Merge branch 'main' into setup.py-to-pyproject.toml

README.md

@@ -126,7 +126,7 @@ audio = whisper.load_audio("audio.mp3")
 audio = whisper.pad_or_trim(audio)
 
 # make log-Mel spectrogram and move to the same device as the model
-mel = whisper.log_mel_spectrogram(audio).to(model.device)
+mel = whisper.log_mel_spectrogram(audio, n_mels=model.dims.n_mels).to(model.device)
 
 # detect the spoken language
 _, probs = model.detect_language(mel)

whisper/audio.py

@@ -122,7 +122,7 @@ def log_mel_spectrogram(
         The path to audio or either a NumPy array or Tensor containing the audio waveform in 16 kHz
 
     n_mels: int
-        The number of Mel-frequency filters, only 80 is supported
+        The number of Mel-frequency filters, only 80 and 128 are supported
 
     padding: int
         Number of zero samples to pad to the right
@@ -132,7 +132,7 @@ def log_mel_spectrogram(
 
     Returns
     -------
-    torch.Tensor, shape = (80, n_frames)
+    torch.Tensor, shape = (n_mels, n_frames)
         A Tensor that contains the Mel spectrogram
     """
     if not torch.is_tensor(audio):

whisper/transcribe.py

@@ -214,6 +214,8 @@ def decode_with_fallback(segment: torch.Tensor) -> DecodingResult:
             if (
                 no_speech_threshold is not None
                 and decode_result.no_speech_prob > no_speech_threshold
+                and logprob_threshold is not None
+                and decode_result.avg_logprob < logprob_threshold
             ):
                 needs_fallback = False  # silence
             if not needs_fallback: