Add logging to evaluation.py

yourbench_space/evaluation.py

@@ -4,7 +4,7 @@ import asyncio
 from pathlib import Path
 
 from yourbench_space.leaderboard_space.env import INIT_MODELS
-
+from loguru import logger
 
 ON_SPACES = os.environ.get("system") == "spaces"
 OUTPUT_DIR = "/data" if ON_SPACES else "."
@@ -49,6 +49,11 @@ async def run_evaluations(eval_ds_name: str, org: str) -> list:
         tasks.append(run_process(args))
     # Will capture the task if failed
     processes = await asyncio.gather(*tasks, return_exceptions=True)
+    for process in processes:
+        logger.info("Logs for process:")
+        logger.info(process["stdout"])
+        logger.info(process["stderr"])
+
     if all(not isinstance(result, Exception) for result in processes):
         return "✅"
     return "At least one model failed"