import gradio as gr
from transformers import pipeline, AutoModelForCausalLM, AutoTokenizer
from PIL import Image

# Load the Image-to-Text (OCR) model
ocr_model = pipeline("image-to-text", model="nlpconnect/vit-gpt2-image-captioning")

# Load the Text Generation model
story_model_name = "EleutherAI/gpt-neo-2.7B"
story_tokenizer = AutoTokenizer.from_pretrained(story_model_name)
story_model = AutoModelForCausalLM.from_pretrained(story_model_name)

# Function to extract text description from an image
def extract_description(image_array):
    try:
        # Convert the NumPy array to a PIL image
        image = Image.fromarray(image_array)
        # Use the OCR model to extract a caption/description from the image
        result = ocr_model(image)
        return result[0]["generated_text"]
    except Exception as e:
        return f"Error extracting description: {e}"

# Function to generate a story based on the extracted description
def generate_story(description):
    try:
        # Format the input prompt for the story
        prompt = f"Create a creative story based on this description: {description}"
        
        # Use the story model to generate text
        inputs = story_tokenizer(prompt, return_tensors="pt", truncation=True)
        outputs = story_model.generate(inputs["input_ids"], max_length=300, num_return_sequences=1, temperature=0.8)
        story = story_tokenizer.decode(outputs[0], skip_special_tokens=True)
        return story
    except Exception as e:
        return f"Error generating story: {e}"

# Main function to process the image and generate a story
def create_story(image):
    try:
        # Step 1: Extract a description from the image
        description = extract_description(image)
        if not description or "Error" in description:
            return description

        # Step 2: Generate a story from the extracted description
        story = generate_story(description)

        # Combine the description and story for the output
        output = f"📷 Extracted Description:\n{description}\n\n📖 Generated Story:\n{story}"
        return output
    except Exception as e:
        return f"Error processing the image: {e}"

# Gradio interface
interface = gr.Interface(
    fn=create_story,
    inputs=gr.Image(label="Upload an Image (PNG, JPG, JPEG)"),
    outputs=gr.Textbox(label="Generated Story"),
    title="Text-Based Story Creator",
    description=(
        "Upload an image, and this app will generate a creative story based on the description of the image. "
        "It uses advanced AI models for image-to-text conversion and story generation."
    ),
    allow_flagging="never"
)

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