Metadata-Version: 2.1
Name: oat-llm
Version: 0.0.1
Summary: Online AlignmenT (OAT) for LLMs.
Home-page: https://github.com/sail-sg/oat
Author: Zichen Liu
Author-email: Zichen Liu <liuzc@sea.com>, Changyu Chen <chency@sea.com>
License: Apache-2.0
Project-URL: Documentation, https://github.com/sail-sg/oat#readme
Project-URL: Issues, https://github.com/sail-sg/oat/issues
Project-URL: Source, https://github.com/sail-sg/oat
Keywords: rlhf,llm,ai-alignment,rl,bandit,ai,sample-efficiency
Classifier: Development Status :: 4 - Beta
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: Implementation :: CPython
Classifier: Programming Language :: Python :: Implementation :: PyPy
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: accelerate
Requires-Dist: bitsandbytes
Requires-Dist: datasets
Requires-Dist: deepspeed ==0.15.0
Requires-Dist: dm-launchpad[tensorflow]
Requires-Dist: einops
Requires-Dist: flash-attn ==2.6.3
Requires-Dist: httpx
Requires-Dist: jsonlines
Requires-Dist: llm-blender ==0.0.2
Requires-Dist: loralib
Requires-Dist: ml-collections
Requires-Dist: mosec
Requires-Dist: msgspec
Requires-Dist: optimum
Requires-Dist: packaging
Requires-Dist: peft
Requires-Dist: protobuf ==3.20.0
Requires-Dist: pyarrow <12,>=0.6.1
Requires-Dist: python-Levenshtein ==0.25.1
Requires-Dist: scipy
Requires-Dist: torchmetrics
Requires-Dist: tqdm
Requires-Dist: transformers ==4.45.2
Requires-Dist: transformers-stream-generator
Requires-Dist: vllm ==0.6.2
Requires-Dist: wandb
Requires-Dist: wheel

# Efficient Online LLM Alignment

[![PyPI - Version](https://img.shields.io/pypi/v/oat.svg)](https://pypi.org/project/oat)
[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/oat.svg)](https://pypi.org/project/oat)

-----


- [Efficient Online LLM Alignment](#efficient-online-llm-alignment)
  - [Installation :wrench:](#installation-wrench)
  - [License](#license)

## Installation :wrench:
`online-align` requires Python 3.7 or above. Install the latest PyPI package for Linux x86_64 or macOS x86_64/ARM64 with:
1. Prepare a python environment with `python>=3.10`.
2. Clone this repository and install it in develop mode:
```console
git clone git@github.com:sail-sg/online-align.git
pip install -r requirements.txt
pip install flash-attn==2.6.3
pip install -e .
```

## License

`oat` is distributed under the terms of the [Apache2](https://www.apache.org/licenses/LICENSE-2.0) license.
