Llama freezes when loading model Llama-4-Scout-17B-16E-Instruct GGUF with Multimodal support

[CalculonPrime]

I filed this bug again Ooba's Web UI, but diving into his code a bit, it looks like he's just calling into llama to do the loading, so the bug is likely in llama. Should I file an issue here then?

Copying from the bug reported on Ooba's website:

You'll hit the problem with either of these Unsloth quants of Llama-4-Scout-17B-16E-Instruct:

• https://huggingface.co/unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF/tree/main/UD-Q6_K_XL or
• https://huggingface.co/unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF/tree/main/UD-Q8_K_XL

You can use either of the mmproj files:

• https://huggingface.co/unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF/blob/main/mmproj-BF16.gguf or
• https://huggingface.co/unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF/blob/main/mmproj-F16.gguf

All files were verified via SHA256 hashes.

The quants work fine when multimodal is not loaded.

If I run Filemon to see what's happening when the bug occurs, I see that when the main model file is done loading, it starts to load the mmproj file, loads about 25K of it, and then closes the file and hangs.