Resolving Interference When Merging Models
Paper
• 2306.01708 • Published
• 17
This is a merge of pre-trained language models created using mergekit.
This model was merged using the TIES merge method using 01-ai/Yi-1.5-34B-32K as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: TheSkullery/Y1.5-MS-Mysteria-34b
parameters:
density: 0.5
weight: 0.5
- model: anthracite-org/magnum-v3-34b
parameters:
density: 0.5
weight: 0.5
merge_method: ties
base_model: 01-ai/Yi-1.5-34B-32K
parameters:
normalize: false
int8_mask: true
dtype: float16