Alelcv27/llama3-8b-dare-linear-v3

llama3-8b-dare-linear-v4

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

Merge Details

Merge Method

This model was merged using the linear DARE merge method using meta-llama/Llama-3.1-8B as a base.

Models Merged

The following models were included in the merge:

• Alelcv27/llama3-8b-math-dpo
• Alelcv27/llama3-8b-code-dpo

Configuration

The following YAML configuration was used to produce this model:

base_model: meta-llama/Llama-3.1-8B
dtype: float16
merge_method: dare_linear
parameters:
  normalize: 1.0
slices:
- sources:
  - layer_range: [0, 32]
    model: Alelcv27/llama3-8b-code-dpo
    parameters:
      weight: 0.7
  - layer_range: [0, 32]
    model: Alelcv27/llama3-8b-math-dpo
    parameters:
      weight: 1.0
  - layer_range: [0, 32]
    model: meta-llama/Llama-3.1-8B