Novaciano/Novaciano-The_Pervert.RUS-3.2-1B

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:1BQuant:BF16Ctx Length:32kPublished:Sep 9, 2025Architecture:Transformer Warm

Novaciano/Novaciano-The_Pervert.RUS-3.2-1B is a 1 billion parameter instruction-tuned language model developed by Novaciano, based on the Vikhr-Llama-3.2-1B-Instruct architecture. This model is specifically fine-tuned for immersive role-playing and narrative generation, with a focus on dark and complex themes. It integrates the Vikhrmodels/GrandMaster-PRO-MAX dataset for enhanced Russian language capabilities, while also supporting Spanish and English. Its primary strength lies in generating consistent and adaptable role-play scenarios despite its compact size.

Loading preview...

Novaciano The Pervert 3.2 - 1B (Russian Version)

This model, developed by Novaciano, is a 1 billion parameter instruction-tuned language model designed for immersive role-playing and narrative generation. It is a Russian version of the "The Pervert" series, also supporting Spanish and English, and is built upon the Vikhr-Llama-3.2-1B-Instruct base model.

Key Capabilities

  • Role-Playing Focus: Specifically engineered for creating engaging and consistent role-playing scenarios, particularly those with dark and complex themes.
  • Multilingual Support: Primarily optimized for Russian, it also offers functionality in Spanish and English.
  • Compact yet Adaptable: Despite its 1B parameter size, the model demonstrates surprising consistency and adaptability in generating narrative content.
  • Dataset Integration: Incorporates the Vikhrmodels/GrandMaster-PRO-MAX dataset, enhancing its Russian language performance.

Merge Details

This model was created using the Arcee Fusion merge method, combining Vikhrmodels/Vikhr-Llama-3.2-1B-Instruct and Novaciano/Novaciano-The_Pervert-NSFW-RP-3.2-1B.

Usage Notes

Users might encounter initial refusals to generate explicit content due to underlying dataset influences. Repeated attempts may be necessary to achieve desired outputs for specific role-playing scenarios.