Resolving Interference When Merging Models
Paper
•
2306.01708
•
Published
•
16
This is a merge of pre-trained language models created using mergekit.
This model was merged using the TIES merge method using Qwen/Qwen3-8B-Base as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model: Qwen/Qwen3-8B-Base
models:
- model: allura-forge/q3-8b-ft-ep2-merged
parameters:
weight: 0.75
density: 0.9
- model: allura-org/remnant-qwen3-8b
parameters:
weight: 0.25
density: 0.5
merge_method: ties
dtype: bfloat16