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

[Bert] Feature: Custom Model Outputs #1

Closed
wants to merge 1 commit into from

Conversation

bkonkle
Copy link
Owner

@bkonkle bkonkle commented Apr 17, 2024

No description provided.

@bkonkle bkonkle force-pushed the feature/22-custom-bert-model-outputs branch 3 times, most recently from 75e57fc to 74930eb Compare April 17, 2024 05:14
@bkonkle bkonkle force-pushed the feature/22-custom-bert-model-outputs branch from a860d95 to 5d3c24a Compare April 25, 2024 21:57
@bkonkle bkonkle changed the base branch from chore/upgrade-bert to main April 25, 2024 21:57
@bkonkle bkonkle force-pushed the feature/22-custom-bert-model-outputs branch 2 times, most recently from 1c5977f to 06b188b Compare April 25, 2024 22:31
@bkonkle bkonkle changed the title [Bert] Custom Bert Model Outputs [Bert] Feature: Custom Model Outputs Apr 25, 2024
@bkonkle bkonkle marked this pull request as ready for review May 3, 2024 01:31
Copy link

@nathanielsimard nathanielsimard left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just a small change. I think it's a good idea to have all initialization related attributes in the model config, but thanks a lot for the contribution!

}

impl BertModelConfig {
/// Initializes a Bert model with default weights
pub fn init<B: Backend>(&self, device: &B::Device) -> BertModel<B> {
let embeddings = BertEmbeddingsConfig {
pub fn init<B: Backend>(&self, device: &B::Device, with_pooling_layer: bool) -> BertModel<B> {

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think I would put the pooling_layer in the config.

Copy link
Owner Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍 I went back and forth on it, and I ended up going in this direction because I reasoned that this was a Burn-specific modifier rather than an actual model config parameter. I can see how it's more obvious in the config, though, so that distinction doesn't really matter. 😄 I'll make the change, and stage the PR against the tracel-ai repo for final review. Thanks for taking a look!

@bkonkle bkonkle force-pushed the feature/22-custom-bert-model-outputs branch from 55315ba to 605f98b Compare May 7, 2024 00:11
@bkonkle
Copy link
Owner Author

bkonkle commented May 7, 2024

Closing this PR in favor of one targeting the upstream repo: tracel-ai#31

@bkonkle bkonkle closed this May 7, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants