feat: add userdata to audioframe constructor

This was raised here: livekit/plugins-ai-coustics-internal#30 (comment)

Author: 1egoman

livekit-rtc/livekit/rtc/audio_frame.py

@@ -33,6 +33,7 @@ def __init__(
         sample_rate: int,
         num_channels: int,
         samples_per_channel: int,
+        userdata: dict[str, Any] = {},
     ) -> None:
         """
         Initialize an AudioFrame instance.
@@ -64,7 +65,7 @@ def __init__(
         self._sample_rate = sample_rate
         self._num_channels = num_channels
         self._samples_per_channel = samples_per_channel
-        self._userdata: dict[str, Any] = {}
+        self._userdata: dict[str, Any] = userdata
 
     @staticmethod
     def create(sample_rate: int, num_channels: int, samples_per_channel: int) -> "AudioFrame":