VoErik/cypher-gemma
VoErik/cypher-gemma is a 0.3 billion parameter instruction-tuned language model, fine-tuned from Google's gemma-3-270m-it, specifically designed to convert natural language queries into Cypher Query Language. With a context length of 32768 tokens, this model excels at generating Cypher queries for graph databases based on user input and provided schema. It was trained using the TRL framework on the text2cypher-2025v1 dataset, making it highly specialized for graph database interactions.
Loading preview...
VoErik/cypher-gemma: Natural Language to Cypher Query Generation
VoErik/cypher-gemma is a specialized 0.3 billion parameter language model, fine-tuned from Google's gemma-3-270m-it architecture. Its core function is to translate natural language questions into executable Cypher Query Language (CQL) statements, making it a valuable tool for interacting with graph databases like Neo4j.
Key Capabilities
- Natural Language to Cypher Conversion: Directly transforms user questions into precise Cypher queries.
- Graph Database Interaction: Optimized for use cases involving querying graph data structures.
- Schema-Aware Query Generation: Designed to incorporate provided database schemas to generate accurate and contextually relevant Cypher queries.
- Efficient for Specific Tasks: With its smaller parameter count, it offers an efficient solution for its targeted use case.
Training Details
The model was trained using the TRL (Transformer Reinforcement Learning) framework. It underwent approximately 3500 steps of Supervised Fine-Tuning (SFT) on the text2cypher-2025v1 dataset, which is specifically curated for text-to-Cypher conversion tasks by Neo4j.
Good For
- Developers building applications that require natural language interfaces for graph databases.
- Automating the generation of Cypher queries from user input.
- Integrating with systems that benefit from programmatic Cypher query construction.