issues
search
kensho-technologies
/
pyctcdecode
A fast and lightweight python-based CTC beam search decoder for speech recognition.
Apache License 2.0
416
stars
89
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Expand testing and linting to cover 3.6-3.9
#15
poneill
closed
2 years ago
0
fix typo in README
#14
poneill
closed
3 years ago
0
`hotwords_weight` should be `hotword_weight` in README
#13
rbracco
closed
3 years ago
1
TypeError: max() received an invalid combination of arguments
#12
rbracco
closed
3 years ago
3
ModuleNotFoundError: No module named 'numpy.typing'
#11
lyjzsyzlt
closed
3 years ago
2
Finish mypyification
#10
poneill
closed
3 years ago
0
Using Nemo with BPE models
#9
pehonnet
closed
3 years ago
5
How to get the timestamps?
#8
khursani8
closed
3 years ago
2
increase test coverage
#7
poneill
closed
3 years ago
1
Make linters fix by default
#6
poneill
closed
3 years ago
4
Transcription being concatenated oddly
#5
usmanfarooq619
closed
3 years ago
10
Add robustness to automatic decoder instantiation
#4
gkucsko
closed
2 years ago
6
Alphabet conversion from Hugging Faces do not work
#3
flariut
closed
3 years ago
13
Modify kenlm dependency for pypi compatibility
#2
gkucsko
closed
3 years ago
1
Set up CI
#1
gkucsko
closed
3 years ago
0
Previous