update task auto scoring

lmms_eval/tasks/olympiadbench/olympiadbench.yaml

@@ -1,5 +1,6 @@
 group: olympiadbench
 task:
 - olympiadbench_test
+- olympiadbench_test_auto_scoring
 metadata:
   - version: 0.0

lmms_eval/tasks/olympiadbench/olympiadbench_test_auto_scoring.yaml

@@ -0,0 +1,25 @@
+dataset_path: lmms-lab/OlympiadBench
+dataset_kwargs:
+  token: True
+task : "olympiadbench_test_auto_scoring"
+test_split: test_auto_scoring
+output_type: generate_until
+doc_to_visual: !function utils.olympiadbench_doc_to_visual
+doc_to_text: !function utils.olympiadbench_doc_to_text
+doc_to_target: "answer"
+generation_kwargs:
+  until:
+    - "ASSISTANT:"
+  max_new_tokens: 1024
+  temperature: 0
+  top_p: 0
+  num_beams: 1
+  do_sample: false
+process_results: !function utils.olympiadbench_process_results
+metric_list:
+  - metric: submission
+    aggregation: !function utils.olympiadbench_aggregate_results
+    higher_is_better: true
+  - metric: exact_match
+    aggregation: mean
+    higher_is_better: true