adds-evalyaml (#2)

- adds eval.yaml (23df8f1505faa2cb90b077d391cdbbfb4cedabb8)


Co-authored-by: Nathan Habib <Linker1907@users.noreply.huggingface.co>

eval.yaml

@@ -0,0 +1,24 @@
+name: Evasion Bench
+description: >
+  vasionBench is a benchmark dataset for detecting evasive answers in earnings
+  call Q&A sessions. The task is to classify how directly corporate management
+  addresses questions from financial analysts.
+tasks:
+  - id: evasion_bench
+    config: default
+    split: train
+
+    epochs: 1
+
+    shuffle_choices: true
+
+    field_spec:
+      input: question
+      target: "eva4b_label"
+      choices: ["direct", "intermediate", "fully_evasive"]
+
+    solvers:
+      - name: multiple_choice
+
+    scorers:
+      - name: choice