System: Home Assistant Green
CORE: core-2026.2.1
OS: 6.12.63-haos
In trying to add speaker training data, the file picker would not see the wav files that i added. i verified they were in the correct sample rate (16kHz). What i found was that their mimetypes were not being looked for in the custom component's config_flow.py file. The audio editor i used (Audacity on Ubuntu) exported the files w/ the audio/x-wav MIME type, which caused the file picker to gloss over them. I made a change to add that MIME type to config_flow.py, but then it would not allow me to submit the selected file(s).
It gave an error: Expected a dictionary
i had to give up on adding training data via the integration, and did it on the server side. (i am not using the add-on, and am running the speaker recognition on my server which is running the STT, LLM, and TTS of my voice pipeline)
System: Home Assistant Green
CORE: core-2026.2.1
OS: 6.12.63-haos
In trying to add speaker training data, the file picker would not see the wav files that i added. i verified they were in the correct sample rate (16kHz). What i found was that their mimetypes were not being looked for in the custom component's
config_flow.pyfile. The audio editor i used (Audacity on Ubuntu) exported the files w/ theaudio/x-wavMIME type, which caused the file picker to gloss over them. I made a change to add that MIME type toconfig_flow.py, but then it would not allow me to submit the selected file(s).It gave an error:
Expected a dictionaryi had to give up on adding training data via the integration, and did it on the server side. (i am not using the add-on, and am running the speaker recognition on my server which is running the STT, LLM, and TTS of my voice pipeline)