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

tune learning rate bahdanau-attention #2104

Merged
merged 1 commit into from
May 16, 2022

Conversation

AnirudhDagar
Copy link
Member

@AnirudhDagar AnirudhDagar commented Apr 18, 2022

Description of changes:
This PR should partially fix #2099 but the TF hyperparams/implementation of the section on seq2seq/transformers needs a follow up to fix. I tried different sets of hyperparams for TF but those didn't yield the expected results.

By submitting this pull request, I confirm that you can use, modify,
copy, and redistribute this contribution, under the terms of your
choice.

@AnirudhDagar
Copy link
Member Author

@d2l-bot please rebuild. It was failing earlier due to a broken notebook in master.

@AnirudhDagar AnirudhDagar force-pushed the hparam_tune branch 2 times, most recently from 5680c1f to f7bcd3b Compare April 21, 2022 13:54
@d2l-bot
Copy link
Member

d2l-bot commented Apr 21, 2022

Job d2l-en/PR-2104/5 is complete.
Check the results at http://preview.d2l.ai/d2l-en/PR-2104/

@astonzhang
Copy link
Member

Is this still WIP while transformer TF needs fix?

@d2l-bot
Copy link
Member

d2l-bot commented May 16, 2022

Job d2l-en/PR-2104/7 is complete.
Check the results at http://preview.d2l.ai/d2l-en/PR-2104/

@AnirudhDagar AnirudhDagar changed the title [MXNet][TF] Fix #2099: tune learning rate hyperparam tune learning rate bahdanau-attention May 16, 2022
@d2l-bot
Copy link
Member

d2l-bot commented May 16, 2022

Job d2l-en/PR-2104/8 is complete.
Check the results at http://preview.d2l.ai/d2l-en/PR-2104/

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.

Need to tune performance for MXNet & TensorFlow for seq2seq
3 participants