Update app.py

app.py

@@ -50,16 +50,16 @@ def transcribe(audio_path):
     audio_cnt, sr = librosa.load(audio_path, sr=16000)
 
     # 将音频数据传递给 processor
-    input_values = processor(audio_cnt, sampling_rate=16000, return_tensors="pt")
+    input_features = processor(audio_cnt, sampling_rate=16000, return_tensors="pt").input_features
 
-    print(input_values)
+    print(input_features)
     
     # 模型推理
     with torch.no_grad():
-        predicted_ids = model.generate(input_values)
+        generated_ids = model.generate(input_features)
     
     # 解码得到转录结果
-    transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)[0]
+    transcription = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
     
     return transcription