Stars
TraceRL & TraDo-8B: Revolutionizing Reinforcement Learning Framework for Diffusion Large Language Models
verl: Volcano Engine Reinforcement Learning for LLMs
DiffusionNFT: Online Diffusion Reinforcement with Forward Process
Repo for SeedVR2 & SeedVR (CVPR2025 Highlight)
🚀 SenseFlow: Scaling Distribution Matching for Flow-based Text-to-Image Distillation
Official implementation of Inductive Moment Matching
Official repo for VGen: a holistic video generation ecosystem for video generation building on diffusion models
[ICLR 2025] Distilled Decoding 1: One-step Sampling of Image Auto-regressive Models with Flow Matching
[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models
A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations
xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism
Development repository for the Triton language and compiler
Generative Models by Stability AI
FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.
Model Compression Toolbox for Large Language Models and Diffusion Models
[MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention
QLoRA: Efficient Finetuning of Quantized LLMs
[ACL 2024] A novel QAT with Self-Distillation framework to enhance ultra low-bit LLMs.
Code Repository of Evaluating Quantized Large Language Models
A pytorch quantization backend for optimum
General technology for enabling AI capabilities w/ LLMs and MLLMs
GPT-Fathom is an open-source and reproducible LLM evaluation suite, benchmarking 10+ leading open-source and closed-source LLMs as well as OpenAI's earlier models on 20+ curated benchmarks under al…
Skywork series models are pre-trained on 3.2TB of high-quality multilingual (mainly Chinese and English) and code data. We have open-sourced the model, training data, evaluation data, evaluation me…
AKShare is an elegant and simple financial data interface library for Python, built for human beings! 开源财经数据接口库