diff --git a/README.md b/README.md
index 533394f83..403cae4d6 100644
--- a/README.md
+++ b/README.md
@@ -326,6 +326,8 @@ tokenizer_type: AutoTokenizer
 trust_remote_code:
 # use_fast option for tokenizer loading from_pretrained, default to True
 tokenizer_use_fast:
+# Whether to use the legacy tokenizer setting, defaults to True
+tokenizer_legacy:
 # resize the model embeddings when new tokens are added to multiples of 32
 # this is reported to improve training speed on some models
 resize_token_embeddings_to_32x: