TenzinGayche
/

whisper-small-3

Automatic Speech Recognition

Inference Endpoints

Model card Files Files and versions Community

TenzinGayche commited on Sep 21, 2023

Commit

6212570

•

1 Parent(s): e205326

Update handler.py

Files changed (1) hide show

handler.py +1 -1

handler.py CHANGED Viewed

@@ -27,7 +27,7 @@ class EndpointHandler():
         inputs = data.pop("inputs", data)
         audio_nparray = ffmpeg_read(inputs, SAMPLE_RATE)
         audio_tensor= torch.from_numpy(audio_nparray)
-        text = self.pipe(audio_tensor)["text"]
         # run inference pipeline
         result = converter.toUnicode(text)

         inputs = data.pop("inputs", data)
         audio_nparray = ffmpeg_read(inputs, SAMPLE_RATE)
         audio_tensor= torch.from_numpy(audio_nparray)
+        text = self.pipe(audio_tensor.numpy())["text"]
         # run inference pipeline
         result = converter.toUnicode(text)