Include the small model in the Nix package

[onechat] / README.rst

diff --git a/README.rst b/README.rst
index bbb4475d3cc89d416bc62dbf898ca00b7bc3e8bc..0148495daa28f76c6fc9ec8327c0acc6e535c769 100644 (file)
--- a/README.rst
+++ b/README.rst
@@ -198,6 +198,13 @@ Images
    Its memory keeps what the image shows, and it can look at an older
    image again. Ask it to show an image or send one to a WhatsApp chat.
 
    Its memory keeps what the image shows, and it can look at an older
    image again. Ask it to show an image or send one to a WhatsApp chat.
 

+Voice notes
+   The agent hears Telegram and WhatsApp voice notes through local
+   Whisper. Nix includes the small model. Outside Nix, set
+   ``voice.model`` in ``config.json`` and install ``whisper-cli``.
+   ``voice.transcriber`` selects another executable. No audio goes to a
+   transcription service.
+

 Schedules
    Ask for reminders and timed tasks, such as "every weekday at 8, tell
    me what my groups talked about". Each run uses the quota, so runs are
 Schedules
    Ask for reminders and timed tasks, such as "every weekday at 8, tell
    me what my groups talked about". Each run uses the quota, so runs are