This is a gradient blockmerge (0.8,0.2) of two Mistral models.
The logic model is a SLERP merge of https://huggingface.co/teknium/OpenHermes-2.5-Mistral-7B and https://huggingface.co/openchat/openchat_3.5
The prose model is a SLERP merge of https://huggingface.co/NousResearch/Nous-Capybara-7B-V1.9 and https://huggingface.co/HuggingFaceH4/zephyr-7b-beta
- Downloads last month
- 11