Update app.py

app.py

@@ -1,36 +1,39 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 from datasets import load_dataset
+import json
+import os
 
 # Load the Spider dataset
 spider_dataset = load_dataset("spider", split='train')  # Load a subset of the dataset
 
-# Extract schema information from the dataset
-db_table_names = set()
-column_names = set()
-for item in spider_dataset:
-    db_id = item['db_id']
-    for table in item['table_names']:
-        db_table_names.add((db_id, table))
-    for column in item['column_names']:
-        column_names.add(column[1])
+# Load the database schemas
+db_schemas = {}
+database_dir = 'path/to/database/folder'
+for filename in os.listdir(database_dir):
+    if filename.endswith('.json'):
+        with open(os.path.join(database_dir, filename), 'r') as file:
+            db_schema = json.load(file)
+            db_schemas[db_schema['db_id']] = db_schema
 
 # Load tokenizer and model
 tokenizer = AutoTokenizer.from_pretrained("mrm8488/t5-base-finetuned-wikiSQL")
 model = AutoModelForSeq2SeqLM.from_pretrained("mrm8488/t5-base-finetuned-wikiSQL")
 
-def post_process_sql_query(sql_query):
+def post_process_sql_query(sql_query, db_id):
     # Modify the SQL query to match the dataset's schema
-    for db_id, table_name in db_table_names:
-        if "TABLE" in sql_query:
-            sql_query = sql_query.replace("TABLE", table_name)
-            break  # Assuming only one table is referenced in the query
-    for column_name in column_names:
-        if "COLUMN" in sql_query:
-            sql_query = sql_query.replace("COLUMN", column_name, 1)
+    if db_id in db_schemas:
+        db_schema = db_schemas[db_id]
+        for table_name in db_schema['table_names_original']:
+            if "TABLE" in sql_query:
+                sql_query = sql_query.replace("TABLE", table_name)
+                break  # Assuming only one table is referenced in the query
+        for column_name in db_schema['column_names_original']:
+            if "COLUMN" in sql_query:
+                sql_query = sql_query.replace("COLUMN", column_name[1], 1)
     return sql_query
 
-def generate_sql_from_user_input(query):
+def generate_sql_from_user_input(query, db_id):
     # Generate SQL for the user's query
     input_text = "translate English to SQL: " + query
     inputs = tokenizer(input_text, return_tensors="pt", padding=True)
@@ -38,13 +41,13 @@ def generate_sql_from_user_input(query):
     sql_query = tokenizer.decode(outputs[0], skip_special_tokens=True)
 
     # Post-process the SQL query to match the dataset's schema
-    sql_query = post_process_sql_query(sql_query)
+    sql_query = post_process_sql_query(sql_query, db_id)
     return sql_query
 
 # Create a Gradio interface
 interface = gr.Interface(
-    fn=generate_sql_from_user_input,
-    inputs=gr.Textbox(label="Enter your natural language query"),
+    fn=lambda query, db_id: generate_sql_from_user_input(query, db_id),
+    inputs=[gr.Textbox(label="Enter your natural language query"), gr.Dropdown(label="Select Database ID", choices=list(db_schemas.keys()))],
     outputs=gr.Textbox(label="Generated SQL Query"),
     title="NL to SQL using Spider Dataset",
     description="This interface generates an SQL query from your natural language input based on the Spider dataset."