---
license: apache-2.0
base_model: unsloth/gemma-3-27b-it-bnb-4bit
tags:
- text-to-sql
- gemma-3
- unsloth
- trl
- finance
- real-estate
- nlp
datasets:
- dan-text2sql/seoul-realestate-sql-v1
library_name: transformers
---

# seoul-realestate-sql-agent-v2

**Developed by:** dan-text2sql  
**License:** apache-2.0  
**Finetuned from model:** [unsloth/gemma-3-27b-it-bnb-4bit](https://huggingface.co/unsloth/gemma-3-27b-it-bnb-4bit)

This model is a **Text-to-SQL agent** specialized in **Korean Real Estate (Seoul)** data.  
It was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.

## Model Description (v2)
This is the **v2** version of the Seoul Real Estate SQL Agent.
* **Base Model:** Gemma-3 27B (IT)
* **Improvement:** Unlike v1 (Mistral-7B), this model leverages the massive 27B parameter size of Gemma-3.
* **Objective:** Translate natural language queries about Seoul apartment real estate data into executable SQL queries.

## Usage Example

```python
from unsloth import FastLanguageModel

# Load the model
model, tokenizer = FastLanguageModel.from_pretrained(
    model_name = "dan-text2sql/seoul-realestate-sql-agent-v2",
    max_seq_length = 2048,
    dtype = None,
    load_in_4bit = True,
)
FastLanguageModel.for_inference(model)

# Test Prompt
prompt = """아래 질문에 대한 올바른 SQL 쿼리를 작성해주세요.

### 질문:
서울시 강남구 삼성동의 20억 이하 아파트 매물을 찾아줘.

### SQL:
"""

inputs = tokenizer([prompt], return_tensors = "pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens = 128, use_cache = True)
print(tokenizer.batch_decode(outputs)[0])