I self-host it in a docker container. You will have to download abour 4 gigabytes of "n-gram" data. And there are no AI features in the self-hosted version.
Here's my podman-compose.yml file (you can use this as you docker-compose file as well):
version: "3"
services:
languagetool:
image: erikvl87/languagetool:latest
container_name: languagetool
ports:
- 8010:8010 # Using default port from the image
environment:
- langtool_languageModel=/ngrams # OPTIONAL: Using ngrams data
- Java_Xms=512m # OPTIONAL: Setting a minimal Java heap size of 512 mib
- Java_Xmx=1g # OPTIONAL: Setting a maximum Java heap size of 1 Gib
volumes:
- path/to/ngrams:/ngrams
restart: always
Just download the ngrams from this link and change the path in the file. You can use this with a Firefox or Thunderbird addon. Go to the advanced settings in the addon preferences and add http(s)://YOUR_IP_OR_DOMAIN/v2 in the other server option.
6 Comments
Helix@feddit.de · 11 pts · 2y
Yes, it's still open source: https://github.com/languagetool-org/languagetool
And yes, it's worth it. However I think that LLaMA based ML models supplementing LanguageTool will be the future.
Even_Adder@lemmy.dbzer0.com · -1 pts · 2y
Very cool.
nutbutter@discuss.tchncs.de · 3 pts · 2y
I self-host it in a docker container. You will have to download abour 4 gigabytes of "n-gram" data. And there are no AI features in the self-hosted version.
gonzoknowsdotcom1@monero.town · 2 pts · 2y
do you have a guide?
nutbutter@discuss.tchncs.de · 1 pts · 2y
Here's my podman-compose.yml file (you can use this as you docker-compose file as well):
Just download the ngrams from this link and change the path in the file. You can use this with a Firefox or Thunderbird addon. Go to the advanced settings in the addon preferences and add http(s)://YOUR_IP_OR_DOMAIN/v2 in the other server option.
References:
PeachMan@lemmy.one · 2 pts · 2y
Worth what? It's free! And yes, it's open source. It can also be self-hosted if you're paranoid.