Update README.md

README.md

@@ -11,11 +11,11 @@ Multilingual speaker-invariant speech content encoder based on [SPIN method](htt
 Usage:
 ```py
 from transformers import AutoProcessor, HubertModel
-processor = AutoProcessor.from_pretrained("hr16/multilingual_spin")
+feature_extractor = AutoFeatureExtractor.from_pretrained("hr16/multilingual_spin")
 model = HubertModel.from_pretrained("hr16/multilingual_spin")
 
 audio_input = torch.randn(16000)  # 1 second of audio at 16kHz
-inputs = processor(audio_input, sampling_rate=16000, return_tensors="pt")
+inputs = feature_extractor(audio_input, sampling_rate=16000, return_tensors="pt")
 with torch.no_grad():
     outputs = model(**inputs)