Generate waifu2x/web/public_html
.
python -m waifu2x.web.webgen
You can also generate minimal WebUI by specifying the --template minimal
option.
minimal UI is intended for use when waifu2x.web is used as an embedded iframe in Google Colab for example.
python -m waifu2x.web.webgen --template minimal
The following line starts the Web Server.
python -m waifu2x.web
The web server starts at http://localhost:8812/ .
Show help.
python -m waifu2x.web -h
With TTA, debug log print
python -m waifu2x.web --tta --debug
or
DEBUG=1 python -m waifu2x.web --tta --debug
Specify HTTP port number and GPU ID.
python -m waifu2x.web --port 8813 --gpu 1
Remove all size limits for the private server.
python -m waifu2x.web --no-size-limit
Copy waifu2x/web/config.ini.sample
to waifu2x/web/config.ini
cp waifu2x/web/config.ini.sample waifu2x/web/config.ini
Edit site_key
and secret_key
in waifu2x/web/config.ini
.
Run
python -m waifu2x.web --enable-recaptcha --config waifu2x/web/config.ini
--bind-addr 0.0.0.0
option will allow requests from all IP addresses.
python -m waifu2x.web --bind-addr 0.0.0.0 --port 8812
However, this method is not suitable for large scale access. It is suitable for small-scale public access, such as Local Area Network.
waifu2x.udp.jp
is deployed using nginx reverse proxy.
The advantage of this method is that static files can be served directly from public_html/
file system.
Examples of configuration files can be found in https://github.com/nagadomi/nunif/tree/master/waifu2x/web/appendix .
Also, in most cases, /recaptcha_state.json
can be a static file. You can reduce the load on your application server if you download /recaptcha_state.json
and place it in public_html/
.