Open source. Open possibilities.

A tiny GPT language model (~540M parameters) designed to be trained from scratch on consumer laptop hardware, with pretraining on Fineweb-edu and chat finetuning capabilities.

vLLM is a high-throughput, memory-efficient library designed for Large Language Model (LLM) inference and serving.

A high-performance C/C++ implementation for inference of OpenAI's Whisper automatic speech recognition (ASR) model, designed for lightweight and cross-platform deployment.

PaddleSeg is an end-to-end image segmentation development kit under the PaddlePaddle ecosystem, featuring 45+ model algorithms and 140+ pretrained models, covering semantic segmentation, interactive segmentation, panoptic segmentation, image matting, and 3D medical segmentation, with config-driven and API usage, spanning annotation, training, compression, and deployment.

A fast reimplementation of OpenAI's Whisper model using CTranslate2, offering improved transcription speed and reduced memory usage.

Hugging Face Transformers is a model-definition framework for machine learning across text, vision, audio, and multimodal tasks. It provides a unified API for inference and training with over 1 million pretrained checkpoints.

SGLang is a high-performance serving framework for large language models (LLMs) and multimodal models, designed for low-latency and high-throughput inference.

Collection of simple PyTorch implementations of neural networks and deep learning papers with documented side-by-side explanations, covering transformers, GANs, diffusion models, RL, optimizers, and more.