remove chunk window for gpu services

matrix_pygmalion_bot/ai/model_helpers.py

@@ -136,7 +136,10 @@ async def get_full_prompt(simple_prompt: str, bot, chat_history, model_name: str
     #prompt += f"{ai_name}:"
 
     MAX_TOKENS = 2048
-    WINDOW = 600
+    if bot.service_text == "koboldcpp":
+        WINDOW = 600
+    else:
+        WINDOW = 0
     max_new_tokens = 200
     total_num_tokens = await num_tokens(prompt, model_name)
     input_num_tokens = await num_tokens(f"{user_name}: {simple_prompt}\n{ai_name}:", model_name)