Update app.py

app.py

@@ -44,11 +44,7 @@ def predict(text,
         begin_length = len(prompt)
     input_ids = inputs["input_ids"][:,-max_context_length_tokens:].to(device)
     torch.cuda.empty_cache()
-    global total_count
-    total_count += 1
-    print(total_count)
-    if total_count % 50 == 0 :
-        os.system("nvidia-smi")
+
     with torch.no_grad():
         for x in greedy_search(input_ids,model,tokenizer,stop_words=["[|Human|]", "[|AI|]"],max_length=max_length_tokens,temperature=temperature,top_p=top_p):
             if is_stop_word_or_prefix(x,["[|Human|]", "[|AI|]"]) is False: