Metadata-Version: 2.3
Name: pyligent
Version: 0.2.3
Summary: Diligent framework for python
Keywords: reasoning,llm,transformers,ml
Author: Dmitry Beresnev, Petr Anokhin, Vladimir Makharev, Roman Khalikov
Author-email: Dmitry Beresnev <d.beresnev.work@gmail.com>, Petr Anokhin <anokhin@airi.net>
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Dist: bitsandbytes>=0.48.1
Requires-Dist: datasets>=4.0.0
Requires-Dist: ipykernel>=6.30.1
Requires-Dist: ipywidgets>=8.1.7
Requires-Dist: loguru>=0.7.3
Requires-Dist: peft>=0.17.1
Requires-Dist: python-dotenv>=1.1.1
Requires-Dist: tqdm>=4.67.1
Requires-Dist: transformers[torch]>=4.56.0
Requires-Dist: trl>=0.22.2
Requires-Dist: tensorboard>=2.20.0
Requires-Dist: networkx[default]>=3.5
Requires-Dist: nvidia-ml-py>=13.580.82
Requires-Dist: orjson>=3.11.5
Requires-Dist: accelerate>=1.12.0
Requires-Dist: gpustat>=1.1.1
Requires-Dist: pyarrow>=22.0.0
Requires-Dist: vllm>=0.13.0
Requires-Dist: numpy>=2.2.6
Requires-Dist: torch-c-dlpack-ext>=0.1.5
Requires-Dist: opencv-python-headless
Requires-Dist: torch>=2.8.0 ; extra == 'cpu'
Requires-Dist: torch>=2.8.0 ; extra == 'cu128'
Requires-Dist: torch>=2.8.0 ; extra == 'cu129'
Requires-Python: >=3.13
Provides-Extra: cpu
Provides-Extra: cu128
Provides-Extra: cu129
Description-Content-Type: text/markdown

# Pyligent

![Python](https://img.shields.io/badge/python-3.13-blue.svg)
[![PyPI version](https://badge.fury.io/py/pyligent.svg)](https://badge.fury.io/py/pyligent)

**Pyligent** is a framework for reinforcement learning through search-based reasoning, implementing concepts from [From Reasoning to Super-Intelligence: A Search-Theoretic Perspective](https://arxiv.org/abs/2507.15865).

## Installation

**Note:** Pyligent requires [PyTorch](https://pytorch.org/). Please install the version appropriate for your hardware before installing Pyligent.

### 1. Install PyTorch

**Using `uv`:**

Check the [official guide](https://docs.astral.sh/uv/guides/integration/pytorch/#using-a-pytorch-index)


**For CPU-only environments:**

```bash
pip install torch --index-url https://download.pytorch.org/whl/cpu
```

**For CUDA 12.8 environments:**

```bash
pip install torch --index-url https://download.pytorch.org/whl/cu128
```

*(For other CUDA versions, please refer to the [PyTorch Get Started](https://pytorch.org/get-started/locally/) guide.)*

### 2. Install Pyligent

Once PyTorch is installed:

```bash
pip install pyligent
```

or

```bash
uv add pyligent
```

## Quick Start

```python
import pyligent

# Your quick start example here
```

## Key Features

- Search-based reasoning for transformer models
- PyTorch-based implementation with PEFT support
- GPU acceleration support (via CUDA-enabled PyTorch)
- Fine-tuning and evaluation pipelines

## Requirements

- Python ≥ 3.13
- PyTorch ≥ 2.8.0 (must be installed manually, see Installation above)

<!-- ## Documentation

For complete documentation, examples, and usage guides, visit our [GitHub repository](https://github.com/crogs-foundation/diligent-learner).

## Citation

If you use Pyligent in your research, please cite:

```bibtex
@article{yourpaper2026,
  title={Your Paper Title},
  author={Your Name},
  journal={arXiv preprint},
  year={2026}
}
```

## Support

- 📝 [Documentation](https://github.com/crogs-foundation/diligent-learner)
- 🐛 [Issue Tracker](https://github.com/crogs-foundation/diligent-learner/issues)
- 💬 [Discussions](https://github.com/crogs-foundation/diligent-learner/discussions) -->
