chanwit/flux-base-optimized

1---2license: apache-2.03language:4- en5---6 7# Flux-Base-Optimized8 9`flux-base-optimized` is the base model for finetuning the series of `flux-7b` models.10It is hierarchical SLERP merged from the following models11 12  * mistralai/Mistral-7B-v0.1 (Apache 2.0)13  * teknium/OpenHermes-2.5-Mistral-7B (Apache 2.0)14  * Intel/neural-chat-7b-v3-3 (Apache 2.0)15  * meta-math/MetaMath-Mistral-7B (Apache 2.0)16  * openchat/openchat-3.5-0106 was openchat/openchat-3.5-1210 (Apache 2.0)17 18Here's how we did the hierarchical SLERP merge.19```20                [flux-base-optimized]21                         ↑22                         |23               [stage-1]-+-[openchat]24                   ↑25                   |26         [stage-0]-+-[meta-math]27             ↑28             |29[openhermes]-+-[neural-chat]30```31 32