This repository was archived by the owner on Nov 22, 2022. It is now read-only.
Beam Search + Top K Metrics #247
Closed
Conversation
This file contains 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
facebook-github-bot
added
the
CLA Signed
emilysw
force-pushed
the
export-D13219339
branch
2 times, most recently
from
f0045b3 to
52cc381
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 27, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Differential Revision: D13219339 fbshipit-source-id: ef9cbe257f7f3444cb72e4aa5719e6de33a747a3
emilysw
force-pushed
the
export-D13219339
branch
from
52cc381 to
bbf2b40
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 28, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Differential Revision: D13219339 fbshipit-source-id: 2c1fc51c06a764d746f31135fe8b76a9f976ca05
emilysw
force-pushed
the
export-D13219339
branch
from
bbf2b40 to
e506bed
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 29, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Differential Revision: D13219339 fbshipit-source-id: b2307ebe171658f482aa936e91e02ff8f04832a5
emilysw
force-pushed
the
export-D13219339
branch
from
e506bed to
fb8ed46
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 29, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Differential Revision: D13219339 fbshipit-source-id: 923a1550d72745a0d49a6bff0280458ebe3dec77
emilysw
force-pushed
the
export-D13219339
branch
from
fb8ed46 to
d96f122
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 29, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Differential Revision: D13219339 fbshipit-source-id: 5edf7ff1ad1777d1c7f0ff6a89058d2ee648f418
emilysw
force-pushed
the
export-D13219339
branch
from
d96f122 to
5b9db48
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 30, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Reviewed By: wowitsmrinal Differential Revision: D13219339 fbshipit-source-id: f8bddddb8ec9809506e3c97d3d4bd3f0e7615fbc
emilysw
force-pushed
the
export-D13219339
branch
from
5b9db48 to
79efb93
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 30, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Reviewed By: wowitsmrinal Differential Revision: D13219339 fbshipit-source-id: 0490b4358a2d89ae67ffb5ddc5ba199dd056de62
emilysw
force-pushed
the
export-D13219339
branch
from
79efb93 to
a122567
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 30, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Reviewed By: wowitsmrinal Differential Revision: D13219339 fbshipit-source-id: 8105283f0c6bf51305865ddf6b330a19a9d81e02
emilysw
force-pushed
the
export-D13219339
branch
from
a122567 to
87a0382
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 30, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Reviewed By: wowitsmrinal Differential Revision: D13219339 fbshipit-source-id: ba4d2e34230380f259ae84a568c278a43aa1f325
emilysw
force-pushed
the
export-D13219339
branch
from
87a0382 to
ca97ed2
Compare
emilysw
pushed a commit
to emilysw/pytext
that referenced
this pull request
Jan 30, 2019
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Reviewed By: wowitsmrinal Differential Revision: D13219339 fbshipit-source-id: 4c9574eb127bd3df1407be92080dbf00eb51fd22
Summary: Pull Request resolved: facebookresearch#247 Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors. Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions. With higher beam size, top K frame accuracy increases for RNNG, but latency also increases. Reviewed By: wowitsmrinal Differential Revision: D13219339 fbshipit-source-id: 4749d485ce1df2678aa5a69426f3fc0eb13d76d9
emilysw
force-pushed
the
export-D13219339
branch
from
ca97ed2 to
e22e9e0
Compare
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
Summary:
Implemented beam search (non-parallelized) for RNNG C++. Previously, the Parser's forward pass returned a single vector {predictedActionsIdx_, action_scores}. Now, the Parser's forward pass returns a vector of top k {predictedActionsIdx_, action_scores} vectors.
Also, implemented metrics for evaluating the top K frame accuracy. For each test example, top k frame accuracy is defined as at least 1 exact match between the expected frame and one of the top k frame predictions.
With higher beam size, top K frame accuracy increases for RNNG, but latency also increases.
Differential Revision: D13219339