use correct hf subset for ifbench multiturn (#1061)

Author: sam-paech-liquid

src/lighteval/tasks/tasks/ifbench/main.py

@@ -168,7 +168,7 @@ async def score(state: TaskState, target: Target):
     name="ifbench_multiturn",
     prompt_function=ifbench_prompt,
     hf_repo="allenai/IFBench_multi-turn",
-    hf_subset="default",
+    hf_subset="ifbench_constraints",
     metrics=[ifbench_metrics],
     hf_avail_splits=["test"],
     evaluation_splits=["test"],