Spaces:
Running
Running
| # Description: Gradio UI for Creating and Testing new Prompts | |
| # | |
| # Imports | |
| import logging | |
| import gradio as gr | |
| from App_Function_Libraries.Chat.Chat_Functions import chat | |
| from App_Function_Libraries.DB.DB_Manager import add_or_update_prompt | |
| from App_Function_Libraries.Prompt_Engineering.Prompt_Engineering import generate_prompt, test_generated_prompt | |
| from App_Function_Libraries.Utils.Utils import format_api_name, global_api_endpoints, default_api_endpoint | |
| # | |
| # Local Imports | |
| # | |
| ######################################################################################################################## | |
| # | |
| # Functions | |
| # Gradio tab for prompt suggestion and testing | |
| def create_prompt_suggestion_tab(): | |
| try: | |
| default_value = None | |
| if default_api_endpoint: | |
| if default_api_endpoint in global_api_endpoints: | |
| default_value = format_api_name(default_api_endpoint) | |
| else: | |
| logging.warning(f"Default API endpoint '{default_api_endpoint}' not found in global_api_endpoints") | |
| except Exception as e: | |
| logging.error(f"Error setting default API endpoint: {str(e)}") | |
| default_value = None | |
| with gr.TabItem("Prompt Suggestion/Creation", visible=True): | |
| gr.Markdown("# Generate and Test AI Prompts with the Metaprompt Approach") | |
| with gr.Row(): | |
| with gr.Column(): | |
| # Task and variable inputs | |
| task_input = gr.Textbox(label="Task Description", | |
| placeholder="E.g., Draft an email responding to a customer complaint") | |
| variables_input = gr.Textbox(label="Variables (comma-separated)", | |
| placeholder="E.g., CUSTOMER_COMPLAINT, COMPANY_NAME") | |
| # API-related inputs | |
| # Refactored API selection dropdown | |
| api_name_input = gr.Dropdown( | |
| choices=["None"] + [format_api_name(api) for api in global_api_endpoints], | |
| value=default_value, | |
| label="API for Analysis (Optional)" | |
| ) | |
| api_key_input = gr.Textbox(label="API Key", placeholder="Enter your API key (if required)", | |
| type="password") | |
| # Temperature slider for controlling randomness of generation | |
| temperature_input = gr.Slider(minimum=0, maximum=1, step=0.01, value=0.7, label="Temperature") | |
| # Button to generate the prompt | |
| generate_prompt_button = gr.Button("Generate Prompt") | |
| with gr.Column(): | |
| # Output for the generated prompt | |
| generated_prompt_output = gr.Textbox(label="Generated Prompt", interactive=False) | |
| # FIXME - figure this out | |
| # copy_button = gr.HTML(""" | |
| # <button onclick="copyToClipboard()">Copy</button> | |
| # <script> | |
| # function copyToClipboard() { | |
| # const textBox = document.querySelector('textarea'); // Select the textarea | |
| # textBox.select(); // Select the text | |
| # document.execCommand('copy'); // Copy it to clipboard | |
| # alert('Copied to clipboard!'); | |
| # } | |
| # </script> | |
| # """) | |
| # Section to test the generated prompt | |
| with gr.Row(): | |
| with gr.Column(): | |
| # Input to test the prompt with variable values | |
| variable_values_input = gr.Textbox(label="Variable Values (comma-separated)", | |
| placeholder="Enter variable values in order, comma-separated") | |
| test_prompt_button = gr.Button("Test Generated Prompt") | |
| with gr.Column(): | |
| # Output for the test result | |
| test_output = gr.Textbox(label="Test Output", interactive=False) | |
| # Section to save the generated prompt to the database | |
| with gr.Row(): | |
| with gr.Column(): | |
| prompt_title_input = gr.Textbox(label="Prompt Title", placeholder="Enter a title for this prompt") | |
| prompt_author_input = gr.Textbox(label="Author", | |
| placeholder="Enter the author's name") # New author field | |
| prompt_description_input = gr.Textbox(label="Prompt Description", placeholder="Enter a description", lines=3) | |
| save_prompt_button = gr.Button("Save Prompt to Database") | |
| save_prompt_output = gr.Textbox(label="Save Prompt Output", interactive=False) | |
| # Callback function to generate prompt | |
| def on_generate_prompt(api_name, api_key, task, variables, temperature): | |
| # Generate the prompt using the metaprompt approach and API | |
| generated_prompt = generate_prompt(api_name, api_key, task, variables, temperature) | |
| return generated_prompt | |
| # Callback function to test the generated prompt | |
| def on_test_prompt(api_name, api_key, generated_prompt, variable_values, temperature): | |
| # Test the prompt by filling in variable values | |
| test_result = test_generated_prompt(api_name, api_key, generated_prompt, variable_values, temperature) | |
| return test_result | |
| # Callback function to save the generated prompt to the database | |
| def on_save_prompt(title, author, description, generated_prompt): | |
| if not title or not generated_prompt: | |
| return "Error: Title and generated prompt are required." | |
| # Add the generated prompt to the database | |
| result = add_or_update_prompt(title, author, description, system_prompt="", user_prompt=generated_prompt, keywords=None) | |
| return result | |
| # Connect the button to the function that generates the prompt | |
| generate_prompt_button.click( | |
| fn=on_generate_prompt, | |
| inputs=[api_name_input, api_key_input, task_input, variables_input, temperature_input], | |
| outputs=[generated_prompt_output] | |
| ) | |
| # Connect the button to the function that tests the generated prompt | |
| test_prompt_button.click( | |
| fn=on_test_prompt, | |
| inputs=[api_name_input, api_key_input, generated_prompt_output, variable_values_input, temperature_input], | |
| outputs=[test_output] | |
| ) | |
| # Connect the save button to the function that saves the prompt to the database | |
| save_prompt_button.click( | |
| fn=on_save_prompt, | |
| inputs=[prompt_title_input, prompt_author_input, prompt_description_input, generated_prompt_output], | |
| outputs=[save_prompt_output] | |
| ) | |
| # Example chat function based on your API structure | |
| def chat_api_call(api_endpoint, api_key, input_data, prompt, temp, system_message=None): | |
| # Here you will call your chat function as defined previously | |
| response = chat(message=input_data, history=[], media_content={}, selected_parts=[], | |
| api_endpoint=api_endpoint, api_key=api_key, prompt=prompt, temperature=temp, | |
| system_message=system_message) | |
| return response | |
| # | |
| # End of Functions | |
| ######################################################################################################################## | |