This is playground and utils libraries for AI stuff.
It is not a framework, but a collection of useful tools and examples. It's not a pinned repo, not a featured and supported, and in general violates all the best practices of software development. But it's a good place to start. At the beginning and the end of all, this repo contains stuff that I mentioned in my articles, thus, refer to my blog posts first.
- examples - a collection of examples of AI projects, including:
- image_generation - see article first, this is example of image generation with LLM via API
- speak_and_hear - see article first, this is LLM speech recognition and TTS example
- test_generator - see article first, this is QA automatic tests generator
- llm_api_comparison - see article first, this is LLM API comparison for 2023
- utils - a collection of useful tools for AI development, in general them all of them used in example projects:
- article_extractor - limbo for article extraction from web pages
- audio_recorder - a simple audio recorder, used in speech recognition / TTS examples
- discord_interations - a simple discord interactions wrapper, used to fire self-bot commands
- discord_watcher - a simple discord watcher bot, used to watch for messages in discord channels and get content (urls)
- llm_timer_wrapper - a simple timer wrapper for LLM APIs, used for benchmarking of models
- logger_config - general logger
- other - all that doesn't fit in other files, i.e. env checkers
- page_retriever - web page retriever and parser
- transcriptors - custom transcriptors wrappers for speech recognition
- translators - custom translators for text translation wrappers
- tts - custom TTS engines wrappers
Once again, refer to my blog posts first. If you do so, you may try exec some top-level scripts from parent directory:
git clone [email protected]:wwakabobik/ai_engines.git # clone this repo
cd ai_engines # go to repo directory
# It's highly recommended to use python 3.11+ and venv
python -m venv venv # create virtual environment
source venv/bin/activate # activate virtual environment
pip install -r requirements.txt # install dependencies
# then you may run some top-level script, like:
PYTHONPATH=. python -m examples.test_generator.generator_test
PYTHONPATH=. python -m examples.speak_and_hear.test_gpt
Well, in most cases you need to create cred file to run examples, so, create creds.py
under examples
directory and fill it with your credentials, like:
# -*- coding: utf-8 -*-
"""Creds file"""
oai_token = "alksjdlksajlkdjlajiouoieuoqijnc"
oai_organization = "slkahkdjshakjhfkjafs"
cohere_token = "saklljlkdjsaldjljasldjlsak"
llama_token="asiuoiduaosudouasosuoduoqoihicdhzch"
ablt_token="sadhsakhdajskhaskdkja"
claude_token="dkshjjsdhkjdshkjhskj"
openweathermap_appid = "salkdjaslkjldasjlkdasl"
midjourney_cookie="skjaklshkldfhkjsahjhfkjbkfsa"
GL&HF!