stefan-it commited on
Commit
25d4aa5
1 Parent(s): 1335cb7

readme: add final result on test set

Browse files
Files changed (1) hide show
  1. README.md +4 -4
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  metrics:
22
  - name: Test WER
23
  type: wer
24
- value: ???
25
  ---
26
 
27
  # Wav2Vec2-Large-XLSR-53-Basque
@@ -104,7 +104,7 @@ model = Wav2Vec2ForCTC.from_pretrained("stefan-it/wav2vec2-large-xlsr-53-basque"
104
 
105
  model.to("cuda")
106
 
107
- chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\%\‘\”\�]'
108
 
109
  resampler = torchaudio.transforms.Resample(48_000, 16_000)
110
 
@@ -148,13 +148,13 @@ print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"],
148
 
149
  ```
150
 
151
- **Test Result**: %
152
 
153
  ## Training
154
 
155
  The Common Voice `train`, `validation` datasets were used for training.
156
 
157
- The script used for training can be found ???
158
 
159
  ## Acknowledgements
160
 
 
21
  metrics:
22
  - name: Test WER
23
  type: wer
24
+ value: 18.272625
25
  ---
26
 
27
  # Wav2Vec2-Large-XLSR-53-Basque
 
104
 
105
  model.to("cuda")
106
 
107
+ chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“\\%\\‘\\”\\�]'
108
 
109
  resampler = torchaudio.transforms.Resample(48_000, 16_000)
110
 
 
148
 
149
  ```
150
 
151
+ **Test Result**: 18.272625%
152
 
153
  ## Training
154
 
155
  The Common Voice `train`, `validation` datasets were used for training.
156
 
157
+ The script used for training can be found here, hopefully very soon!
158
 
159
  ## Acknowledgements
160