giannisanni

giannisanni

GitHub

Repositories

1

Total stars

239

Top repositories

SSD-streaming inference engine for giant MoE models (Rust + CUDA). GLM 5.2 743B at 2 tok/s and Hy3 295B at 7 tok/s on two consumer 16GB GPUs. Zero-config multi-GPU: measures PCIe bandwidth, places attention and hot experts where they fit.

Created Jul 2026

cuda gguf glm inference-engine llm
Total239