Skip to content

Incorporate minibatch-training #11

Open
hsmaan wants to merge 1 commit intomainfrom
minibatch_Training
Open

Incorporate minibatch-training #11
hsmaan wants to merge 1 commit intomainfrom
minibatch_Training

Conversation

@hsmaan
Copy link
Copy Markdown
Member

@hsmaan hsmaan commented Jan 22, 2024

Currently, minibatch training is not incorporated because of limitations in the graph-base training setup.

Aim of PR is to:

  • Enable minibatch training in the model
  • Test stability of minibatch training, especially with current hyperparameter setup of the model
  • After stability ensured, incorporate explicit parallelization (DataParallel already available in BaseTrainer, but DDP not implemented) for multi-gpu training

@hsmaan hsmaan added the enhancement New feature or request label Jan 22, 2024
@hsmaan hsmaan requested a review from subercui January 22, 2024 21:14
@hsmaan hsmaan mentioned this pull request Jan 22, 2024
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@subercui subercui subercui approved these changes

Assignees

@subercui subercui

@hsmaan hsmaan

Labels

enhancement New feature or request

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@hsmaan @subercui