Training neural models with structured signals.
APACHE-2.0 License
Bot releases are visible (Hide)
params
as an optional third argument to the embedding_fn
argument ofnsl.estimator.add_graph_regularization
. This is similar to the params
model_fn
, which allows users to pass arbitraryembedding_fn
will allowembedding_fn
.nsl.keras.AdversarialRegularization
andnsl.keras.GraphRegularization
now support the save
method which willnsl.keras.AdversarialRegularization
now supports a tf.keras.Sequential
tf.keras.layers.DenseFeatures
layer.nsl.configs.AdvNeighborConfig
has a new field random_init
. If set toTrue
, a random perturbation will be performed before FGSM/PGD steps.nsl.lib.gen_adv_neighbor
now has a new parameter use_while_loop
. If setTrue
, the PGD steps are done in a tf.while_loop
which is potentiallynsl.lib.random_in_norm_ball
for generating random tensors in a normnsl.lib.project_to_ball
for projecting tensors onto a norm ball.nsl.keras.AdversarialRegularization
and nsl.lib.gen_adv_neighbor
willdtype
. This doesn’t change the functionality, but only suppresses excessestimator/adversarial_regularization.py
andestimator/graph_regularization.py
explicitly import estimator
fromtensorflow
as a separate import instead of accessing it via tf.estimator
estimator
target.workshops
directory contains presentation materials fromusage.md
page describes featured usage of NSL, external talks,examples
directory:
examples/notebooks
directory:
This release contains contributions from many people at Google Research and from
TF community members: @angela-wang1 , @dipanjanS, @joshchang1112, @SamuelMarks,
@sayakpaul, @wangbingnan136, @zoeyz101
Published by aheydon-google about 4 years ago
None.
lsh_rounds
when lsh_splits
< 1. Bylsh_rounds
has been changedThis release contains contributions from many people at Google.
Published by aheydon-google about 4 years ago
nsl.configs.GraphBuilderConfig
class wasnsl.tools.build_graph_from_config
function.lsh_rounds
lsh_splits
.nsl.tools.add_edge
to return a boolean result indicating if a newnsl.tools.read_tsv_graph
that was incrementing the reportednsl.estimator.add_adversarial_regularization
andnsl.estimator.add_graph_regularization
so that the UPDATE_OPS
can bensl.estimator.add_graph_regularization
andnsl.estimator.add_adversarial_regularization
respectively.nsl.keras.GraphRegularization
and nsl.keras.AdversarialRegularization
This release contains contributions from many people at Google.
Published by arjung over 4 years ago
nsl.tools.build_graph(...)
to be more efficient and use far lessnsl.tools.build_graph(...)
nsl.lib.strip_neighbor_features
, a function to remove graphkeep_rank
from False
to True
as well asnsl.keras.layers.NeighborFeatures.call
andnsl.utils.unpack_neighbor_features
.clip_value_min
and clip_value_max
in nsl.configs.AdvNeighborConfig
.nsl.lib.adversarial_neighbor.gen_adv_neighbor
API.nsl.AdvNeighborConfig.feature_mask
field.nsl.tools.build_graph
andnsl.tools.pack_nbrs
utilities as binaries.nsl.lib.gen_adv_neighbor
.nsl.lib.maximize_within_unit_norm
.base_with_labels_in_features
tonsl.keras.AdversarialRegularization
for passing label features to the basensl.keras.AdversarialRegularization
This release contains contributions from many people at Google as well as
@mzahran001.
Published by arjung about 5 years ago
Introduces nsl.tools.build_graph
, a function for graph building.
Introduces nsl.tools.pack_nbrs
, a function to prepare input for
graph-based NSL.
Adds tf.estimator.Estimator
support for NSL. In particular, this release
introduces two new wrapper functions named
nsl.estimator.add_graph_regularization
and
nsl.estimator.add_adversarial_regularization
to wrap existing
tf.estimator.Estimator
-based models with NSL. These APIs are currently
supported only for TF 1.x.
Adds version information to the NSL package, which can be queried as
nsl.__version__
.
Fixes loss computation with Loss
objects in AdversarialRegularization
.
Adds a new parameter to nsl.keras.adversarial_loss
which can be used to
pass additional arguments to the model.
Fixes typos in documentation and notebooks.
Updates notebooks to use the release version of TF 2.0.
This release contains contributions from many people at Google.
Published by DualityGap about 5 years ago
Adds make_graph_reg_config
, a new API to help construct a nsl.configs.GraphRegConfig
object
Updates the package description on PyPI
Fixes metric computation with Metric
objects in AdversarialRegularization
Fixes typos in documentation and notebooks
This release contains contributions from many people at Google, as well as:
@joaogui1, @aspratyush.
Published by csferng about 5 years ago
Initial release of Neural Structured Learning.