This repository was archived by the owner on Mar 20, 2026. It is now read-only.
adding search.PrefixConstrainedBeamSearch#2646
Closed
nicola-decao wants to merge 5 commits intofacebookresearch:masterfrom
nicola-decao:add_PrefixConstrainedBeamSearch
Closed
adding search.PrefixConstrainedBeamSearch#2646nicola-decao wants to merge 5 commits intofacebookresearch:masterfrom nicola-decao:add_PrefixConstrainedBeamSearch
nicola-decao wants to merge 5 commits intofacebookresearch:masterfrom
nicola-decao:add_PrefixConstrainedBeamSearch
Conversation
Contributor
Author
|
The test failed on something that is not part of the pull request |
myleott
suggested changes
Sep 24, 2020
myleott
left a comment
myleott
left a comment
There was a problem hiding this comment.
You can ignore the test_translation_multi_simple_epoch test failure (the psutil import failure has been fixed in trunk).
But the test_ensemble_sequence_generator (tests.test_sequence_generator.TestJitSequeneceGenerator) failures seems related (see comment below)
myleott
reviewed
Sep 25, 2020
| if num_remaining_sent == 0: | ||
| break | ||
| if isinstance(self.search, search.PrefixConstrainedBeamSearch) and step >= max_len: | ||
| if self.search.stop_on_max_len and step >= max_len: |
Contributor
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@myleott has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
Contributor
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@myleott has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
Contributor
Author
I am not a Facebook employee so I cannot see the warnings and why this fails. |
I'm taking care of this :) |
5 tasks
Contributor
jinyiyang-jhu
pushed a commit
to jinyiyang-jhu/fairseq-jyang
that referenced
this pull request
Feb 26, 2021
Summary: # Before submitting - [ ] Was this discussed/approved via a Github issue? (no need for typos, doc improvements) - [x] Did you read the [contributor guideline](https://github.com/pytorch/fairseq/blob/master/CONTRIBUTING.md)? - [x] Did you make sure to update the docs? - [x] Did you write any new necessary tests? ## What does this PR do? This adds a new decoding strategy `search.PrefixConstrainedBeamSearch` that limits the vocabulary of the next token generation given a prefix (that is the previously generated tokens during beam search). An end user has just to give the optional argument `prefix_allowed_tokens_fn` to `.generate` or `.sample` to activate `PrefixConstrainedBeamSearch`. `prefix_allowed_tokens_fn(batch_id, tokens)` is a callback function that given the `batch_id` and `tokens` returns the list of allowed token for the next generation step. ## Did you have fun? YES! � Pull Request resolved: facebookresearch/fairseq#2646 Reviewed By: fabiopetroni Differential Revision: D24006805 Pulled By: myleott fbshipit-source-id: 40b1a866c6ea9f936272db27e2a020b18dbf8164
caltia
pushed a commit
to caltia/fairseq
that referenced
this pull request
Jul 8, 2025
Summary: # Before submitting - [ ] Was this discussed/approved via a Github issue? (no need for typos, doc improvements) - [x] Did you read the [contributor guideline](https://github.com/pytorch/fairseq/blob/master/CONTRIBUTING.md)? - [x] Did you make sure to update the docs? - [x] Did you write any new necessary tests? ## What does this PR do? This adds a new decoding strategy `search.PrefixConstrainedBeamSearch` that limits the vocabulary of the next token generation given a prefix (that is the previously generated tokens during beam search). An end user has just to give the optional argument `prefix_allowed_tokens_fn` to `.generate` or `.sample` to activate `PrefixConstrainedBeamSearch`. `prefix_allowed_tokens_fn(batch_id, tokens)` is a callback function that given the `batch_id` and `tokens` returns the list of allowed token for the next generation step. ## Did you have fun? YES! � Pull Request resolved: facebookresearch/fairseq#2646 Reviewed By: fabiopetroni Differential Revision: D24006805 Pulled By: myleott fbshipit-source-id: 40b1a866c6ea9f936272db27e2a020b18dbf8164
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Before submitting
What does this PR do?
This adds a new decoding strategy
search.PrefixConstrainedBeamSearchthat limits the vocabulary of the next token generation given a prefix (that is the previously generated tokens during beam search). An end user has just to give the optional argumentprefix_allowed_tokens_fnto.generateor.sampleto activatePrefixConstrainedBeamSearch.prefix_allowed_tokens_fn(batch_id, tokens)is a callback function that given thebatch_idandtokensreturns the list of allowed token for the next generation step.Did you have fun?
YES! 🙃