A Datacenter Scale Distributed Inference Serving Framework
-
Updated
Oct 31, 2026 - Rust
A Datacenter Scale Distributed Inference Serving Framework
Constrained Decoding of Diffusion LLMs with Context-Free Grammars.
Local-first AI image generation toolkit. Pull models, train LoRAs, generate images. One CLI, no glue code.
Pixel art sprite generator. Three diffusion models (1M-17M params). Pure Rust, Metal/CUDA/CPU. No Python, no cloud.
Train conditional vector fields with flow matching
QIS — Local-first generative AI studio (Text-to-Image, Image-to-Image, Turbo diffusion, WD14 Danbooru interrogator, Qwen3-VL vision reverse-engineering, Civitai explorer, SIMD Lanczos resizer, ComfyUI backend).
Native Rust engine for neural network inference and training: LLM, image, video, speech and music models on hand-written CUDA kernels JIT-compiled via NVRTC for any NVIDIA sm_80+ GPU (native NVFP4/MXFP8 on Blackwell sm_120+). .syn bundles in NVFP4/MXFP8/SQ1–SQ8, GGUF loads directly. No PyTorch, no Python.
EAFAR — Event-Driven Asynchronous Fuzzy Automata Runtime in Rust: sparse worlds where only changed regions wake up. Bit-exact determinism, zero dependencies.
The latest generation of Claude is Anthropic's state-of-the-art AI model series (headlined by Claude 5 / Claude Opus 5 and Claude Sonnet 5), engineered for advanced reasoning, agentic coding, multi-step problem solving, and complex enterprise workflows with an expanded 1-million-token context window.
A collection of tools for working with generative models. This is a work in progress as I need more tools for my own use cases
Diffusion of Quantity (Heat / Chemicals) Along a One-Dimensional Line. High Concentration in the Center — Spreads Out Similarly to Noise in a Diffusion Model
Open-source video generation and motion-guided neural synthesis pipeline in Rust
Pure-Rust LLM, diffusion and embedding inference: Gemma 4/3, Qwen 3.x, MLX 4-bit, Stable Diffusion with LoRA, sentence-transformers embedders. CPU-first with Metal, CUDA and WebGPU (wasm) backends.
A simple gas diffusion simulation built in Rust
ComfyUI LTX-Video generation runner and node pipeline on Apple Silicon in Rust
High-Performance Fluid Dynamics Solver
A simple genetic algorithm to search for good constants in a diffusion function
Native Apple Silicon MLX Video generation and inference in Rust (Wan2.1, Wan2.2, LTX-2)
To associate your repository with the diffusion topic, visit your repo's landing page and select "manage topics."