Code Generation LM Evaluation Harness

Tasks | Usage | Contribution | Documentation | BigCode

Features

This is a framework for the evaluation of code generation models. This is a work in progress part of the BigCode project, and is inspired from EleutherAI/lm-evaluation-harness for evaluating language models in general. We welcome contributions to fix issues, enhance features and add new benchmarks. You can find a contribution guides in docs/guide.md and CONTRIBUTING.md and more documentation in docs/README.md.

Below are the features and tasks of this framework:

Any autoregressive model available on Hugging Face hub can be used, but we recommend using code generation models trained specifically on Code such as CodeParrot, InCoder and CodeGen.
3 code generation Python tasks (with unit tests): HumanEval, APPS and MBPP.
CoNaLa for Python code generation (2-shot setting and evaluation with BLEU score)
Concode for Java code generation (2-shot setting and evaluation with BLEU score)
Code to text task from CodeXGLUE (zero-shot & fine-tuning) for 6 languages: Python, Go, Ruby, Java, JavaScript and PHP.
3 multilingual downstream classification tasks: Java Complexity prediction, Java code equivalence prediction, C code defect prediction.

Setup

git clone https://github.com/bigcode-project/bigcode-evaluation-harness.git
cd bigcode-evaluation-harness

Install torch based on your device type and the other packages using:

pip install -r requirements.txt

To run the DS-1000 benchmark, additional constraints must be resolved.

# python version must be 3.7.10
pip install -e ".[ds1000]" # installs all additional dependencies except PyTorch
# torch==1.12.1 required. Download version with relevant GPU support etc., e.g.,
pip install torch==1.12.1+cu116 --extra-index-url https://download.pytorch.org/whl/cu116

# to suppress any tensorflow optimization warnings, 
# precede call to "accelerate launch" with "TF_CPP_MIN_LOG_LEVEL=3"

# on some systems, tensorflow will attempt to allocate all GPU memory
# to its process at import which will raise a CUDA out-of-memory error
# setting "export TF_FORCE_GPU_ALLOW_GROWTH=true" resolves this

Also make sure you have git-lfs installed and are logged in the Hub

huggingface-cli login

We use accelerate to generate code/text in parallel when multiple GPUs are present (multi-GPU mode). You can configure it using:

accelerate config

This evaluation harness can also be used in an evaluation only mode, you can use a Multi-CPU setting. For this mode you can also find an example of setup instructions in evaluation_setup.sh, where we configure the environment and evaluate some MBPP generations donwloaded from the hub.

Usage

You can use this evaluation harness to generate text solutions to code benchmarks with your model, to evaluate (and execute) the solutions or to do both. While it is better to use GPUs for the generation, the evaluation only requires CPUs. So it might be beneficial to separate these two steps. By default both generation and evaluation are performed.

For more details on how to evaluate on the tasks, please refer to the documentation in docs/README.md.

Generation and evaluation

Below are some examples to generate and evaluate on some tasks.

accelerate launch  main.py \
  --model <MODEL_NAME> \
  --tasks <TASK_NAME> \
  --limit <NUMBER_PROBLEMS> \
  --max_length_generation <MAX_LENGTH> \
  --temperature <TEMPERATURE> \
  --do_sample True \
  --n_samples 100 \
  --batch_size 10 \
  --allow_code_execution \
  --save_generations

limit represents the number of problems to solve, if it's not provided all problems in the benchmark are selected.
allow_code_execution is for executing the generated code: it is off by default, read the displayed warning before calling it to enable execution.
Some models with custom code on the HF hub like SantaCoder require calling --trust_remote_code, for private models add --use_auth_token.
save_generations saves the post-processed generations in a json file. You can also save references by calling --save_references

Some tasks don't require code execution such as codexglue_code_to_text-<LANGUAGE>/codexglue_code_to_text-python-left/conala/concode that use BLEU evaluation. In addition, we generate one candidate solution for each problem in these tasks, so use n_samples=1 and batch_size=1. (Note that batch_size should always be equal or less than n_samples).

For APPS tasks, you can use n_samples=1 for strict and average accuracies (from the original APPS paper) and n_samples>1 for pass@k.

Generation only

If you want to generate solutions without executing and evaluating the code, call --generation_only, in addition to the instructions above. This will save the solutions in a json file in the working directory.

This can be useful if you don't want to execute code in the machine you're using for generations for security or efficiency reasons. For instance, you can do the generations on multiple GPUs, but switch to a multiple workers CPU machine for the execution, which can save money and time.

Evaluation only

If you already have the generations in a json file from this evaluation harness and want to evaluate them, specify the path of the generations via the generation_path argument. You may need to reconfigure accelerate to use multiple CPUs. For this mode, you can also find an example of setup instructions in evaluation_setup.sh.

Below is an example, be mind of specifying arguments proper to the task you are evaluating on, and note that model value here only serves for documenting the experiment.

accelerate launch  main.py   --tasks mbpp  --allow_code_execution  --generations_path generations.json  --model incoder-temperature-08

Implementing new tasks

To implement a new task in this evaluation harness, see the guide in docs/guide. The are also contribution guidelines in this CONTRIBUTING.md

Documentation

We provide documentation for the existing benchmarks and how we make the evaluation in docs/README.md.

Remarks

Currenltly, we use parallel evaluation across multiple GPUs using accelerate, this assumes that you can fit the model in one GPU.
Please note this evaluation harness tries to cover a wide set of models, but there could still be room for improvement based on each model, some might require different prompt engineering or post-processing of the code generations.
For some scores of ongoing experiments please refer to example_scores/README.md.

Acknowledgements

We thank EleutherAI for their work on the lm-evaluation harness from which this repository is inspired.

Cite as

@software{bigcode-evaluation-harness,
  author       = {Ben Allal, Loubna and
                  Muennighoff, Niklas and
                  Von Werra, Leandro},
  title = {A framework for the evaluation of code generation models},
  howpublished = {\url{https://github.com/bigcode-project/bigcode-evaluation-harness}},
  year = 2022,
  month = December
}

ddkang1 / bigcode-evaluation-harness