evalplus
Rigourous evaluation of LLM-synthesized code - NeurIPS 2023 & COLM 2024
File Explorer
Download Latest Version (.zip)- buggy_contract.yml
- buggy_test.yml
- config.yml
- model_eval_request.yml
- ci.yml
- pre-commit.yml
- cli.md
- evalperf.md
- execution.md
- ollama.md
- __init__.py
- humaneval.py
- mbpp.py
- utils.py
- __init__.py
- _special_oracle.py
- utils.py
- __init__.py
- anthropic_request.py
- ollama_request.py
- openai_request.py
- __init__.py
- chatgpt_gen.py
- mut_gen.py
- type_mut.py
- __init__.py
- profile.py
- sampling.py
- sas.py
- select_pe_inputs.py
- select_pe_tasks.py
- __init__.py
- anthropic.py
- base.py
- bedrock.py
- google.py
- gptqmodel.py
- hf.py
- hf_gaudi.py
- ollama.py
- openai.py
- utility.py
- vllm.py
- __init__.py
- codegen.py
- config.py
- evalperf.py
- evaluate.py
- inputgen.py
- lecacy_sanitize.py
- sanitize.py
- syncheck.py
- utils.py
- overview.png
- render.gif
- requirements.txt
- test_legacy_sanitizer.py
- test_treesitter_sanitizer.py
- evaluate_coverage.py
- evaluate_runtime.py
- generate_big_input.py
- README.md
- set_cover.py
- topset_distill.py
- type_mut_for_eff.py
- hf_upload.py
- intra_model_viz.py
- pairwise_heatmap.py
- viz_by_params.py
- check_ground_truth.py
- filter_extreme.py
- fix_utils.py
- fix_v011.py
- fix_v012.py
- fix_v013.py
- fix_v014.py
- fix_v015.py
- fix_v016.py
- fix_v017.py
- fix_v018.py
- fix_v019.py
- init_ground_truth.py
- init_plus.py
- to_original_fmt.py
- check_ground_truth.py
- filter_extreme.py
- fix_v010.py
- fix_v020.py
- init_ground_truth.py
- init_plus.py
- to_original_fmt.py
- .gitignore
- __init__.py
- coverage_init.py
- minimization.py
- mutation_init.py
- README.md
- requirements.txt
- run.py
- sample_init.py
- utils.py
- checker.py
- collect_valid_solutions.py
- directory_to_jsonl.py
- filter_inputs.py
- merge_dataset.py
- render.py
- requirements.txt
- sanitize.py
- stat_plus.py
- viz_passrate.py
- zip_solutions.py
- .dockerignore
- .gitignore
- .pre-commit-config.yaml
- CITATION.cff
- Dockerfile
- LICENSE
- MANIFEST.in
- pyproject.toml
- README.md
- release.sh
- requirements-evalperf.txt
- requirements-gaudi.txt
- requirements.txt
- setup.cfg
# Installation Guide
git clone https://github.com/evalplus/evalplus
Downloads the entire project code from GitHub to your computer.
cd evalplus
Moves into the project folder you just downloaded.
2. Official Install Script
Easy Recommended- Python 3 Python is required to use pip.
pip install --upgrade "evalplus[vllm] @ git+https://github.com/evalplus/evalplus"
Installs the package published on PyPI directly โ no need to clone the source.
pip install --upgrade "evalplus[perf,vllm] @ git+https://github.com/evalplus/evalplus"
Installs the package published on PyPI directly โ no need to clone the source.
pip install packaging ninja
Installs the package published on PyPI directly โ no need to clone the source.
pip install flash-attn --no-build-isolation
Installs the package published on PyPI directly โ no need to clone the source.
Pulled directly from this repo's README.
3. Docker
Easy- Git Needed to download the project code from GitHub.
- Docker Desktop Needed to build and run containers. Install it and keep it running in the background.
docker run --rm --pull=always -v $(pwd)/evalplus_results:/app ganler/evalplus:latest \
Runs the built image as an actual container.
docker run --cap-add PERFMON --rm --pull=always -v $(pwd)/evalplus_results:/app ganler/evalplus:latest \
Runs the built image as an actual container.
<a href="https://hub.docker.com/r/ganler/evalplus" title="Docker"><img src="https://img.shields.io/docker/image-size/ganler/evalplus"></a>
Type this command into your terminal and run it.
Pulled directly from this repo's README.
4. Python
Easypip install --upgrade "evalplus[vllm] @ git+https://github.com/evalplus/evalplus"
Installs the package published on PyPI directly โ no need to clone the source.
pip install --upgrade "evalplus[perf,vllm] @ git+https://github.com/evalplus/evalplus"
Installs the package published on PyPI directly โ no need to clone the source.
pip install packaging ninja
Installs the package published on PyPI directly โ no need to clone the source.
pip install flash-attn --no-build-isolation
Installs the package published on PyPI directly โ no need to clone the source.
pip install git+https://github.com/huggingface/optimum-habana.git
Installs the package published on PyPI directly โ no need to clone the source.
Pulled directly from this repo's README.