cahya commited on
Commit
2fd961b
1 Parent(s): 6b1b272

fixed the chars and tabs

Browse files
Files changed (1) hide show
  1. README.md +10 -10
README.md CHANGED
@@ -87,31 +87,31 @@ processor = Wav2Vec2Processor.from_pretrained("indonesian-nlp/wav2vec2-large-xls
87
  model = Wav2Vec2ForCTC.from_pretrained("indonesian-nlp/wav2vec2-large-xlsr-indonesian-baseline")
88
  model.to("cuda")
89
 
90
- chars_to_ignore_regex = '[\\\\,\\\\?\\\\.\\\\!\\\\-\\\\;\\\\:\\\\"\\\\“\\\\%\\\\‘\\\\'\\\\”]'
91
 
92
  resampler = torchaudio.transforms.Resample(48_000, 16_000)
93
 
94
  # Preprocessing the datasets.
95
  # We need to read the aduio files as arrays
96
  def speech_file_to_array_fn(batch):
97
- \\tbatch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
98
- \\tspeech_array, sampling_rate = torchaudio.load(batch["path"])
99
- \\tbatch["speech"] = resampler(speech_array).squeeze().numpy()
100
- \\treturn batch
101
 
102
  test_dataset = test_dataset.map(speech_file_to_array_fn)
103
 
104
  # Preprocessing the datasets.
105
  # We need to read the audio files as arrays
106
  def evaluate(batch):
107
- \\tinputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
108
 
109
- \\twith torch.no_grad():
110
- \\t\\tlogits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
111
 
112
  pred_ids = torch.argmax(logits, dim=-1)
113
- \\tbatch["pred_strings"] = processor.batch_decode(pred_ids)
114
- \\treturn batch
115
 
116
  result = test_dataset.map(evaluate, batched=True, batch_size=8)
117
 
 
87
  model = Wav2Vec2ForCTC.from_pretrained("indonesian-nlp/wav2vec2-large-xlsr-indonesian-baseline")
88
  model.to("cuda")
89
 
90
+ chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\%\‘\'\”\�]'
91
 
92
  resampler = torchaudio.transforms.Resample(48_000, 16_000)
93
 
94
  # Preprocessing the datasets.
95
  # We need to read the aduio files as arrays
96
  def speech_file_to_array_fn(batch):
97
+ batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
98
+ speech_array, sampling_rate = torchaudio.load(batch["path"])
99
+ batch["speech"] = resampler(speech_array).squeeze().numpy()
100
+ return batch
101
 
102
  test_dataset = test_dataset.map(speech_file_to_array_fn)
103
 
104
  # Preprocessing the datasets.
105
  # We need to read the audio files as arrays
106
  def evaluate(batch):
107
+ inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
108
 
109
+ with torch.no_grad():
110
+ logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
111
 
112
  pred_ids = torch.argmax(logits, dim=-1)
113
+ batch["pred_strings"] = processor.batch_decode(pred_ids)
114
+ return batch
115
 
116
  result = test_dataset.map(evaluate, batched=True, batch_size=8)
117