NVIDIA
NVIDIA
STT Fr Conformer-CTC Large
Model
NVIDIA
NVIDIA
STT Fr Conformer-CTC Large

Conformer-CTC-Large model for French Automatic Speech Recognition, Trained on NeMo ASRSET 2.0

2 Versions
1.5.1Selected
12/14/2021 11:08 PM UTC983.3 MB500 EpochsBatch Size: 16
ACCURACY (BEAM SEARCH + LM, size=128, alpha=.5, beta=1)
KeyValue
MCV7.0 DEV7.95
MCV7.0 DEV-no hyphen7.48
MCV7.0 TEST9.16
MCV7.0 TEST-no hyphen8.46
MLS DEV5.57
MLS DEV-no hyphen5.51
MLS TEST4.66
MLS TEST-no hyphen4.40
ACCURACY (GREEDY)
KeyValue
MCV7.0 DEV8.35
MCV7.0 DEV-no hyphen7.88
MCV7.0 TEST9.63
MCV7.0 TEST-no hyphen9.01
MLS DEV5.88
MLS DEV-no hyphen5.90
MLS TEST4.91
MLS TEST-no hyphen4.63
MODEL
KeyValue
ARCHITECTURECITRINET
INPUTS16000 KHZ MONO-CHANNEL AUDIO (WAV FILES)
OUTPUTSTRANSCRIBED SPEECH
11/16/2021 9:11 PM UTC491.6 MB500 EpochsBatch Size: 16
ACCURACY (BEAM size=128, alpha=.5, beta=1)
KeyValue
MCV7.0 DEV7.95
MCV7.0 TEST9.16
MLS DEV5.57
MLS TEST4.66
ACCURACY (GREEDY)
KeyValue
MCV7.0 DEV8.35
MCV7.0 TEST9.63
MLS DEV5.88
MLS TEST4.91
MODEL
KeyValue
ARCHITECTURECONFORMER-CTC-LARGE
INPUTS16000 KHZ MONO-CHANNEL AUDIO (WAV FILES)
OUTPUTSTRANSCRIBED SPEECH

NVIDIA uses cookies to improve your experience on our web site. We and our third-party partners also use cookies and other tools to collect and record information you provide as well as information about your interactions with our websites for performance improvement, analytics, and to assist in marketing efforts. By clicking "Accept All", you consent to our use of cookies and other tools as described in our Cookie Policy. You can manage your cookie settings by clicking on "Manage Settings." By continuing to use this site or by clicking one of the buttons below, you agree to our Terms of Service (which contains important waivers). Please see our Privacy Policy for more information on our privacy practices.