v1.0.0.dev1

Early developmental release of tensorflow-gnn 1.0.0 code; docs still outdated.

v0.6.0

What's Changed

• Typo fix is_writen -> is_written by @copybara-service in #304
• For examples/notebooks, install the latest release (not release candidate). by @copybara-service in #306
• Add OGBN-MAG colab links by @copybara-service in #305
• Partially re-enable some tests for TF2.9. by @copybara-service in #316
• Expose integrated gradient function to enable building custom exporter with it. by @copybara-service in #319
• Add causal masking when training with labels. by @copybara-service in #321
• Fixes double-install of tf-nightly and tensorflow in nightly test script. by @copybara-service in #323
• GCN model: Different normalizations of the adjacency matrix is added. The value of 'degree_normalization' controls how the edges are normalized. The default value of 'degree_normalization' is set to 'in_out'. Note that the default value gives a different output for asymmetric graphs compared to the previous default value. To reinstate the old behaviour of the code for asymmetric graphs, set 'degree_normalization' to 'in_in'. The 'normalize' flag is removed. To make sure the edges are not normalized, set the 'degree_normalization' to 'none'. by @copybara-service in #326
• For eventual multi-task support: by @copybara-service in #335
• This change is a no-op by @copybara-service in #262
• Raise exception if a key has been provided, but no dictionary is available for ResidualNextState. by @copybara-service in #353
• Hooks for HGT hyperparameter tuning by @copybara-service in #356
• Update TF-GNN models to best practices: by @copybara-service in #358
• Remove the unnecessary epsilon parameter. by @copybara-service in #363
• Call validate_graph_tensor_spec_for_readout() from validate_schema(). by @copybara-service in #364
• Remove the function random_graph_tensor_from_schema(). by @copybara-service in #367
• Fix typo: s/contined/continued/ by @copybara-service in #379
• Merge DGI layers with others. by @copybara-service in #385
• Add code for upcoming generalized pool() and broadcast() operations. by @copybara-service in #384
• Add more granular shuffling. by @copybara-service in #387
• Unit-test broadcast_node_to_edges() for Adjacency and HyperAdjacency. by @copybara-service in #394
• Expose RunResult. by @copybara-service in #395
• Merge DGI layers with others. by @copybara-service in #402
• Updates TF version for newest stable tests to 2.12 by @copybara-service in #410
• Allow passing collection of receiver tags to ConvGNNBuilder by @copybara-service in #412
• Fix the behavior of shuffle_tensor for 0 shapes. This happens with make_empty_feature and shuffling rate not in {0, 1}. by @copybara-service in #416
• Correctly handle auxiliary graph pieces in convert_to_line_graph by @copybara-service in #422
• TFLite tests for vanilla_mpnn and mt_albis by @copybara-service in #425
• Allows specialization how composite tensors are converted for the bulk executor. by @copybara-service in #432
• Silence some pytype errors. by @copybara-service in #439
• Update attribution.py testing: the preprocessing model created by orchestration returns a Sequence[GraphTensor]. by @copybara-service in #442
• Add Keras wrappers ReadoutNamedIntoFeature and AddReadoutFromFirstNode by @copybara-service in #452
• Adjust maxval += 1 in the case of an integer random counterfactual. by @copybara-service in #457
• Switch tfgnn.pool() to implementation v2 and remove v1. by @copybara-service in #464
• Enable test case: subgraph sampling model, saving and loading, as tf.keras.Model. by @copybara-service in #468
• Utils to read unigraph-formatted data into sampler v2 input format and a driver program used like graph_sampler.py by @copybara-service in #469
• Align context indices dtype resolution between GraphTensor and GraphTensorSpec by @copybara-service in #477
• Redefines indices_dtype as two concepts: by @copybara-service in #486
• Add convenience function. by @copybara-service in #495
• Fix type annotation: SimpleConv(..., receiver_feature=None) is supported by @copybara-service in #506
• Fixes type annotation syntax in runner link prediction by @copybara-service in #511
• Disable test that fails for unrelated reasons. by @copybara-service in #518
• Sampler pieces are removed from tfgnn core package. by @copybara-service in #519
• Re-enables edge sampling tests. by @copybara-service in #529
• Implement the Coherence metric. by @copybara-service in #533
• Fix for inconsistent ragged row partitions of edges and indices to select unique edges. by @copybara-service in #535

Full Changelog: v0.5.1...v0.6.0

v0.6.0rc1

Full Changelog: v0.6.0rc0...v0.6.0rc1

v0.6.0rc0

What's Changed

• Typo fix is_writen -> is_written by @copybara-service in #304
• For examples/notebooks, install the latest release (not release candidate). by @copybara-service in #306
• Add OGBN-MAG colab links by @copybara-service in #305
• Partially re-enable some tests for TF2.9. by @copybara-service in #316
• Expose integrated gradient function to enable building custom exporter with it. by @copybara-service in #319
• Add causal masking when training with labels. by @copybara-service in #321
• Fixes double-install of tf-nightly and tensorflow in nightly test script. by @copybara-service in #323
• GCN model: Different normalizations of the adjacency matrix is added. The value of 'degree_normalization' controls how the edges are normalized. The default value of 'degree_normalization' is set to 'in_out'. Note that the default value gives a different output for asymmetric graphs compared to the previous default value. To reinstate the old behaviour of the code for asymmetric graphs, set 'degree_normalization' to 'in_in'. The 'normalize' flag is removed. To make sure the edges are not normalized, set the 'degree_normalization' to 'none'. by @copybara-service in #326
• For eventual multi-task support: by @copybara-service in #335
• This change is a no-op by @copybara-service in #262
• Raise exception if a key has been provided, but no dictionary is available for ResidualNextState. by @copybara-service in #353
• Hooks for HGT hyperparameter tuning by @copybara-service in #356
• Update TF-GNN models to best practices: by @copybara-service in #358
• Remove the unnecessary epsilon parameter. by @copybara-service in #363
• Call validate_graph_tensor_spec_for_readout() from validate_schema(). by @copybara-service in #364
• Remove the function random_graph_tensor_from_schema(). by @copybara-service in #367
• Fix typo: s/contined/continued/ by @copybara-service in #379
• Merge DGI layers with others. by @copybara-service in #385
• Add code for upcoming generalized pool() and broadcast() operations. by @copybara-service in #384
• Add more granular shuffling. by @copybara-service in #387
• Unit-test broadcast_node_to_edges() for Adjacency and HyperAdjacency. by @copybara-service in #394
• Expose RunResult. by @copybara-service in #395
• Merge DGI layers with others. by @copybara-service in #402
• Updates TF version for newest stable tests to 2.12 by @copybara-service in #410
• Allow passing collection of receiver tags to ConvGNNBuilder by @copybara-service in #412
• Fix the behavior of shuffle_tensor for 0 shapes. This happens with make_empty_feature and shuffling rate not in {0, 1}. by @copybara-service in #416
• Correctly handle auxiliary graph pieces in convert_to_line_graph by @copybara-service in #422
• TFLite tests for vanilla_mpnn and mt_albis by @copybara-service in #425
• Allows specialization how composite tensors are converted for the bulk executor. by @copybara-service in #432
• Silence some pytype errors. by @copybara-service in #439
• Update attribution.py testing: the preprocessing model created by orchestration returns a Sequence[GraphTensor]. by @copybara-service in #442
• Add Keras wrappers ReadoutNamedIntoFeature and AddReadoutFromFirstNode by @copybara-service in #452
• Adjust maxval += 1 in the case of an integer random counterfactual. by @copybara-service in #457
• Switch tfgnn.pool() to implementation v2 and remove v1. by @copybara-service in #464
• Enable test case: subgraph sampling model, saving and loading, as tf.keras.Model. by @copybara-service in #468
• Utils to read unigraph-formatted data into sampler v2 input format and a driver program used like graph_sampler.py by @copybara-service in #469
• Align context indices dtype resolution between GraphTensor and GraphTensorSpec by @copybara-service in #477
• Redefines indices_dtype as two concepts: by @copybara-service in #486
• Add convenience function. by @copybara-service in #495
• Fix type annotation: SimpleConv(..., receiver_feature=None) is supported by @copybara-service in #506
• Fixes type annotation syntax in runner link prediction by @copybara-service in #511
• Disable test that fails for unrelated reasons. by @copybara-service in #518
• Sampler pieces are removed from tfgnn core package. by @copybara-service in #519
• Re-enables edge sampling tests. by @copybara-service in #529
• Implement the Coherence metric. by @copybara-service in #533
• Fix for inconsistent ragged row partitions of edges and indices to select unique edges. by @copybara-service in #535

Full Changelog: v0.5.1...v0.6.0rc0

v0.5.1

Full Changelog: v0.5.0...v0.5.1

v0.5.0

Full Changelog: v0.5.0rc0...v0.5.0

v0.5.0rc0

What's Changed

• Code Tutorial #1: Running TF-GNN on single machine. by @copybara-service in #215
• Minor tutorial site changes. by @copybara-service in #218
• Add link to slides by @copybara-service in #220
• Add GPT-GNN folder for the new autoregressive edge-prediction Task. by @copybara-service in #232
• Add compatibility tests for oldest supported Python and TensorFlow version (in this case Python 3.7 and TensorFlow 2.8) by @copybara-service in #234
• Add MultiHeadAttention model option to xm_launcher by @copybara-service in #235
• Satisfy pytype by having kwarg Convolve(extra_receiver_ops=...) where unused. by @copybara-service in #237
• Fix embedded paths to point to LoG location instead. by @copybara-service in #240
• Small colab fixes. by @copybara-service in #242
• Fix link and copy as_dataset implementation that is not yet in pip package. by @copybara-service in #246
• Use Keras layers (vs. tensor functions) for pooling. by @copybara-service in #247
• Move interfaces from orchestration.py to interfaces.py. by @copybara-service in #248
• Exhaustively test Tasks cross distribution strategy. by @copybara-service in #250
• Add save_options to ModelExporters by @copybara-service in #254
• fixes tag filters in test script by @copybara-service in #256
• Fix overriding return type. by @copybara-service in #265
• Apply DGI Task random seed for feature shuffling operations. Also add the missing tests for distributed strategy to ensure model correctness is deterministically validated. by @copybara-service in #182
• Expose check_compatible_with_schema_pb to the public interface. by @copybara-service in #271
• Enabling GCNConv layer to work with weighted adjacency. by @copybara-service in #274
• Add a tensorflow_gnn/models/constrastive_losses/... with a Deep Graph Infomax implementation. by @copybara-service in #285
• Negative sampling operation that returns a tensor of unconnected nodes for each node-id in the adjacency list. by @copybara-service in #158

Full Changelog: v0.4.0...v0.5.0rc0

v0.4.1

Full Changelog: v0.4.0...v0.4.1

v0.4.0

Full Changelog: v0.4.0rc2...v0.4.0

v0.4.0rc2

Bugfix for python 3.7 compatibility

Full Changelog: v0.4.0rc1...v0.4.0rc2