https://github.com/andrewdarnall/the-code-unmasker
SpringBoot-based microserviced web app which unmasks, using CodeBERT MLM, a code prompt
https://github.com/andrewdarnall/the-code-unmasker
codebert docker language-model masked-language-models microservices spring-boot
Last synced: 2 months ago
JSON representation
SpringBoot-based microserviced web app which unmasks, using CodeBERT MLM, a code prompt
- Host: GitHub
- URL: https://github.com/andrewdarnall/the-code-unmasker
- Owner: AndrewDarnall
- Created: 2024-02-22T10:29:30.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-10-14T22:53:04.000Z (7 months ago)
- Last Synced: 2025-01-20T21:57:52.754Z (4 months ago)
- Topics: codebert, docker, language-model, masked-language-models, microservices, spring-boot
- Language: Jupyter Notebook
- Homepage:
- Size: 71.3 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# The Code Unmasker
Large Language Models (LLMs) have made a significant impact on the technology industry, fundamentally transforming the landscape. However, this technology is not yet positioned to replace software engineers. Instead, it can serve as a powerful tool to assist engineers in various tasks, enhancing their daily workflows.
A notable example of such LLMs is Masked Language Models (MLMs), with `CodeBERT` being one prominent instance. Fine-tuned on a diverse dataset of programming languages, CodeBERT is capable of predicting masked tokens in a given text prompt, although its accuracy may vary depending on the input.
This project deploys a simple masked language model as a microservice. It operates alongside another microservice, which accepts code input from a client, processes it, and transmits it to the MLM microservice via a message queue system, specifically `RabbitMQ`.
-----
## Requirements
| Component | Version |
|---------------|------------|
| Docker | `20.10.5` |
-----
## Project Structure
-----
## Usage