Update app.py

app.py

@@ -1,8 +1,9 @@
 import gradio as gr
 import librosa
-from transformers import Wav2Vec2Processor, SpeechEncoderDecoderModel
+from transformers import AutoFeatureExtractor, AutoTokenizer, SpeechEncoderDecoderModel
     
-processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-xls-r-300m-en-to-15", use_auth_token="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")
+feature_extractor = AutoFeatureExtractor.from_pretrained("facebook/wav2vec2-xls-r-300m-en-to-15", use_auth_token="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")
+tokenizer = AutoTokenizer.from_pretrained("facebook/wav2vec2-xls-r-300m-en-to-15", use_auth_token="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")
 model = SpeechEncoderDecoderModel.from_pretrained("facebook/wav2vec2-xls-r-300m-en-to-15", use_auth_token="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")
 
 def process_audio_file(file):
@@ -10,7 +11,7 @@ def process_audio_file(file):
     if sr != 16000:
         data = librosa.resample(data, sr, 16000)
     print(data.shape)
-    input_values = processor(data, return_tensors="pt").input_values
+    input_values = feature_extractor(data, return_tensors="pt").input_values
     return input_values
     
 def transcribe(file):
@@ -18,7 +19,7 @@ def transcribe(file):
     
     sequences = model.generate(input_values, num_beams=1, max_length=30)
     
-    transcription = processor.batch_decode(sequences)
+    transcription = tokenizer.batch_decode(sequences)
     return transcription[0]
     
 iface = gr.Interface(