Skip to content
View inikishev's full-sized avatar
3 followers · 9 following

Achievements

Achievement: QuickdrawAchievement: Pull Shark

Achievements

Achievement: QuickdrawAchievement: Pull Shark

Block or report inikishev

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
inikishev/README.md

HI

If you are into optimization, you might find pytorch image descent useful for debugging and quickly visualizing momentum or generally how an optimizer works. I've also implemented some gradient-free PyTorch optimizers in torchzero (pretty sure its the first library to do so!), which is useful when you need to optimize a function or a model without having access to gradients.

My first every project was a music beat swapping and automatic mixing/remixing tool with AI-based beat detection which you can try in your browser https://huggingface.co/spaces/dpe1/beat_manipulator.

Pinned Loading

  1. beat-manipulator beat-manipulator Public

    beat swapping powered by AI

    Python 8 2

  2. image-descent image-descent Public

    Pytorch library to test optimizers by visualizing how they descend on a your images. You can draw your own custom loss landscape and see what different optimizers do.

    Jupyter Notebook

  3. torchzero torchzero Public

    0th order optimizers, gradient chaining, random gradient approximation

    Python

  4. Saber2Osu Saber2Osu Public

    Converts BeatSaber beatmaps to osu! beatmaps . Opposite of https://github.com/tmokmss/Osu2Saber

    Python 3