src/interface.py

import gradio as gr

# Gradio application setup
def create_demo():

    with gr.Blocks(title="LLAMA 3 Rag on Fly", theme="Monochrome") as demo:
        # App Description
        gr.Markdown(
            """
            ## LLAMA 3 Rag on Fly App

            This application allows you to experiment with a LLAMA 3 model for RAG powered Chatbot. 
            You can adjust various parameters to control the model's output. 
            """
        )
        with gr.Column():
            with gr.Row():
                chat_history = gr.Chatbot(value=[], elem_id='chatbot', height=480)
                show_img = gr.Image(label='Overview', height=480)

                # Sliders in the Rightmost Column
                with gr.Column():
                    with gr.Row():
                        slider_chunk_size = gr.Slider(
                            minimum=256, maximum=1024, value=256, label="Chunk Size", elem_id='slider1'
                        )
                        gr.Markdown(
                            """
                            **Chunk Size:** Controls the size of text chunks processed by the model. Larger chunks 
                            provide more context but may increase processing time.
                            """
                        )

                    with gr.Row():
                        slider_overlap_percentage = gr.Slider(
                            minimum=0, maximum=100, value=50, label="Chunk Overlap Percentage", elem_id='slider2'
                        )
                        gr.Markdown(
                            """
                            **Chunk Overlap Percentage:** Determines the percentage of overlap between consecutive 
                            chunks. Higher overlap can improve coherence but may also increase redundancy. 
                            """
                        )

                    with gr.Row():
                        slider_temp = gr.Slider(
                            minimum=0, maximum=1, value=0.5, label="Model Temperature", elem_id='slider3'
                        )
                        gr.Markdown(
                            """
                            **Model Temperature:** Controls the randomness of the generated text. Higher temperature 
                            leads to more creative and unpredictable outputs, while lower temperature results 
                            in more conservative and deterministic text. 
                            """
                        )

                    with gr.Row():
                        slider_k = gr.Slider(
                            minimum=1, maximum=5, value=2, label="Max Chunks in Context", elem_id='slider2'
                        )
                        gr.Markdown(
                            """
                            **Max Chunks in Context:** Determines the number of previous chunks considered by the 
                            model when generating text. More chunks provide more context but may also increase 
                            processing time.
                            """
                        )

        with gr.Row():
            with gr.Column(scale=0.60):
                text_input = gr.Textbox(
                    show_label=False,
                    placeholder="Type here to ask your PDF",
                container=False)
            with gr.Column(scale=0.20):
                submit_button = gr.Button('Send')
            with gr.Column(scale=0.20):
                uploaded_pdf = gr.UploadButton("📁 Upload PDF", file_types=[".pdf"], elem_id='upload_pdf')

    return demo, chat_history, show_img, text_input, submit_button, uploaded_pdf, slider_chunk_size,slider_overlap_percentage,slider_temp,slider_k

if __name__ == '__main__':
    demo, chatbot, show_img, text_input, submit_button, uploaded_pdf, slider_chunk_size,slider_overlap_percentage,slider_temp,slider_k = create_demo()
    demo.queue()
    demo.launch()