Add text mode (#321)

* Pretty good spot

* Working draft

* Fix other mode

* Add js to git

* Working

* Add code

* fix

* Fix

* Add code

* Fix submit race condition

* demo

* fix

* Fix

* Fix

backend/fastrtc/utils.py

@@ -17,6 +17,7 @@ import av
 import librosa
 import numpy as np
 from fastapi import WebSocket
+from gradio.data_classes import GradioModel, GradioRootModel
 from numpy.typing import NDArray
 from pydub import AudioSegment
 
@@ -31,6 +32,16 @@ class AudioChunk(TypedDict):
     end: int
 
 
+class WebRTCData(GradioModel):
+    webrtc_id: str
+    textbox: str = ""
+    audio: Any | None = None
+
+
+class WebRTCModel(GradioRootModel):
+    root: WebRTCData | str
+
+
 class AdditionalOutputs:
     def __init__(self, *args) -> None:
         self.args = args