Update app.py

app.py

@@ -11,6 +11,7 @@ import torch.nn as nn
 from transformers import CLIPModel, AutoModel
 from torchvision.transforms import CenterCrop, ConvertImageDtype, Normalize, Resize
 from torchvision.transforms.functional import InterpolationMode
+from torchvision import transforms
 
 from huggingface_hub import hf_hub_download
 from safetensors.torch import load_model
@@ -126,7 +127,8 @@ image_processor = AutoImageProcessor.from_pretrained("openai/clip-vit-base-patch
 
 def predict_sentiment(text, image):
     print(text, image)
-    image = transforms.ToTensor()(image).unsqueeze(0)
+    image = torchvision.io.read_image(image)
+    # image = transforms.ToTensor()(image).unsqueeze(0)
     
     text_inputs = tokenizer(
             text,