Update app.py

app.py

@@ -24,209 +24,4 @@ app = FastAPI()
 app.mount("/static", StaticFiles(directory="static"), name="static")
 
 # Load a multimodal model for image captioning and visual question answering
-multimodal_pipeline = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
-
-# Load a text-based model for summarization and text question answering
-text_pipeline = pipeline("text2text-generation", model="t5-small")
-
-# Load a translation model (initialized dynamically based on target language)
-translation_models = {
-    "fr": "Helsinki-NLP/opus-mt-en-fr",
-    "es": "Helsinki-NLP/opus-mt-en-es",
-    "de": "Helsinki-NLP/opus-mt-en-de"
-}
-
-@app.get("/")
-def read_root():
-    # Redirect to the static HTML file
-    return RedirectResponse(url="/static/index.html")
-
-@app.post("/summarize")
-async def summarize_text(
-    file: UploadFile = File(None),  # Optional file upload
-    text: str = Form(None)  # Optional manual text input
-):
-    if file:
-        logger.info(f"Received document for summarization: {file.filename}")
-        try:
-            # Extract text from the document
-            text = await extract_text_from_file(file)
-        except Exception as e:
-            logger.error(f"Error extracting text from file: {e}")
-            raise HTTPException(status_code=400, detail=str(e))
-    elif text:
-        logger.info("Received manual text for summarization")
-    else:
-        raise HTTPException(status_code=400, detail="No file or text provided")
-
-    try:
-        # Use the text pipeline to summarize the text
-        summary = text_pipeline(f"summarize: {text}", max_length=100)
-        logger.info(f"Generated summary: {summary[0]['generated_text']}")
-        return {"summary": summary[0]['generated_text']}
-    except Exception as e:
-        logger.error(f"Error during summarization: {e}")
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.post("/caption")
-async def caption_image(file: UploadFile = File(...)):
-    logger.info(f"Received image for captioning: {file.filename}")
-    try:
-        # Read the image file
-        image_data = await file.read()
-        image = Image.open(io.BytesIO(image_data))
-
-        # Use the multimodal pipeline to generate a caption for the image
-        caption = multimodal_pipeline(image)
-        logger.info(f"Generated caption: {caption[0]['generated_text']}")
-        return {"caption": caption[0]['generated_text']}
-    except Exception as e:
-        logger.error(f"Error during image captioning: {e}")
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.post("/answer")
-async def answer_question(
-    file: UploadFile = File(None),  # Optional file upload
-    text: str = Form(None),  # Optional manual text input
-    question: str = Form(...)  # Required question
-):
-    if file:
-        logger.info(f"Received document for question answering: {file.filename}")
-        try:
-            # Extract text from the document
-            text = await extract_text_from_file(file)
-        except Exception as e:
-            logger.error(f"Error extracting text from file: {e}")
-            raise HTTPException(status_code=400, detail=str(e))
-    elif text:
-        logger.info("Received manual text for question answering")
-    else:
-        raise HTTPException(status_code=400, detail="No file or text provided")
-
-    try:
-        # Use the text pipeline to answer the question
-        answer = text_pipeline(f"question: {question} context: {text}")
-        logger.info(f"Generated answer: {answer[0]['generated_text']}")
-        return {"answer": answer[0]['generated_text']}
-    except Exception as e:
-        logger.error(f"Error during question answering: {e}")
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.post("/vqa")
-async def visual_question_answering(file: UploadFile = File(...), question: str = Form(...)):
-    logger.info(f"Received image for visual question answering: {file.filename}")
-    logger.info(f"Received question: {question}")
-    try:
-        # Read the image file
-        image_data = await file.read()
-        image = Image.open(io.BytesIO(image_data))
-
-        # Use the multimodal pipeline to answer the question about the image
-        answer = multimodal_pipeline(image, question=question)
-        logger.info(f"Generated answer: {answer[0]['generated_text']}")
-        return {"answer": answer[0]['generated_text']}
-    except Exception as e:
-        logger.error(f"Error during visual question answering: {e}")
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.post("/visualize")
-async def visualize_data(
-    file: UploadFile = File(...),
-    request: str = Form(...)
-):
-    logger.info(f"Received Excel file for visualization: {file.filename}")
-    logger.info(f"Received visualization request: {request}")
-
-    try:
-        # Read the Excel file
-        df = pd.read_excel(io.BytesIO(await file.read()))
-
-        # Generate visualization code based on the request
-        if "bar" in request.lower():
-            code = f"""
-import matplotlib.pyplot as plt
-plt.bar(df['{df.columns[0]}'], df['{df.columns[1]}'])
-plt.xlabel('{df.columns[0]}')
-plt.ylabel('{df.columns[1]}')
-plt.title('Bar Chart')
-plt.show()
-"""
-        elif "line" in request.lower():
-            code = f"""
-import matplotlib.pyplot as plt
-plt.plot(df['{df.columns[0]}'], df['{df.columns[1]}'])
-plt.xlabel('{df.columns[0]}')
-plt.ylabel('{df.columns[1]}')
-plt.title('Line Chart')
-plt.show()
-"""
-        else:
-            code = f"""
-import seaborn as sns
-sns.pairplot(df)
-plt.show()
-"""
-
-        # Save the generated code to a file (optional)
-        code_filename = f"visualization_{uuid.uuid4()}.py"
-        with open(code_filename, "w") as f:
-            f.write(code)
-
-        return {"code": code, "filename": code_filename}
-    except Exception as e:
-        logger.error(f"Error during visualization code generation: {e}")
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.post("/translate")
-async def translate_document(
-    file: UploadFile = File(...),
-    target_language: str = Form(...)
-):
-    logger.info(f"Received document for translation: {file.filename}")
-    logger.info(f"Target language: {target_language}")
-
-    try:
-        # Extract text from the document
-        text = await extract_text_from_file(file)
-
-        # Load a translation model based on the target language
-        if target_language in translation_models:
-            model_name = translation_models[target_language]
-        else:
-            model_name = "Helsinki-NLP/opus-mt-en-de"  # Default to German
-
-        tokenizer = MarianTokenizer.from_pretrained(model_name)
-        model = MarianMTModel.from_pretrained(model_name)
-
-        # Translate the text
-        translated = model.generate(**tokenizer(text, return_tensors="pt", truncation=True))
-        translated_text = tokenizer.decode(translated[0], skip_special_tokens=True)
-
-        return {"translated_text": translated_text}
-    except Exception as e:
-        logger.error(f"Error during document translation: {e}")
-        raise HTTPException(status_code=500, detail=str(e))
-
-# Helper function to extract text from files
-async def extract_text_from_file(file: UploadFile):
-    try:
-        if file.filename.endswith(".pdf"):
-            doc = fitz.open(stream=await file.read(), filetype="pdf")
-            text = ""
-            for page in doc:
-                text += page.get_text()
-            return text
-        elif file.filename.endswith(".docx"):
-            doc = Document(io.BytesIO(await file.read()))
-            text = "\n".join([para.text for para in doc.paragraphs])
-            return text
-        else:
-            raise ValueError("Unsupported file format. Please upload a PDF or DOCX file.")
-    except Exception as e:
-        logger.error(f"Error extracting text from file: {e}")
-        raise HTTPException(status_code=400, detail=str(e))
-
-# Hugging Face Spaces expects the app to be served on port 7860
-if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+multimodal