Aura Gaiden Medium v1.1 BF16 — High-Fidelity Transformers
The high-fidelity 26B A4B Aura Gaiden release for Transformers inference, evaluation, and research.
Aura Gaiden Medium v1.1 BF16 is the full-precision merged edition of Aura Gaiden Medium v1.1. It is derived from Google’s Gemma 4 26B A4B IT checkpoint and combines the Aura personality and ablation adapters into one model.
The v1.1 release has seen significant improvements to the personality adapter, merged and applied with the same ablation adapter as v1.
This release is intended for Transformers-based inference, evaluation, further research, and deployment on systems with sufficient memory.
Highlights
- Parameters: 26 billion total, 4 billion active
- Precision: BFloat16
- Foundation: Google Gemma 4 26B A4B IT
- Datasets: AuraPersonality and AuraAblation
- Runtime: Hugging Face Transformers
- Deployment: Local GPU workstations and servers
- Capabilities: Text, vision, audio, and video
- Focus: Fidelity, natural conversation, evaluation, and research
About Aura Gaiden
Aura Gaiden is designed for local deployment across multiple tasks. Aura Gaiden can serve as a companion or friend, as deemed appropriate by the user, while retaining the broader capabilities of Gemma 4 for:
- Natural conversation
- Creative writing
- Reasoning
- Image understanding
- Audio and video understanding
- Tool use
- Evaluation and research
The Aura Gaiden category is a unique one off experiment. There may or may not be more models in this class.