Extend context to 8192

app.py

@@ -11,7 +11,7 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStream
 
 MAX_MAX_NEW_TOKENS = 2048
 DEFAULT_MAX_NEW_TOKENS = 1024
-MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
+MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "8192"))
 
 if torch.cuda.is_available():
     model_id = "Rijgersberg/GEITje-7B-chat"