fix: pass the use_cache false in the header

app.py

@@ -99,7 +99,8 @@ stop_sequences = ["<|endoftext|>"] # ":پایان","@","#","$",
 
 client = InferenceClient(
     API_URL,
-    headers={"Authorization": f"Bearer {HF_TOKEN}"},
+    headers={"Authorization": f"Bearer {HF_TOKEN}",
+             "use_cache": False},
 )
 
 def asynchronous_push(f_stop):
@@ -151,7 +152,6 @@ def generate(
         repetition_penalty=repetition_penalty,
         stop_sequences=stop_sequences,
         do_sample=True,
-        use_cache=False,
         #best_of=2,
         top_k=top_k,
         #typical_p=0.9,