@lucasberg/ai-translator v1.0.2

AI-Translator

This is a console-based application that uses AI to translate phrases into different languages.

Running the application though npx:

• Run npx @lucasberg/ai-translator and provide some of the following arguments as comma-separated lists in quotation marks:
• -l for languages
• -p for phrases
• -s for starting language
• -c for context
• -j to output a json at the end

• -r to regenerate the translation and not take it from a database

• The translation automatically saves to a local json file

To import this as a JavaScript project:

• Add import { translator } from ‘@lucasberg/ai-translator' to the top of the file
• The two avalible functions are:
  1. translate
  2. saveTranslation

Examples of proper use:

translate("Hello, how are you doing?", "Spanish", "Make it formal", "¿Hola cómo estás?", "English") saveTranslation("Hello, how are you doing?", "English", "¿Hola, qué tal?", "Spanish")

Running the application locally:

1. Set up the AI

• Set the AUTH_TOKEN enviornment varible to your key
• Change the endpoint if required, right now it is set up for an Open AI Key from https://platform.openai.com/settings/organization/api-keys (you may need to add credtis to the token, as creating a key from this website does not give any free credits)
• Change the model if wanted, right now it is set to the cheapest available

2. Install files

• Run npm install into the terminal

3. Run program

• Run the program with node index.mjs, add the phrases in -p, and add the output languages in -l (put both as a comma-seperated list in quotation marks)

4. Follow the prompts

• Regenerate all or one language as many times as you want
• To keep the translations, entre none or just press entre to see the final answers

Additional features:

• Ability to add context with -c
• Ability to change the starting language with -s, default is English
• Ability to output a JSON with -j
• Ability to store context independently for each language, ensuring that adding context to one translation does not affect the others when regenerating all languages
• Ability to remember all incorrect answers separately for each language
• Ability to store to a database, currently using a JSON file
• Contains unit tests and error handling