jaspionjader/kstc-3-8b

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/kstc-2-8b
• jaspionjader/Kosmos-EVAA-Franken-Immersive-v41-8B

Configuration

The following YAML configuration was used to produce this model:

slices:
- sources:
  - model: jaspionjader/Kosmos-EVAA-Franken-Immersive-v41-8B
    layer_range:
    - 0
    - 32
  - model: jaspionjader/kstc-2-8b
    layer_range:
    - 0
    - 32
merge_method: slerp
base_model: jaspionjader/Kosmos-EVAA-Franken-Immersive-v41-8B
parameters:
  t:
  - filter: self_attn
    value:
    - 0.16
    - 0.63
    - 0.75
    - 0.58
    - 0.14
  - filter: mlp
    value:
    - 0.14
    - 0.37
    - 0.65
    - 0.32
    - 0.16
  - value: 0.41
dtype: bfloat16