Fix generate process

spanish_medica_llm.py

@@ -737,8 +737,9 @@ def generate_response(query):
         tokenizer = loadSpanishTokenizer()
         model =  loadBaseModel(HUB_MODEL_ID)
 
-        system = f"[INST]\nYou are a helpful coding assistant.[/INST]\n"
-        prompt = f"{system}\n{query}\n \n"
+        #system = f"[INST]\nYou are a helpful coding assistant.[/INST]\n"
+        query = get_chat_format({'raw_text':query, 'topic':''})
+        prompt = tokenizer.apply_chat_template(query, tokenize=False)
         print(prompt)
         inputs = tokenizer(prompt, return_tensors="pt")
         input_ids = inputs["input_ids"].to("cuda")