easy-open-ai 0.2.0

This is my first REAL package.

Welcome to my OpenAI Playground!

Making Open AI even easier.

STILL ACTIVELY DEVELOPING

How to play

• pip install easy-open-ai

• in your working folder, create .env file with OPENAI_API_KEY=, or make such environment variable

• enjoy! from easy_open_ai import ...

  Version 0.1.6

  • get_picture(text, save=True) (as b64_json string)

  • get_n_pictures(text, n=3)

  • get_answer(question)

  • correct_grammar(text)

  Version 0.1.7

  • get_picture_url(text) (url lives 1 hour)

  • get_n_pictures_urls(text, n=3)

  • is_harmful_text(text) (sometimes ignores censorship escaping: '@' instead of 'a', etc)

  • how_many_text_tokens(text) (most of this package works with text no longer than 1024 text-tokens, it's recommended for validating the text length)

  • sum_up_as_haiku(text)

  • get_poem(text)

  • get_answer_as_poem(question)

  • autocomplete_text(text)

  • translate_text(text, language='Ukrainian')

  • get_answer_with_instruction(question,instruction, chaos_coefficient=0.5, model="gpt-3.5-turbo", max_tokens=1024)

  Version 0.1.8

  ADDED LIVE TEXT DISPLAYING

  • get_answer_with_instruction_stream(question) (it's a generator)

  to get chatGPT-like live text generation, add _stream to the name, for example, aget_answer_as_poem_stream(question)

  !not all of the available text functions have this yet

Version 0.1.9

• Automatic documentation for your function and FastAPI endpoints in a python file with docstring file_name.py terminal command (or docstring path\to\file_name.py ). If the docstring is already present, it will NOT overwrite it.

More easy functions coming soon!

For async version of a function, add a in front, for example, aget_picture(text) instead of get_picture(text).