Data min max

app.py

@@ -104,12 +104,12 @@ def fit_to_length(text, min_length=3, max_length=100):
     return text
 
 
-def get_model_reponse(text: str):
+def get_model_reponse(query: str):
     """Process the text content."""
 
     # Prepare the model input
     messages_think = [
-        {"role": "user", "content": text}
+        {"role": "user", "content": query}
     ]
     text = tokenizer.apply_chat_template(
         messages_think,
@@ -144,7 +144,7 @@ async def completion(data: Completion):
         raise HTTPException(status_code=503, detail="Model not loaded")
 
     try:
-        text = fit_to_length(input_data.text, input_data.max_length)
+        text = fit_to_length(data.prompt, 3, data.max_tokens)
 
         result = get_model_reponse(text, model)
 
@@ -179,7 +179,7 @@ async def predict(q: str):
 
         input_data = TextInput(text=q)
 
-        text = fit_to_length(input_data.text, input_data.max_length)
+        text = fit_to_length(input_data.text, input_data.min_length, input_data.max_length)
 
         result = get_model_reponse(text, model)