Safetensors
mistral

First impressions

#1
by pilchardus - opened

Compared to V3 with the same prompts, it seems like responses are shorter on average and the model seems overall less proactive. I may need to adjust my prompting style.

Sign up or log in to comment