Spaces:
Runtime error
Runtime error
File size: 2,756 Bytes
639e3fa 45ab2ce 639e3fa 32e3531 45ab2ce 32e3531 639e3fa 45ab2ce 639e3fa 45ab2ce 639e3fa |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
import streamlit as st
from PIL import Image
import torch
from transformers import Qwen2VLForConditionalGeneration, AutoProcessor, pipeline
from colpali_engine.models import ColPali, ColPaliProcessor
import os
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
hf_token = os.getenv('HF_TOKEN')
try:
model = pipeline("image-to-text", model="google/paligemma-3b-mix-448", use_auth_token=hf_token)
except Exception as e:
st.error(f"Error loading image-to-text model: {e}")
st.stop()
try:
model_colpali = ColPali.from_pretrained("vidore/colpali-v1.2", torch_dtype=torch.bfloat16).to(device)
processor_colpali = ColPaliProcessor.from_pretrained("google/paligemma-3b-mix-448")
except Exception as e:
st.error(f"Error loading ColPali model or processor: {e}")
st.stop()
try:
model_qwen = Qwen2VLForConditionalGeneration.from_pretrained("Qwen/Qwen2-VL-7B-Instruct").to(device)
processor_qwen = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-7B-Instruct")
except Exception as e:
st.error(f"Error loading Qwen model or processor: {e}")
st.stop()
st.title("OCR and Document Search Web Application")
st.write("Upload an image containing text in both Hindi and English for OCR processing and keyword search.")
uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])
if uploaded_file is not None:
try:
image = Image.open(uploaded_file)
st.image(image, caption='Uploaded Image.', use_column_width=True)
st.write("")
conversation = [{"role": "user", "content": [{"type": "image"}, {"type": "text", "text": "Describe this image."}]}]
text_prompt = processor_qwen.apply_chat_template(conversation, add_generation_prompt=True)
inputs = processor_qwen(text=[text_prompt], images=[image], padding=True, return_tensors="pt").to(device)
with torch.no_grad():
output_ids = model_qwen.generate(**inputs, max_new_tokens=128)
generated_ids = [output_ids[len(input_ids):] for input_ids, output_ids in zip(inputs.input_ids, output_ids)]
output_text = processor_qwen.batch_decode(generated_ids, skip_special_tokens=True, clean_up_tokenization_spaces=True)
st.write("Extracted Text:")
st.write(output_text)
keyword = st.text_input("Enter a keyword to search in the extracted text:")
if keyword:
if keyword.lower() in output_text[0].lower():
st.write(f"Keyword '{keyword}' found in the text.")
else:
st.write(f"Keyword '{keyword}' not found in the text.")
except Exception as e:
st.error(f"An error occurred: {e}")
if __name__ == "__main__":
st.write("Deploying the web application...")
|