[Misc] Add network timeout to eval dataset downloads (#21873)

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

sgl-model-gateway/e2e_test/infra/simple_eval_common.py

@@ -457,7 +457,7 @@ def download_dataset(path: str, url: str) -> None:
     """Download a dataset from URL to path."""
     logger.info("Downloading dataset from %s", url)
     try:
-        response = requests.get(url, stream=True)
+        response = requests.get(url, stream=True, timeout=30)
         response.raise_for_status()
 
         total_size = int(response.headers.get("content-length", 0))

sgl-model-gateway/e2e_test/infra/simple_eval_mmlu.py

@@ -93,7 +93,10 @@ class MMLUEval(Eval):
     """MMLU benchmark evaluation."""
 
     def __init__(self, filename: str, num_examples: int | None, num_threads: int):
-        df = pandas.read_csv(filename)
+        if "://" in filename:
+            df = pandas.read_csv(filename, storage_options={"timeout": 30})
+        else:
+            df = pandas.read_csv(filename)
         examples = [row.to_dict() for _, row in df.iterrows()]
         if num_examples:
             examples = random.Random(0).sample(examples, num_examples)