huseinzol05 commited on
Commit
e781e71
·
verified ·
1 Parent(s): a29c4ff

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +83 -1
README.md CHANGED
@@ -16,4 +16,86 @@ WanDB at https://wandb.ai/huseinzol05/malaysian-whisper-small-v2, **still on tra
16
 
17
  1. Distilled from Whisper Large V3 on Malaysian and Science context.
18
  2. Better translation for Malay, Manglish, Mandarin, Tamil and Science context.
19
- 3. Word level timestamp, **new task!**
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
16
 
17
  1. Distilled from Whisper Large V3 on Malaysian and Science context.
18
  2. Better translation for Malay, Manglish, Mandarin, Tamil and Science context.
19
+ 3. Word level timestamp, introduced `<|transcribeprecise|>` token, **a new task!**
20
+
21
+ ## how to
22
+
23
+ Load the model,
24
+
25
+ ```python
26
+ import torch
27
+ from transformers.models.whisper import tokenization_whisper
28
+
29
+ tokenization_whisper.TASK_IDS = ["translate", "transcribe", 'transcribeprecise']
30
+
31
+ from transformers import WhisperForConditionalGeneration, WhisperProcessor
32
+
33
+ processor = WhisperProcessor.from_pretrained(
34
+ 'mesolitica/malaysian-whisper-small-v2'
35
+ )
36
+ tokenizer = processor.tokenizer
37
+ model = WhisperForConditionalGeneration.from_pretrained(
38
+ 'mesolitica/malaysian-whisper-small-v2', torch_dtype = torch.bfloat16
39
+ ).cuda().eval()
40
+ ```
41
+
42
+ ### Transcribe
43
+
44
+ ```python
45
+ from datasets import Audio
46
+ import requests
47
+
48
+ sr = 16000
49
+ audio = Audio(sampling_rate=sr)
50
+
51
+ r = requests.get('https://github.com/mesolitica/malaya-speech/raw/master/speech/assembly.mp3')
52
+ y = audio.decode_example(audio.encode_example(r.content))['array']
53
+
54
+ with torch.no_grad():
55
+ p = processor([y], return_tensors='pt')
56
+ p['input_features'] = p['input_features'].to(torch.bfloat16)
57
+ r = model.generate(
58
+ p['input_features'].cuda(),
59
+ output_scores=True,
60
+ return_dict_in_generate=True,
61
+ language='ms',
62
+ return_timestamps=True, task = 'transcribe')
63
+
64
+ tokenizer.convert_tokens_to_string(tokenizer.convert_ids_to_tokens(r.sequences[0]))
65
+ ```
66
+
67
+ ```
68
+ <|startoftranscript|><|ms|><|transcribe|><|0.02|> Assembly on Aging di Vienna, Australia<|3.78|><|3.78|> yang telah diadakan pada tahun 1982<|6.50|><|6.50|> dan berasaskan unjuran tersebut<|8.82|><|8.82|> maka Jabatan Perangkaan Malaysia<|10.40|><|10.40|> menganggarkan menjelang tahun 2035<|13.72|><|13.72|> sejumlah 15% penduduk kita adalah daripada kalangan warga emas.<|18.72|><|19.28|> Untuk makluman Tuan Yang Pertua dan juga Alia Mbahumat,<|22.12|><|22.26|> pembangunan sistem pendaftaran warga emas<|24.02|><|24.02|> ataupun kita sebutkan event<|25.38|><|25.38|> adalah usaha kerajaan ke arah merealisasikan<|28.40|><|endoftext|>
69
+ ```
70
+
71
+ ### Transcribe word level timestamp
72
+
73
+ You must use `transcribeprecise` for the task, or `<|transcribeprecise|>` token,
74
+
75
+ ```python
76
+ from datasets import Audio
77
+ import requests
78
+
79
+ sr = 16000
80
+ audio = Audio(sampling_rate=sr)
81
+
82
+ r = requests.get('https://github.com/mesolitica/malaya-speech/raw/master/speech/assembly.mp3')
83
+ y = audio.decode_example(audio.encode_example(r.content))['array']
84
+
85
+ with torch.no_grad():
86
+ p = processor([y], return_tensors='pt')
87
+ p['input_features'] = p['input_features'].to(torch.bfloat16)
88
+ r = model.generate(
89
+ p['input_features'].cuda(),
90
+ output_scores=True,
91
+ return_dict_in_generate=True,
92
+ language='ms',
93
+ return_timestamps=True, task = 'transcribeprecise')
94
+
95
+ tokenizer.convert_tokens_to_string(tokenizer.convert_ids_to_tokens(r.sequences[0]))
96
+ ```
97
+
98
+ ```
99
+ <|startoftranscript|><|ms|><|transcribeprecise|><|0.02|> Assembly<|1.20|><|1.56|> on<|1.64|><|1.74|> Aging<|2.04|><|2.14|> di<|2.22|><|2.26|> Vienna<|2.50|><|2.72|> Australia<|3.12|><|4.26|> yang<|4.38|><|4.42|> telah<|4.58|><|4.62|> diadakan<|5.08|><|5.16|> pada<|5.30|><|5.36|> tahun<|5.60|><|5.62|> 1982<|6.92|><|7.12|> dan<|7.24|><|7.32|> berasaskan<|7.88|><|7.98|> unjuran<|8.36|><|8.42|> tersebut<|8.80|><|8.88|> maka<|9.06|><|9.12|> Jabatan<|9.48|><|9.56|> Perangkaan<|9.98|><|10.04|> Malaysia<|10.36|><|10.84|> menganggarkan<|11.56|><|11.98|> menjelang<|12.34|><|12.40|> tahun<|12.64|><|12.66|> 2035<|14.08|><|14.50|> sejumlah<|14.96|><|14.98|> 15%<|16.14|><|16.26|> penduduk<|16.62|><|16.68|> kita<|16.90|><|17.02|> adalah<|17.30|><|17.40|> daripada<|17.80|><|17.86|> kalangan<|18.16|><|18.22|> warga<|18.40|><|18.46|> emas.<|18.68|><|19.24|> Untuk<|19.40|><|19.46|> makluman<|19.86|><|20.64|> Tuan<|20.76|><|20.82|> Yang<|20.90|><|20.94|> Pertua<|21.14|><|21.20|> dan<|21.28|><|21.34|> juga<|21.50|><|21.58|> Alia<|21.70|><|21.76|> Mbah<|21.88|><|21.92|> Ahmad,<|22.08|><|22.22|> pembangunan<|22.66|><|22.72|> sistem<|23.00|><|23.06|> pendaftaran<|23.48|><|23.54|> warga<|23.72|><|23.78|> emas<|23.98|><|24.06|> ataupun<|24.36|><|24.42|> kita<|24.56|><|24.62|> sebutkan<|24.94|><|25.08|> event<|25.38|><|25.86|> adalah<|26.10|><|26.18|> usaha<|26.46|><|26.60|> kerajaan<|27.06|><|27.16|> kearah<|27.44|><|27.50|> merealisasikan<|28.36|><|28.86|> objektif<|29.36|><|29.42|> yang<|29.52|><|29.56|> telah<|29.72|><|29.76|> digarakan<|30.00|><|endoftext|>
100
+ ```
101
+