Stars
High-performance safetensors model loader
Official Python inference and LoRA trainer package for the LTX-2 audio–video generative model.
Block Diffusion for Ultra-Fast Speculative Decoding
Produce your own Dynamic 3.0 Quants and achieve optimum accuracy & SOTA quantization performance! Input your VRAM and RAM and the toolchain will create a GGUF recipe tuned to your system within sec…
Sharp Monocular View Synthesis in Less Than a Second
TurboDiffusion: 100–200× Acceleration for Video Diffusion Models
A framework for few-shot evaluation of language models.
slime is an LLM post-training framework for RL Scaling.
REAP: Router-weighted Expert Activation Pruning for SMoE compression
A unified library of SOTA model optimization techniques like quantization, pruning, distillation, speculative decoding, etc. It compresses deep learning models for downstream deployment frameworks …
OpenTUI is a library for building terminal user interfaces (TUIs)
Soprano: Instant, Ultra-Realistic Text-to-Speech
MCP server for Godot Engine - AI assistant integration
GRPO training code which scales to 32xH100s for long horizon terminal/coding tasks. Base agent is now the top Qwen3 agent on Stanford's TerminalBench leaderboard.
Android Virtual Keyboard Input via ADB (Useful for Test Automation)
A high-performance observability data pipeline.
MoE training for Me and You and maybe other people
Native and Compact Structured Latents for 3D Generation
A LLM trained only on data from certain time periods to reduce modern bias
An Open Phone Agent Model & Framework. Unlocking the AI Phone for Everyone
A Claude Skill to give your agent the ability to use a web browser