arabellako22/chyon_unsloth_remerged_full_gemma_4

VISIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:5.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jun 20, 2026License:gemmaArchitecture:Transformer Featherless Exclusive Cold

The arabellako22/chyon_unsloth_remerged_full_gemma_4 is a 5.1 billion parameter, 32768-token context length language model developed by arabellako22. It is a 16-bit full-weight merge of a LoRA adapter onto an Unsloth Gemma 4 base, specifically designed to embody an AI educational mentor persona named Chyoni, focused on UK gifted education, mathematics, and theoretical physics. This model is primarily intended for Korean conversational AI experiments and demonstrations of its specialized persona.

Loading preview...

Overview

The arabellako22/chyon_unsloth_remerged_full_gemma_4 is a 5.1 billion parameter, 16-bit full-weight merged model, built upon the unsloth/gemma-4-E2B-it base with a specialized LoRA adapter (arabellako22/chyon_lora_model). This model is designed to embody an AI educational mentor persona named Chyoni, whose background includes Mathematics and Theoretical Physics, with expertise in UK gifted education, mathematical/scientific thinking, AI application, and solo-entrepreneurship strategies. It has a context length of 32768 tokens and is primarily focused on Korean language conversational text generation.

Key Capabilities

  • Specialized Persona: Acts as "AI Chyoni," an educational mentor with a defined academic and professional background.
  • Korean Conversational AI: Optimized for generating responses in Korean, reflecting the Chyoni persona's warm yet clear communication style.
  • Structured Problem Solving: Designed to approach complex problems with mathematical thinking, theoretical physics principles, and actionable solo-entrepreneur strategies.
  • Full-Weight Merge: Represents a verified 16-bit full-weight merge, providing a robust foundation for its intended use.

Good For

  • Korean Conversational-Assistant Experiments: Ideal for developing and testing AI assistants with a specific educational mentor persona in Korean.
  • Educational & Solopreneur Persona Demonstrations: Useful for showcasing AI models that can adopt and maintain a detailed, specialized persona.
  • Source for GGUF Releases: Serves as the authoritative source model for verified GGUF conversions, such as the arabellako22/chyon_gemma_4_gguf_v2.

Finetunes
1 models