-
Is there any update coming up on the English model? (voxforge/tri2b_mmi_b0.05)
According to [history](https://github.com/alumae/kaldi-gstreamer-server/tree/master/test/models/english/voxforge/tri2b…
-
I am trying to test my own custom Kaldi model.
With the following yaml configure file, I got an error which was written in worker.log file:
```
decoder:
use-threaded-decoder: true
mo…
hcnoh updated
5 years ago
-
First off thanks for the docker file. It has really saved me a lot of time, while truing to get Kaldi up and running quickly.
I tried the practical end to end example, but I keep getting a 'No decoder…
-
Are their guides on how to deploy this to heroku? I've set up the server and It's working fine.
-
I have a 10 second silence timeout configured, which kicks in as expected, however when the worker is sending the final status message to the client the web socket seems to already be closed. It looks…
-
i noticed the demo http://bark.phon.ioc.ee/dictate/ could get audio from microphone, but current README only showed the recognization by sending audio file. which api could i use to get that.
by th…
-
To support client side VAD, the server should send FINAL hyp after a specified timeout, even if it's not been recieving any audio (so not much silence is detected)
-
i have tried this and got the same type of problem in decoder. and i am trying: "Using the 'onlinegmmdecodefaster' based worker"
python kaldigstserver/worker.py -u ws://localhost:8888/worker/ws/spe…
-
**wget https://phon.ioc.ee/~tanela/tedlium_nnet_ms_sp_online.tgz**
converted 'https://phon.ioc.ee/~tanela/tedlium_nnet_ms_sp_online.tgz' (ANSI_X3.4-1968) -> 'https://phon.ioc.ee/~tanela/tedlium_nnet…
-
For services that support this (e.g. the service that is based on kaldi-gstreamer-server) present the status of the service (number of available slots, latency, etc.) somewhere / under certain conditi…