 Jong Wook Kim
|
a84191faae
rename GitHub workflow
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
b1d213c0c7
allow test_transcribe to run on CPU when CUDA is not available
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
493dfffa37
add github action to run pytest
|
%!s(int64=2) %!d(string=hai) anos |
 Mikko Vedru
|
0f39c89d92
Update README.md (#804)
|
%!s(int64=2) %!d(string=hai) anos |
 Markus Hennerbichler
|
6df3ea1fb5
Support batch-dimension in log_mel_spectogram (#839)
|
%!s(int64=2) %!d(string=hai) anos |
 adamreis
|
70861c7ce3
Fix tiny transcribe() docstring typo (#857)
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
f82bc59f5e
torch.concatenate -> torch.cat for compatibility
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
28769fcfe5
word-level timestamps in Multilingual_ASR notebook
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
53807677fe
MultiHeadAttention to return qk as well
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
9323b2526c
Revert "saving the qk matrix in the attention module for convenience"
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
68e44bd83c
saving the qk matrix in the attention module for convenience
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
0b5dcfdef7
large-v2 figure and arxiv url update
|
%!s(int64=2) %!d(string=hai) anos |
 altryne
|
b9265e5796
Update Hebrew language code to he per IANA registry (#401)
|
%!s(int64=2) %!d(string=hai) anos |
 Paul Harter
|
fd8f80c8b8
Explicitly closing model file after reading it (#630)
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
4179ed2475
add large-v2 model
|
%!s(int64=2) %!d(string=hai) anos |
 jumon
|
ec1b34bb90
fix compression ratio function (#561)
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
eff383b27b
invoking __call__ instead of forward()
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
02aa851a49
fix to return only the text token ids
|
%!s(int64=2) %!d(string=hai) anos |
|
jumon
|
76148a56c5
suppress generating non-timestamp tokens at the beginning (#532)
|
%!s(int64=2) %!d(string=hai) anos |
 Vicki Anand
|
9f70a352f9
Fix attention caching to make it actually work (#370)
|
%!s(int64=2) %!d(string=hai) anos |
 Sumana Harihareswara
|
7f3e408e09
Add package metadata to setup.py (#315)
|
%!s(int64=2) %!d(string=hai) anos |
 Michael Monashev
|
f680570016
Fix bug (#305)
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
d18e9ea5dd
transcribe() on English-only model won't complain when language="en" is not given
|
%!s(int64=2) %!d(string=hai) anos |
 David Marx
|
82725cea9c
infer download_root from XDG_CACHE_HOME if avail (#257)
|
%!s(int64=2) %!d(string=hai) anos |
 eudoxos
|
35713c66e0
Add --threads option to transcribe (#278)
|
%!s(int64=2) %!d(string=hai) anos |
 Corentin Jemine
|
9e653bd0ea
Fixed CoW RuntimeError in DecodingTask.run() (#240)
|
%!s(int64=2) %!d(string=hai) anos |
 Tom Stuart
|
02b74308ff
Fix timestamps and strip extraneous whitespace in WebVTT output (#219)
|
%!s(int64=2) %!d(string=hai) anos |
 Jibin Mathew
|
0b1ba3d46e
Add model_dir to arguments (#202)
|
%!s(int64=2) %!d(string=hai) anos |
 Caleb McQuillin
|
60132ade70
Use , character instead of . for SRT output. (#197)
|
%!s(int64=2) %!d(string=hai) anos |
|
Jong Wook Kim
|
7cb4cc21bf
allowing nonzero initial temperature
|
%!s(int64=2) %!d(string=hai) anos |
