---
license: apache-2.0  # (Or another license like MIT)
datasets: ["custom-dataset", "neo4j-cypher"]  # Mention relevant datasets
language: ["en"]  # Language(s) the model supports
metrics: ["bleu", "exact_match"]  # Metrics used for evaluation
base_model: "meta-llama/Meta-Llama-3-8B"  # If fine-tuned from an existing model
new_version: "1.0.0"  # Specify the version number
pipeline_tag: "text2text-generation"  # Pipeline type
library_name: "transformers"  # Library used for inference
tags:
  - cypher
  - graph-database
  - neo4j
  - text-generation
---

# 🚀 My Graph Query Model (LLM for Cypher Queries)

This model generates **Cypher queries** for **Neo4j** based on text-based instructions.

## 📚 Datasets
- The model was trained on a custom dataset containing **structured graph data** and **Cypher queries**.

## ⚙️ Model Details
- **Base Model**: `Meta-Llama-3-8B`
- **Fine-Tuned On**: Custom dataset of graph queries
- **Tokenization**: SentencePiece
- **Training Framework**: `transformers` with `UnsLoT` optimization

## 🏆 Evaluation Metrics
- **BLEU Score**: `0.80`
- **Exact Match Score**: `66.57%`

## 📥 How to Use the Model

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("thilaksai04/text2cypher-llama3.1-8b")
tokenizer = AutoTokenizer.from_pretrained("thilaksai04/text2cypher-llama3.1-8b")

input_text = "Find all guidelines related to diabetes treatment"
inputs = tokenizer(input_text, return_tensors="pt")
output = model.generate(**inputs)
print(tokenizer.decode(output[0], skip_special_tokens=True))