https://github.com/zdevito/single_controller

https://github.com/zdevito/single_controller

Last synced: about 1 month ago
JSON representation

• Host: GitHub
• URL: https://github.com/zdevito/single_controller
• Owner: zdevito
• Created: 2023-08-30T21:55:38.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-06-25T03:51:58.000Z (5 months ago)
• Last Synced: 2024-06-25T19:43:38.209Z (5 months ago)
• Language: Python
• Size: 1.98 MB
• Stars: 1
• Watchers: 4
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
Getting Started With Examples

-----------------------------

Install the package:

    pip install -e .

For supervisor

--------------

Run the example_train application, which simulates training.

There is some stuff commented out which can be enabled to inject failures.

    python example_train/launch.py

The other files in the folder are the different parts of the job (health_check, train, supervisor script).

For single_controller tensors

-----------------------------

Run the tests (kinda noisy at this point)

    python test/simple.py

Prepare data for nanoGPT example:

    cd nanoGPT

    python data/shakespeare_char/prepare.py

Run the nanoGPT example:

    cd nanoGPT

    python train.py config/train_shakespeare_char.py

See flags in `single_controller/config.py` for changing behaviors.