patrickvonplaten commited on
Commit
cbeb18d
1 Parent(s): ac5d2e5

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +4 -4
README.md CHANGED
@@ -34,7 +34,7 @@ Fine-tuned facebook/wav2vec2-large-xlsr-53 on Hindi and Marathi using the OpenSL
34
  ## Usage
35
  The model can be used directly (without a language model) as follows, assuming you have a dataset with Marathi text and audio_path fields:
36
 
37
- ```
38
  import torch
39
  import torchaudio
40
  import librosa
@@ -65,8 +65,8 @@ print("Prediction:", processor.batch_decode(predicted_ids))
65
  print("Reference:", test_data["text"][:2])
66
  Evaluation
67
  The model can be evaluated as follows on 10% of the Marathi data on OpenSLR.
68
- ```
69
- ```
70
  import torchaudio
71
  from datasets import load_metric
72
  from transformers import Wav2Vec2Processor,Wav2Vec2ForCTC
@@ -81,7 +81,7 @@ model = Wav2Vec2ForCTC.from_pretrained("tanmaylaud/wav2vec2-large-xlsr-hindi-mar
81
 
82
  model.to("cuda")
83
 
84
- chars_to_ignore_regex = '[\\\\,\\\\?\\\\.\\\\!\\\\-\\\\;\\\\:\\\\"\\\\“\\\\%\\\\‘\\\\”\\\\�\\\\।]'
85
 
86
  # Preprocessing the datasets.
87
  # We need to read the audio files as arrays
34
  ## Usage
35
  The model can be used directly (without a language model) as follows, assuming you have a dataset with Marathi text and audio_path fields:
36
 
37
+ ```python
38
  import torch
39
  import torchaudio
40
  import librosa
65
  print("Reference:", test_data["text"][:2])
66
  Evaluation
67
  The model can be evaluated as follows on 10% of the Marathi data on OpenSLR.
68
+
69
+ ```python
70
  import torchaudio
71
  from datasets import load_metric
72
  from transformers import Wav2Vec2Processor,Wav2Vec2ForCTC
81
 
82
  model.to("cuda")
83
 
84
+ chars_to_ignore_regex = '[\\\\\\\\,\\\\\\\\?\\\\\\\\.\\\\\\\\!\\\\\\\\-\\\\\\\\;\\\\\\\\:\\\\\\\\"\\\\\\\\“\\\\\\\\%\\\\\\\\‘\\\\\\\\”\\\\\\\\�\\\\\\\\।]'
85
 
86
  # Preprocessing the datasets.
87
  # We need to read the audio files as arrays