Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time
Paper
•
2203.05482
•
Published
•
7
This is a merge of pre-trained language models created using mergekit.
This model was merged using the Linear merge method.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: canopylabs/3b-hi-ft-research_release
parameters:
weight: 1.0
- model: canopylabs/3b-zh-ft-research_release
parameters:
weight: 1.0
- model: canopylabs/3b-de-ft-research_release
parameters:
weight: 1.0
- model: canopylabs/3b-fr-ft-research_release
parameters:
weight: 1.0
- model: canopylabs/orpheus-3b-0.1-ft
parameters:
weight: 1.0
merge_method: linear
dtype: float16