dsantiago

def resize_image(image, input_size):
    w, h = image.size
    scale = input_size / max(w, h)
    new_w = int(w * scale)
    new_h = int(h * scale)
    image = image.resize((new_w, new_h))
    return image

dsantiago

from PIL import Image
import numpy as np
from io import BytesIO
from pathlib import Path
from google.cloud import storage
import functions_framework

SIZE = (768, 1024)

# Triggered by a change in a storage bucket

dsantiago

To use RoBERTa for classifying text into 32 types of documents, you can leverage the Hugging Face transformers library, which provides a pre-trained RoBERTa model that you can fine-tune for your specific classification task.

Here’s a step-by-step guide with code snippets on how to achieve this:

1. Install Necessary Libraries

First, you need to install the transformers and datasets libraries from Hugging Face, as well as torch for PyTorch.

pip install transformers datasets torch

dsantiago

dsantiago

dsantiago

dsantiago

dsantiago

dsantiago

dsantiago