Nubinu/Gemma4-E2B-MicroFantasy-V1

VISIONConcurrency Cost:1Model Size:5.1BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:Jun 8, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

Nubinu/Gemma4-E2B-MicroFantasy-V1 is a 5.1 billion parameter language model, fine-tuned from MuXodious/gemma-4-E2B-it-ARA-heresy. This model is specifically optimized for role-playing (RP) scenarios, focusing on narrative pacing and character adherence. It excels at generating detailed, in-character responses for collaborative storytelling, particularly within platforms like SillyTavern.

Loading preview...

Overview

Nubinu/Gemma4-E2B-MicroFantasy-V1 is a 5.1 billion parameter model, derived from a 4-bit LoRA fine-tune of the MuXodious/gemma-4-E2B-it-ARA-heresy base model. It is designed with a 32768 token context length, making it suitable for extended narrative generation. The primary focus of this model is to enhance role-playing experiences, providing detailed and consistent character embodiment.

Key Capabilities

  • Optimized for Role-Playing: Specifically fine-tuned for generating narrative-driven, in-character responses in collaborative storytelling environments.
  • Character Adherence: Trained to maintain character voice, personality, flaws, and behavior consistently throughout interactions.
  • Narrative Pacing: Emphasizes appropriate narrative pacing to prevent repetition and drive scenes forward naturally.
  • Structured Input: Benefits from a specific Markdown-based character card format for optimal personality and lore adherence, ensuring rich character context.

Recommended Use Cases

  • SillyTavern Integration: Ideal for use with SillyTavern for text completion, with recommended sampler settings and character card structures provided for best performance.
  • Collaborative Storytelling: Excels in scenarios where the model acts as a character in a shared narrative, writing in third-person limited narration and combining speech with physical actions.
  • Detailed Character Generation: Suitable for users who require models that can deeply embody complex character profiles and backstories.