louiecerv/gemini2_file_multimodal

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

This app provides a single-turn multimodal chat interface, allowing you to interact with an AI model using both text and visual inputs.

Key Features:

• —Multimodal Input: Engage in conversations by providing text prompts, images, and even PDF files.

• —Single-Turn Interactions: Each interaction with the AI model is treated as a fresh query, ensuring responses are focused on the current input rather than the conversation history.
• —Persistent Prompts: Your text prompts are saved in a JSON file for easy retrieval and reuse.

• —Conversation History: While the AI model doesn't use conversation history, your interactions are displayed for your reference. This helps maintain context and track your exploration.

How it Works:

1. 1.Input your query: Type a text prompt and/or upload an image or PDF file.
2. 2.Submit: Send your input to the AI model for processing.
3. 3.Receive a response: The AI model will generate a response based on your input.
4. 4.View history: Your interaction is added to the conversation history for your reference.

Technical Details:

• —The app utilizes a powerful AI model for understanding and responding to your queries.
• —A JSON file stores your text prompts for persistence.
• —The conversation history is displayed for user convenience but not used as context for the AI model's responses.

Getting Started:

1. 1.Clone the repository.
2. 2.Install the required dependencies.
3. 3.Run the Streamlit app.

Contributing:

Contributions are welcome! Feel free to open issues or submit pull requests.