ML monitoring helper

Telegram bot made to simplify management of Machine Learning and Deep Learing routines. You may find it useful to aggregate information about learning processes from multiple remote machines into a single telegram channel. If so checkout this repo

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development, testing or usage purposes.

Prerequisites

• docker 18.06.0+ ✅
• docker-compose 1.22.0+ ✅

Installing

1. Clone this repository on your local machine:
   $ git clone https://github.com/snk4tr/ML-monitoring-helper.git
2. Follow a few simple steps to get access token. You may have telegram site blocked for some reason. If so please follow instructions provided here.
3. Create config.yaml file based on config-template.yaml. Fill it with you personal info and set of parameters you prefer.
4. Run the docker container, it will do all the work:
   $ docker-compose up

Commands

• /help - informs about available commands.
• /gpu - provides gpustat-like info about current work machine load.
• /ls - performs plain old ls command (lists files in current directory).
• /watch - enables and disables watching for learning on target machine.
• /stop - deactivates the bot (does not stop the bot app itself❗).

Built with

• python-telegram-bot - wrapper over Telegram Bot API.
• parallel-ssh - asynchronous parallel SSH client library.

Versioning

SemVer is used for versioning. For the versions available, see the tags on this repository.

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Authors

• Sergey Kastryulin - Initial work - snk4tr