Silero V5
We now support the newest Silero model, Silero v5. It is not activated unless you set the model
configuration parameter to the string "v5"
. You can test it out here. The reason it is not the default is that if you have tweaked some of the parameters such as positiveSpeechThreshold
, redemptionMs
, etc, you will likely have to find new values that work better with v5.
The frame size is automatically set based on the model: 512 samples for the "v5" model and 1536 samples for the "legacy" model. You no longer need to configure frameSamples
manually.