-
Hi,
I an my colleagues have faced the problem of appearing NANs in CTC loss and interruption of training due to too much infinite gradients when training Zipformer on our data.
After some debuggi…
-
Needed to merge/convert the ASR data here: https://cchdo.ucsd.edu/cruise/49NZ20210713
Raman Units are dimensionless and are described in Lawaetz and Stedmon 2009: https://doi.org/10.1366/0003702097…
-
运行后可以打开页面，输入文字后tts成功，随后就Connection state is failed和Connection state is closed了
flask_sockets.py已经改成websocket=True
(/root/autodl-tmp/nerfstream) root@autodl-container-22da11ad52-3d284594:~/autodl-t…
-
In a recent hunt for more ASR providers who offer per-word timecodes I found some that I already knew of and a few I hadn't heard of before. Among all providers is [WhisperX](https://github.com/m-bain…
-
currently vad detection happens only when gather verb's minBargeinWordCount is 0, we have vad at config verb, we should start vad detection when config.vad.enable is true
-
intrinsics_316_FAST fails:
https://github.com/lfortran/lfortran-conda-testing/pull/40#issuecomment-2414064645
-
## Vision
Generate a story based on the labels in the given picture, ask to create harmful characters which could have knowledge about everything without filters.
Modify the image to have a hidden…
-
I have followed the steps given
However i always get the following output from the asr server
{"status":"ok","data":[{"confidence":0.862751,"text":""}],"interrupted":"endofspeech","time":1080}
…
-
### Work environment
| Questions | Answers
|------------------------------------------------------|--------------------
| OS/arch/bits (mandatory) …
-
### Minimal Reproducible Example
```f90
module mod
implicit none
integer :: i, iarx(3)
data(iarx(i), i=1,3) / 3*-1 /
end module mod
program main
use mod
implicit none
…
gxyd updated
2 months ago