google-deepmind/clrs v1.0.0

CLRS 1.0.0

Repository: google-deepmind/clrs

Tag: v1.0.0

Published: 2022-06-01T16:07:23Z

Prerelease: no

Release notes:

Main changes

• Extended the benchmark from 21 to 30 tasks by adding the following:
• Activity selection (Gavril, 1972)
• Longest common subsequence
• Articulation points
• Bridges
• Kosaraju's strongly connected components algorithm (Aho et al., 1974)
• Kruskal's minimum spanning tree algorithm (Kruskal, 1956)
• Segment intersection
• Graham scan convex hull algorithm (Graham, 1972)
• Jarvis' march convex hull algorithm (Jarvis, 1973)
• Added new baseline processors:
• Deep Sets (Zaheer et al., NIPS 2017) and Pointer Graph Networks (Veličković et al., NeurIPS 2020) as particularisations of the existing Message-Passing Neural Network processor.
• End-to-End Memory Networks (Sukhbaatar et al., NIPS 2015)
• Graph Attention Networks v2 (Brody et al., ICLR 2022)

Detailed changes

• Add PyPI installation instructions. by @copybara-service in https://github.com/deepmind/clrs/pull/6
• Fix README typo. by @copybara-service in https://github.com/deepmind/clrs/pull/7
• Expose Sampler base class in public API. by @copybara-service in https://github.com/deepmind/clrs/pull/8
• Add dataset reader. by @copybara-service in https://github.com/deepmind/clrs/pull/12
• Patch imbalanced samplers for DFS-based algorithms. by @copybara-service in https://github.com/deepmind/clrs/pull/15
• Disk-based samplers for convex hull algorithms. by @copybara-service in https://github.com/deepmind/clrs/pull/16
• Avoid dividing by zero in F_1 score computaton. by @copybara-service in https://github.com/deepmind/clrs/pull/18
• Sparsify the graphs generated for Kruskal. by @copybara-service in https://github.com/deepmind/clrs/pull/20
• Option to add an lstm after the processor. by @copybara-service in https://github.com/deepmind/clrs/pull/19
• Include dataset class and creation using tensorflow_datasets format. by @copybara-service in https://github.com/deepmind/clrs/pull/23
• Change types of DataPoint and DataPoint members. by @copybara-service in https://github.com/deepmind/clrs/pull/22
• Remove unnecessary data loading procedures. by @copybara-service in https://github.com/deepmind/clrs/pull/24
• Modify example to run with the tf.data.Datasets dataset. by @copybara-service in https://github.com/deepmind/clrs/pull/25
• Expose processors in CLRS by @copybara-service in https://github.com/deepmind/clrs/pull/21
• Update CLRS-21 to CLRS-30. by @copybara-service in https://github.com/deepmind/clrs/pull/26
• Update README with new algorithms. by @copybara-service in https://github.com/deepmind/clrs/pull/27
• Add dropout to example. by @copybara-service in https://github.com/deepmind/clrs/pull/28
• Make example download dataset. by @copybara-service in https://github.com/deepmind/clrs/pull/30
• Force full dataset pipeline to be on the CPU. by @copybara-service in https://github.com/deepmind/clrs/pull/31
• Set default dropout to 0.0 for now. by @copybara-service in https://github.com/deepmind/clrs/pull/32
• Added support for GATv2 and masked GATs. by @copybara-service in https://github.com/deepmind/clrs/pull/33
• Pad memory in MemNets and disable embeddings. by @copybara-service in https://github.com/deepmind/clrs/pull/34
• baselines.py refactoring (2/N) by @copybara-service in https://github.com/deepmind/clrs/pull/36
• baselines.py refactoring (3/N). by @copybara-service in https://github.com/deepmind/clrs/pull/38
• Update readme. by @copybara-service in https://github.com/deepmind/clrs/pull/37
• Generate more samples in tasks where the number of signals is small. by @copybara-service in https://github.com/deepmind/clrs/pull/40
• Fix MemNet embeddings by @copybara-service in https://github.com/deepmind/clrs/pull/41
• Supporting multiple attention heads in GAT and GATv2. by @copybara-service in https://github.com/deepmind/clrs/pull/42
• Use GATv2 + add option to use different number of heads. by @copybara-service in https://github.com/deepmind/clrs/pull/43
• Fix GAT processors. by @copybara-service in https://github.com/deepmind/clrs/pull/44
• Fix samplers_test by @copybara-service in https://github.com/deepmind/clrs/pull/47
• Update requirements.txt by @copybara-service in https://github.com/deepmind/clrs/pull/45
• Bug in hint loss for CATEGORICAL type. The number of unmasked datapoints (jnp.sum(unmasked_data)) was computed over the whole time sequence instead of the pertinent time slice. by @copybara-service in https://github.com/deepmind/clrs/pull/53
• Use internal rng for batch selection. Makes batch sampling deterministic given seed. by @copybara-service in https://github.com/deepmind/clrs/pull/49
• baselines.py refactoring (6/N) by @copybara-service in https://github.com/deepmind/clrs/pull/52
• Time-chunked datasets. by @copybara-service in https://github.com/deepmind/clrs/pull/48
• Potential bug in edge diff decoding. by @copybara-service in https://github.com/deepmind/clrs/pull/54
• Losses for chunked data. by @copybara-service in https://github.com/deepmind/clrs/pull/55
• Changes to hint losses, mostly for decode_diffs=True. Before, only one of the terms of the MASK type loss was masked by gt_diff. Also, the loss was averaged over all time steps, including steps without diffs and therefore contributing 0 to the loss. Now we average only over the non-zero-diff steps. by @copybara-service in https://github.com/deepmind/clrs/pull/57
• Adapt baseline model to process multiple algorithms with a single processor. by @copybara-service in https://github.com/deepmind/clrs/pull/59
• Explicitly denote a hint learning mode, to delimit the tasks of interest to CLRS. by @copybara-service in https://github.com/deepmind/clrs/pull/60
• Give names to encoder and decoder params. This facilitates analysis, especially in multi-algorithm training. by @copybara-service in https://github.com/deepmind/clrs/pull/63
• Symmetrise the weights of sampled weighted undirected Erdos-Renyi graphs. by @copybara-service in https://github.com/deepmind/clrs/pull/62
• Fix dataset size for augmented validation + test sets. by @copybara-service in https://github.com/deepmind/clrs/pull/65
• Bug when hint mode is 'none': the multi-algorithm version needs something in the list diff decoders. by @copybara-service in https://github.com/deepmind/clrs/pull/66
• Change requirements to a fixed tensorflow datasets nightly build. by @copybara-service in https://github.com/deepmind/clrs/pull/68
• Patch KMP algorithm to...