evaluation result return 0 #2
Description
Hello, yesterday I used evaluate_all_results.py to test Gemini-3-Pro on the 70 problems and got a score of 9.2. However, today when I ran the same command with the same API key on the same dataset, all results returned 0. Could this be due to evaluation instability?