v 0.1

- добавлен системный промт;
- add_generation_prompt=False

app.py

@@ -14,7 +14,7 @@ logger.info("Запускаемся... 🥳🥳🥳")
 
 app = FastAPI()
 model_url = "https://huggingface.co/Qwen/Qwen2.5-7B-Instruct-GGUF/resolve/main/qwen2.5-7b-instruct-q3_k_m.gguf"
-
+SYSTEM_PROMPT = ''
 tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-7B-Instruct")
 
 
@@ -27,9 +27,12 @@ def messages_to_prompt(messages):
 
 
 def completion_to_prompt(completion):
-    messages = [{"role": "user", "content": completion}]
+    messages = [
+        {"role": "system", "content": SYSTEM_PROMPT or "Ответ должен быть точным, кратким и с юмором."},
+        {"role": "user", "content": completion},
+    ]
     prompt = tokenizer.apply_chat_template(
-        messages, tokenize=False, add_generation_prompt=True
+        messages, tokenize=False, add_generation_prompt=False
     )
     return prompt