ERNIE-Turbo-NF4 | Debloated

This repository contains the ERNIE-Turbo weights, debloated and quantized to 4-bit NF4 using bitsandbytes. This is the high-speed distilled variant optimized for 1-4 step inference within the Asset Editor ecosystem.

🚀 Model Details

• Architecture: ERNIE Image Transformer (Distilled Turbo)
• Backbone: Mistral-3 Text Encoder
• Quantization: NF4 (NormalFloat 4-bit)
• Primary Use: Lightning-fast image synthesis (1-4 steps).

🛠 Integration

Optimized for the Asset Editor Zerodrag pipeline.

🔗 Links

• Core Engine: Veetance Asset Editor
• Triage & Development: Asset Editor GitHub

---

Note: This is a specialized manifold optimized for the Veetance sovereign resource governor.