Update app.py

app.py

@@ -3,7 +3,7 @@ from PIL import Image
 from torchvision import transforms, models
 from transformers import AutoModelForCausalLM, AutoTokenizer
 import pandas as pd
-import open_clip
+from sentence_transformers import SentenceTransformer
 import random
 import urllib.parse
 import torch.nn as nn
@@ -84,11 +84,8 @@ model_resnet = DualOutputResNet(num_styles, num_artists).to(device)
 optimizer = torch.optim.Adam(model_resnet.parameters(), lr=0.001, weight_decay=1e-5)
 scheduler = ReduceLROnPlateau(optimizer, mode='min', factor=0.5, patience=3, verbose=True)
 
-# Load GPT-Neo and CLIP
-model_clip = open_clip.create_model('ViT-B/32', pretrained='openai').to(device)
-preprocess_clip = open_clip.image_transform((224, 224), is_train=False)
-tokenizer_clip = open_clip.get_tokenizer('ViT-B/32')
-model_clip.eval()
+# Load SentenceTransformer model
+clip_model = SentenceTransformer('sentence-transformers/clip-ViT-B-32-multilingual-v1').to(device)
 
 model_name = "EleutherAI/gpt-neo-1.3B"
 tokenizer = AutoTokenizer.from_pretrained(model_name)