aboutsummaryrefslogtreecommitdiff
path: root/benchmark/agbenchmark/challenges/deprecated/3_formatting/data.json
diff options
context:
space:
mode:
Diffstat (limited to 'benchmark/agbenchmark/challenges/deprecated/3_formatting/data.json')
-rw-r--r--benchmark/agbenchmark/challenges/deprecated/3_formatting/data.json30
1 files changed, 30 insertions, 0 deletions
diff --git a/benchmark/agbenchmark/challenges/deprecated/3_formatting/data.json b/benchmark/agbenchmark/challenges/deprecated/3_formatting/data.json
new file mode 100644
index 000000000..72c1e15a9
--- /dev/null
+++ b/benchmark/agbenchmark/challenges/deprecated/3_formatting/data.json
@@ -0,0 +1,30 @@
+{
+ "category": [
+ "scrape_synthesize"
+ ],
+ "cutoff": 60,
+ "dependencies": [
+ "TestRevenueRetrieval1.1"
+ ],
+ "eval_id": "838128f9-79ee-45cf-8a8f-c19b0d576a76",
+ "ground": {
+ "answer": "It was $81.462 billion in 2022. In millions the answer is 81,462.",
+ "eval": {
+ "type": "file"
+ },
+ "files": [
+ ".txt"
+ ],
+ "should_contain": [
+ "81,462"
+ ],
+ "should_not_contain": []
+ },
+ "info": {
+ "description": "Advanced version of the r2.1 challenge that also asks for specific formatting.",
+ "difficulty": "intermediate",
+ "side_effects": []
+ },
+ "name": "DeprecatedRevenueRetrieval1.2",
+ "task": "Write tesla's exact revenue in 2022 into a .txt file. Use the US notation, with a precision rounded to the nearest million dollars (for instance, $31,578 billion)."
+}