hotfix 3.0

app.py

@@ -31,8 +31,15 @@ def transcribe(inputs, task):
     if inputs is None:
         raise gr.Error("No audio file submitted! Please upload or record an audio file before submitting your request.")
     
+    waveform, sample_rate = torchaudio.load(inputs)
+    if sample_rate != 16000:
+        transform = torchaudio.transforms.Resample(orig_freq=sample_rate, new_freq=16000)
+        waveform = transform(waveform)
+    
     segments, _ = model.transcribe(inputs, task=task)
-    text = " ".join([segment.text for segment in segments])
+    transcription = "\n".join([segment.text for segment in segments])
+    
+    # Diarisation avec le pipeline de pyannote
     diarization = pipeline(inputs)
     speaker_segments = []
     for segment, _, speaker in diarization.itertracks(yield_label=True):
@@ -40,9 +47,11 @@ def transcribe(inputs, task):
     
     # Associer les segments de transcription aux locuteurs
     speaker_texts = []
-    #diarization = pipeline({"uri": "audio", "audio": audio_path})
-    #speaker_segments = []
-    return text
+    for start, end, speaker in speaker_segments:
+        spoken_text = " ".join([seg.text for seg in segments if seg.start >= start and seg.end <= end])
+        if spoken_text:
+            speaker_texts.append(f"{speaker}: {spoken_text}")
+    return "\n".join(speaker_texts)
 
 def _return_yt_html_embed(yt_url):
     video_id = yt_url.split("?v=")[-1]