DELLA-Merging: Reducing Interference in Model Merging through Magnitude-Based Sampling
Paper
• 2406.11617 • Published
• 10
This is a merge of pre-trained language models created using mergekit.
This model was merged using the DELLA merge method using tinycompany/Tamed-Sigma-hi as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model: tinycompany/Tamed-Sigma-hi
merge_method: della
dtype: bfloat16
models:
- model: tinycompany/Tamed-Sigma
parameters:
weight: 0.3
- model: tinycompany/Tamed-Sigma-hi
parameters:
weight: 0.4
- model: tinycompany/SigmaBoi-Instruct-nomic
parameters:
weight: 0.3