fixed the resampler with the correct sampling rate which improved

README.md

@@ -23,7 +23,7 @@ model-index:
     metrics:
        - name: Test WER
          type: wer
-         value: 20.67
+         value: 18.32
 ---
 
 # Wav2Vec2-Large-XLSR-Indonesian
@@ -46,12 +46,12 @@ test_dataset = load_dataset("common_voice", "id", split="test[:2%]")
 processor = Wav2Vec2Processor.from_pretrained("indonesian-nlp/wav2vec2-large-xlsr-indonesian")
 model = Wav2Vec2ForCTC.from_pretrained("indonesian-nlp/wav2vec2-large-xlsr-indonesian")
 
-resampler = torchaudio.transforms.Resample(48_000, 16_000)
 
 # Preprocessing the datasets.
 # We need to read the aduio files as arrays
 def speech_file_to_array_fn(batch):
   speech_array, sampling_rate = torchaudio.load(batch["path"])
+  resampler = torchaudio.transforms.Resample(sampling_rate, 16_000)
   batch["speech"] = resampler(speech_array).squeeze().numpy()
   return batch
 
@@ -117,7 +117,7 @@ result = test_dataset.map(evaluate, batched=True, batch_size=8)
 print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
 ```
 
-**Test Result**: 20.67 %
+**Test Result**: 18.32 %
 
 ## Training