High-Resolution Image Synthesis with Latent Diffusion Models

dLLM: Simple Diffusion Language Modeling

verl: Volcano Engine Reinforcement Learning for LLMs

[NeurIPS' 25] Benchmark for evaluating TTS models on complex prosodic, expressiveness, and linguistic challenges.

Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

Fast and memory-efficient exact kmeans

Compute WER and SER for speech recognition evaluation

MiMo-Audio: Audio Language Models are Few-Shot Learners

Qwen3-omni is a natively end-to-end, omni-modal LLM developed by the Qwen team at Alibaba Cloud, capable of understanding text, audio, images, and video, as well as generating speech in real time.

个人构建MoE大模型：从预训练到DPO的完整实践

[TMLR 2025🔥] A survey for the autoregressive models in vision.

FlashCosyVoice: A lightweight vLLM implementation built from scratch for CosyVoice.

Official code for "F5R-TTS: Improving Flow-Matching based Text-to-Speech with Group Relative Policy Optimization"

Text-audio foundation model from Boson AI

Step-Audio 2 is an end-to-end multi-modal large language model designed for industry-strength audio understanding and speech conversation.

[CVPR'2025] VoCo-LLaMA: This repo is the official implementation of "VoCo-LLaMA: Towards Vision Compression with Large Language Models".

Update ASR paper everyday

CUDA Templates and Python DSLs for High-Performance Linear Algebra

A native-PyTorch library for large scale M-LLM (text/audio) training with tp/cp/dp.

Bert-VITS2项目bug多且教程不友好。本proj尽可能修复了Bert-vits2项目的bug，并且可一键启动训练。仅需50条目标说话人语音，获得稳定、快速的TTS模型。

An easy-to-use, fast, and easily integrable tool for evaluating audio LLM

A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

（WIP）long form speech generatoins

SGLang is a fast serving framework for large language models and vision language models.

A high-throughput and memory-efficient inference and serving engine for LLMs

DH long context env in jax