Upload folder using huggingface_hub

README.md

@@ -1,3 +1,3 @@
----
-license: mit
----
+# FTANch
+
+A benchmark for evaluating the robustness of toxic content classifiers against various types of text distortions, mutations, and obfuscation.

benchmark.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3dc0e9d41d54969d99067ca1069902dd5db46cbc81ab4331e3c614c4131c297
+size 4230818

eval.yaml

@@ -0,0 +1,39 @@
+id: ваш_юзернейм/имя_вашего_датасета  # Замените на реальный ID репозитория
+name: "FTANch Offensive Language Benchmark"
+description: "A benchmark for evaluating the robustness of toxic content classifiers against various types of text distortions, mutations, and obfuscation"
+
+primary_task: test
+
+tasks:
+  - id: overall
+    name: "Overall"
+    metrics: &standard_metrics
+      - id: acc
+        name: "Accuracy"
+        higher_is_better: true
+      - id: f1
+        name: "F1-Score"
+        higher_is_better: true
+      - id: p
+        name: "Precision"
+        higher_is_better: true
+      - id: r
+        name: "Recall"
+        higher_is_better: true
+
+  - id: test
+    name: "Standard Test Split"
+    metrics: *standard_metrics
+
+  - id: test_obfuscated
+    name: "Obfuscated Test Split"
+    metrics: *standard_metrics
+
+  - id: plain
+    name: "Plain Text Split"
+    metrics: *standard_metrics
+
+  - id: mutated
+    name: "Mutated Text Split"
+    metrics: *standard_metrics
+