WekaDeeplearning4j - v1.7.2

New Features

• Added the IsGPUAvailable tool (docs), a simple tool to check whether your GPU has been detected by WDL4J which can help with diagnosing GPU driver issues.

Other

• Merged example usage docs with I2K tutorial docs
• Add unit tests for recently added class
• Add Javadocs to code
• Update Zoo Model URLs to use WekaDeeplearning4j Release URL

SHA256 sums

• wekaDeeplearning4j-1.7.2.zip
  3051c65b6aebfb54a12a48f33d01b07470ac0db54f02deb3359aa13751408e6f
• wekaDeeplearning4j-cuda-10.0-1.7.2-linux-x86_64.zip
  caadaa8faf5cf386f0fd3a6e288809fa2bcc0360b99fdfc282c7c1f40d3dec96
• wekaDeeplearning4j-cuda-10.0-1.7.2-macosx-x86_64.zip
  42e30265a0e8a684d5645510c7c7d1181dc4546021d2a6c1270f1e6b1e615903
• wekaDeeplearning4j-cuda-10.0-1.7.2-windows-x86_64.zip
  6a02862c6e0cf6954f2f499b58ce715e8c098c8f7e18bc0ea750662f28ebe194
• wekaDeeplearning4j-cuda-10.1-1.7.2-linux-x86_64.zip
  56d56f45e7a631a53c8220f8cccbcff354407079b40702d54865c409445cdfa3
• wekaDeeplearning4j-cuda-10.1-1.7.2-macosx-x86_64.zip
  fdccca9e147ffa807af2624ba2caca767fcfa4bb229d1198a397c5df67aae1d3
• wekaDeeplearning4j-cuda-10.1-1.7.2-windows-x86_64.zip
  a391c5d58d9061e5b2987675711b97abc614d0a3bdafcff615fd40d7db39e31a
• wekaDeeplearning4j-cuda-10.2-1.7.2-linux-x86_64.zip
  8b03875df8866838dadad07de20230ac1e3e801dbc3890693d7e1b51d2a57619
• wekaDeeplearning4j-cuda-10.2-1.7.2-macosx-x86_64.zip
  b72192755ceb3e4238ee3c6496021f228d1cd9da7135c7c061027a6bad5e57db
• wekaDeeplearning4j-cuda-10.2-1.7.2-windows-x86_64.zip
  c1d574475a471fa41872a6cc2ef0701793d4dd585bbe83be3c6f25e75985a3aa

Keras Model Zoo Files - DL4J Format

Moving the keras zoo model files from the basedrhys fork to this fork. This keeps the data all together and limits dependencies on other repos.

WekaDeeplearning4j Examples - Asset Pack

Moving the tutorial asset pack from the I2K Tutorial Repo, to keep all data together and limit dependencies on other repos.

This contains all the assets required to follow through the usage examples.

WekaDeeplearning4j - v1.7.1

Release Notes

• Fix bug when using a custom-trained Dl4jMlpClassifier model with the Dl4jCNNExplorer.
  • Wasn't applying the same image value normalisation as is done during training.
• Fix ImageDirectoryLoader not loading .tiff datasets
  • Wasn't recognizing .tiff files as images
• Allow loading of Dl4jInference class map from CSV file
• Add extra error checking when using Darknet19 with an incorrect class map

SHA256 sums

• wekaDeeplearning4j-1.7.1.zip
  a87f01185a3b417567943968609d95ec8afd320bd57c6c0f7307df8dfee15502
• wekaDeeplearning4j-cuda-10.0-1.7.1-linux-x86_64.zip
  e51068ae5caf395cd6d1b8276bb1a0f39818a52858fec3113758d21ffd0ccc4b
• wekaDeeplearning4j-cuda-10.0-1.7.1-macosx-x86_64.zip
  065b2557981844c5cd44b37b1a754218507814e47e4647f9b287138238ecbf97
• wekaDeeplearning4j-cuda-10.0-1.7.1-windows-x86_64.zip
  a1baba033eab73f82b835130fa92c4be86fb347b8ca6d8e8c91cb1c78eb4f035
• wekaDeeplearning4j-cuda-10.1-1.7.1-linux-x86_64.zip
  1b5d317dce67742a7e8772c5e4d1ddce82816b1835a14da786f0e0cdda44dac0
• wekaDeeplearning4j-cuda-10.1-1.7.1-macosx-x86_64.zip
  ec198f090fbc01b5edc8e439b8c06d8de1c69443f53c5efcd86df7dc4ca2e0ed
• wekaDeeplearning4j-cuda-10.1-1.7.1-windows-x86_64.zip
  161e3cc19cb63709700a9abe760c7474eb9ab30c90cf2c72df950050259a02d0
• wekaDeeplearning4j-cuda-10.2-1.7.1-linux-x86_64.zip
  3a4e6bc39e8b7d5f27a88f985013d452e08cfdc0e047869307a80e5536ab6827
• wekaDeeplearning4j-cuda-10.2-1.7.1-macosx-x86_64.zip
  b2b101e228799b3c467b97fe783ede35c220209f984f0f701153487101ec271c
• wekaDeeplearning4j-cuda-10.2-1.7.1-windows-x86_64.zip
  1c1cf470808048ac594fa9ab09a195f87fe21ae86e151672f69d8beb82faa9dc

WekaDeeplearning4j v1.7.0

Changelog

• Introduce the new Dl4jCNNExplorer class (usage examples), which allows you to use zoo models or custom trained models to perform prediction on an individual image, without loading a .arff dataset.
• Implement the ScoreCAM saliency map generation method and integrate it with the Dl4jCNNExplorer (usage examples).
• Implement a progress bar (for command line and GUI) to aid usability with long-running tasks.

SHA256 sums

• wekaDeeplearning4j-1.7.0.zip
  0d03b8702268fa7c6a904d587bf3f4356d0ad5edca8f4a778bfb313863e6b319
• wekaDeeplearning4j-cuda-10.0-1.7.0-linux-x86_64.zip
  e06b70490b5caeab8ae1f11ac9afcd49ca8e671401183cbe562cc42858be63a6
• wekaDeeplearning4j-cuda-10.0-1.7.0-windows-x86_64.zip
  8e1076c8e9a534b96f4314da0d8dcfbf6982b1e0206770ea51e8c994f669a15b
• wekaDeeplearning4j-cuda-10.1-1.7.0-linux-x86_64.zip
  153fce7c627fccad6ddf13d0b1edf7ffb7d8d3b6e17db6947fb1da6bcc86e6ee
• wekaDeeplearning4j-cuda-10.1-1.7.0-windows-x86_64.zip
  97b3140f4641288d4ebd56b9c398afc2cb9fccc11036ab57b9db6808f0fef9f0
• wekaDeeplearning4j-cuda-10.2-1.7.0-linux-x86_64.zip
  a013a616b52bfda3541e10ecf720a9ee6e98842f599972e10634883ddcd697b3
• wekaDeeplearning4j-cuda-10.2-1.7.0-windows-x86_64.zip
  f94418525efde2488d9483abc8685bf6051061a8d72a65279bbbc0d29bf5b9c1

wekaDeeplearning4j-v1.6.0

This release primarily focuses on improvements and new features for the image-classification side of WekaDeeplearning4j.

Changelog

• Pretrained Models - Add DL4J and Keras pre-trained models into the model zoo. WekaDeeplearning4j now has over 30 pre-trained models, including big names like ResNet, InceptionV3, and EfficientNet
• Improved Feature Extraction - Enable Dl4jMlpFilter to use a pre-trained model as feature extractor, as well as use multiple feature layers, and different pooling methods
• Add new usage examples for feature extraction and training with a pre-trained model
• Add ImageDirectoryLoader - an easy way to load 'folder-organised' image datasets into WEKA (usage example)
• Update Deeplearning4j backend to 1.0.0-beta7
• Add CUDA 10.2 support
• Drop CUDA 9.2 support

SHA256 sums

• wekaDeeplearning4j-1.6.0.zip
  cc528b0e8bbfd869f731661a599e49cf880eac94fdba0c6d8d529c0c80ad625b
• wekaDeeplearning4j-cuda-10.0-1.6.0-linux-x86_64.zip
  77793662457d3550e1aeebcd6d943a8dac7d3cb687dfd6d7193b7c1bf9f5c609
• wekaDeeplearning4j-cuda-10.0-1.6.0-windows-x86_64.zip
  84e5d1459d7e00fa5664a1e35eba6fc1c05e635f4e7d3df5b793ba3f4d33976c
• wekaDeeplearning4j-cuda-10.1-1.6.0-linux-x86_64.zip
  19fef0f8aba6ffb9cd5543712fc15a812ae7836b6b8ef8ff60c8b531c20e33e5
• wekaDeeplearning4j-cuda-10.1-1.6.0-windows-x86_64.zip
  82aa13dd25d8b46acfa7083a498861a1fdcd3eb9e25d341a4756be626880f23f
• wekaDeeplearning4j-cuda-10.2-1.6.0-linux-x86_64.zip
  da3a7d2b049816acdc53ac6e9738b74b118e01ae8af0c0c43fd77359052a21fa
• wekaDeeplearning4j-cuda-10.2-1.6.0-windows-x86_64.zip
  7e02cf449fd96d92f032385659f7f46039d7c3c1aa0c402626f2fd28a9e8fd1b

wekaDeeplearning4j v1.5.14

Changelog

• Update backend to Deeplearning4j 1.0.0-beta5
• Dropped support for CUDA 8.0 and 9.0
• Added support for CUDA 10.0 and 10.1

Notes

• This release does not support cuda on macosx (see deeplearning4j/deeplearning4j#8245 for updates)
• Nadam optimizer will not support learning rate schedules until the next Deeplearning4j release

SHA256 sums

• wekaDeeplearning4j-1.5.14.zip
  2cf5d49b5e9025a1d94922260446b4e38472819ef8e22f70953d9feb08691f43
• wekaDeeplearning4j-cuda-10.0-1.5.14-linux-x86_64.zip
  d5711f4478cde860ea4de0048aa3df16c6c37258221e06b1ba77bd1d965016df
• wekaDeeplearning4j-cuda-10.0-1.5.14-windows-x86_64.zip
  83bddee29432ded4d62d15d4c120919fec0808ffd9831473755cf9af47b6bd8d
• wekaDeeplearning4j-cuda-10.1-1.5.14-linux-x86_64.zip
  31f2f72d4b75d730efd1dbb7264702f0a99a9565d7e12d88552299de983a5c7a
• wekaDeeplearning4j-cuda-10.1-1.5.14-windows-x86_64.zip
  0447e88729d80aa8b81dc64e85a7725d96d9fa56d6ba57c77800d2a0df8e9035
• wekaDeeplearning4j-cuda-9.2-1.5.14-linux-x86_64.zip
  90a7fcf1e0de5fee3c8211f33309da33a99fd962d5868c8ff924a5d5e12695dc
• wekaDeeplearning4j-cuda-9.2-1.5.14-windows-x86_64.zip
  a2a20cb6afad832f37ffc987887861d39f211806e169e785a25fb26b0d6eacbe

wekaDeeplearning4j v1.5.13

Changelog

• Fix critical bug when serializing RnnSequenceClassifier

SHA256 sums

• wekaDeeplearning4j-1.5.13.zip
  5260fcafea9605106ab11b309a30a66a69cd0b635fe109cd58004326d1192479
• wekaDeeplearning4j-cuda-8.0-1.5.13-linux-x86_64.zip
  2b0ca3a01488b0e89475e0e56218030883e561087720c07c921948013eba95b9
• wekaDeeplearning4j-cuda-8.0-1.5.13-macosx-x86_64.zip
  de959e949c6c0bee65bc239cf2db092d8176dd6290dda9bf7c915815a4e53bd6
• wekaDeeplearning4j-cuda-8.0-1.5.13-windows-x86_64.zip
  ea40933b9b1b3959d75262f5884770cc5309e275c11e1f2b07b3b11b18c5603d
• wekaDeeplearning4j-cuda-9.0-1.5.13-linux-x86_64.zip
  db72cc14c7c6a45d85d4156c5516af868843d3ce719a00f045b915f4c5b505a9
• wekaDeeplearning4j-cuda-9.0-1.5.13-macosx-x86_64.zip
  f40b6d1ec665571994032308a28f95c18138f91e29d8f283d23e53951bd8552e
• wekaDeeplearning4j-cuda-9.0-1.5.13-windows-x86_64.zip
  e5d3ad0655d2ebe11541c0710da5d93f195a6db7c26882591ee633e349cf91ee
• wekaDeeplearning4j-cuda-9.2-1.5.13-linux-x86_64.zip
  57c9e5fa54f742eda2ba21ea023ef9d8be67dad8a42bf85e29dc122c9648e5d2
• wekaDeeplearning4j-cuda-9.2-1.5.13-macosx-x86_64.zip
  6556b35bd660153655537df18c026d417e408a91d6819d207dbea0957d755399
• wekaDeeplearning4j-cuda-9.2-1.5.13-windows-x86_64.zip
  459a7dad303c029b6fb6dc425a5fc2305337670433e16bc2ba4aaade398f1139

wekaDeeplearning4j v1.5.12

Changelog

• Critical hotfix for inverted stopwords filtering in CnnText(Files)EmbeddingInstanceIterator

SHA256 sums

• wekaDeeplearning4j-1.5.12.zip
  82b3373862d987d53868e538b3f051c2245a44ce61e776a835dd2c969124c142
• wekaDeeplearning4j-cuda-8.0-1.5.12-linux-x86_64.zip
  f5c6186622331cd9403a5a5ab8037736b7faa9f4495b91a27c9f089cfdd057e6
• wekaDeeplearning4j-cuda-8.0-1.5.12-macosx-x86_64.zip
  2fe2e74fac7faa17f7aa2300bb63dc21bce4e6e6ebcbf55ed86170610a2bc016
• wekaDeeplearning4j-cuda-8.0-1.5.12-windows-x86_64.zip
  ab44f3bc0b638a2904724360ee3008b3908e88b7f06d5e1869d2e40fdcc911c9
• wekaDeeplearning4j-cuda-9.0-1.5.12-linux-x86_64.zip
  2173c434109b80011ecff8112aa1acb2b306a2962fd202f111c1d10f662e9d87
• wekaDeeplearning4j-cuda-9.0-1.5.12-macosx-x86_64.zip
  63f10b9a321d7d349657033efc36bf6fb25246412db68d95cb9d7cd9322d9386
• wekaDeeplearning4j-cuda-9.0-1.5.12-windows-x86_64.zip
  c60d8e48d258deb8a776408a1dc561318c368ba062f97662ae2977e3dc35a400
• wekaDeeplearning4j-cuda-9.2-1.5.12-linux-x86_64.zip
  f8d493492efa97b5881797a424f32a1e160ff01154e6d64eca3ba27f13c3a01b
• wekaDeeplearning4j-cuda-9.2-1.5.12-macosx-x86_64.zip
  5b21ba25dca4d42d728e5ccf75c0998f3d709cd1943460ee7ffaf56b0b5358ab
• wekaDeeplearning4j-cuda-9.2-1.5.12-windows-x86_64.zip
  f72452d53255d1387fc33e350b4f84c0f596bd80314c40e1ab7881a0b2329b33

wekaDeeplearning4j v1.5.11

Changelog

• Add resumable feature #43
• Fix LogConfiguration for command line usage c805a8e

SHA256 sums

• wekaDeeplearning4j-1.5.11.zip
  a0952692981ffb82dd067124e55b7840de3a158c035b1dbc6fded616906fbebf
• wekaDeeplearning4j-cuda-8.0-1.5.11-linux-x86_64.zip
  708aaa8cd685d6a047b8fab448563b734dfc9644fead1ff75ae59353cba6ae05
• wekaDeeplearning4j-cuda-8.0-1.5.11-macosx-x86_64.zip
  632290e3ade126434333979a04c66b9a139f0e056cc4688efc398bfa90851856
• wekaDeeplearning4j-cuda-8.0-1.5.11-windows-x86_64.zip
  dc253822b7fa5b4869cd6f83d981f7f960ef7d7278243c1ca7a64db60f04afa8
• wekaDeeplearning4j-cuda-9.0-1.5.11-linux-x86_64.zip
  cdb11e3b803282bb94b6263004daec8865596772c5d093bff0b922ce7bfc0c1d
• wekaDeeplearning4j-cuda-9.0-1.5.11-macosx-x86_64.zip
  fe98655b97c144a6ef63ec0e136cbfeb178b554a1b5a39f48f4f5abec924f8c4
• wekaDeeplearning4j-cuda-9.0-1.5.11-windows-x86_64.zip
  da7eb7c4ef29e6041db6dcb55d5c5305b632e48ac958f9665427e99d90b44fe9
• wekaDeeplearning4j-cuda-9.2-1.5.11-linux-x86_64.zip
  78d1b16ef4fd21fddf78421ad8510bb3d35956a48bc239993bc3cb3291996eed
• wekaDeeplearning4j-cuda-9.2-1.5.11-macosx-x86_64.zip
  2096a16cb5b0e231c0ae447d7f5c3d6e8d7da5b067ba66e613a2c4278142ca9b
• wekaDeeplearning4j-cuda-9.2-1.5.11-windows-x86_64.zip
  9e50fcc16a0133b0c19513a839df48c78317e58848bd01042b0dc55185a91204