Update README.md

README.md

@@ -14,7 +14,7 @@ Refer to the [original model card](https://huggingface.co/jackboot/uwu-qwen-32b)
 
 ## Use with llama.cpp
 
-I hope that it's an up to date enough l.cpp. In ooba you can swap the tokenizers around using llama.cpp HF. This was created with the default mergekit tokenizer which uses the QWQ bos token.
+I hope that it's an up to date enough l.cpp. In ooba you can swap the tokenizers around using llama.cpp HF. This was created with the default mergekit tokenizer which uses the QWQ bos/eos token.
 
 ### CLI:
 ```bash