-
Hey guys! When I run the binary compiled using code from the master branch on Windows 11, during interaction with the model, in situations involving Latin characters such as á, ã, é as a response, the characters are displayed incorrectly. My system information: Is this behaviour expected? |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
That output suggests to me that your terminal is not using the UTF-8 encoding. Make sure it’s configured to do that — LLaMA was trained to output UTF-8 text. |
Beta Was this translation helpful? Give feedback.
That output suggests to me that your terminal is not using the UTF-8 encoding. Make sure it’s configured to do that — LLaMA was trained to output UTF-8 text.