Mistral-7B-Neural-Story-mix / mergekit_config.yml
pcuenq's picture
pcuenq HF staff
Duplicate from julien-c/Mistral-7B-Neural-Story-mix
b8e00bf
raw
history blame
219 Bytes
models:
- model: mistralai/Mistral-7B-Instruct-v0.2
parameters:
weight: 1.0
- model: NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
parameters:
weight: 0.75
merge_method: linear
dtype: float16