unsloth

/

Llama-4-Maverick-17B-128E-Instruct-GGUF

like 45

Follow

Unsloth AI 19.7k

Using the 4.5-bit Q4_K_XL 243GB model with the mmproj-f16.gguf on LM studio 0.3.32, I get various issues:

1. Placed the mmproj-f16.gguf in the same location as the model, the model loaded and LM studio indicated it had vision, but would not accept images. Error:

"This message contains no content. The AI has nothing to say." and "model does not support images. please use a model that does."

2. Now I am getting "clip_init: failed to load model C:/.../Llama-4-Maverick-17B-128E-Unsloth/mmproj-F16.gguf': load_tensors: failed to seek for tensor v.positon_embd.weight

mtmd_init_from_file: error: Failed to load LCIP model from C:/.../Llama-4-Maverick-17B-128E-Unsloth/mmproj-F16.gguf

Any advice?