EAGLE-4

Estimate whether a Hugging Face model fits and fine-tunes on your local GPU.

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

cuda-oxide is an experimental Rust-to-CUDA compiler that lets you write (SIMT) GPU kernels in safe(ish), idiomatic Rust. It compiles standard Rust code directly to PTX — no DSLs, no foreign language bindings, just Rust.

DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms

DFlash: Block Diffusion for Flash Speculative Decoding

分享AI Infra知识&代码练习：PyTorch、vLLM/SGLang、slime/vime框架入门⚡️、性能加速🚀、大模型基础🧠、AI软硬件🔧等