Update app.py

app.py

@@ -1,64 +1,77 @@
+# app.py
+
 import gradio as gr
-from huggingface_hub import InferenceClient
+import pandas as pd
+import duckdb
+import requests
+import re
+import io
+import os
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+# Load Together API Key from environment variable or fallback
+TOGETHER_API_KEY = os.getenv("TOGETHER_API_KEY", "your-api-key-here")  # Replace with HF Secret
 
+# Function to generate SQL from a prompt
+def generate_sql_from_prompt(prompt, df):
+    schema = ", ".join([f"{col} ({str(dtype)})" for col, dtype in df.dtypes.items()])
+    full_prompt = f"""
+You are a SQL expert. Here is a table called 'df' with the following schema:
+{schema}
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+User question: "{prompt}"
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+Write a valid SQL query using the 'df' table. Return only the SQL code.
+"""
 
-    messages.append({"role": "user", "content": message})
+    url = "https://api.together.xyz/v1/chat/completions"
+    headers = {
+        "Authorization": f"Bearer {TOGETHER_API_KEY}",
+        "Content-Type": "application/json"
+    }
+    payload = {
+        "model": "mistralai/Mixtral-8x7B-Instruct-v0.1",
+        "messages": [{"role": "user", "content": full_prompt}],
+        "temperature": 0.2,
+        "max_tokens": 200
+    }
 
-    response = ""
+    response = requests.post(url, headers=headers, json=payload)
+    response.raise_for_status()
+    result = response.json()
+    return result['choices'][0]['message']['content'].strip("```sql").strip("```").strip()
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+# Function to clean SQL for DuckDB
+def clean_sql_for_duckdb(sql, df_columns):
+    sql = sql.replace("`", '"')
+    for col in df_columns:
+        if " " in col and f'"{col}"' not in sql:
+            pattern = r'\b' + re.escape(col) + r'\b'
+            sql = re.sub(pattern, f'"{col}"', sql)
+    return sql
 
-        response += token
-        yield response
+# Combined Gradio function
+def chatbot_interface(file, question):
+    try:
+        df = pd.read_excel(file)
+        sql = generate_sql_from_prompt(question, df)
+        cleaned_sql = clean_sql_for_duckdb(sql, df.columns)
+        result = duckdb.query(cleaned_sql).to_df()
+        return f"📜 SQL Query:\n```sql\n{sql}\n```", result
+    except Exception as e:
+        return f"❌ Error: {str(e)}", pd.DataFrame()
 
+# Gradio UI
+with gr.Blocks() as demo:
+    gr.Markdown("## 📊 Excel SQL Chatbot with Together API")
+    file_input = gr.File(label="📂 Upload Excel File (.xlsx)")
+    question = gr.Textbox(label="🧠 Ask a question about your data", placeholder="e.g., Show me the average sales by region")
+    submit = gr.Button("🚀 Generate & Query")
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
+    sql_output = gr.Markdown()
+    result_table = gr.Dataframe()
 
+    submit.click(fn=chatbot_interface, inputs=[file_input, question], outputs=[sql_output, result_table])
 
+# Launch app
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()