metadata

library_name: transformers
license: cc-by-nc-4.0
pipeline_tag: text-generation
tags:
  - text-to-sql
  - reinforcement-learning

SLM-SQL: An Exploration of Small Language Models for Text-to-SQL

Important Links

News

July 31, 2025: Upload model to modelscope and huggingface.
July 30, 2025: Publish the paper to arxiv

Abstract

Large language models (LLMs) have demonstrated strong performance in translating natural language questions into SQL queries (Text-to-SQL). In contrast, small language models (SLMs) ranging from 0.5B to 1.5B parameters currently underperform on Text-to-SQL tasks due to their limited logical reasoning capabilities. However, SLMs offer inherent advantages in inference speed and suitability for edge deployment. To explore their potential in Text-to-SQL applications, we leverage recent advancements in post-training techniques. Specifically, we used the open-source SynSQL-2.5M dataset to construct two derived datasets: SynSQL-Think-916K for SQL generation and SynSQL-Merge-Think-310K for SQL merge revision. We then applied supervised fine-tuning and reinforcement learning-based post-training to the SLM, followed by inference using a corrective self-consistency approach. Experimental results validate the effectiveness and generalizability of our method, SLM-SQL. On the BIRD development set, the five evaluated models achieved an average improvement of 31.4 points. Notably, the 0.5B model reached 56.87% execution accuracy (EX), while the 1.5B model achieved 67.08% EX.

Framework

Main Results

Performance Comparison of different Text-to-SQL methods on BIRD dev and test dataset.

How to Use

You can easily use this model with the Hugging Face transformers library. Below is a general example for inference:

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

# Load the model and tokenizer
model_name = "cycloneboy/SLM-SQL-1.5B" # Example: You can choose other models from the table below
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16, # or torch.float16, adjust based on your GPU
    device_map="auto" # Automatically map model to available devices
)
model.eval()

# Example prompt for Text-to-SQL
# Replace this with your natural language query for a specific database schema
prompt = """
[Instruction]: Given the following database schema, generate a SQL query that answers the question.
[Schema]:
CREATE TABLE Student (StuID INT, Name TEXT, Age INT, Sex TEXT, Major TEXT, Advisor INT, Graduated BOOL);
CREATE TABLE Course (CrsID INT, Title TEXT, Dept TEXT, Credits INT);
CREATE TABLE Enrollment (StuID INT, CrsID INT, Grade REAL);
CREATE TABLE Advisor (AdvID INT, Name TEXT, Dept TEXT);
[Question]: What is the average age of students who are taking 'Database' course?
"""

inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

# Generate SQL query
outputs = model.generate(
    **inputs,
    max_new_tokens=256,
    num_beams=1, # Adjust for different decoding strategies
    do_sample=False,
    temperature=0.0,
    top_p=1.0,
    eos_token_id=tokenizer.eos_token_id
)

generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(generated_text)

# The output will contain the prompt and the generated SQL.
# You might need to parse the generated_text to extract only the SQL query.

Model

Model	Base Model	Train Method	Modelscope	HuggingFace
SLM-SQL-Base-0.5B	Qwen2.5-Coder-0.5B-Instruct	SFT	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-0.5B	Qwen2.5-Coder-0.5B-Instruct	SFT + GRPO	🤖 Modelscope	🤗 HuggingFace
CscSQL-Merge-Qwen2.5-Coder-0.5B-Instruct	Qwen2.5-Coder-0.5B-Instruct	SFT + GRPO	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-Base-1.5B	Qwen2.5-Coder-1.5B-Instruct	SFT	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-1.5B	Qwen2.5-Coder-1.5B-Instruct	SFT + GRPO	🤖 Modelscope	🤗 HuggingFace
CscSQL-Merge-Qwen2.5-Coder-1.5B-Instruct	Qwen2.5-Coder-1.5B-Instruct	SFT + GRPO	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-Base-0.6B	Qwen3-0.6B	SFT	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-0.6B	Qwen3-0.6B	SFT + GRPO	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-Base-1.3B	deepseek-coder-1.3b-instruct	SFT	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-1.3B	deepseek-coder-1.3b-instruct	SFT + GRPO	🤖 Modelscope	🤗 HuggingFace
SLM-SQL-Base-1B	Llama-3.2-1B-Instruct	SFT	🤖 Modelscope	🤗 HuggingFace

Dataset

Dataset	Modelscope	HuggingFace
SynsQL-Think-916k	🤖 Modelscope	🤗 HuggingFace
SynsQL-Merge-Think-310k	🤖 Modelscope	🤗 HuggingFace
bird train and dev dataset	🤖 Modelscope	🤗 HuggingFace

TODO

Release inference code
Upload Model
Release training code
Fix bug
Update doc

Thanks to the following projects

Citation


@misc{sheng2025slmsqlexplorationsmalllanguage,
      title={SLM-SQL: An Exploration of Small Language Models for Text-to-SQL}, 
      author={Lei Sheng and Shuai-Shuai Xu},
      year={2025},
      eprint={2507.22478},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2507.22478}, 
}

@misc{sheng2025cscsqlcorrectiveselfconsistencytexttosql,
      title={CSC-SQL: Corrective Self-Consistency in Text-to-SQL via Reinforcement Learning}, 
      author={Lei Sheng and Shuai-Shuai Xu},
      year={2025},
      eprint={2505.13271},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2505.13271}, 
}