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

[sujitvasanth]

Overview

When the Gemma 4 assistant GGUF is loaded via llama_model_load_mtp_from_file, its block tensors (blk.0-3.*), token_embd, output_norm and rope_freqs share identical names with the target model's tensors. This makes it impossible to uniquely target MTP assistant tensors via -ot rules for GPU placement.

Fix: after loading the assistant from file, rename all tensors not already prefixed with 'mtp.' to 'mtp.<original_name>'. This is done purely in-memory on the tensors_by_name vector and the ggml_tensor name field — the GGUF file and published arch names are unchanged.

After this change, all MTP assistant tensors are addressable as mtp.blk.N.*, mtp.token_embd.weight, mtp.output_norm.weight etc, and can be pinned with:

-ot 'mtp..*=CUDA0'
this keads to speedups on multi GPU systems

This is important on dual GPU systems as splitting MTP head slows down the inference.

Requirements

• I have read and agree with the contributing guidelines
• AI usage disclosure: Yes, co-wrote with Claude, manually edited the code and I have reviewed the code, compiled and tested and all works on ubuntu 20.04, giving further speedup.