Gunulhona/Gemma-3-4B
Gunulhona/Gemma-3-4B is a 4.3 billion parameter language model created by merging several specialized Gemma-based models using the Karcher Mean method. This merge combines capabilities from models focused on text-to-Cypher conversion, medical applications, translation, and document writing. It is designed to offer a versatile instruction-tuned model with a 32768 token context length, integrating diverse functionalities into a single compact package.
Loading preview...
Overview
Gunulhona/Gemma-3-4B is a 4.3 billion parameter instruction-tuned language model, developed by merging multiple pre-trained Gemma-based models. This model was created using the Karcher Mean merge method, with huihui-ai/gemma-3-4b-it-abliterated serving as the base model.
Key Capabilities
This merged model integrates diverse functionalities from its constituent models, aiming for a broad range of applications. The merge includes models specialized in:
- Graph Database Interaction:
neo4j/text-to-cypher-Gemma-3-4B-Instruct-2025.04.0contributes to text-to-Cypher query generation. - Medical Applications:
google/medgemma-4b-itandgoogle/medgemma-1.5-4b-itenhance its performance in medical contexts. - Translation:
google/translategemma-4b-itprovides robust translation capabilities. - Document Generation:
ZySec-AI/gemma-3-4b-document-writersupports document writing tasks.
Good For
This model is suitable for use cases requiring a combination of specialized skills, particularly in areas like:
- Generating Cypher queries from natural language.
- Assisting with medical text processing or information retrieval.
- Performing multilingual translation tasks.
- Aiding in the creation of various document types.
- General instruction-following tasks where a compact, multi-purpose model is beneficial.