jaspionjader/Kosmos-EVAA-Franken-v38-8B

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the SLERP merge method.

Models Merged

The following models were included in the merge:

• jaspionjader/fct-18-8b
• jaspionjader/fct-14-8b

Configuration

The following YAML configuration was used to produce this model:

slices:
- sources:
  - model: jaspionjader/fct-18-8b
    layer_range:
    - 0
    - 32
  - model: jaspionjader/fct-14-8b
    layer_range:
    - 0
    - 32
merge_method: slerp
base_model: jaspionjader/fct-18-8b
parameters:
  t:
  - filter: self_attn
    value:
    - 0.07
    - 0.07
    - 0.07
    - 0.07
    - 0.07
  - filter: mlp
    value:
    - 0.07
    - 0.07
    - 0.07
    - 0.07
    - 0.07
  - value: 0.07
dtype: bfloat16