Chat with boxwrench/atomic-llama-cpp-turboquant — AI on the actual code · gitaskhub

boxwrench/atomic-llama-cpp-turboquant↗

llama.cpp fork with TurboQuant WHT-rotated KV cache & weight compression + Gemma 4 MTP and Qwen 3.6 NextN speculative decoding (+30-50% throughput).

Stars · 1

Language · C++

License · MIT

Ask anything about this repo to start.