Merge pull request #316 from makaveli10/fix_data_incosistency

Add lock to thread shared variables updates/reads.

setup.py

@@ -43,7 +43,7 @@ setup(
     ),
     install_requires=[
         "PyAudio",
-        "faster-whisper==1.0.1",
+        "faster-whisper==1.1.0",
         "torch",
         "torchaudio",
         "websockets",
@@ -52,9 +52,10 @@ setup(
         "scipy",
         "websocket-client",
         "numba",
-        "openai-whisper",
+        "openai-whisper==20240930",
         "kaldialign",
         "soundfile",
+        "tokenizers==0.20.3"
     ],
     python_requires=">=3.8"
 )

whisper_live/server.py

@@ -479,6 +479,7 @@ class ServeClientBase(object):
         Clip audio if the current chunk exceeds 30 seconds, this basically implies that
         no valid segment for the last 30 seconds from whisper
         """
+        with self.lock:
             if self.frames_np[int((self.timestamp_offset - self.frames_offset)*self.RATE):].shape[0] > 25 * self.RATE:
                 duration = self.frames_np.shape[0] / self.RATE
                 self.timestamp_offset = self.frames_offset + duration - 5
@@ -497,6 +498,7 @@ class ServeClientBase(object):
                 - input_bytes (np.ndarray): The next chunk of audio data to be processed.
                 - duration (float): The duration of the audio chunk in seconds.
         """
+        with self.lock:
             samples_take = max(0, (self.timestamp_offset - self.frames_offset) * self.RATE)
             input_bytes = self.frames_np[int(samples_take):].copy()
         duration = input_bytes.shape[0] / self.RATE
@@ -715,6 +717,8 @@ class ServeClientTensorRT(ServeClientBase):
             self.transcript.append({"text": last_segment + " "})
         elif self.transcript[-1]["text"].strip() != last_segment:
             self.transcript.append({"text": last_segment + " "})
+        
+        with self.lock():
             self.timestamp_offset += duration
 
     def speech_to_text(self):
@@ -1067,6 +1071,7 @@ class ServeClientFasterWhisper(ServeClientBase):
             for i, s in enumerate(segments[:-1]):
                 text_ = s.text
                 self.text.append(text_)
+                with self.lock:
                     start, end = self.timestamp_offset + s.start, self.timestamp_offset + min(duration, s.end)
 
                 if start >= end:
@@ -1080,6 +1085,7 @@ class ServeClientFasterWhisper(ServeClientBase):
         # only process the last segment if it satisfies the no_speech_thresh
         if segments[-1].no_speech_prob <= self.no_speech_thresh:
             self.current_out += segments[-1].text
+            with self.lock:
                 last_segment = self.format_segment(
                     self.timestamp_offset + segments[-1].start,
                     self.timestamp_offset + min(duration, segments[-1].end),
@@ -1098,6 +1104,7 @@ class ServeClientFasterWhisper(ServeClientBase):
         if self.same_output_count > self.same_output_threshold:
             if not len(self.text) or self.text[-1].strip().lower() != self.current_out.strip().lower():
                 self.text.append(self.current_out)
+                with self.lock:
                     self.transcript.append(self.format_segment(
                         self.timestamp_offset,
                         self.timestamp_offset + duration,
@@ -1113,6 +1120,7 @@ class ServeClientFasterWhisper(ServeClientBase):
 
         # update offset
         if offset is not None:
+            with self.lock:
                 self.timestamp_offset += offset
 
         return last_segment