FuseChat: Knowledge Fusion of Chat Models
Paper
• 2408.07990 • Published
• 14
This is a merge of pre-trained language models created using mergekit.
This model was merged using the SCE merge method using NewEden/ckpt-44 as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
base_model: NewEden/ckpt-44
merge_method: sce
dype: bfloat16
out_dtype: bfloat16
Parameters:
select_topk: 0.15
models:
- model: NewEden/checkpoint-30
- model: NewEden/ckpt-36
- model: NewEden/ckpt-42