Wikisql

🚀 PEFT

This library, peft, is designed for SQL generation based on a given table and question, leveraging the mistralai/Mistral-7B-v0.1 base model.

✨ Features

• SQL Generation: Generate SQL code according to a specific table and question.
• Free Query: You can try querying this adapter for free in Lora Land at https://predibase.com/lora-land.
• Reasoning Adapter: The adapter category is Reasoning, named WikiSQL (SQL Generation).

📦 Installation

No installation steps are provided in the original document, so this section is skipped.

💻 Usage Examples

Basic Usage

# Try using this adapter yourself!
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "mistralai/Mistral-7B-v0.1"
peft_model_id = "predibase/wikisql"

model = AutoModelForCausalLM.from_pretrained(model_id)
model.load_adapter(peft_model_id)

Sample Input and Output

• Sample Input: Considering the provided database schema and associated query, produce SQL code to retrieve the answer to the query.

Database Schema: {'header': ['Game', 'Date', 'Opponent', 'Score', 'Decision', 'Location/Attendance', 'Record'], 'types': ['real', 'real', 'text', 'text', 'text', 'text', 'text'], 'rows': [array(['77', '1', 'Pittsburgh Penguins', '1-6', 'Brodeur',

   'Mellon Arena - 17,132', '47-26-4'], dtype=object), array(['78', '3', 'Tampa Bay Lightning', '4-5 (OT)', 'Brodeur',
   'Prudential Center - 17,625', '48-26-4'], dtype=object), array(['79', '4', 'Buffalo Sabres', '3-2', 'Brodeur',
   'HSBC Arena - 18,690', '49-26-4'], dtype=object), array(['80', '7', 'Toronto Maple Leafs', '4-1', 'Brodeur',
   'Prudential Center - 15,046', '49-27-4'], dtype=object), array(['81', '9', 'Ottawa Senators', '3-2 (SO)', 'Brodeur',
   'Scotiabank Place - 20,151', '50-27-4'], dtype=object), array(['82', '11', 'Carolina Hurricanes', '2-3', 'Brodeur',
   'Prudential Center - 17,625', '51-27-4'], dtype=object)]}

Query: What is Score, when Game is greater than 78, and when Date is "4"?

SQL:

• Sample Output:

SELECT Score FROM table WHERE Game > 78 AND Date = 4

📚 Documentation

• Original Dataset: https://huggingface.co/datasets/wikisql

🔧 Technical Details

No technical details are provided in the original document, so this section is skipped.

📄 License

No license information is provided in the original document, so this section is skipped.