edit: model.py

modify clip tokenizer in function generate_images

model.py

@@ -462,7 +462,7 @@ def generate_images(
         if exists(clip):
             from transformers import AutoTokenizer
 
-            clip_tokenizer = AutoTokenizer.from_pretrained("monologg/distilkobert") # clip에 사용된 tokenizer
+            clip_tokenizer = AutoTokenizer.from_pretrained("klue/roberta-base") # clip에 사용된 tokenizer
             input_text = clip_tokenizer(
                 raw_text,
                 padding=True,