Update app.py

app.py

@@ -22,10 +22,10 @@ For more details, please check [our post](https://huggingface.co/blog/llama32).
 access_token = os.getenv('HF_TOKEN')
 # Download the Base model
 #model_id = "./models/Llama-32-3B-Instruct"
-model_id = "meta-llama/Llama-3.1-8B-Instruct"
-MAX_MAX_NEW_TOKENS = 2048
-DEFAULT_MAX_NEW_TOKENS = 1024
-MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
+model_id = "nvidia/Llama-3_1-Nemotron-51B-Instruct"
+MAX_MAX_NEW_TOKENS = 6144
+DEFAULT_MAX_NEW_TOKENS = 6144
+MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "6144"))
 
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 
@@ -81,7 +81,7 @@ def generate(
 
 
 
-    streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
+    streamer = TextIteratorStreamer(tokenizer, timeout=2000.0, skip_prompt=True, skip_special_tokens=True)
     generate_kwargs = dict(
         input_ids=input_ids,
         streamer=streamer,