[dataset] support pad or trim for whisper decoding (#2378)

* [dataset] support pad or trim for whisper decoding

* refactor

* refactor

wenet/dataset/processor.py

@@ -287,13 +287,16 @@ def compute_log_mel_spectrogram(sample,
                                 n_fft=400,
                                 hop_length=160,
                                 num_mel_bins=80,
-                                padding=0):
+                                padding=0,
+                                pad_or_trim: bool = False,
+                                max_duration: int = 30):
     """ Extract log mel spectrogram, modified from openai-whisper, see:
         - https://github.com/openai/whisper/blob/main/whisper/audio.py
         - https://github.com/wenet-e2e/wenet/pull/2141#issuecomment-1811765040
 
         Args:
             sample: {key, wav, sample_rate, ...}
+            max_duration: valid when pad_or_trim is True (orign whisper style)
 
         Returns:
             {key, feat, wav, sample_rate, ...}
@@ -305,6 +308,13 @@ def compute_log_mel_spectrogram(sample,
     waveform = sample['wav'].squeeze(0)  # (channel=1, sample) -> (sample,)
     if padding > 0:
         waveform = F.pad(waveform, (0, padding))
+    if pad_or_trim:
+        length = max_duration * sample_rate
+        if waveform.size(0) >= length:
+            waveform = waveform[:length]
+        else:
+            waveform = F.pad(waveform, (0, length - waveform.size(0)))
+
     window = torch.hann_window(n_fft)
     stft = torch.stft(waveform,
                       n_fft,