Added comment

app.py

@@ -19,7 +19,8 @@ DEFAULT_MAX_NEW_TOKENS = 128
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 if torch.cuda.is_available():
-    model_id = "bandhit/typhoon-7b-q4-bnb_cuda-ts-1703352224"
+    model_id = "scb10x/typhoon-7b" # original model
+    model_id = "bandhit/typhoon-7b-q4-bnb_cuda-ts-1703352224" # quantized model
     model = AutoModelForCausalLM.from_pretrained(model_id)
     tokenizer = AutoTokenizer.from_pretrained(model_id)
     tokenizer.pad_token_id = tokenizer.eos_token_id