Resolving Interference When Merging Models
Paper
•
2306.01708
•
Published
•
15
This is a merge of pre-trained language models created using mergekit.
This model was merged using the TIES merge method using coffiee/s31 as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: coffiee/s31
# no parameters necessary for base model
- model: rwh/s1
parameters:
density: 0.5
weight: 0.5
- model: tomaszki/stablelm-5-copy
parameters:
density: 0.5
weight: 0.3
merge_method: ties
base_model: coffiee/s31
parameters:
normalize: true
dtype: bfloat16
Base model
tomaszki/stablelm-5-copy