| 73928ed | Curtis "Fjord" Hawthorne | 28 February 2018, 23:22:04 UTC | bump version (#1079) | 28 February 2018, 23:22:04 UTC |
| bad6882 | Adam Roberts | 28 February 2018, 20:20:58 UTC | Update @magenta/music-vae API with NoteSequence-like I/O. (#1073) * Add NoteSequence-like interface to MusicVAE-js * rename package to music_vae for consistency * rename package to music_vae for consistency * Add missing files. * Replace tabs. * Replace tabs. * Rename package to music-vae * Add documentation to model.ts * add javadoc | 28 February 2018, 20:20:58 UTC |
| 63ad504 | Ian Simon | 28 February 2018, 18:58:27 UTC | support control sequences in split and hierarchical decoders (#1075) | 28 February 2018, 18:58:27 UTC |
| 2e7bc85 | Adam Roberts | 27 February 2018, 04:45:54 UTC | Clean up the code, fix tidying to avoid memory leaks, and use async properly. (#1070) * Tidying up a bit. * Tidying up a bit. * more tidying * more tidying * Remove bundle.js * Moved init to separate line for clarity * Responding to reviewer comments. * Responding to reviewer comments. | 27 February 2018, 04:45:54 UTC |
| 91ab4ed | Adam Roberts | 26 February 2018, 08:17:36 UTC | Fixes to dl 0.5 update. Added more complete demo. (#1069) * Add initial version of MusicVAE deeplearn implementation. * Add readme * Add readme * Add readme * Incorporating Kyle's suggestions. * Add @magenta prefix to package * 0.5 * Add more useful demo page. * Fixes to dl 0.5 update. Added more complete demo. * fix indent * fix indent | 26 February 2018, 08:17:36 UTC |
| 7a601e1 | Torin Blankensmith | 26 February 2018, 08:06:59 UTC | Updated MusicVAE to deeplearn 0.5.0 (#1068) * updated to deeplearn 0.5.0 * updated demo to deeplearn 0.5 | 26 February 2018, 08:06:59 UTC |
| 909758c | Curtis "Fjord" Hawthorne | 23 February 2018, 21:53:53 UTC | Initial @magenta/core npm javascript package (#1065) | 23 February 2018, 21:53:53 UTC |
| c0d99f6 | Adam Roberts | 23 February 2018, 21:38:35 UTC | Add initial version of MusicVAE deeplearn implementation. (#1067) * Add initial version of MusicVAE deeplearn implementation. * Add readme * Add readme * Add readme * Incorporating Kyle's suggestions. * Add @magenta prefix to package | 23 February 2018, 21:38:35 UTC |
| 7240ec1 | Ian Simon | 23 February 2018, 01:08:24 UTC | allow truncating performances by time when extracting (#1066) | 23 February 2018, 01:08:24 UTC |
| 949f1ec | Adam Roberts | 22 February 2018, 21:40:12 UTC | Fix MusicVAE __init__ and set pyfuncs to stateless. (#1064) | 22 February 2018, 21:40:12 UTC |
| 173e0c0 | Adam Roberts | 22 February 2018, 19:53:49 UTC | Add new hierarchical encoders and decoders. (#1056) * Add new hierarchical components. * Handle Nones in DecodeResults * Add lstm_utils_test file * Add configs * Fix references to HierarchicalMultiOut decoder * Put segment metrics under their own group * Fix sum * Update docs * Rename cat-mel_16bar to flat-mel_16bar * Rename cat-trio_16bar to flat-trio_16bar * Improve documentation. * Lint * Lint * rename state conversion * lint | 22 February 2018, 19:53:49 UTC |
| 92b523f | Ian Simon | 22 February 2018, 18:27:47 UTC | use default number of chords per bar when unspecified in chord inference (#1063) * use default number of chords per bar when unspecified in chord inference * add ChordInferenceException superclass * simplify chord inference usage in MusicVAE | 22 February 2018, 18:27:47 UTC |
| 8515d02 | Ian Simon | 22 February 2018, 17:52:41 UTC | fix bug in checking control sequence existence (#1062) | 22 February 2018, 17:52:41 UTC |
| 1dbc95f | Ian Simon | 22 February 2018, 06:50:31 UTC | add chord-conditioning to melody and trio MusicVAE (#1059) * add chord conditioning to melody and trio MusicVAE * add output_velocity constant * move chord_utils out of music_vae and into magenta.music.chords_lib * fix typo | 22 February 2018, 06:50:31 UTC |
| 9794130 | Adam Roberts | 22 February 2018, 00:36:30 UTC | Add variable name substitution to TrainedModel. (#1058) * Add variable name substitution to TrainedModel. * Add missing import | 22 February 2018, 00:36:30 UTC |
| f8203d9 | Adam Roberts | 22 February 2018, 00:35:52 UTC | Remove conditional hparam and use z_size=0 instead. (#1057) | 22 February 2018, 00:35:52 UTC |
| cc4ff1d | Sergey Bokhnyak | 22 February 2018, 00:21:02 UTC | Fixed IllegalFlagValueError: flag value must be string caused by line 28 in nsynth_generate.py because of TF Flags expecting string but getting Boolean (#1040) | 22 February 2018, 00:21:02 UTC |
| 2759022 | Curtis "Fjord" Hawthorne | 21 February 2018, 19:31:44 UTC | Upgrade to librosa 0.6.0 api (#1060) * upgrade to librosa 0.6.0 api * switch to ref | 21 February 2018, 19:31:44 UTC |
| bc00859 | Ian Simon | 20 February 2018, 18:27:40 UTC | extract performances for multiple instruments (#1055) | 20 February 2018, 18:27:40 UTC |
| e5a45e3 | Adam Roberts | 20 February 2018, 17:38:26 UTC | Shuffle order of dataset files when training. (#1054) * Fix shuffle * Shuffle in interleave * Push some dataset creation code into data lib.: * lint * simplify * Pass None when control_depth is 0. | 20 February 2018, 17:38:26 UTC |
| e813ca6 | Ian Simon | 16 February 2018, 00:17:22 UTC | add chord_utils.py with functions for extracting event-aligned chords and adding chords to a sequence (#1052) | 16 February 2018, 00:17:22 UTC |
| 5e4adcd | Ian Simon | 15 February 2018, 17:26:18 UTC | allow MusicVAE to condition on a control sequence (#1051) * allow MusicVAE to condition on a control sequence * some MusicVAE documentation/style fixes * add ConverterTensors struct to be returned by to_tensors in MusicVAE | 15 February 2018, 17:26:18 UTC |
| 632236b | Ian Simon | 14 February 2018, 22:24:22 UTC | add heuristic chord inference (#1050) * add heuristic chord inference * python 3 bugfix | 14 February 2018, 22:24:22 UTC |
| fa9c9e6 | Curtis "Fjord" Hawthorne | 14 February 2018, 00:22:43 UTC | Removes unused test notesequences.tfrecord. (#1049) Fixes #1042 | 14 February 2018, 00:22:43 UTC |
| 29572d6 | Ian Simon | 13 February 2018, 21:45:06 UTC | transpose chords and not drums in MusicVAE NoteSequenceAugmenter (#1048) * also transpose chords in NoteSequenceAugmenter * add unit tests for NoteSequenceAugmenter in MusicVAE * don't transpose drums in MusicVAE * fix lint error | 13 February 2018, 21:45:06 UTC |
| 4d18691 | Ian Simon | 13 February 2018, 00:39:02 UTC | support metric quantization for performances (#1047) | 13 February 2018, 00:39:02 UTC |
| 8a902fa | Ian Simon | 05 February 2018, 23:30:47 UTC | add steps method to EventSequence, to be used for indexing into time-based control sequences (#1039) | 05 February 2018, 23:30:47 UTC |
| 39b407f | Ian Simon | 05 February 2018, 22:01:41 UTC | move performance_lib and performance_encoder_decoder to magenta.music (#1038) | 05 February 2018, 22:01:41 UTC |
| e6597d7 | Curtis "Fjord" Hawthorne | 30 January 2018, 17:29:42 UTC | Factor out preprocessing and tensor combining code from BaseNoteSequenceConverter (#1036) * Factor out preprocessing and tensor combining code from BaseNoteSequenceConverter * py3 fix | 30 January 2018, 17:29:42 UTC |
| 6b31521 | Adam Roberts | 29 January 2018, 23:12:12 UTC | Fix Seq2SeqLstmDecoder output_size. (#1035) * Fix Seq2SeqLstmDecoder output_size. * Fix Seq2SeqLstmDecoder output_size. | 29 January 2018, 23:12:12 UTC |
| 8190b87 | Adam Roberts | 26 January 2018, 22:38:49 UTC | Created separate lstm_utils and cleaned up lstm_models. (#1034) * Add lstm utils. * Add lstm_models changes * lint * lint * lint | 26 January 2018, 22:38:49 UTC |
| 80167c6 | Curtis "Fjord" Hawthorne | 26 January 2018, 00:03:21 UTC | lint (#1033) | 26 January 2018, 00:03:21 UTC |
| fdcdfd4 | Curtis "Fjord" Hawthorne | 25 January 2018, 22:56:51 UTC | https links (#1032) | 25 January 2018, 22:56:51 UTC |
| c4390f4 | Douglas Eck | 25 January 2018, 19:51:12 UTC | Update README.md (#1031) * Update README.md * Oxford commas everywhere * Comma splice | 25 January 2018, 19:51:12 UTC |
| 5964752 | Curtis "Fjord" Hawthorne | 25 January 2018, 18:05:26 UTC | update docker to 0.3.2 release (#1030) | 25 January 2018, 18:05:26 UTC |
| 4a5e55e | Adam Roberts | 25 January 2018, 01:06:30 UTC | Add length shape specification to tf_to_tensors. (#1029) * Add length shape specification to tf_to_tensors. * fix | 25 January 2018, 01:06:30 UTC |
| 082c1aa | Adam Roberts | 24 January 2018, 23:38:34 UTC | Initial changes required for support of variable-length hierarchy. (#1028) * Initial changes required for support of variable-length hierarchy. * lint * Revert accidental change. * Add changes to _remove_pad_fn * Additional changes to generalize for different shape lengths. * lint * lint * lin | 24 January 2018, 23:38:34 UTC |
| e19f844 | Chen Shen | 24 January 2018, 18:49:44 UTC | provide_batch() got an unexpected keyword argument 'examples_path' (#1026) | 24 January 2018, 18:49:44 UTC |
| 1d37759 | Adam Roberts | 23 January 2018, 00:20:08 UTC | Add generic BaseConverter class. (#1024) * Add BaseConverter class. * Add str_to_item_fn accessor. * lint * lint * lint * lint * lint * Fix typo. * Fix maybe_sample | 23 January 2018, 00:20:08 UTC |
| 7475172 | Curtis "Fjord" Hawthorne | 22 January 2018, 23:06:19 UTC | Switch to global norm clipping by default, update grad_norm_clip_to_zero to a more reasonable default (#1025) * update grad_norm_clip_to_zero to a more reasonable default * make global_norm the default clip mode | 22 January 2018, 23:06:19 UTC |
| 37fcd37 | Adam Roberts | 19 January 2018, 18:13:21 UTC | Set dropout keep probability to 1.0 for non-training jobs inside of rnn_cell for consistency. (#1022) * Set dropout keep probability to 1.0 for non-training jobs inside of rnn_cell. * lint * lint | 19 January 2018, 18:13:21 UTC |
| 7dd72d0 | czhuang | 19 January 2018, 04:31:59 UTC | Coconet prep (#1020) * coconet and build files! * fix build deps * adding bazel scripts * changes and checks to flags * fix lint errors * load checkpoint from absolute path * remove commented out lines | 19 January 2018, 04:31:59 UTC |
| b74f92d | Adam Roberts | 18 January 2018, 18:25:47 UTC | Return final_state in reconstuction_loss and sample. (#1021) * Return final_state in reconstuction_loss and sample. * Return final_state in reconstuction_loss and sample. * Add doc | 18 January 2018, 18:25:47 UTC |
| d0edf93 | Adam Roberts | 17 January 2018, 17:36:45 UTC | Add data loading performance optimizations. (#1019) * Add data loading performance optimizations. * Lint * fix | 17 January 2018, 17:36:45 UTC |
| 2eab3bf | Adam Roberts | 17 January 2018, 03:21:26 UTC | Fix bad return values in TrioConverter. Automatically set shuffle buffer size based on batch size. (#1018) | 17 January 2018, 03:21:26 UTC |
| a998062 | Curtis "Fjord" Hawthorne | 16 January 2018, 21:46:29 UTC | Add prefetch to end of data pipeline (#1017) * prefetch * add flag * fix * lint | 16 January 2018, 21:46:29 UTC |
| d931faa | Adam Roberts | 16 January 2018, 17:59:16 UTC | Correct order of biases in BackwardCompatibleCudnnLSTMSaveable. (#1015) * Correct order of biases in BackwardCompatibleCudnnLSTMSaveable. * Fix index ordering in _get_final | 16 January 2018, 17:59:16 UTC |
| 672d0db | Adam Roberts | 16 January 2018, 17:56:35 UTC | rebase (#1014) | 16 January 2018, 17:56:35 UTC |
| 6ca81a4 | Adam Roberts | 13 January 2018, 01:21:20 UTC | Fix backwards compatibility issues between CudnnLSTM and LSTMBlockCell. Correct LstmEncoder behavior for variable length sequences. (#1013) * Fix incorrect input to backwards CudnnLSTM in bidi encoder. * Fix backwards compatibility issues between CudnnLSTM and LSTMBlockCell * lint * lint * lint * Respond to reviewer. | 13 January 2018, 01:21:20 UTC |
| 5cc3bc4 | Adam Roberts | 11 January 2018, 22:20:38 UTC | Fix incorrect input to backwards CudnnLSTM in bidi encoder. (#1012) | 11 January 2018, 22:20:38 UTC |
| 685620f | Adam Roberts | 11 January 2018, 18:29:15 UTC | Add support for variable-length cudnn encoder. Fix minor bugs. (#1010) * Add support for variable-length cudnn encoder. Fix minor bugs. * add back bidirectional name prefix support * other required changes | 11 January 2018, 18:29:15 UTC |
| d4aaa9a | Curtis "Fjord" Hawthorne | 11 January 2018, 18:02:26 UTC | Global norm clipping for MusicVAE (#1011) * global norm gradient clipping * fix * fix * fix * fix * fix * fix | 11 January 2018, 18:02:26 UTC |
| b153bb3 | Curtis "Fjord" Hawthorne | 11 January 2018, 18:02:12 UTC | save every 1 hour (#1008) | 11 January 2018, 18:02:12 UTC |
| 8341aeb | Curtis "Fjord" Hawthorne | 11 January 2018, 00:35:39 UTC | Fix typo (#1009) | 11 January 2018, 00:35:39 UTC |
| 0cbd5da | Adam Roberts | 05 January 2018, 22:55:57 UTC | Fix incorrect use CudnnLSTM with scheduled sampling. Simplify related code and add documentation. (#1004) * Fix incorrect use CudnnLSTM with scheduled sampling. Simplify related code and add documentation. * lint * Add warning when cudnn isn't being used | 05 January 2018, 22:55:57 UTC |
| 837ff44 | Adam Roberts | 05 January 2018, 05:39:48 UTC | Add beam search to CategoricalLstmDecoder sampling. (#1000) * Add beam search to CategoricalLstmDecoder sampling. * Add some comments. * lint * lint * Fixes to hierarchical beam search * Remove stray logging message | 05 January 2018, 05:39:48 UTC |
| a5618e0 | Adam Roberts | 05 January 2018, 00:01:45 UTC | Minor updates to MusicVAE. (#1001) * Minor changes to MusicVAE. * Remove stray line | 05 January 2018, 00:01:45 UTC |
| c47f404 | Adam Roberts | 04 January 2018, 19:42:38 UTC | Add support for CudnnLSTM and remove hparams for unsupported/unused options. (#999) * Add support for CudnnLSTM remove hparams for unsupported/unused options. * Add TODO * Add doc * lint * Add cudnn support for hiearchical decoder. * Fix indentation. * Lint * fix directions * Fix default cudnn name * Respond to reviewer comments | 04 January 2018, 19:42:38 UTC |
| de1c8e8 | Curtis "Fjord" Hawthorne | 21 December 2017, 21:49:19 UTC | Only output trial summary if the task is_chief (#991) | 21 December 2017, 21:49:19 UTC |
| 7a92b1a | Curtis "Fjord" Hawthorne | 20 December 2017, 23:29:05 UTC | fix is_training (#990) | 20 December 2017, 23:29:05 UTC |
| c7a9126 | Adam Roberts | 20 December 2017, 19:08:28 UTC | Add additional MusicVAE documentation and logging flag. (#989) * Add additional MusicVAE documentation and add logging. * Fix typo * lint * lint * lint | 20 December 2017, 19:08:28 UTC |
| 9e4f4cd | Adam Roberts | 20 December 2017, 17:02:28 UTC | Fix accidental removal of TrioConverterTest (#988) | 20 December 2017, 17:02:28 UTC |
| b3e51fe | Adam Roberts | 20 December 2017, 05:33:14 UTC | Add support for modeling drum pianorolls with a NADE. (#987) * Move NADE to common to make it available to other models. * lint * Add support for drum pianorolls * update configs * Add missing definitions in data.py * lint * Fix config name * Only define NADE bias variables if they will be used. * Add temperature to NADE * lint * lint | 20 December 2017, 05:33:14 UTC |
| d51396c | Michael Ma | 20 December 2017, 04:44:59 UTC | Add music_vae_train as entry point (#986) * Add music_vae_train as entry point * Moved to correct position | 20 December 2017, 04:44:59 UTC |
| 8a20594 | Adam Roberts | 18 December 2017, 22:03:17 UTC | Move NADE to common to make it available to other models, and add optional internal bias. (#984) * Move NADE to common to make it available to other models. * lint * Add back accidentally deleted docstring * lint * lint * Add NADE tests. | 18 December 2017, 22:03:17 UTC |
| 1a01ccb | Adam Roberts | 13 December 2017, 22:01:26 UTC | Version bump for ephemeral play support. (#981) | 13 December 2017, 22:01:26 UTC |
| e711cfa | Adam Roberts | 13 December 2017, 21:50:05 UTC | Add ephemeral play support for Colab. (#980) * Add ephemeral playback to notebook_utils for colab * lint * lint * lint * lint * Add python3 support for stringio * Import cStringIO only in Colab since it is not Python3 compatible * Switch to BytesIO for py2/3 compatibility. | 13 December 2017, 21:50:05 UTC |
| 1153b71 | Jesse Engel | 13 December 2017, 01:30:37 UTC | NSynth multigpu generate (#952) * fix bug for folder full of .npy files (array -> vstack). Allow generation with multiple gpus. * Only use one gpu at a time. Allow memory growth * lint * lint | 13 December 2017, 01:30:37 UTC |
| 28f268a | Curtis "Fjord" Hawthorne | 13 December 2017, 01:16:09 UTC | Fix markdown typo (#979) | 13 December 2017, 01:16:09 UTC |
| 1ba8e2f | Chen Shen | 13 December 2017, 01:00:38 UTC | modify Polyphony_RNN README.txt (#963) * modify README.txt * Update README.md * Update README.md | 13 December 2017, 01:00:38 UTC |
| 7c29d79 | Curtis "Fjord" Hawthorne | 13 December 2017, 00:46:33 UTC | Add ability to decode sequences with different lengths (#978) * fix * fix * fix | 13 December 2017, 00:46:33 UTC |
| 1757477 | Ian Simon | 13 December 2017, 00:33:27 UTC | refactor RNN graph building to support distributed training (#972) * refactor RNN graph building to support distributed training * rename events_rnn graph-building function for clarity | 13 December 2017, 00:33:27 UTC |
| e70f262 | Adam Roberts | 12 December 2017, 20:45:29 UTC | Add merge_hparams to common. (#977) | 12 December 2017, 20:45:29 UTC |
| c6ee6d2 | Adam Roberts | 12 December 2017, 18:21:03 UTC | Fix support for internal datasets. (#976) * Add support for internal datasets. * Add support for internal datasets. * lint | 12 December 2017, 18:21:03 UTC |
| 6340958 | Curtis "Fjord" Hawthorne | 08 December 2017, 17:18:01 UTC | Add link to audio examples (#975) llvmlite error is unrelated to change, so will submit anyway. | 08 December 2017, 17:18:01 UTC |
| c23fe08 | Curtis "Fjord" Hawthorne | 08 December 2017, 02:13:12 UTC | Add colab link to Onsets and Frames (#974) | 08 December 2017, 02:13:12 UTC |
| d0ddc88 | Adam Roberts | 07 December 2017, 23:17:33 UTC | Add MusicVAE README. (#973) * Add MusicVAE readme * Add MusicVAE README * Add md extension to README * Add Colab link | 07 December 2017, 23:17:33 UTC |
| c06d695 | Curtis "Fjord" Hawthorne | 06 December 2017, 16:46:29 UTC | Switch Onsets and Frames to use Dataset (#971) * first step * second step * fixes * lint * lint | 06 December 2017, 16:46:29 UTC |
| 243c16e | Adam Roberts | 06 December 2017, 16:39:21 UTC | Fix MusicVAE config names for colab compatibility and bump version. (#969) * Fix config names * Bump to 0.3.4 | 06 December 2017, 16:39:21 UTC |
| b01da24 | Adam Roberts | 06 December 2017, 08:11:19 UTC | Version bump to 0.3.3 (#968) | 06 December 2017, 08:11:19 UTC |
| a91c7df | Adam Roberts | 06 December 2017, 05:47:25 UTC | Add MusicVAE to models. (#966) * Add MusicVAE model. * lint * lint * lint * lint * lint * lint * lint * lint * lint * lint * lint * Fix failing data_test for python3 * Address fjord review * lint * lint * lint * Fix data_test * Update configs * lint | 06 December 2017, 05:47:25 UTC |
| b71a414 | Curtis "Fjord" Hawthorne | 05 December 2017, 21:39:26 UTC | Add new ways of running Performance RNN (#967) | 05 December 2017, 21:39:26 UTC |
| 1dba37b | Stepan Parunashvili | 27 November 2017, 21:19:29 UTC | include hparams in generate example (#959) | 27 November 2017, 21:19:29 UTC |
| 631ce4f | Curtis "Fjord" Hawthorne | 22 November 2017, 00:13:00 UTC | audio_io: add crop_samples, remove unused exception (#954) * add crop_samples, remove unused exception * lint | 22 November 2017, 00:13:00 UTC |
| b434291 | Adam Roberts | 14 November 2017, 16:34:06 UTC | Switch leading NOTE_OFFs to NO_EVENTs when slicing Melody EventLists. (#946) * Switch leading NOTE_OFFs to NO_EVENTs when slicing Melody EventLists. * Expand test. * Remove deprecated __getslice__ method. | 14 November 2017, 16:34:06 UTC |
| f7cd428 | Curtis "Fjord" Hawthorne | 06 November 2017, 19:07:04 UTC | Increase tensorflow version requirement to 1.4 (#942) Required for training Onsets and Frames model. | 06 November 2017, 19:07:04 UTC |
| f10c6a5 | Curtis "Fjord" Hawthorne | 03 November 2017, 00:21:52 UTC | Bump version for new pip package (#940) | 03 November 2017, 00:21:52 UTC |
| 9752638 | Curtis "Fjord" Hawthorne | 02 November 2017, 20:01:39 UTC | Onsets and Frames documentation (#939) * Update README.md * Add training instructions | 02 November 2017, 20:01:39 UTC |
| 464c48b | Curtis "Fjord" Hawthorne | 01 November 2017, 18:44:02 UTC | Add Onsets and Frames to the pip package (#938) * pip * lint * lint | 01 November 2017, 18:44:02 UTC |
| a8c47e2 | Curtis "Fjord" Hawthorne | 31 October 2017, 23:33:56 UTC | fixes for inference (#937) | 31 October 2017, 23:33:56 UTC |
| 79e93f1 | Curtis "Fjord" Hawthorne | 31 October 2017, 16:48:16 UTC | Fixes for Onsets and Frames training (#936) * Fixes for onsets and frames training. * fix * fix | 31 October 2017, 16:48:16 UTC |
| 4112def | ekelsen | 31 October 2017, 00:54:28 UTC | add create_dataset (#931) address review comments, add test fix lint errors fix naming, console entry point add record field add id field to train as well as test add Performance RNN config where conditioning can be disabled at generation time (#930) * add encoder with disable flag * minor simplification of Performance RNN conditioning code * add Performance RNN config where conditioning can be disabled at generation time * duplicate dataset (with conditioning enabled/disabled for Performance RNN with optional conditioning TranspositionPipeline can no longer ignore out-of-range notes (#929) | 31 October 2017, 00:54:28 UTC |
| 633b2c6 | Ian Simon | 31 October 2017, 00:28:02 UTC | fix lint errors (#932) | 31 October 2017, 00:28:02 UTC |
| 8b324b5 | Ian Simon | 30 October 2017, 23:21:53 UTC | TranspositionPipeline can no longer ignore out-of-range notes (#929) | 30 October 2017, 23:21:53 UTC |
| ac45b5a | Ian Simon | 30 October 2017, 22:53:56 UTC | add Performance RNN config where conditioning can be disabled at generation time (#930) * add encoder with disable flag * minor simplification of Performance RNN conditioning code * add Performance RNN config where conditioning can be disabled at generation time * duplicate dataset (with conditioning enabled/disabled for Performance RNN with optional conditioning | 30 October 2017, 22:53:56 UTC |
| f84b5c7 | golnazg | 30 October 2017, 20:30:25 UTC | update readme of arbitrary_image_stylization (#928) * update readme * minor change | 30 October 2017, 20:30:25 UTC |
| ece5880 | Curtis "Fjord" Hawthorne | 30 October 2017, 16:50:59 UTC | Add ABC parser to music API and converter script (#927) * abc fixes * fix | 30 October 2017, 16:50:59 UTC |
| 585bade | golnazg | 30 October 2017, 15:50:39 UTC | Open sourcing arbitrary style transfer method (#926) * open sourcing arbitrary image stylization * update readme * minor change * update readme. | 30 October 2017, 15:50:39 UTC |
| 7dbad22 | golnazg | 27 October 2017, 19:17:58 UTC | Fix the preprocessing of imagenet data (#925) | 27 October 2017, 19:17:58 UTC |
| 1e9ccf7 | Douglas Eck | 27 October 2017, 17:02:55 UTC | Remove deprecated resize attribute for bokeh (#919) | 27 October 2017, 17:02:55 UTC |
