t5 small awesome text to sql

Introduction

The T5-Small-Awesome-Text-to-SQL model is designed to generate SQL queries from natural language input. It is based on the T5-small architecture and fine-tuned to handle SQL generation across various tables with "CREATE TABLE" statements. This lightweight model is suitable for analytical applications that require SQL query generation.

Architecture

The model is a language model fine-tuned from the T5-small model. It operates within the PyTorch framework and uses the Hugging Face Transformers library. The model is trained to work with English language inputs, utilizing datasets such as Clinton/Text-to-sql-v1 and b-mc2/sql-create-context.

Training

The model was trained using a combination of datasets focusing on text-to-SQL tasks. The training process employed one A100-80 GPU and leveraged PyTorch along with the Hugging Face Transformers library.

Guide: Running Locally

1. Install Dependencies: Ensure you have Python and PyTorch installed. Install the Transformers library via pip:

   pip install transformers
   

2. Initialize the Model: Use the provided Python code snippet to load and run the model:

   import torch
   from transformers import T5Tokenizer, T5ForConditionalGeneration
   
   tokenizer = T5Tokenizer.from_pretrained('t5-small')
   device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
   model = T5ForConditionalGeneration.from_pretrained('cssupport/t5-small-awesome-text-to-sql').to(device)
   model.eval()
   
   def generate_sql(input_prompt):
       inputs = tokenizer(input_prompt, padding=True, truncation=True, return_tensors="pt").to(device)
       with torch.no_grad():
           outputs = model.generate(**inputs, max_length=512)
       return tokenizer.decode(outputs[0], skip_special_tokens=True)
   

3. Test the Model: Use a sample input prompt to generate SQL:

   input_prompt = "tables:\n CREATE TABLE student_course_attendance (student_id VARCHAR); CREATE TABLE students (student_id VARCHAR)\n query for: List the id of students who never attends courses?"
   print("The generated SQL query is:", generate_sql(input_prompt))
   

4. Cloud GPUs: For intensive workloads, consider using cloud solutions such as AWS, GCP, or Azure, which offer GPU instances like Nvidia A100.

License

The T5-Small-Awesome-Text-to-SQL model is licensed under the Apache 2.0 License, allowing for wide use and modification with proper attribution.