Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
A lightweight, rollbackable, and visual Long-Term Memory Server for MCP Agents. Say goodbye to Vector RAG and amnesia. Empower your AI with persistent, graph-like structured memory across any model, session, or tool. Drop-in replacement for OpenClaw.
WeeLLM runs large diffusion models with as little as 4 GB of VRAM, without any quantization. It dynamically determines how many layers can fit within the available VRAM and streams the text encoder and transformer layers to the GPU layer by layer, enabling inference on hardware with limited VRAM. It supports both safetensors and GGUF models.
Low-latency collective transport and distributed LLM serving stack for switchless GB10 clusters.
Video masking tools for ComfyUI: process a masked subject at high resolution inside a moving crop, then paste it back without jitter or visible seams
β
This is Meta's fork of the CPython runtime. The name "cinder" here is historical, see https://github.com/facebookincubator/cinderx for the Python extension / JIT compiler.
Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.
Probabilistic programming with NumPy powered by JAX for autograd and JIT compilation to GPU/TPU/CPU.
A lightweight LLVM python binding for writing JIT compilers
π Ansible playbook which helps you host various FOSS services as Docker containers on your own server
DSL and compiler framework for automated finite-differences and stencil computation
reverse engineering docs.
JIT-Agent
Zero-copy MPI communication of JAX arrays, for turbo-charged HPC applications in Python :zap: