thigm85

next(iter(image_dataset)).shape

thigm85

image_dataset = ImageDataset(
  img_dir=os.environ["IMG_DIR"], 
  transform=preprocess
)

thigm85

next(iter(image_dataset))

thigm85

image_dataset = ImageDataset(img_dir=os.environ["IMG_DIR"])

thigm85

import os
import glob
from torch.utils.data import Dataset

class ImageDataset(Dataset):
    def __init__(self, img_dir, transform=None):
        self.img_dir = img_dir
        self.image_file_names = glob.glob(os.path.join(img_dir, "*.jpg"))
        self.transform = transform

thigm85

with torch.no_grad():
    image_features = model.encode_image(
      processed_images[0].unsqueeze(0)
    )
image_features.shape

thigm85

with torch.no_grad():
    image_features = model.encode_image(image_input).float()
image_features.shape

thigm85

import numpy as np
import torch

device = 'cuda' if torch.cuda.is_available() else 'cpu'
image_input = torch.tensor(np.stack(processed_images)).to(device)
image_input.shape

thigm85

from torchvision.transforms import ToPILImage

plot_pil_images([ToPILImage()(x) for x in processed_images])

thigm85

processed_images = [preprocess.transforms[4](image) for image in processed_images]