Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add support for L4 optimizer #21

Open
lukostaz opened this issue Mar 7, 2019 · 1 comment
Open

Add support for L4 optimizer #21

lukostaz opened this issue Mar 7, 2019 · 1 comment
Labels
enhancement New feature or request model All discussions and activities about new adding new models into AmplIGra
Milestone

Comments

@lukostaz
Copy link
Contributor

lukostaz commented Mar 7, 2019

Description

Integrate L4 optimizer and allow its use as drop-in replacement for adam, adagrad as optimizer argument in EmbeddingModel constructor.

https://github.com/martius-lab/l4-optimizer
https://arxiv.org/abs/1802.05074v2

Run experiments on FB15k-237 on ComplEx to assess:

  • i) predictive power
  • ii) training speed
@lukostaz lukostaz added this to the 1.0 milestone Mar 7, 2019
@sumitpai
Copy link
Contributor

sumitpai commented Mar 13, 2019

Added support for L4 but committed on a different branch in commit 2d330fb . We are getting error related to issue 3 on their repo.

@sumitpai sumitpai modified the milestones: 1.0, 1.1 Mar 15, 2019
@sumitpai sumitpai added the enhancement New feature or request label Jun 10, 2019
@lukostaz lukostaz added the model All discussions and activities about new adding new models into AmplIGra label Jun 12, 2019
@lukostaz lukostaz removed this from the 1.1 milestone Jun 12, 2019
@sumitpai sumitpai removed their assignment Sep 13, 2019
@lukostaz lukostaz added this to the 2.1 milestone Feb 14, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request model All discussions and activities about new adding new models into AmplIGra
Projects
None yet
Development

No branches or pull requests

2 participants