This repository has been archived by the owner on Nov 22, 2022. It is now read-only.
pytext multi-label support (#729) #731
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
Summary: Pull Request resolved: facebookresearch#729 Add multi-label support to Pytext training workflow including * LabelListTensorizer to read label list * MultiLabelSoftMarginLoss with n_hot_encoding encoding to calculate the loss of multi-label task (need to take care of the padded -1 in n_hot_encoding) * MultiLabelOutputLayer with predictions of all potential labels for each example * LabelListPrediction(NamedTuple) for an example including * label_scores: List[float] * predicted_label: List[int] * expected_label: List[int] * MultiLabelClassificationMetricReporter * compute_multi_label_classification_metrics with both predicted and expected labels in lists * compute_multi_label_soft_metrics with both predicted and expected labels in lists * handle the label / label list inputs at the same time in channel.py * In input arguments, users are able to choose * LabelTensorizer / LabelListTensorizer * BinaryClassificationOutputLayer/ MultiLabelOutputLayer / MulticlassOutputLayer * loss including BinaryCrossEntropyLoss, MultiLabelSoftMarginLoss, etc. * associated with loss, user can choose ClassificationMetricReporter / MultiLabelClassificationMetricReporter * define@register_adapter(from_version=12) v11_to_v12 in config_adapter.py to make ClassificationMetricReporter expansible * keep RECALL_AT_PRECISION_THREHOLDS as it was, users can change the values to test their data set (for example, add 0.7 into the list of thresholds) It has been tested on single-label and multi-label examples for DocNN and BERT models Differential Revision: D15777482 fbshipit-source-id: 7eff3b27eff076d6c36a0feaef223573608b0d5d
haowu666
force-pushed
the
export-D15777482
branch
from
4e66c6d to
ba6ea88
Compare
|
This pull request has been merged in 1a44019. |
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:
Pull Request resolved: #729
Add multi-label support to Pytext training workflow including
LabelListTensorizer to read label list
MultiLabelSoftMarginLoss with n_hot_encoding encoding to calculate the loss of multi-label task
(need to take care of the padded -1 in n_hot_encoding)
MultiLabelOutputLayer with predictions of all potential labels for each example
LabelListPrediction(NamedTuple) for an example including
MultiLabelClassificationMetricReporter
handle the label / label list inputs at the same time in channel.py
In input arguments, users are able to choose
define@register_adapter(from_version=12) v11_to_v12 in config_adapter.py to make ClassificationMetricReporter expansible
keep RECALL_AT_PRECISION_THREHOLDS as it was, users can change the values to test their data set (for example, add 0.7 into the list of thresholds)
It has been tested on single-label and multi-label examples for DocNN and BERT models
Differential Revision: D15777482