Dazzling Star Aurora 32b V0.0 Experimental 1130

🚀 Dazzling-Star-Aurora-32b-v0.0

If somewhere amid that aimlessly drifting sky,There was a planet where our wishes could flow free... would we try to make it there? I wonder what we'd wish for if we did...~

This project is a model merge, which combines multiple models to create a new one. It provides a new choice for model application.

Listen to the song on youtube: https://www.youtube.com/watch?v=e1EExQiRhC0

Story behind it: Bored at midnight, decided to create a merge I guess, resulting in this model, I like it, so try it out?

Models:

• EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2
• ArliAI/Qwen2.5-32B-ArliAI-RPMax-v1.3
• Qwen/Qwen2.5-32B

🚀 Quick Start

Instruct Format

The instruct format of this model is ChatML.

Acknowledgment

Thank you to AuriAetherwiing for providing compute and helping merge the models.

🔧 Technical Details

Merge Details

Merge Method

This model was merged using the TIES merge method using Qwen_Qwen2.5-32B as a base.

Models Merged

The following models were included in the merge:

• ArliAI/Qwen2.5-32B-ArliAI-RPMax-v1.3
• EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2
    parameters:
      weight: 0.3
      density: 0.7
  - model: ArliAI/Qwen2.5-32B-ArliAI-RPMax-v1.3
    parameters:
      weight: 0.4
      density: 0.8
base_model: Qwen/Qwen2.5-32B
parameters:
  epsilon: 0.05
  lambda: 1
  int8_mask: true
  normalize: true
merge_method: ties
dtype: bfloat16