This folder contains evaluation harness for evaluating agents on the logic reasoning benchmark ProntoQA and ProofWriter.
Create a config.toml file if it does not exist at the root of the workspace.
Add the following configurations:
[core]
max_iterations = 100
cache_dir = "/tmp/cache"
ssh_hostname = "localhost"
[sandbox]
enable_auto_lint = true
# TODO: Change these to the model you want to evaluate
[llm.eval_gpt4_1106_preview_llm]
model = "gpt-4-1106-preview"
api_key = "XXX"
temperature = 0.0
[llm.eval_some_openai_compatible_model_llm]
model = "openai/MODEL_NAME"
base_url = "https://OPENAI_COMPATIBLE_URL/v1"
api_key = "XXX"
temperature = 0.0The following code will run inference on the first example of the ProntoQA dataset, using OpenDevin 0.6.2 version.
./evaluation/logic_reasoning/scripts/run_infer.sh ProntoQA eval_gpt4_1106_preview_llm 0.6.2 1