A "Ho̍k tsè bûn" / "copypasta" / "複製文" bot for telegram
👉Start using the bot👈

• Post related copypasta for you whenever the bot detects matching keyword
• Generate summarization (or caption, for media) automaticly by utlizing top-of-the-line DL models
  • Model used for text summarization: csebuetnlp/mT5_multilingual_XLSum
  • Model used for image captioning: OFA-Sys/OFA-base

#ImageCaptioning
#TextGeneration
#TextSummarization

and more......

Supported commands atm:

1. new/add: insert new Ho̍k tsè bûn into the database
2. random: select a Ho̍k tsè bûn randomly and post it
3. search: fuzzy search every file in database
4. Automatically insert new media into the database whenever an media with caption is sent
5. delete: delete copypasta
6. example: show tutorial

1. Import the repo into replit
2. Setup your config file following section "Config Setup" and move it to /HokSeBunBot
3. Run

go mod init HokSeBunBot  
go mod tidy
bash run.sh

4. (Optional) In order to make the bot work in group chat, turn off privacy mode for you bot (by using BotFother).
5. Profit!

Environment Variable: API.TG.TOKEN
Description: API token for your telegram bot
Default value: "YOUR_TELEGRAM_API_TOKEN" (no this does not work)

Environment Variable: API.HT.TOKENs
Description: A list of huggingface tokens, bot switchs token whenever it fails (ex: quota exceeded)
Default value= ["YOUR_HUGGINGFACETOKEN1", "YOUR_HUGGINGFACETOKEN2",]

Setting: LOG_FILE
Description: Name of your log file
Default value: "../log.log"

Setting: SUM_MODEL
Description: The desired model to use for summarization, any model which supports summarization in your language should work
Default value = "csebuetnlp/mT5_multilingual_XLSum"
Note: The model should support inference api to work.

Setting: MT_MODEL
Description: The desired model to use for translation, any model which supports translation in your language should work
Default value = "Helsinki-NLP/opus-mt-en-zh"
Note: This setting isn't working since the bot uses google translate

Setting: DIR
Description: The location to store your clover database
Default Value: "../HokSeBun_db"

Setting: COLLECTION
Description: The collection name in your clover database
Default Value: "Copypasta"