Modelpublic

CultriX/Qwen2.5-14B-ReasoningMerge

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face

3likes31downloads

Model Card

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:

—Sakalti/Saka-14B
—RDson/WomboCombo-R1-Coder-14B-Preview

Configuration

The following YAML configuration was used to produce this model:

yaml

base_model: Sakalti/Saka-14B
merge_method: slerp
dtype: bfloat16
parameters:
  t:
    - filter: self_attn
      value: 0.4    # Favor RDson/WomboCombo for self-attention
    - filter: mlp
      value: 0.7    # Favor Sakalti/Saka-14B for MLP layers
    - value: 0.5
models:
  - model: Sakalti/Saka-14B
  - model: RDson/WomboCombo-R1-Coder-14B-Preview
tokenizer_source: Sakalti/Saka-14B
chat_template: chatml
name: reasoning_blend