Update app.py

app.py

@@ -41,8 +41,7 @@ model = Llama4ForConditionalGeneration.from_pretrained(
     # quantization_config=bnb_config,
     device_map="auto",
 )
-# processor = AutoProcessor.from_pretrained(model_name, cache_dir = cache_dir)
-tokenizer = AutoProcessor.from_pretrained(model_name
+processor = AutoProcessor.from_pretrained(model_name, cache_dir = cache_dir)
     # , gguf_file=filename
     # , subfolder=subfolder
     )
@@ -81,7 +80,7 @@ def generate(prompt, history):
     
     # response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
     # return response
-    inputs = tokenizer.apply_chat_template(
+    inputs = processor.apply_chat_template(
         messages,
         add_generation_prompt=True,
         tokenize=True,
@@ -92,7 +91,7 @@ def generate(prompt, history):
         **inputs,
         max_new_tokens=512,
     )
-    response = tokenizer.batch_decode(outputs[:, inputs["input_ids"].shape[-1]:])[0]
+    response = processor.batch_decode(outputs[:, inputs["input_ids"].shape[-1]:])[0]
 
 
 chat_interface = gr.ChatInterface(