---
base_model:
- Sela223/Repose-Marlin-12B
- Sela223/Captain-Foxfire-12B
library_name: transformers
tags:
- mergekit
- merge

---
# merged

This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).

## Merge Details
### Merge Method

This model was merged using the [SLERP](https://en.wikipedia.org/wiki/Slerp) merge method.

### Models Merged

The following models were included in the merge:
* [Sela223/Repose-Marlin-12B](https://huggingface.co/Sela223/Repose-Marlin-12B)
* [Sela223/Captain-Foxfire-12B](https://huggingface.co/Sela223/Captain-Foxfire-12B)

### Configuration

The following YAML configuration was used to produce this model:

```yaml
base_model: Sela223/Captain-Foxfire-12B
dtype: bfloat16
merge_method: slerp
tokenizer_source: base

slices:
  - sources:
      - model: Sela223/Captain-Foxfire-12B
        layer_range: [0, 40]
      - model: Sela223/Repose-Marlin-12B
        layer_range: [0, 40]

parameters:
  rescale: true
  t:
    - filter: ".*(q_proj|k_proj|v_proj).*"
      value: [0.0, 0.1, 0.25, 0.4, 0.5, 0.5, 0.5, 0.5, 0.4, 0.25, 0.1, 0.0]
    - filter: ".*o_proj.*"
      value: [0.0, 0.1, 0.2, 0.35, 0.5, 0.5, 0.5, 0.5, 0.35, 0.2, 0.1, 0.0]
    - filter: self_attn
      value: [0.0, 0.1, 0.25, 0.4, 0.5, 0.5, 0.5, 0.5, 0.4, 0.25, 0.1, 0.0]

    - filter: ".*(gate_proj|up_proj|down_proj).*"
      value: [0.0, 0.15, 0.3, 0.45, 0.5, 0.5, 0.5, 0.5, 0.45, 0.3, 0.15, 0.0]
    - filter: mlp
      value: [0.0, 0.15, 0.3, 0.45, 0.5, 0.5, 0.5, 0.5, 0.45, 0.3, 0.15, 0.0]

    - filter: ".*(input_layernorm|post_attention_layernorm|layernorm).*"
      value: [0.0, 0.3, 0.5, 0.6, 0.4, 0.0, 0.0, 0.4, 0.6, 0.5, 0.3, 0.0]
    
    - filter: "^(embed_tokens|lm_head)$"
      value: 0.5

    - value: [0.0, 0.3, 0.5, 0.6, 0.4, 0.0, 0.0, 0.4, 0.6, 0.5, 0.3, 0.0]
```