Closed LukeHearon closed 11 months ago
Trying to re-analyze training data to see what the confidence scores look like; buzzdetect won't run through CLI. No time to troubleshoot right now, but I'm guessing the issue arises from the default 1h chunk length applied to 5s audio files.
fixed in latest commit to preprocess.py
Trying to re-analyze training data to see what the confidence scores look like; buzzdetect won't run through CLI. No time to troubleshoot right now, but I'm guessing the issue arises from the default 1h chunk length applied to 5s audio files.