An application posting generated text from posts on Mastodon using the Markov chain.
$ docker compose build app
$ docker compose run --rm app /app/bot run
Run docker compose run --rm app /app/bot run --help to view help.
You can also pass arguments as environment variables.
This application requires a configuration file to run. The format of the configuration file is as follows:
input:
platform: "mastodon"
origin: ""
access_token: ""
output:
platform: "mastodon"
origin: ""
access_token: ""
post_visibility: "unlisted"
fetch_status_count: 100
state_size: 3
min_words_count: 3See config/bot_config.go for details.
- Create ECR repository to upload container image and make note of the repository url.
- Enter to
cmd/lambda. - Run
sam buildto build container image. - Run
sam deploy --guidedand pass the repository url to create new lambda function.
To use the function, follow steps below:
- Set event to specify which S3 bucket and key path will be used to place configuration and model file.
- See
PostEventstruct incmd/lambda/main.go.
- See
- Put configration file on S3.
- See
ConfigFilestruct inconfig/bot_config.go.
- See