llama: prefix MTP assistant tensors with 'mtp.' on load allowing use of -ot 'mtp..*=CUDA0' flag - #7

llama: prefix MTP assistant tensors with 'mtp.' on load allowing use of -ot 'mtp..*=CUDA0' flag#7

Ooooze merged 1 commit into

AtomicBot-ai:feature/turboquant-kv-cacheAtomicBot-ai/atomic-llama-cpp-turboquant:feature/turboquant-kv-cachefrom

sujitvasanth:fix/mtp-assistant-tensor-prefixsujitvasanth/atomic-llama-cpp-turboquant:fix/mtp-assistant-tensor-prefixCopy head branch name to clipboard