Actual test 7

code/app.py

@@ -16,7 +16,7 @@ model_path = "./starling-lm-7b-alpha.Q6_K.gguf"
 
 mdlpath = hf_hub_download(repo_id="TheBloke/Starling-LM-7B-alpha-GGUF", filename=model_path, local_dir="./")
 
-lcpp_model = Llama(model_path=model_path)
+lcpp_model = Llama(model_path=model_path, n_ctx=16768)
 global otxt, txtinput, txtoutput, running, result
 otxt = ""
 running = False
@@ -61,7 +61,7 @@ def talk(txt, jsn):
     if not jsn:
         jsn = txt
     if not running:
-        result = lcpp_model.create_chat_completion(messages=txt,stream=True, n_ctx=16768)
+        result = lcpp_model.create_chat_completion(messages=txt,stream=True)
         running = True
     for r in result:
         txt2 = None