Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/AlgorithmicResearchGroup/ML-Research-Agent-Public
Public, general purpose agent for ML Research Benchmark. This agent provides a foundation for comparing and evaluating machine learning research and development tasks that agents can perform.
https://github.com/AlgorithmicResearchGroup/ML-Research-Agent-Public
Last synced: 3 days ago
JSON representation
Public, general purpose agent for ML Research Benchmark. This agent provides a foundation for comparing and evaluating machine learning research and development tasks that agents can perform.
- Host: GitHub
- URL: https://github.com/AlgorithmicResearchGroup/ML-Research-Agent-Public
- Owner: AlgorithmicResearchGroup
- License: agpl-3.0
- Created: 2024-10-18T15:56:11.000Z (3 months ago)
- Default Branch: master
- Last Pushed: 2024-11-12T16:03:56.000Z (about 2 months ago)
- Last Synced: 2024-11-12T16:35:47.698Z (about 2 months ago)
- Language: Python
- Homepage:
- Size: 3.72 MB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome_ai_agents - Ml-Research-Agent-Public - Public, general purpose agent for ML Research Benchmark. This agent provides a foundation for comparing and evaluating machine learning r… (Building / Benchmarks)
- awesome_ai_agents - Ml-Research-Agent-Public - Public, general purpose agent for ML Research Benchmark. This agent provides a foundation for comparing and evaluating machine learning r… (Building / Benchmarks)
README
# ML Research Benchmark Baseline Agent
This is our public baseline research and development agent. It is an agentic system designed to serve as a baseline for various AI and machine learning tasks. This agent provides a foundation for comparing and evaluating machine learning research and development tasks that agents can perform. This agent is a simple, single-agent system that uses a task planner and a tools to perform machine learning tasks.
## Features
- Supports multiple AI/ML tasks
- Compatible with different LLM providers (OpenAI, Anthropic)
- Dockerized for easy deployment and reproducibility
[](https://www.youtube.com/watch?v=Xhpe8MHk56w)
## Available Tools
The AI Research Benchmark Baseline Agent comes equipped with a variety of tools to assist in different AI and machine learning tasks:
1. **Bash Tool**: Executes bash commands and scripts.
2. **Code Tool**: Manages code operations including writing, inserting, replacing, and deleting code.
3. **GitHub Tool**: Interacts with GitHub repositories to get README files, list files, and retrieve file contents.
4. **Semantic Scholar Tool**: Searches for academic papers, retrieves paper details, citations, and downloads papers.
5. **Python Tool**: Executes Python code.
6. **Return Function Tool**: Handles task completion.
7. **Scratchpad Tool**: Provides a scratchpad for experiment note-taking and temporary storage.
8. **Thought Tool**: Allows the agent to process and record thoughts.
9. **Long-Term Memory Tool**: Manages long-term memory storage and retrieval.
These tools can be used individually or in combination to tackle a wide range of AI research and benchmark tasks. The agent can seamlessly switch between tools as needed for complex operations.
## Prerequisites
- Python 3.x
- Docker (for containerized execution)
## Installation
1. Clone this repository:
```bash
git clone https://github.com/AlgorithmicResearchGroup/ML-Research-Agent-Public.git
cd ML-Research-Agent-Public
```
2. Install dependencies:
```bash
pip install -r requirements.txt
```
## Usage
Step 1: Create a .env file with the following environment variables:
```bash
OPENAI =
ANTHROPIC =
YOU_API_KEY =
GITHUB_ACCESS_TOKEN =
```
### Running without Docker
Step 2a: Run the agent:
To run the agent without Docker, use the following command:
```bash
python3 run.py --prompt "" --provider ""
```
### Running with Docker
Step 2b: Run the agent with Docker:
Build for CPU:
```
docker build --build-arg BASE_IMAGE=ubuntu:22.04 -t .
```
Build for GPU:
```
docker build --build-arg BASE_IMAGE=nvidia/cuda:12.2.2-cudnn8-runtime-ubuntu22.04 -t .
```
```bash
bash run.sh \
\
\
<"cpu" or gpu_ids eg. 0> \
\
```
Example on CPU:
```bash
bash run.sh ghcr.io/algorithmicresearchgroup/ml-research-agent-public \
"train an mlp on the mnist dataset" \
openai \
"cpu" \
\
/root/ML-Research-Agent-Public/.env
```
Example on GPU:
```bash
bash run.sh ghcr.io/algorithmicresearchgroup/ml-research-agent-public \
"train an mlp on the mnist dataset" \
openai \
0 \
\
/path/to/.env
```
## Contributing
Contributions to improve the baseline agent or add new tasks are welcome. Please submit a pull request or open an issue to discuss proposed changes.
## License
AGPL-3.0
## Contact
For questions or support, please contact Algorithmic Research Group at [email protected]