Update app.py

app.py

@@ -1,6 +1,7 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForCausalLM, AutoConfig
 from peft import PeftModel, LoraConfig
+from unsloth.chat_templates import get_chat_template
 
 # Define the path where the model and adapters are saved
 model_path = "yentinglin/Llama-3-Taiwan-8B-Instruct"  # Update this to your model path
@@ -20,8 +21,14 @@ base_model = AutoModelForCausalLM.from_pretrained(model_path, config=config, ign
 model = PeftModel.from_pretrained(base_model, adapter_path)
 
 def generate_text(input_text):
-    input_ids = tokenizer.encode(input_text, return_tensors='pt')
-    outputs = model.generate(input_ids, max_length=50, num_return_sequences=1)
+    inputs = tokenizer.apply_chat_template(
+        messages,
+        tokenize = True,
+        add_generation_prompt = True, # Must add for generation
+        return_tensors = "pt",
+    ).to("cuda")
+    #input_ids = tokenizer.encode(input_text, return_tensors='pt')
+    outputs = model.generate(inputs, max_length=50, num_return_sequences=1)
     generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
     return generated_text