Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
trowngon
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
trowngon
5y ago
Vosk https://github.com/alphacep/vosk-api
2.
▲
by
trowngon
5y ago
You don't even need to compare accuracy, you can just check the technology. Facebook model is trained on 256 GPU cards and you can fine-tune it to your domain in a day or two. The release was 2 month ago. There is no way any cloud star
3.
▲
by
trowngon
5y ago
I believe most people already moved to offline engines. No need to send the data to some random guys like this Assembly. Nemo Conformer from Nvidia, Robust Wav2Vec from Facebook, Vosk. There are dozen options. And the cost is $0.01 per hour
4.
▲
by
trowngon
5y ago
Movies are no longer just an imagination. Chances are they are shaping the world for real superhero superhuman arrival ;)
5.
▲
Google Assistant’s upcoming 'Quick phrases' will let you skip 'Hey Google'
(9to5google.com)
2 points
by
trowngon
5y ago
|
0 comments
6.
▲
Apple scans your phone (and how to evade it)
(youtube.com)
1 points
by
trowngon
5y ago
|
0 comments
7.
▲
by
trowngon
5y ago
Deepspeech project is closed by Mozilla. Developers fired. Now they are Coqui.
8.
▲
by
trowngon
5y ago
Looks like after nvidia 1.5m grant devs returned back ;)
9.
▲
by
trowngon
5y ago
Looks like a fork of Mozilla DeepSpeech by former DeepSpeech developers. What is the relation to the original project?
10.
▲
by
trowngon
6y ago
Irrespective of the subject Deepspeech is very old archtecture with suboptimal results. You'd better try any recent conformer implementations (flashlight, nemo, wenet, etc) or wav2vec.
11.
▲
by
trowngon
6y ago
Assembly AI is at $0.5 per hour, not extremely reasonable these days. With open source models like Facebook RASR or Vosk you can get self-hosted solution with even better accuracy and cost of $0.05 per hour, 10 times cheaper. Once any of yo
12.
▲
by
trowngon
6y ago
Are there open source projects like this?
13.
▲
by
trowngon
6y ago
Vosk supports both Italian and French. French model is trained by Linto project, pretty good one.
14.
▲
Future of DeepSpeech / STT after recent changes at Mozilla
(discourse.mozilla.org)
147 points
by
trowngon
6y ago
|
74 comments