Hugging Face's logo

ctl
/
wav2vec2-large-xlsr-cantonese

Automatic Speech Recognition Transformers PyTorch JAX
Yue Chinese wav2vec2 audio speech xlsr-fine-tuning-week Eval Results Inference Endpoints
Model card Files Community
3
ctl commited on
Commit
ae2e3a0
•
1 Parent(s): 72f3c91

update README

Browse files
Files changed (1) hide show
  1. README.md +1 -3
README.md CHANGED
@@ -77,8 +77,6 @@ The model can be evaluated as follows on the Chinese (Hong Kong) test data of Co
77
 
78
 
79
  ```python
80
- !mkdir cer
81
- !wget -O cer/cer.py https://huggingface.co/ctl/wav2vec2-large-xlsr-cantonese/raw/main/cer.py
82
  !pip install jiwer
83
  import torch
84
  import torchaudio
@@ -93,7 +91,7 @@ model_id = "ctl/wav2vec2-large-xlsr-cantonese"
93
  chars_to_ignore_regex = '[\,\?\.\!\-\;\:"\“\%\‘\”\�\．\⋯\！\－\：\–\。\》\,\）\,\？\；\～\~\…\︰\，\（\」\‧\《\﹔\、\—\／\,\「\﹖\·\']'
94
 
95
  test_dataset = load_dataset("common_voice", f"{lang_id}", split="test")
96
- cer = load_metric("./cer")
97
 
98
  processor = Wav2Vec2Processor.from_pretrained(f"{model_id}")
99
  model = Wav2Vec2ForCTC.from_pretrained(f"{model_id}")
77
 
78
 
79
  ```python
 
 
80
  !pip install jiwer
81
  import torch
82
  import torchaudio
91
  chars_to_ignore_regex = '[\,\?\.\!\-\;\:"\“\%\‘\”\�\．\⋯\！\－\：\–\。\》\,\）\,\？\；\～\~\…\︰\，\（\」\‧\《\﹔\、\—\／\,\「\﹖\·\']'
92
 
93
  test_dataset = load_dataset("common_voice", f"{lang_id}", split="test")
94
+ cer = load_metric("cer")
95
 
96
  processor = Wav2Vec2Processor.from_pretrained(f"{model_id}")
97
  model = Wav2Vec2ForCTC.from_pretrained(f"{model_id}")