CodeBERT-C Summary

Introduction

CodeBERT-C is a model based on microsoft/codebert-base-mlm, specifically trained for masked language modeling on C code. It is designed to enhance tasks such as CodeBERTScore, which evaluates code generation.

Architecture

The model architecture is based on the RoBERTa transformer, utilizing PyTorch as its underlying framework. It is compatible with Hugging Face's inference endpoints.

Training

The model was trained for 1,000,000 steps with a batch size of 32. The training dataset was codeparrot/github-code-clean, focusing on C code using the masked language modeling task.

Guide: Running Locally

To run CodeBERT-C locally:

1. Set up your environment: Ensure you have Python and PyTorch installed.
2. Install Hugging Face Transformers: Use pip install transformers.
3. Load the model: Use the Hugging Face API to load the model with from transformers import AutoModel.
4. Execute tasks: Apply the model to your desired C code tasks.

For enhanced performance, consider using cloud GPUs such as those provided by AWS or Google Cloud.

License

The documentation does not specify a license for CodeBERT-C. Users should refer to the original repository or contact the authors for licensing details.