Dolphin Mistral 24B Venice Edition: Version 1.2 Update

#2425
by douganger - opened

Dolphin Mistral 24B Venice Edition has been updated to version 1.2.

Link: https://huggingface.co/dphn/Dolphin-Mistral-24B-Venice-Edition/commit/337ce042026e598256f4bbb8178116b59c9e302d

I would love to see either a new quant for this version, or an update to the existing repo at mradermacher/Dolphin-Mistral-24B-Venice-Edition-i1-GGUF.

Thank you so much for all the work you do making AI models accessible!

@simonko912 Can you please clone, version and queue this one? This version 1.2 update is such a massive change compared to the original model that I personally find it crazy they even replaced the original model. They added vision, increased context length and improved tool calling. We definitely don't want to replace our existing version with the new one and instead offer it separately.

In case anyone is still looking for this, I quantized version 1.2 here. This has been a learning project for me and I'm open to feedback from your team if there are ways I could improve it.

Sign up or log in to comment