Update app.py

app.py

@@ -38,6 +38,7 @@ VQ_HUB = "BAAI/Emu3-VisionTokenizer"
 
 # Prepare models and processors
 # Emu3-Gen model and processor
+"""
 gen_model = AutoModelForCausalLM.from_pretrained(
     EMU_GEN_HUB,
     device_map="cpu",
@@ -45,6 +46,7 @@ gen_model = AutoModelForCausalLM.from_pretrained(
     attn_implementation="flash_attention_2",
     trust_remote_code=True,
 )
+"""
 
 # Emu3-Chat model and processor
 chat_model = AutoModelForCausalLM.from_pretrained(
@@ -67,10 +69,11 @@ processor = Emu3Processor(
 )
 
 print(device)
-gen_model.to(device)
+# gen_model.to(device)
 chat_model.to(device)
 image_tokenizer.to(device)
 
+"""
 @spaces.GPU(duration=120)
 def generate_image(prompt):
     POSITIVE_PROMPT = " masterpiece, film grained, best quality."
@@ -130,6 +133,7 @@ def generate_image(prompt):
         if isinstance(im, Image.Image):
             return im
     return None
+"""
 
 def vision_language_understanding(image, text):
     inputs = processor(