update README.md

README.md

@@ -27,7 +27,9 @@ Download your tokenizer, model checkpoints, and optionally the training logs (`e
 Optionally, test model using the MLM task:
 ```sh
 pip install pya0
-python test.py --test_file test.txt
+python test.py \
+    --model_name_or_path ./ckpt/to/tokenizer \
+    --tokenizer_name_or_path ./ckpt/to/tokenizer
 ```
 > **Note**  
 > Modify the test examples in `test.txt` to play with it.