Releases: deepchem/deepchem
2.8.0 Initial Release
Initial release of 2.8.0. Human-written release notes to be added soon. We will be doing stability checks over the next few weeks with bugfixes going in 2.8.1
What's Changed
- Bug Fix #3086 by @advikavs in #3096
- Update by @BalajiAI in #3110
- Molnet added example and w matrix explanation by @tonydavis629 in #3143
- flake8 configuration update by @arunppsg in #3136
- Fixing issue 3090 by @advikavs in #3097
- Updating RDKitDescriptors featurizer to support custom descriptors by @arunppsg in #3135
- Update hyperlink for "datasets already available in MolNet" in docs by @ARY2260 in #3152
- Refactoring DMPNN: removing _MapperDMPNN class by @arunppsg in #3158
- Changing indent width to 4 - setup.cfg by @advikavs in #3160
- Merge dataset when the dimension of y is 1 by @advikavs in #3163
- PEP8 for layers by @tonydavis629 in #3167
- Yapf fixes on pytorch_datasets.py by @advikavs in #3165
- yapf on datasets.py by @arunppsg in #3166
- yapf fixes deepchem/data/data_loader.py by @advikavs in #3171
- Feedforward by @tonydavis629 in #3164
- yapf fixes to data/tests by @advikavs in #3173
- yapf fixes on torch_model.py by @arunppsg in #3172
- yapf fixes to trans/tests directory by @maithili232 in #3178
- yapf fixes to featurizers by @advikavs in #3177
- type fixes for pep 484 compliance by @arunppsg in #3181
- yapf fixes deepchem/models by @advikavs in #3186
- Adding grover featurizer by @arunppsg in #3138
- yapf on deepchem/hyper by @maithili232 in #3188
- yapf fixes deepchem keras models by @advikavs in #3192
- adding dqc to deepchem by @advikavs in #3196
- yapf fixes to deepchem featurizers by @advikavs in #3198
- yapf fixes on deepchem/metalearning by @maithili232 in #3197
- yapf on deepchem/utils part one by @maithili232 in #3203
- yapf on deepchem/utils/test by @maithili232 in #3202
- DeepChem package version update in init.py by @arunppsg in #3194
- yapf on deepchem/metrics by @maithili232 in #3199
- Yapf fixes on molnet directory by @maithili232 in #3169
- yapf fixes on deepchem/utils part three by @maithili232 in #3205
- changed indentation width to 4 spaces in deepchem/models/layers.py by @brianpzaide in #3210
- yapf fixes on deepchem/trans directory by @maithili232 in #3170
- Changing indent width to 4 by @GreatRSingh in #3207
- Yapf fixes on test directories by @arunppsg in #3212
- yapf on deepchem/utils part two by @maithili232 in #3204
- DQC Pytest by @advikavs in #3220
- Changing indent width to 4 spaces on deepchem/models dir by @arunppsg in #3213
- Yapf fixes on deepchem.feat module by @arunppsg in #3211
- Changing indent to 4 spaces on rl, split, molnet, dock dir by @arunppsg in #3216
- DFT initial pr - Adding utilities by @advikavs in #3190
- fixing dqc pytest by @advikavs in #3232
- Fix flake8 errors by @brianpzaide in #3235
- Fixing CI by @arunppsg in #3244
- Adding attention layers by @arunppsg in #3183
- update quickstart installation procedure by @vsaravind01 in #3249
- Adding Usage Examples to Splitters by @GreatRSingh in #3241
- ModularTorchModel by @tonydavis629 in #3242
- Adding grover layers by @arunppsg in #3179
- Refactor MLP by @tonydavis629 in #3257
- Infograph by @tonydavis629 in #3254
- added base vocabulary builder by @arunppsg in #3265
- Adding readout layers for use in grover by @arunppsg in #3269
- Fix: flake8 link by @gauthamk02 in #3276
- Adding yapf ignore by @advikavs in #3279
- Adding hugging-face tokenizer by @arunppsg in #3270
- Adding more grover layers by @arunppsg in #3277
- ci fix by @advikavs in #3286
- Adding DFT data classes by @advikavs in #3284
- Graphdata.numpy_to_torch() by @tonydavis629 in #3283
- Grover vocabulary builder by @arunppsg in #3281
- Batchgraph edge_index bug fix by @tonydavis629 in #3291
- New Code formatting CI by @arunppsg in #3278
- Global and Local Mutual Information Loss by @tonydavis629 in #3292
- add to device for numpy to torch by @tonydavis629 in #3299
- Snap featurizer by @tonydavis629 in #3266
- pin torch-sparse to 0.6.16 by @tonydavis629 in #3303
- align torch device in MI Loss by @tonydavis629 in #3301
- removed embedding output type arg in grover by @arunppsg in #3300
- Surpress assignment errors in graph_data by @tonydavis629 in #3304
- Rename Infograph to InfoGraphStar, add documentation and fixes by @tonydavis629 in #3282
- removing scipy pin by @arunppsg in #3309
- PR on nnlda layer by @advikavs in #3237
- Batching kwargs in batch graph by @arunppsg in #3294
- Infograph and InfographStar by @tonydavis629 in #3280
- Wrapper to use hugging face algorithms for building vocabulary by @arunppsg in #3271
- Revert pull request 3271 - wrapper to use hugging face algorithms for vocabulary building by @arunppsg in #3314
- Docs fix by @tonydavis629 in #3313
- DFTYaml Loader by @advikavs in #3295
- Added grover loss functions by @arunppsg in #3297
- Docs fix remove torch sparse and scatter by @tonydavis629 in #3317
- Modular loading from pretrained by @tonydavis629 in #3305
- removed torch-scatter and torch-sparse from dependency by @arunppsg in #3319
- extracting attributes for grover model from batch graph by @arunppsg in #3312
- InfographStar Multitask Classification and Regression by @tonydavis629 in #3318
- Cleaning docstring for grover layers by @arunppsg in #3333
- Removing duplicate grover test by @arunppsg in #3329
- SCF iterations by @advikavs in #3320
- adding grover pretrain module by @arunppsg in #3334
- adding grover finetune model by @arunppsg in #3335
- Infograph test fix by @tonydavis629 in #3340
- fixing dqc docs by @advikavs in #3332
- update docstring of IRVLayer by @shoaib6174 in #3324
- Descriptive names for ci runs by @arunppsg in #3336
- adding grover pretrain model as ModularTorchModel by @arunppsg in #3272
- remnants of #3336 by @arunppsg in #3348
- GNNModular by @tonydavis629 in #3339
- Wrapper to use hugging face algorith...
2.8.0.pre
2.7.1
DeepChem 2.7.0
Highlights
- DeepChem adds support for new models including DMPNNs, and MEGNet
- We have ported NormalizingFlows to PyTorch
- Added support for multi-gpu training via pytorch lightning.
- Utilities to run hhsearch multisequence alignment search on a dataset
- We have ported several layers to pytorch
Porting Models to PyTorch
The following models/layers have been ported to pytorch: GRU, InterAtomicL2Distance, WeightedLinearCombo, CombineMeanStd, AtomicConvolution layer, NeighborList, CNN, LSTMStep
New Features
- Fake graph data generator to generate random graphs
- FASTQ Loader to load biological sequences of data
- Added top_k_accuracy_score metric for evaluating model performances
- Extracting molecular coordinates from QM9 dataset
- Support for Random hyperparameter tuning
Featurizers
- DMPNN Featurizer
- Sparse matrix one hot featurizer
- Position Frequency Matrix Featurizer implements a featurizer for position frequency matrices on a list of multisequence alignments to return a list of position frequency matrices.
New Layers
- MEGNet Layer
Deprecations
- dc.evaluate.utils.relative_difference is being deprecated. A deprecation warning to use
math.isclose
,np.isclose
,np.allclose
has been put in place.
Examples and Tutorials
- Using hydra config system with pytorch-lightning system
- New tutorial have been added to DeepQMC, SCVI and ScanPy, HierVAE, molGAN, hyper-parameter optimization, neural ODE, gaussian process, pytorch lightning, training a normalising flow on qm9 model, grover.
Documentation
- Documentation has been improved with wider examples, using deepchem with docker, model cheat sheets.
- Citations have been added to some of the tutorials to make them citable.
Improvements
- Speed up in atomic convolution model
- Utilities in deepchem disk dataset to convert it to a csv file.
- Added file storage of validation and train scores during hyperparameter optimization.
- Modified GraphData to support kwargs for storing additional attributes
- Made it possible to run DeepChem in offline mode by removing default download call from CGCNN
Refactors
- Mol2vec_fingerprints to directly use method from gensim library rather than mol2vec sub-package.
Bug Fixes
- Retrieving shape of disk dataset when task names are not specified
- Improvements in k-fold split when the number of data points is not exactly divisible by k
- Fix a bug in SmilesToSeq featurizer when the padding length is 0.
- A bug in which LogTransformer fails on data without an explicit task dimension has been fixed.
Maintenance
- Adding type hints.
- CI pipeline to consume less time
What's Changed
- Bump to dev version by @rbharath in #2825
- add top_k_accuracy metric by @tonydavis629 in #2818
- Update models.rst by @neerajanand321 in #2828
- adding Docker tutorial by @shaipranesh2 in #2814
- Added Graph Networks by @arunppsg in #2843
- added molGAN tutorial by @saithat in #2773
- adding hyperopt tutorials by @shaipranesh2 in #2851
- Improvements to GraphData by @arunppsg in #2860
- Update to Documentation for Using DeepChem in Jupyter Notebook by @iherath in #2856
- Removing some obsolete code by @arunppsg in #2855
- Removing package pins by @arunppsg in #2783
- Fake graph data generator by @arunppsg in #2865
- removing package pin by @arunppsg in #2873
- adding jax dependencies by @arunppsg in #2877
- Sparse matrix one hot featurizer by @davidRFB in #2870
- added neural ode tutorial 📚 by @shivance in #2859
- Fixing Colab Links by @rbharath in #2883
- Adding batch processing to GraphNet layer by @arunppsg in #2874
- MEGNet layer implementation by @arunppsg in #2837
- Fixing broken CI on windows - Jax and Vina by @arunppsg in #2886
- Fresh gp tutorial by @TheRealSalmon in #2864
- Fix to log transformer by @rbharath in #2887
- Best score callback by @TheRealSalmon in #2866
- fix for bug(issue #2106) by @shaipranesh2 in #2857
- deepchem pytorch lightning tutorial by @Chahalprincy in #2826
- sequence_utils for sequence homology search by @tonydavis629 in #2890
- Speed up AtomicConv model, improvements to AtomicConv tutorial by @juliusgeo in #2888
- Update Training_a_Normalizing_Flow_on_QM9.ipynb by @JoseAntonioSiguenza in #2885
- Resolve the bug issue of loading the .sdf files by @JoseAntonioSiguenza in #2795
- Position Frequency Matrix Featurizer by @tonydavis629 in #2896
- Adding a tutorial for GROVER. by @atreyamaj in #2901
- Update tutorials.rst by @BalajiAI in #2902
- Adding FutureWarning to depreciate deepchem.utils.evaluate.relative_difference by @arunppsg in #2909
- Module dl dependancies by @Nozziel in #2908
- Extracting molecular coordinates for QM9 dataset from sdf files by @arunppsg in #2903
- Add bibtex citation to first tutorial by @paupaiz in #2912
- Update torchvision version requirement by @Matthew-Hostetler in #2916
- add citation tutorial 4 by @paupaiz in #2921
- add citation tutorial 5 by @paupaiz in #2922
- Fixing CI errors by @arunppsg in #2931
- Update scientists.rst by @shivance in #2932
- resolved deprecation warning by @ARY2260 in #2937
- atom features function and helper functions for DMPNN Featurizer by @ARY2260 in #2929
- NormalizingFlow class and Affine transformation created using Pytorch by @JoseAntonioSiguenza in #2918
- add bond_features and reaction mapping with suitable tests for DMPNN by @ARY2260 in #2942
- Adding DeepQMC tutorial by @shaipranesh2 in #2914
- Intro8 bibtex by @paupaiz in #2925
- add bibtex tutorial 7 by @paupaiz in #2924
- Update model cheatsheet by @j-frie in #2947
- first steps in fixing docker build by @Nozziel in #2949
- Added torch equivalent of InterAtomicL2Distances in torch_layers.py + YAPF changes by @atreyamaj in #2934
- added HierVAE tutorial by @saithat in #2904
- modify molecular featurizer base class and suitable tests by @ARY2260 in #2960
- added _MapperDMPNN class and suitable tests by @ARY2260 in #2962
- implement DCLightningModule by @Chahalprincy in #2945
- Normalizing Flow Torch Model by @JoseAntonioSiguenza in #2944
- Adding Metropolis Hasting sampler by @shaipranesh2 in #2935
- Minor fix for utils.rst by @shaipranesh2 in #2973
- add global feature generator and suitable unit tests by @ARY2260 in #2971
- Updated Tutorial File with TorchModel example by @allesrebel in #2977
- Fixed potential bug in deepchem's CNN implementation by @shivance in #2964
- fix bug in GraphData class and add suitable unit test by @ARY2260 in #2979
- Gcn by @Chahalprincy in #2958
- python v3.8 in readthedocs.yml by @arunppsg in #2984
- Pinning sphinx version in docs/requirements.txt by @arunppsg in #2985
- Fixing #2986 by @shivance in #2988
- PEP-008 style corrections for electron sampler by @shaipranesh2 in #2987
- add count-based morgan fingerprint featurizer and suitable unit tests by @ARY2260 in #2980
- removed css theme by @arunppsg in #2991
- modify RDKitDescriptors class for normalized features by @ARY2260 in #2983
- Add Trident Chemwidgets tutorial to `examples/tutorials by @ts...
Minor version bump for numpy
This release is a minor version bump that increases the required version of numpy.
What's Changed
- fix gpu installation by @Chahalprincy in #2806
- Release 2.6.0 by @arunppsg in #2812
- Dependecy version fixes by @arunppsg in #2815
- tighten test_regression_overfit bound by @austereantelope in #2796
- Update uv_datasets.py by @neerajanand-coder in #2817
- Minor Release 2.6.1 by @arunppsg in #2821
New Contributors
- @Chahalprincy made their first contribution in #2806
- @austereantelope made their first contribution in #2796
- @neerajanand-coder made their first contribution in #2817
Full Changelog: 2.6.0...2.6.1
DeepChem 2.6.0
DeepChem 2.6.0 adds a range of new features (detailed below) along with significant improvements to the robustness of our testing infrastructure.
What's Changed
- Version bump to 2.6.0.dev by @rbharath in #2447
- Revised first tutorial by @Suzukazole in #2435
- Add save reload to atomicconv by @ncfrey in #2450
- Fix import issue in 22:Chemberta tutorial by @seyonechithrananda in #2445
- Fixed error in computing Pearson correlation coefficient by @peastman in #2463
- [WIP] Generalize OneHotFeaturizer to Support Arbitrary Strings by @alat-rights in #2458
- Run all tests even if some fail by @peastman in #2461
- Adds zip handling to SDFLoader by @NinadBhat in #2446
- Use Log Transform for Clearance and HPPB by @mufeili in #2462
- Improve SmilesToImage Error Message for long molecules by @PascalIversen in #2442
- Added updating website to release instructions by @alat-rights in #2469
- Fix pymatgen subclasses import error by @JainSamyak8840 in #2460
- Update issues.rst by @alat-rights in #2471
- Fixed Documentation for Graph Convolution Featurizers - MolGanFeaturizer by @atreyamaj in #2473
- Add img_size parameter to ChemCeption by @PascalIversen in #2466
- Fixing the failing test case - test_mol2vec_fingerprint by @VIGNESHinZONE in #2486
- MEGNET model and evaluate the design by @VIGNESHinZONE in #2485
- Update infra.rst by @alat-rights in #2491
- Update release.md to reflect migration of deepchem.io repo by @alat-rights in #2492
- Fix String Comparison to use POSIX standard #2481 by @peterskipper in #2482
- Adding essential Molecular Utils by @VIGNESHinZONE in #2484
- Basic MolGAN model by @MiloszGrabski in #2426
- Adding the Huber Loss function by @atreyamaj in #2479
- Update moleculenet.rst by @ncfrey in #2503
- PAGTN featurisation support for Molecular Graph Conv by @VIGNESHinZONE in #2496
- Added Sparse/Lazy Adam optimizer by @atreyamaj in #2493
- Adding the AdamW optimizer by @atreyamaj in #2488
- n_classes made variable(fixed to 2 classes) by @OmerOzgur271 in #2506
- Added Squared Hinge loss by @atreyamaj in #2497
- Fix DGL Dependency by @mufeili in #2516
- Wrapper function for Pagtn Model by @VIGNESHinZONE in #2508
- Tokenizer fix by @schithranandanurix in #2524
- Adding Pytorch dependency into deepchem by @VIGNESHinZONE in #2541
- added paper reference to tutorial 28 by @MariBerry in #2535
- Fix typo in class name by @alat-rights in #2539
- correct link of flake8 by @autodataming in #2522
- Converted MultitaskRegressor and MultitaskClassifier to PyTorch by @peastman in #2559
- Fixed errors in training models with uncertainty by @peastman in #2538
- Setup for Jax dependency by @VIGNESHinZONE in #2560
- [WIP] Adding the MAT Featurizer by @atreyamaj in #2544
- Weights & Biases Basic Integration by @kshen3778 in #2520
- Fix failing test cases by @peastman in #2568
- Fix: Tokenizer is not able to encode triple bonds by @niklashoelter in #2566
- Flake8 fix for molnet directory by @Suzukazole in #2525
- Tensorflow independent environment by @VIGNESHinZONE in #2567
- deepchem-Torch environment by @VIGNESHinZONE in #2563
- Add DummyFeaturizer by @Suzukazole in #2570
- Corrected Test Case for Smiles Tokenizer (adding to #2572) by @seyonechithrananda in #2578
- Update FASTA Loader to accept arbitrary featurizers by @alat-rights in #2565
- [WIP] Added examples for featurizers in documentation by @arunppsg in #2571
- Fix full CI by @VIGNESHinZONE in #2573
- Add section on setting up symbolic link by @Suzukazole in #2587
- Added 'transformers' argument to ValidationCallback by @kshen3778 in #2584
- USPTO Loader by @Suzukazole in #2546
- Updates docs to Material Featurizers by @arunppsg in #2585
- Adding the Freesolv dataset by @atreyamaj in #2576
- Remove pin for matminer by @peastman in #2590
- JaxModel (For integerating Jax into deepchem) by @VIGNESHinZONE in #2549
- Adding Docs for New CI by @VIGNESHinZONE in #2591
- Improved handling of classification_handling_mode in Metric by @peastman in #2595
- Added MATFeaturizer to website by @atreyamaj in #2599
- Fixed errors caused by new Pandas version by @peastman in #2605
- WandbLogger fixes: removed finish() from Model fit(), ValidationCallbacks fixes by @kshen3778 in #2586
- Added ScaleNorm as the first MAT layer by @atreyamaj in #2601
- Fixed ScaleNorm Test by @atreyamaj in #2610
- Added ToC and Getting Started section in CONTRIBUTING.md by @arunppsg in #2592
- Changed return type to np.ndarray in a few files by @atreyamaj in #2607
- Made behavior of GaussianProcessHyperparamOpt more consistent by @peastman in #2620
- Added defaults for precision_recall_curve by @peastman in #2614
- FirstDraftTutorial30 by @davidRFB in #2483
- Fixed examples for loading perovskite dataset by @arunppsg in #2621
- Reaction Split Transformer by @Suzukazole in #2597
- Update loader to include transformer by @Suzukazole in #2628
- Evaluation stage of JaxModel by @VIGNESHinZONE in #2604
- Tutorial 31 - Introduction to material science by @arunppsg in #2626
- [WIP] MAT: Attention Module by @atreyamaj in #2622
- Adding bash-script for lightweight installs by @VIGNESHinZONE in #2618
- Fixes to robertafeaturizer by @seyonechithrananda in #2581
- pass featurizer kwargs by @walid0925 in #2507
- RobertaFeaturizer by @walid0925 in #2523
- First linear layer from AlphaFold by @rbharath in #2634
- Minor fixes to examples in torch models by @arunppsg in #2630
- [WIP] MAT Layers: Encoder by @atreyamaj in #2623
- Fix doctest error in RxnSplitTransform by @Suzukazole in #2643
- Made a test less flaky by @peastman in #2645
- [WIP] MAT Layers: Embedding + Generator by @atreyamaj in #2624
- Add integrations section to README by @kshen3778 in #2649
- Init parameter modifications for WandbLogger by @kshen3778 in #2648
- Update PR template and docs with doctest, pytest by @Suzukazole in #2651
- Fix splitter errors for datasets without labels by @Suzukazole in #2641
- installation clarification for zsh [skip ci] by @arunppsg in #2664
- fixes attribute error from sklearn in print model by @arunppsg in #2671
- BertFeaturizer (Replaces #2608) by @alat-rights in #2642
...
DeepChem 2.5.0
See full release notes at https://forum.deepchem.io/t/deepchem-2-5-0-release/439
DeepChem 2.4.0
Read the full release notes at https://forum.deepchem.io/t/deepchem-2-4-0-release-notes/340
Keras Models as Default
This release of DeepChem swaps from our home-grown TensorGraph
framework to using Keras as the foundation of our models. This swap leaves us well prepared for the jump to Tensorflow 2.0 which will happen in our next major release. This version also bumps the TensorFlow version to 1.14. This release also includes a number of improvements to MoleculeNet and our transfer learning infrastructure
Remove uses of deprecated APIs #1550
Added attr-slow for the AtomicConvFeaturizer test #1552
Upgrade to TensorFlow 1.13.1 #1553
fix bug of load_pdbbind() and add new features #1561
Replaced Saver with Checkpoint #1566
Replaced uses of deprecated layers #1567
Convert TensorGraph layers to Keras layers #1578
Create KerasModel #1583
Update dependencies for DeepChem 2.2 #1584
Converted multitask models to KerasModel #1587
Remove contagious logger setup #1591
Converted graph models to KerasModel #1594
Construct dataset first time, even with reload set to True #1595
Loading thermosol and hppb datasets #1596
simple install one-liner #1602
Converted more models to Keras #1615
Smiles Based featurizers for ChemNet #1618
Converted progressive multitask models to KerasModel #1620
Swapping Split-Transform order #1621
Added ChemNet models with tests #1623
Swap Split-Transform order - II #1624
Converted GAN to KerasModel #1625
Converted reinforcement learning classes to Keras #1635
Created new MAML API #1636
SmilesToImage featurizer for Tox21, Sampl, HIV datasets #1637
ChemNet Fixes and Additions #1638
First version of pretrained loading #1643
Upgrade to TF 1.14 (Optional) #1645
Custom directories and SmilesToImage for MolNet #1649
ChemNet Fixes #1651
Created ValidationCallback #1652
Moved to Python 3.5 and 3.7 for Travis #1658
Stratified splitters, and minor changes for MolNet #1660
Updated installation instructions #1661
Workaround for bug in TF 1.14 #1662
Reorganized models directory #1664
Move test cases out of tensorgraph module #1666
Fixed broken and out of date examples #1671
Updated version number to 2.3.0 #1672
Update README.md #1682
DiskDataset.move() would not overwrite an existing dataset #1683
Improved Protein Structure, Microscopy, Interpretation Support
DeepChem 2.2 takes large steps towards making DeepChem a general purpose deep learning library for life science applications. Major improvements have been made to support for deep learning on protein structures, and significant support for image-based dataset and model handling has been added. In addition, tooling for interpreting deep models has been improved. A number of improvements to existing models have been added as well, including adding estimator support for a number of model classes. Many bugfixes and small improvements made it in as well. DeepChem 2.2 now depends on TensorFlow 1.12.
PDBBind and Protein Structure Improvements
#1366, #1383, #1411, #1413, #1476 Atomic Convolution Improvements
#1503, #1430, #1432 PDBBind bugfixes
#1497 Using binding pockets to load PDBBind
#1498 DeepMHC for protein peptide binding
#1369, #1360, #1372, #1397 Featurziation Improvements
#1498 DeepMHC for protein peptide binding
Image Handling Improvements
#1516 Image Transformation improvements
#1324 Cell counting dataset added. ImageLoader
added
#1414 Diabetic Retinopathy example model
#1439 ImageDataset
class
Dataset and Splitter Additions, Improvements and fixes
#1507, #1540, #1406 Bugfixes
#1514 Handling verbose=False when transforming data
#1499 Butina splitter improvement
#1347 Adds USPTO dataset.
#1348 BBBC002 dataset addition
#1339 Split datasets on ID
#1416 Molnet loaders for UV/Kinase/Factors datasets
#1327 BBBC001 dataset addition
#1447 SDFLoader improvements
#1425 Binary classification metric improvements
Model and Layer Additions and Improvemetns
#1500 Seq2seq model improvements
#1513 Clean up symmetry functions
#1488 New graph convolution
#1365 Average pooling for conv-nets
#1370 ResNet50 improvements
#1450 Layer output shapes
#1452 Pad batch improvements
#1453 Example distributed multitask classifier
#1335 GraphConv improvements
#1343 Making it easy to pull out neural fingerprints
#1433 TensorGraph get layer weights
#1325 UNet model changes
#1334 First Resnet50 build
#1473, #1142 TextCNN make_estimator support
#1475 DTNN make_estimator support
#1495 ANIRegression, BPSymmetryFunction make_estimator
Better Interpretability
#1393 Saliency Mapping
#1445 Saliency maps for diabetic retinopathy
Tests, Docs, Housekeeping
#1527, #1457 Readme cleanup
#1548 Version bump
#1515 Upgrade to TF 1.12
#1462 Typo fix predict_proba
#1385, #1418 Build Fixes
#1423 Yapf updates
#1408 indentation cleanup
#1344 Python 3.6 updates
#1330: Docs updates
#1337 Large screens tutorial
#1338 Colab notebook version
#1437 Python3 fixes
#1454 Make RDKit a soft requirement
#1455 Make simdna a soft requirement
#1456, #1484 Make six a soft requirement
#1458 Add tutorial section
#1420 genomics code grouping into single file
#1535, #1485, #1487, #1371, #1421, #1479, #1480 Test Improvements and Fixes