hugobloem / wyoming-microsoft-tts

Wyoming protocol server for Microsoft Azure text-to-speech
13 stars 3 forks source link

Add Dockerfile #40

Closed hugobloem closed 7 months ago

nerijus commented 7 months ago

Are docker images available somewhere?

hugobloem commented 7 months ago

Soon, I just need to set up building them for various platforms. However, you can use the Dockerfile in this repo to build your own if you want.

hugobloem commented 7 months ago

Hi, the Docker container is now avaible here:

docker pull ghcr.io/hugobloem/wyoming-microsoft-tts-noha:latest
nerijus commented 7 months ago

Thank you. uri example in Readme should probably be tcp://0.0.0.0:10200 instead of tcp://0.0.0.0:10300?

nerijus commented 7 months ago

I do not see the port open. I use podman, not docker, and run it like this:

podman run \
          --name=wyoming-microsoft-tts \
          --network=host \
          -p 10200:10200 \
          ghcr.io/hugobloem/wyoming-microsoft-tts-noha:latest --service-region northeurope --subscription-key b2d.... \
          --uri tcp://192.168.1.10:10200 --voice lt-LT-OnaNeural --debug

(tried with tcp://0.0.0.0:10200 too). ss -nlp|grep 10020 does not list an open port, telnet 192.168.1.10 10020 gives Connection refused. I have a Rhasspy container with similar setup (it has -p 12101:12101), and I do see 12101 port opened with ss and can telnet into it. Any ideas?

nerijus commented 7 months ago

Oops, I am dumb, used 10020 instead of 10200 :) The port is open.

hugobloem commented 7 months ago

Glad you resolved it 😜

nerijus commented 7 months ago

Thank you, I was able to add both tts and stt to HA and it works!