Leigh L. Klotz, Jr.

Cannot use llamafile models on any input containing HTML '<img src="' text

1

If you are doing text inference and the prompt contains an HTML sequence '

This PR simply removes the feature. Maybe a CLI flag to enable or disable would be appropriate, or perhaps I am misunderstanding the feature and there is a different way...

In https://github.com/Mozilla-Ocho/llamafile/releases/tag/0.6.2 >Other models: > >https://hf.co/jartine/jartine/wizardcoder-13b-python >https://hf.co/jartine/jartine/Nous-Hermes-Llama2-llamafile >https://hf.co/jartine/jartine/dolphin-2.5-mixtral-8x7b-llamafile The userid is duplicated, resulting in 404 on dereference.

--silent-prompt is now the same as --verbose

1

https://github.com/Mozilla-Ocho/llamafile/blob/956fb587b30f14c195465e7d281c5721767677bc/llama.cpp/common.cpp#L932 This may be a merge issue with llama.cpp; it looks like the commit that introduced --silent-prompt got only partially merged. https://github.com/Mozilla-Ocho/llamafile/commit/28d3160145e7595f84a9d7ec7f24e11ba153f796