Update modules/vad/silero_vad.py

modules/vad/silero_vad.py

@@ -9,7 +9,7 @@ import faster_whisper
 from faster_whisper.transcribe import SpeechTimestampsMap
 import gradio as gr
 
-class Segment(BaseModel):
+class Segment(self):
     id: Optional[int] = Field(default=None, description="Incremental id for the segment")
     seek: Optional[int] = Field(default=None, description="Seek of the segment from chunked audio")
     text: Optional[str] = Field(default=None, description="Transcription text of the segment")
@@ -22,6 +22,12 @@ class Segment(BaseModel):
     no_speech_prob: Optional[float] = Field(default=None, description="Probability that it's not speech")
     words: Optional[List['Word']] = Field(default=None, description="List of words contained in the segment")
 
+class Word(self):
+    start: Optional[float] = Field(default=None, description="Start time of the word")
+    end: Optional[float] = Field(default=None, description="Start time of the word")
+    word: Optional[str] = Field(default=None, description="Word text")
+    probability: Optional[float] = Field(default=None, description="Probability of the word")
+
 class SileroVAD:
     def __init__(self):
         self.sampling_rate = 16000