Uncanned/Captain-VioLyra-v0.3-12B

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

I am testing different default values for effective merging. Still, feel free to try out.
ChatML format is recommended.
The two base models were playing nice with Alpaca format as well, but I haven't extensively tried this out as of yet.

Merge Method

This model was merged using the SLERP merge method.

Models Merged

The following models were included in the merge:

• Sao10K/MN-12B-Lyra-v4
• Nitral-AI/Captain-Eris_Violet-V0.420-12B

Configuration

The following YAML configuration was used to produce this model:

slices:
  - sources:
      - model: Nitral-AI/Captain-Eris_Violet-V0.420-12B
        layer_range: [0, 40]
      - model: Sao10K/MN-12B-Lyra-v4
        layer_range: [0, 40]
merge_method: slerp
base_model: Nitral-AI/Captain-Eris_Violet-V0.420-12B
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.3
dtype: bfloat16