app.py

import PIL.Image as Image
import gradio as gr
from ultralytics import YOLO
import os
import time
import uuid
from langchain_groq import ChatGroq
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain.text_splitter import CharacterTextSplitter
from langchain.chains.combine_documents import create_stuff_documents_chain
from langchain_core.prompts import ChatPromptTemplate
from langchain.chains import create_retrieval_chain
from langchain_community.vectorstores import FAISS
from langchain_community.document_loaders import PyPDFLoader
from dotenv import load_dotenv

load_dotenv()
groq_api_key = os.getenv('GROQ_API_KEY')

# Initialize object detection model
model = YOLO("version4c.pt")

# Set default confidence and IoU thresholds
CONF_THRESHOLD = 0.25
IOU_THRESHOLD = 0.45

def predict_image(img):
    # Perform object detection
    results = model.predict(source=img, conf=CONF_THRESHOLD, iou=IOU_THRESHOLD, show_labels=True, show_conf=True, imgsz=640)
    
    # Check if any objects were detected
    if len(results[0].boxes) == 0:
        return None, "Please upload a clearer image, and don't upload images of breeds that are not been used."
    
    # Plot the result
    for r in results:
        im_array = r.plot()
        im = Image.fromarray(im_array[..., ::-1])
    
    # Generate a unique filename
    filename = f"detected_result_{uuid.uuid4()}.jpg"
    
    # Save the image as JPG
    im.save(filename, format='JPEG')
    
    return filename, None

# Initialize chatbot components
llm = ChatGroq(groq_api_key=groq_api_key, model_name="Llama3-70b-8192")
prompt = ChatPromptTemplate.from_template(
    """ Answer the questions based on the provided context only. Please provide the most accurate response based on the question <context> {context} </context> Questions:{input} """
)
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
loader = PyPDFLoader("Document.pdf")
docs = loader.load()
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
final_documents = text_splitter.split_documents(docs)
# Extract text content from the Document instances
doc_texts = [doc.page_content for doc in final_documents]
embeddings_result = embeddings.embed_documents(doc_texts)
if embeddings_result:
    vectors = FAISS.from_documents(final_documents, embeddings)
else:
    raise ValueError("Failed to generate embeddings. Please check your input documents or try a different embedding model.")

document_chain = create_stuff_documents_chain(llm, prompt)
retriever = vectors.as_retriever()
retrieval_chain = create_retrieval_chain(retriever, document_chain)

def print_like_dislike(x: gr.LikeData):
    print(x.index, x.value, x.liked)

def add_message(history, message):
    if message is not None:
        history.append((message, None))
    return history, gr.Textbox(value=None, interactive=False)

stop_generation = False

def bot(history):
    global stop_generation
    stop_generation = False
    message = history[-1][0]
    start_time = time.time()
    response = retrieval_chain.invoke({'input': message})['answer']
    response_time = time.time() - start_time
    if response_time > 6:
        return [(f"Sorry, I couldn't generate a response within 6 seconds. Please try again.", None)]
    history[-1][1] = ""
    for character in response:
        if stop_generation:
            break
        history[-1][1] += character
        time.sleep(0.05)
        yield history

def stop_response(dummy_placeholder):
    global stop_generation
    stop_generation = True

with gr.Blocks() as demo:
    with gr.Row():
        with gr.Column(scale=2):
            model_input = gr.Image(type="pil", label="Upload Image")
            model_output = gr.Image(type="filepath", label="Result")
            caution_message = gr.Textbox(label="Caution", visible=False)

            def process_image(img):
                result, caution = predict_image(img)
                if caution:
                    return None, caution, gr.Image(visible=False), gr.Textbox(visible=True)
                else:
                    return result, "", gr.Image(visible=True), gr.Textbox(visible=False)

            model_btn = gr.Button("Detect Result")
            model_btn.click(process_image, inputs=model_input, outputs=[model_output, caution_message, model_output, caution_message])

        with gr.Column(scale=1):
            chatbot = gr.Chatbot(
                [],
                elem_id="chatbot",
                bubble_full_width=False
            )

            chat_input = gr.Textbox(interactive=True, placeholder="Enter message...", show_label=False)
            chat_msg = chat_input.submit(add_message, [chatbot, chat_input], [chatbot, chat_input])
            bot_msg = chat_msg.then(bot, chatbot, chatbot, api_name="bot_response")
            bot_msg.then(lambda: gr.Textbox(interactive=True), None, [chat_input])
            chatbot.like(print_like_dislike, None, None)

            stop_btn = gr.Button("Stop Generation")
            stop_btn.click(stop_response, None, None)

    demo.queue()

if __name__ == "__main__":
    demo.launch()