Hugging Face's logo

abr-ai
/
asr-19m-v2-en-32b

Automatic Speech Recognition
Transformers
English
asr-19m-v2
custom_code
Eval Results
Model card Files
xet
Community
Daniel Rasmussen commited on
Commit
8da6334
·
1 Parent(s): 31746ab

Add model-index metadata

Browse files
Files changed (1) hide show
  1. README.md +123 -0
README.md CHANGED
@@ -8,6 +8,129 @@ metrics:
8
  - wer
9
  pipeline_tag: automatic-speech-recognition
10
  library_name: transformers
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  ---
12
 
13
  # ABR's asr-19m-v2-en-32b SSM
 
8
  - wer
9
  pipeline_tag: automatic-speech-recognition
10
  library_name: transformers
11
+ model-index:
12
+ - name: abr-ai/asr-19m-v2-en-32b
13
+ results:
14
+ - task:
15
+ type: automatic-speech-recognition
16
+ name: Automatic Speech Recognition
17
+ dataset:
18
+ type: hf-audio/esb-datasets-test-only-sorted
19
+ name: AMI
20
+ config: ami
21
+ split: test
22
+ metrics:
23
+ - type: wer
24
+ value: 18.76
25
+ name: WER%
26
+ source:
27
+ name: Open ASR Leaderboard
28
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
29
+ - task:
30
+ type: automatic-speech-recognition
31
+ name: Automatic Speech Recognition
32
+ dataset:
33
+ type: hf-audio/esb-datasets-test-only-sorted
34
+ name: Earnings22
35
+ config: earnings22
36
+ split: test
37
+ metrics:
38
+ - type: wer
39
+ value: 13.53
40
+ name: WER%
41
+ source:
42
+ name: Open ASR Leaderboard
43
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
44
+ - task:
45
+ type: automatic-speech-recognition
46
+ name: Automatic Speech Recognition
47
+ dataset:
48
+ type: hf-audio/esb-datasets-test-only-sorted
49
+ name: Gigaspeech
50
+ config: gigaspeech
51
+ split: test
52
+ metrics:
53
+ - type: wer
54
+ value: 15.44
55
+ name: WER%
56
+ source:
57
+ name: Open ASR Leaderboard
58
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
59
+ - task:
60
+ type: automatic-speech-recognition
61
+ name: Automatic Speech Recognition
62
+ dataset:
63
+ type: hf-audio/esb-datasets-test-only-sorted
64
+ name: LibriSpeech Clean
65
+ config: librispeech
66
+ split: test.clean
67
+ metrics:
68
+ - type: wer
69
+ value: 4.66
70
+ name: WER%
71
+ source:
72
+ name: Open ASR Leaderboard
73
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
74
+ - task:
75
+ type: automatic-speech-recognition
76
+ name: Automatic Speech Recognition
77
+ dataset:
78
+ type: hf-audio/esb-datasets-test-only-sorted
79
+ name: LibriSpeech Other
80
+ config: librispeech
81
+ split: test.other
82
+ metrics:
83
+ - type: wer
84
+ value: 11.16
85
+ name: WER%
86
+ source:
87
+ name: Open ASR Leaderboard
88
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
89
+ - task:
90
+ type: automatic-speech-recognition
91
+ name: Automatic Speech Recognition
92
+ dataset:
93
+ type: hf-audio/esb-datasets-test-only-sorted
94
+ name: SPGISpeech
95
+ config: spgispeech
96
+ split: test
97
+ metrics:
98
+ - type: wer
99
+ value: 3.94
100
+ name: WER%
101
+ source:
102
+ name: Open ASR Leaderboard
103
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
104
+ - task:
105
+ type: automatic-speech-recognition
106
+ name: Automatic Speech Recognition
107
+ dataset:
108
+ type: hf-audio/esb-datasets-test-only-sorted
109
+ name: Tedlium
110
+ config: tedlium
111
+ split: test
112
+ metrics:
113
+ - type: wer
114
+ value: 7.53
115
+ name: WER%
116
+ source:
117
+ name: Open ASR Leaderboard
118
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
119
+ - task:
120
+ type: automatic-speech-recognition
121
+ name: Automatic Speech Recognition
122
+ dataset:
123
+ type: hf-audio/esb-datasets-test-only-sorted
124
+ name: Voxpopuli
125
+ config: voxpopuli
126
+ split: test
127
+ metrics:
128
+ - type: wer
129
+ value: 9.88
130
+ name: WER%
131
+ source:
132
+ name: Open ASR Leaderboard
133
+ url: https://huggingface.co/spaces/hf-audio/open_asr_leaderboard
134
  ---
135
 
136
  # ABR's asr-19m-v2-en-32b SSM