
       "learning_rate": 3.676022023258643e-05,
-      "loss": 0.0553,
+      "loss": 0.0657,
       "step": 215
     },
     {
       "epoch": 3.93,
       "learning_rate": 3.609625951616623e-05,
-      "loss": 0.0496,
+      "loss": 0.0494,
       "step": 220
     },
     {
       "epoch": 4.0,
-      "eval_CAUSATOR_f1": 0.8172043010752689,
+      "eval_CAUSATOR_f1": 0.8021390374331551,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.8,
-      "eval_CAUSATOR_recall": 0.8351648351648352,
-      "eval_EXPIRIENCER_f1": 0.9458128078817734,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9142857142857143,
-      "eval_EXPIRIENCER_recall": 0.9795918367346939,
-      "eval_INSTRUMENT_f1": 0.8,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6666666666666666,
+      "eval_CAUSATOR_precision": 0.78125,
+      "eval_CAUSATOR_recall": 0.8241758241758241,
+      "eval_EXPIRIENCER_f1": 0.9358974358974359,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9240506329113924,
+      "eval_EXPIRIENCER_recall": 0.948051948051948,
+      "eval_INSTRUMENT_f1": 0.6,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.42857142857142855,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9548872180451128,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.927007299270073,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.16831780970096588,
-      "eval_overall_accuracy": 0.9607558139534884,
-      "eval_overall_f1": 0.9090909090909091,
-      "eval_overall_precision": 0.8815028901734104,
-      "eval_overall_recall": 0.9384615384615385,
-      "eval_runtime": 0.569,
-      "eval_samples_per_second": 173.979,
-      "eval_steps_per_second": 22.846,
+      "eval_PREDICATE_f1": 0.966183574879227,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9523809523809523,
+      "eval_PREDICATE_recall": 0.9803921568627451,
+      "eval_loss": 0.1507418304681778,
+      "eval_overall_accuracy": 0.9637120761451516,
+      "eval_overall_f1": 0.8948306595365421,
+      "eval_overall_precision": 0.8745644599303136,
+      "eval_overall_recall": 0.916058394160584,
+      "eval_runtime": 0.5441,
+      "eval_samples_per_second": 181.955,
+      "eval_steps_per_second": 23.893,
       "step": 224
     },
     {
       "epoch": 4.02,
       "learning_rate": 3.5422403872724686e-05,
-      "loss": 0.0687,
+      "loss": 0.043,
       "step": 225
     },
     {
       "epoch": 4.11,
       "learning_rate": 3.473925420316946e-05,
-      "loss": 0.0588,
+      "loss": 0.0488,
       "step": 230
     },
     {
       "epoch": 4.2,
       "learning_rate": 3.404741969621996e-05,
-      "loss": 0.02,
+      "loss": 0.0228,
       "step": 235
     },
     {
       "epoch": 4.29,
       "learning_rate": 3.334751728517223e-05,
-      "loss": 0.0176,
+      "loss": 0.0117,
       "step": 240
     },
     {
       "epoch": 4.38,
       "learning_rate": 3.264017109775771e-05,
-      "loss": 0.0414,
+      "loss": 0.055,
       "step": 245
     },
     {
       "epoch": 4.46,
       "learning_rate": 3.1926011899586485e-05,
-      "loss": 0.0142,
+      "loss": 0.0133,
       "step": 250
     },
     {
       "epoch": 4.55,
       "learning_rate": 3.120567653167127e-05,
-      "loss": 0.0386,
+      "loss": 0.0342,
       "step": 255
     },
     {
       "epoch": 4.64,
       "learning_rate": 3.0479807342533722e-05,
-      "loss": 0.0294,
+      "loss": 0.0271,
       "step": 260
     },
     {
       "epoch": 4.73,
       "learning_rate": 2.9749051615399637e-05,
-      "loss": 0.0415,
+      "loss": 0.0331,
       "step": 265
     },
     {
       "epoch": 4.82,
       "learning_rate": 2.9014060990993496e-05,
-      "loss": 0.0446,
+      "loss": 0.043,
       "step": 270
     },
     {
       "epoch": 4.91,
       "learning_rate": 2.827549088644748e-05,
-      "loss": 0.0396,
+      "loss": 0.0283,
       "step": 275
     },
     {
       "epoch": 5.0,
       "learning_rate": 2.7533999910842766e-05,
-      "loss": 0.0529,
+      "loss": 0.0699,
       "step": 280
     },
     {
       "epoch": 5.0,
-      "eval_CAUSATOR_f1": 0.8128342245989305,
+      "eval_CAUSATOR_f1": 0.8089887640449438,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.7916666666666666,
-      "eval_CAUSATOR_recall": 0.8351648351648352,
-      "eval_EXPIRIENCER_f1": 0.9468599033816426,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.8990825688073395,
-      "eval_EXPIRIENCER_recall": 1.0,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
+      "eval_CAUSATOR_precision": 0.8275862068965517,
+      "eval_CAUSATOR_recall": 0.7912087912087912,
+      "eval_EXPIRIENCER_f1": 0.9382716049382716,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.8941176470588236,
+      "eval_EXPIRIENCER_recall": 0.987012987012987,
+      "eval_INSTRUMENT_f1": 0.5454545454545454,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.375,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9513108614232209,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9202898550724637,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.15264536440372467,
-      "eval_overall_accuracy": 0.9626937984496124,
-      "eval_overall_f1": 0.9056047197640117,
-      "eval_overall_precision": 0.8696883852691218,
-      "eval_overall_recall": 0.9446153846153846,
-      "eval_runtime": 0.5502,
-      "eval_samples_per_second": 179.946,
-      "eval_steps_per_second": 23.629,
+      "eval_PREDICATE_f1": 0.961904761904762,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9351851851851852,
+      "eval_PREDICATE_recall": 0.9901960784313726,
+      "eval_loss": 0.18301859498023987,
+      "eval_overall_accuracy": 0.9559785841760856,
+      "eval_overall_f1": 0.896797153024911,
+      "eval_overall_precision": 0.875,
+      "eval_overall_recall": 0.9197080291970803,
+      "eval_runtime": 0.5444,
+      "eval_samples_per_second": 181.842,
+      "eval_steps_per_second": 23.878,
       "step": 280
     },
     {
       "epoch": 5.09,
       "learning_rate": 2.6790249277904582e-05,
-      "loss": 0.0221,
+      "loss": 0.0288,
       "step": 285
     },
     {
       "epoch": 5.18,
       "learning_rate": 2.6044902216374496e-05,
-      "loss": 0.0136,
+      "loss": 0.0152,
       "step": 290
     },
     {
       "epoch": 5.27,
       "learning_rate": 2.5298623378585884e-05,
-      "loss": 0.0125,
+      "loss": 0.011,
       "step": 295
     },
     {
       "epoch": 5.36,
       "learning_rate": 2.4552078247770007e-05,
-      "loss": 0.0184,
+      "loss": 0.0151,
       "step": 300
     },
     {
       "epoch": 5.45,
       "learning_rate": 2.3805932544621002e-05,
-      "loss": 0.0111,
+      "loss": 0.0082,
       "step": 305
     },
     {
       "epoch": 5.54,
       "learning_rate": 2.3060851633649245e-05,
-      "loss": 0.0317,
+      "loss": 0.0426,
       "step": 310
     },
     {
       "epoch": 5.62,
       "learning_rate": 2.2317499929852273e-05,
-      "loss": 0.0472,
+      "loss": 0.0476,
       "step": 315
     },
     {
       "epoch": 5.71,
       "learning_rate": 2.1576540306232418e-05,
-      "loss": 0.0183,
+      "loss": 0.0155,
       "step": 320
     },
     {
       "epoch": 5.8,
       "learning_rate": 2.0838633502689545e-05,
-      "loss": 0.0412,
+      "loss": 0.0409,
       "step": 325
     },
     {
       "epoch": 5.89,
       "learning_rate": 2.0104437536815883e-05,
-      "loss": 0.0283,
+      "loss": 0.0257,
       "step": 330
     },
     {
       "epoch": 5.98,
       "learning_rate": 1.9374607117118485e-05,
-      "loss": 0.0419,
+      "loss": 0.0352,
       "step": 335
     },
     {
       "epoch": 6.0,
-      "eval_CAUSATOR_f1": 0.8042328042328042,
+      "eval_CAUSATOR_f1": 0.8148148148148148,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.7755102040816326,
-      "eval_CAUSATOR_recall": 0.8351648351648352,
-      "eval_EXPIRIENCER_f1": 0.9313725490196079,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.8962264150943396,
-      "eval_EXPIRIENCER_recall": 0.9693877551020408,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
+      "eval_CAUSATOR_precision": 0.7857142857142857,
+      "eval_CAUSATOR_recall": 0.8461538461538461,
+      "eval_EXPIRIENCER_f1": 0.9440993788819876,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9047619047619048,
+      "eval_EXPIRIENCER_recall": 0.987012987012987,
+      "eval_INSTRUMENT_f1": 0.5454545454545454,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.375,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9513108614232209,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9202898550724637,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.14020289480686188,
-      "eval_overall_accuracy": 0.9651162790697675,
-      "eval_overall_f1": 0.8980797636632201,
-      "eval_overall_precision": 0.8636363636363636,
-      "eval_overall_recall": 0.9353846153846154,
-      "eval_runtime": 0.5662,
-      "eval_samples_per_second": 174.839,
-      "eval_steps_per_second": 22.959,
+      "eval_PREDICATE_f1": 0.957345971563981,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.926605504587156,
+      "eval_PREDICATE_recall": 0.9901960784313726,
+      "eval_loss": 0.19936639070510864,
+      "eval_overall_accuracy": 0.9571683521713266,
+      "eval_overall_f1": 0.8970331588132635,
+      "eval_overall_precision": 0.8595317725752508,
+      "eval_overall_recall": 0.9379562043795621,
+      "eval_runtime": 0.5401,
+      "eval_samples_per_second": 183.296,
+      "eval_steps_per_second": 24.069,
       "step": 336
     },
     {
       "epoch": 6.07,
       "learning_rate": 1.864979305919248e-05,
-      "loss": 0.0176,
+      "loss": 0.0297,
       "step": 340
     },
     {
       "epoch": 6.16,
       "learning_rate": 1.7930641705365836e-05,
-      "loss": 0.0194,
+      "loss": 0.0216,
       "step": 345
     },
     {
       "epoch": 6.25,
       "learning_rate": 1.721779434833299e-05,
-      "loss": 0.0112,
+      "loss": 0.0069,
       "step": 350
     },
     {
       "epoch": 6.34,
       "learning_rate": 1.651188665929151e-05,
-      "loss": 0.017,
+      "loss": 0.0149,
       "step": 355
     },
     {
       "epoch": 6.43,
       "learning_rate": 1.5813548121091622e-05,
-      "loss": 0.015,
+      "loss": 0.013,
       "step": 360
     },
     {
       "epoch": 6.52,
       "learning_rate": 1.512340146690402e-05,
-      "loss": 0.0132,
+      "loss": 0.0115,
       "step": 365
     },
     {
       "epoch": 6.61,
       "learning_rate": 1.4442062124906763e-05,
-      "loss": 0.0211,
+      "loss": 0.0232,
       "step": 370
     },
     {
       "epoch": 6.7,
       "learning_rate": 1.3770137669486171e-05,
-      "loss": 0.007,
+      "loss": 0.0046,
       "step": 375
     },
     {
       "epoch": 6.79,
       "learning_rate": 1.3108227279441243e-05,
-      "loss": 0.0163,
+      "loss": 0.012,
       "step": 380
     },
     {
       "epoch": 6.88,
       "learning_rate": 1.2456921203674715e-05,
-      "loss": 0.0308,
+      "loss": 0.0322,
       "step": 385
     },
     {
       "epoch": 6.96,
       "learning_rate": 1.1816800234847303e-05,
-      "loss": 0.0156,
+      "loss": 0.0186,
       "step": 390
     },
     {
       "epoch": 7.0,
-      "eval_CAUSATOR_f1": 0.8279569892473119,
+      "eval_CAUSATOR_f1": 0.8555555555555556,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.8105263157894737,
+      "eval_CAUSATOR_precision": 0.8651685393258427,
       "eval_CAUSATOR_recall": 0.8461538461538461,
-      "eval_EXPIRIENCER_f1": 0.9359605911330049,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9047619047619048,
-      "eval_EXPIRIENCER_recall": 0.9693877551020408,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
+      "eval_EXPIRIENCER_f1": 0.9433962264150942,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9146341463414634,
+      "eval_EXPIRIENCER_recall": 0.974025974025974,
+      "eval_INSTRUMENT_f1": 0.5454545454545454,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.375,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9477611940298507,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9136690647482014,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.14980360865592957,
-      "eval_overall_accuracy": 0.9660852713178295,
-      "eval_overall_f1": 0.9050445103857566,
-      "eval_overall_precision": 0.8739255014326648,
-      "eval_overall_recall": 0.9384615384615385,
-      "eval_runtime": 0.5711,
-      "eval_samples_per_second": 173.35,
-      "eval_steps_per_second": 22.763,
+      "eval_PREDICATE_f1": 0.961904761904762,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9351851851851852,
+      "eval_PREDICATE_recall": 0.9901960784313726,
+      "eval_loss": 0.16573503613471985,
+      "eval_overall_accuracy": 0.9672813801308745,
+      "eval_overall_f1": 0.912655971479501,
+      "eval_overall_precision": 0.89198606271777,
+      "eval_overall_recall": 0.9343065693430657,
+      "eval_runtime": 0.5399,
+      "eval_samples_per_second": 183.354,
+      "eval_steps_per_second": 24.077,
       "step": 392
     },
     {
       "epoch": 7.05,
       "learning_rate": 1.1188435191464303e-05,
-      "loss": 0.0158,
+      "loss": 0.0093,
       "step": 395
     },
     {
       "epoch": 7.14,
       "learning_rate": 1.0572386408856552e-05,
-      "loss": 0.0054,
+      "loss": 0.0075,
       "step": 400
     },
     {
       "epoch": 7.23,
       "learning_rate": 9.969203239509592e-06,
-      "loss": 0.0063,
+      "loss": 0.0039,
       "step": 405
     },
     {
       "epoch": 7.32,
       "learning_rate": 9.379423563186652e-06,
-      "loss": 0.0166,
+      "loss": 0.0124,
       "step": 410
     },
     {
       "epoch": 7.41,
       "learning_rate": 8.803573307282209e-06,
-      "loss": 0.0185,
+      "loss": 0.0167,
       "step": 415
     },
     {
       "epoch": 7.5,
       "learning_rate": 8.242165977833974e-06,
-      "loss": 0.0097,
+      "loss": 0.0106,
       "step": 420
     },
     {
       "epoch": 7.59,
       "learning_rate": 7.695702201611338e-06,
-      "loss": 0.0111,
+      "loss": 0.0118,
       "step": 425
     },
     {
       "epoch": 7.68,
       "learning_rate": 7.1646692796888456e-06,
-      "loss": 0.0082,
+      "loss": 0.0044,
       "step": 430
     },
     {
       "epoch": 7.77,
       "learning_rate": 6.649540752902544e-06,
-      "loss": 0.0082,
+      "loss": 0.0086,
       "step": 435
     },
     {
       "epoch": 7.86,
       "learning_rate": 6.150775979576906e-06,
-      "loss": 0.0208,
+      "loss": 0.0207,
       "step": 440
     },
     {
       "epoch": 7.95,
       "learning_rate": 5.668819725898747e-06,
-      "loss": 0.0066,
+      "loss": 0.0052,
       "step": 445
     },
     {
       "epoch": 8.0,
-      "eval_CAUSATOR_f1": 0.8085106382978723,
+      "eval_CAUSATOR_f1": 0.850828729281768,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.7835051546391752,
-      "eval_CAUSATOR_recall": 0.8351648351648352,
-      "eval_EXPIRIENCER_f1": 0.9411764705882353,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9056603773584906,
-      "eval_EXPIRIENCER_recall": 0.9795918367346939,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
+      "eval_CAUSATOR_precision": 0.8555555555555555,
+      "eval_CAUSATOR_recall": 0.8461538461538461,
+      "eval_EXPIRIENCER_f1": 0.949367088607595,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9259259259259259,
+      "eval_EXPIRIENCER_recall": 0.974025974025974,
+      "eval_INSTRUMENT_f1": 0.5454545454545454,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.375,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9477611940298507,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9136690647482014,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.150935560464859,
-      "eval_overall_accuracy": 0.9680232558139535,
-      "eval_overall_f1": 0.9010339734121122,
-      "eval_overall_precision": 0.8664772727272727,
-      "eval_overall_recall": 0.9384615384615385,
-      "eval_runtime": 0.5698,
-      "eval_samples_per_second": 173.753,
-      "eval_steps_per_second": 22.816,
+      "eval_PREDICATE_f1": 0.961904761904762,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9351851851851852,
+      "eval_PREDICATE_recall": 0.9901960784313726,
+      "eval_loss": 0.17156818509101868,
+      "eval_overall_accuracy": 0.9672813801308745,
+      "eval_overall_f1": 0.912655971479501,
+      "eval_overall_precision": 0.89198606271777,
+      "eval_overall_recall": 0.9343065693430657,
+      "eval_runtime": 0.5406,
+      "eval_samples_per_second": 183.131,
+      "eval_steps_per_second": 24.047,
       "step": 448
     },
     {
       "epoch": 8.04,
       "learning_rate": 5.204101769303474e-06,
-      "loss": 0.0068,
+      "loss": 0.0059,
       "step": 450
     },
     {
       "epoch": 8.12,
       "learning_rate": 4.757036515227381e-06,
-      "loss": 0.0055,
+      "loss": 0.0036,
       "step": 455
     },
     {
       "epoch": 8.21,
       "learning_rate": 4.328022627567657e-06,
-      "loss": 0.0071,
+      "loss": 0.0042,
       "step": 460
     },
     {
       "epoch": 8.3,
       "learning_rate": 3.91744267317968e-06,
-      "loss": 0.0061,
+      "loss": 0.0072,
       "step": 465
     },
     {
       "epoch": 8.39,
       "learning_rate": 3.5256627807286085e-06,
-      "loss": 0.017,
+      "loss": 0.0135,
       "step": 470
     },
     {
@@ -842,69 +842,69 @@
     {
       "epoch": 8.57,
       "learning_rate": 2.7998835613573137e-06,
-      "loss": 0.0103,
+      "loss": 0.0139,
       "step": 480
     },
     {
       "epoch": 8.66,
       "learning_rate": 2.4665314374336335e-06,
-      "loss": 0.0056,
+      "loss": 0.004,
       "step": 485
     },
     {
       "epoch": 8.75,
       "learning_rate": 2.153273204306153e-06,
-      "loss": 0.013,
+      "loss": 0.0181,
       "step": 490
     },
     {
       "epoch": 8.84,
       "learning_rate": 1.8603882054194915e-06,
-      "loss": 0.0078,
+      "loss": 0.0089,
       "step": 495
     },
     {
       "epoch": 8.93,
       "learning_rate": 1.5881376166848149e-06,
-      "loss": 0.0084,
+      "loss": 0.0094,
       "step": 500
     },
     {
       "epoch": 9.0,
-      "eval_CAUSATOR_f1": 0.8042328042328042,
+      "eval_CAUSATOR_f1": 0.839779005524862,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.7755102040816326,
+      "eval_CAUSATOR_precision": 0.8444444444444444,
       "eval_CAUSATOR_recall": 0.8351648351648352,
-      "eval_EXPIRIENCER_f1": 0.9359605911330049,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9047619047619048,
-      "eval_EXPIRIENCER_recall": 0.9693877551020408,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
+      "eval_EXPIRIENCER_f1": 0.949367088607595,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9259259259259259,
+      "eval_EXPIRIENCER_recall": 0.974025974025974,
+      "eval_INSTRUMENT_f1": 0.6,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.42857142857142855,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9477611940298507,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9136690647482014,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.15476951003074646,
-      "eval_overall_accuracy": 0.9656007751937985,
-      "eval_overall_f1": 0.8980797636632201,
-      "eval_overall_precision": 0.8636363636363636,
-      "eval_overall_recall": 0.9353846153846154,
-      "eval_runtime": 0.5704,
-      "eval_samples_per_second": 173.572,
-      "eval_steps_per_second": 22.792,
+      "eval_PREDICATE_f1": 0.961904761904762,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9351851851851852,
+      "eval_PREDICATE_recall": 0.9901960784313726,
+      "eval_loss": 0.17145974934101105,
+      "eval_overall_accuracy": 0.966686496133254,
+      "eval_overall_f1": 0.9107142857142856,
+      "eval_overall_precision": 0.8916083916083916,
+      "eval_overall_recall": 0.9306569343065694,
+      "eval_runtime": 0.5348,
+      "eval_samples_per_second": 185.116,
+      "eval_steps_per_second": 24.308,
       "step": 504
     },
     {
       "epoch": 9.02,
       "learning_rate": 1.3367642135800329e-06,
-      "loss": 0.0045,
+      "loss": 0.0039,
       "step": 505
     },
     {
@@ -916,103 +916,103 @@
     {
       "epoch": 9.2,
       "learning_rate": 8.975267816580201e-07,
-      "loss": 0.0123,
+      "loss": 0.0111,
       "step": 515
     },
     {
       "epoch": 9.29,
       "learning_rate": 7.100544363921324e-07,
-      "loss": 0.0107,
+      "loss": 0.0085,
       "step": 520
     },
     {
       "epoch": 9.38,
       "learning_rate": 5.442422945807462e-07,
-      "loss": 0.0123,
+      "loss": 0.0108,
       "step": 525
     },
     {
       "epoch": 9.46,
       "learning_rate": 4.002382167745428e-07,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 530
     },
     {
       "epoch": 9.55,
       "learning_rate": 2.781706165022879e-07,
-      "loss": 0.006,
+      "loss": 0.0048,
       "step": 535
     },
     {
       "epoch": 9.64,
       "learning_rate": 1.7814834575997363e-07,
-      "loss": 0.0115,
+      "loss": 0.0175,
       "step": 540
     },
     {
       "epoch": 9.73,
       "learning_rate": 1.0026059794363551e-07,
-      "loss": 0.0054,
+      "loss": 0.0047,
       "step": 545
     },
     {
       "epoch": 9.82,
       "learning_rate": 4.457682831244259e-08,
-      "loss": 0.0049,
+      "loss": 0.0063,
       "step": 550
     },
     {
       "epoch": 9.91,
       "learning_rate": 1.114669205298513e-08,
-      "loss": 0.0036,
+      "loss": 0.0028,
       "step": 555
     },
     {
       "epoch": 10.0,
       "learning_rate": 0.0,
-      "loss": 0.0083,
+      "loss": 0.0078,
       "step": 560
     },
     {
       "epoch": 10.0,
-      "eval_CAUSATOR_f1": 0.8,
+      "eval_CAUSATOR_f1": 0.8444444444444446,
       "eval_CAUSATOR_number": 91,
-      "eval_CAUSATOR_precision": 0.7676767676767676,
+      "eval_CAUSATOR_precision": 0.8539325842696629,
       "eval_CAUSATOR_recall": 0.8351648351648352,
-      "eval_EXPIRIENCER_f1": 0.9359605911330049,
-      "eval_EXPIRIENCER_number": 98,
-      "eval_EXPIRIENCER_precision": 0.9047619047619048,
-      "eval_EXPIRIENCER_recall": 0.9693877551020408,
-      "eval_INSTRUMENT_f1": 0.7499999999999999,
-      "eval_INSTRUMENT_number": 6,
-      "eval_INSTRUMENT_precision": 0.6,
+      "eval_EXPIRIENCER_f1": 0.949367088607595,
+      "eval_EXPIRIENCER_number": 77,
+      "eval_EXPIRIENCER_precision": 0.9259259259259259,
+      "eval_EXPIRIENCER_recall": 0.974025974025974,
+      "eval_INSTRUMENT_f1": 0.5454545454545454,
+      "eval_INSTRUMENT_number": 3,
+      "eval_INSTRUMENT_precision": 0.375,
       "eval_INSTRUMENT_recall": 1.0,
       "eval_OTHER_f1": 0.0,
       "eval_OTHER_number": 1,
       "eval_OTHER_precision": 0.0,
       "eval_OTHER_recall": 0.0,
-      "eval_PREDICATE_f1": 0.9477611940298507,
-      "eval_PREDICATE_number": 129,
-      "eval_PREDICATE_precision": 0.9136690647482014,
-      "eval_PREDICATE_recall": 0.9844961240310077,
-      "eval_loss": 0.15370479226112366,
-      "eval_overall_accuracy": 0.9660852713178295,
-      "eval_overall_f1": 0.8967551622418879,
-      "eval_overall_precision": 0.8611898016997167,
-      "eval_overall_recall": 0.9353846153846154,
-      "eval_runtime": 0.5646,
-      "eval_samples_per_second": 175.344,
-      "eval_steps_per_second": 23.025,
+      "eval_PREDICATE_f1": 0.961904761904762,
+      "eval_PREDICATE_number": 102,
+      "eval_PREDICATE_precision": 0.9351851851851852,
+      "eval_PREDICATE_recall": 0.9901960784313726,
+      "eval_loss": 0.17230771481990814,
+      "eval_overall_accuracy": 0.966686496133254,
+      "eval_overall_f1": 0.9107142857142856,
+      "eval_overall_precision": 0.8916083916083916,
+      "eval_overall_recall": 0.9306569343065694,
+      "eval_runtime": 0.5403,
+      "eval_samples_per_second": 183.236,
+      "eval_steps_per_second": 24.061,
       "step": 560
     },
     {
       "epoch": 10.0,
       "step": 560,
       "total_flos": 130378685407080.0,
-      "train_loss": 0.14471116384956986,
-      "train_runtime": 174.2063,
-      "train_samples_per_second": 50.859,
-      "train_steps_per_second": 3.215
+      "train_loss": 0.13590615599782074,
+      "train_runtime": 186.3702,
+      "train_samples_per_second": 47.54,
+      "train_steps_per_second": 3.005
     }
   ],
   "max_steps": 560,