diff --git a/benchmark/benchmark.py b/benchmark/benchmark.py index d26bb55ed..b692073d3 100755 --- a/benchmark/benchmark.py +++ b/benchmark/benchmark.py @@ -86,6 +86,10 @@ def main( summarize_results(dirname) return + if "AIDER_DOCKER" not in os.environ: + print("Warning: benchmarking runs unvetted code from GPT, run in a docker container") + return + assert BENCHMARK_DNAME.exists() and BENCHMARK_DNAME.is_dir(), BENCHMARK_DNAME assert ORIGINAL_DNAME.exists() and ORIGINAL_DNAME.is_dir(), ORIGINAL_DNAME diff --git a/benchmark/docker.sh b/benchmark/docker.sh index 51803404a..0182cdd34 100755 --- a/benchmark/docker.sh +++ b/benchmark/docker.sh @@ -6,5 +6,6 @@ docker run \ -v `pwd`/tmp.benchmarks/.:/benchmarks \ -e OPENAI_API_KEY=$OPENAI_API_KEY \ -e HISTFILE=/aider/.bash_history \ + -e AIDER_DOCKER=1 \ aider-benchmark \ bash