Previous Versions

neon v1.9.0

Docs

neon v1.9.0 released May 3, 2017 supporting:

  • Add support for 3D deconvolution
  • Generative Adversarial Networks (GAN) implementation, and MNIST DCGAN example, following GoodFellow 2014 (http://arXiv.org/abs/1406.2661)
  • Implement Wasserstein GAN cost function and make associated API changes for GAN models
  • Add a new benchmarking script with per-layer timings
  • Add weight clipping for GDM, RMSProp, Adagrad, Adadelta and Adam optimizers
  • Make multicost an explicit choice in mnist_branch.py example
  • Enable NMS kernels to work with normalized boxes and offset
  • Fix missing links in api.rst [#366]
  • Fix docstring for –datatype option to neon [#367]
  • Fix perl shebang in maxas.py and allow for build with numpy 1.12 [#356]
  • Replace os.path.join for Windows interoperability [#351]
  • Update aeon to 0.2.7 to fix a seg fault on termination

neon v1.8.2

Docs

neon v1.8.2 released February 23, 2017 supporting:

  • Make the whale calls example stable and shuffle dataset before splitting into subsets
  • Reduce default depth in cifar_msra example to 2
  • Fix the formatting of the conv layer description
  • Fix documentation error in the video-c3d example
  • Support greyscale videos

neon v1.8.1

Docs

neon v1.8.1 released January 17, 2017 supporting:

  • Bug fix: Add dilation to object dict and assign defaults to dil_w = dil_h = 1 [#335, #336]
  • Bug fix: Prevent GPU backend from ignoring non-zero slope in Rectlinclip and change default slope to 0
  • Bug fix: Nesterov momentum was updating velocities incorrectly

neon v1.8.0

Docs

neon v1.8.0 released December 28, 2016 supporting:

  • Skip Thought Vectors (http://arxiv.org/abs/1506.06726) example
  • Dilated convolution support
  • Nesterov Accelerated Gradient option to SGD optimizer
  • MultiMetric class to allow wrapping Metric classes
  • Support for serializing and deserializing encoder-decoder models
  • Allow specifying the number of time steps to evaluate during beam search
  • A new community-contributed Docker image
  • Improved error messages when a tensor is created with an invalid shape or reshaped to an incompatible size
  • Fix bugs in MultiCost support
  • Documentation fixes [#331]

neon v1.7.0

Docs

neon v1.7.0 released November 11 2016 supporting:

  • Update Data Loader to aeon https://github.com/NervanaSystems/aeon
  • Add Neural Machine Translation model
  • Remove Fast RCNN model (use Faster RCNN model instead)
  • Remove music_genres example
  • Fix super blocking for small N with 1D conv
  • Fix update-direct conv kernel for small N
  • Add gradient clipping to Adam optimizer
  • Documentation updates and bug fixes

neon v1.6.0

Docs

neon v1.6.0 released September 21 2016 supporting:

  • Faster RCNN model
  • Sequence to Sequence container and char_rae recurrent autoencoder model
  • Reshape Layer that reshapes the input [#221]
  • Pip requirements in requirements.txt updated to latest versions [#289]
  • Remove deprecated data loaders and update docs
  • Use NEON_DATA_CACHE_DIR envvar as archive dir to store DataLoader ingested data
  • Eliminate type conversion for FP16 for CUDA compute capability >= 5.2
  • Use GEMV kernels for batch size 1
  • Alter delta buffers for nesting of merge-broadcast layers
  • Support for ncloud real-time logging
  • Add fast_style Makefile target
  • Fix Python 3 builds on Ubuntu 16.04
  • Run setup.py for sysinstall to generate version.py [#282]
  • Fix broken link in mnist docs
  • Fix conv/deconv tests for CPU execution and fix i32 data type
  • Fix for average pooling with batch size 1
  • Change default scale_min to allow random cropping if omitted
  • Fix yaml loading
  • Fix bug with image resize during injest
  • Update references to the ModelZoo and neon examples to their new locations

neon v1.5.4

Docs

neon v1.5.4 released July 15 2016 supporting:

neon v1.5.3

Docs

neon v1.5.3 released July 7 2016 supporting:

  • Bug fixes [#267]

neon v1.5.2

Docs

neon v1.5.2 released July 6 2016 supporting:

  • Bug fixes to audio loader

neon v1.5.1

Docs

neon v1.5.1 released June 30 2016 supporting:

  • Bug fixes

neon v1.5.0

Docs

neon v1.5.0 released June 29 2016 supporting:

  • Python2/Python3 compatibility [#191]
  • Support for Pascal GPUs
  • Persistent RNN kernels [#262]
  • Dataloader enhancements (audio loader with examples)
  • HDF5 file data iterator
  • Convolution kernel improvements
  • Winograd kernel for fprop/bprop and 5x5 stride 1 filters
  • API documentation improvements [#234, #244, #263]
  • Cache directory cleanup
  • Reorganization of all unit tests
  • Check for compatible shapes before doing a memcpy [#182, #183]
  • Bug fixes [#231, #241, #253, #257, #259]

neon v1.4.0

Docs

neon v1.4.0 released Apr 29 2016 supporting:

  • VGG16 based Fast R-CNN model using winograd kernels
  • new, backward compatible, generic data loader
  • C3D video loader model trained on UCF101 dataset
  • Deep Dream example
  • make conv layer printout more informative [#222]
  • fix some examples to use new arg override capability
  • improve performance for relu for small N
  • better support for arbitrary batch norm layer placement
  • documentation updates [#210, #213, #236]

neon v1.3.0

Docs

neon v1.3.0 released Mar 3 2016 supporting:

  • Winograd kernels and associated autotuning routines
  • benchmarking scripts
  • deprecation of deterministic argument for backend constructor
  • improve batch norm stability with fp16 backend
  • allow strided support for dimshuffle kernel
  • speed up zero momentum gradient descent

neon v1.2.2

Docs

neon v1.2.2 released Feb 24 2016 supporting:

  • Benchmarking enhancements
  • fast dimshuffle, transpose, other kernel speedups and refactoring
  • batch norm states fix, deterministic updates
  • example fixes for fast rcnn and conv_autoencoder
  • image decoding rescaling method fix
  • deserialization fixes for RNN’s, refactoring
  • caffe compatibility fixes
  • documentation updates

neon v1.2.1

Docs

neon v1.2.1 released Feb 15 2016 supporting:

  • New MergeSum, Colornoise layers
  • support for aspect_ratio scaling augmentation
  • updated IMDB sentiment analysis example
  • generic CSV batchwriter
  • various build and deserialization bugfixes, doc updates

neon v1.2.0

Docs

neon v1.2.0 released Jan 31 2016 supporting:

  • Kepler GPU kernel support [#80]
  • new dataloader format, updated docs [#115, #170]
  • new serialization format
  • FastRCNN implementation, ROI pooling support [#135]
  • deep residual nets implementation and example
  • expanded model zoo
  • Ticker dataset and copy, repeat copy tasks
  • autodiff transpose support [#173]
  • numerous bug fixes and documentation updates.

neon v1.1.5

Docs

neon v1.1.5 released Jan 15 2016 supporting:

  • CUDA kernels for lookuptable layer (up to 4x speedup)
  • support for determinstic Conv layer updatesa
  • LRN layer support
  • custom dataset walkthrough utilizing bAbI data
  • reduced number of threads in deep reduction EW kernels [#171]
  • additional (de)serialization routines [#106]
  • CPU tensor slicing fix
  • corrections for PrecisionRecall, MultiLabelStats [#148]
  • explicitly specify python2.7 for virtualenv [#155]
  • default to SM50 when no working GPU found [#186]
  • Add alpha to ELU activation [#164]
  • deconv callback fix [#162]
  • various documentation updates [#151, #152]

neon v1.1.4

Docs

neon v1.1.4 released Jan 4 2016 supporting:

  • Add support for bidirectional RNNs and LSTMs
  • added ELU, leaky ReLU activations
  • significantly faster GPU kernel builds (using ptx instead of cuda-c)
  • data shuffling enhancements, removal of old data loader code.
  • caffe conv, pool, dropout layer matching and compatibility flags
  • add scheduling support for RMSProp
  • callback enhancements, additional unit tests
  • documentation auditing, added links to introductory video tutorials

neon v1.1.3

Docs

neon v1.1.3 released Dec 1 2015 supporting:

  • deconvolution and weight histogram visualization examples and documentation
  • CPU convolution and pooling layer speedups (~2x faster)
  • bAbI question and answer interactive demo, dataset support.
  • various ImageLoader enhancements.
  • interactive usage improvements (shortcut Callback import, multiple Callbacks init, doc updates, single item batch size support)
  • set default verbosity level to warning
  • CIFAR10 example normalization updates
  • CUDA detection enhancements [#132]
  • only parse batch_writer arguments when used as a script, allow undefined global_mean [#137, #140]

neon v1.1.2

Docs

neon v1.1.2 released Nov 17 2015 supporting:

  • completely re-written C++ multithreaded dataloader
  • new weight initialization options for recurrent layers
  • Added deconvolution visualization support (guided backprop)
  • new bAbI question answering example network
  • Improved performance of cifar10_allcnn, word_lstm examples
  • new CUDA-C max and avg pooling kernels
  • Additional bugfixes and documentation updates

neon v1.1.1

Docs

neon v1.1.1 released Nov 6 2015 supporting:

  • Callback initialization bug fix [#127]
  • IMDB LSTM example bug fix [#130]
  • Added cuda-convnet2 style binary dropout variant
  • Added benchmark function to model (separate fprop, bprop, update timings)
  • Remove h_buffer references in lieu of outputs for recurrent layers
  • Multi-cost output buffer bugfix for inference [#131]
  • New timeseries prediction and generation example
  • Change Callback initialization to re-support named arguments. Separate out these arguments in argparser. [#128]

neon v1.1.0

Docs

neon v1.1.0 released Oct 30 2015 supporting:

  • Sentiment analysis support (LSTM lookupTable based), new IMDB example
  • Support for merge and branch layer stacks via LayerContainers * Sequential, Tree, MergeBroadcast, MergeMultiStream
  • Support for freezing layer stacks
  • Adagrad optimizer support
  • new GPU kernels for fast compounding batch norm, conv and pooling engine updates, new kernel build system and flags.
  • Modifications for Caffe support
    • conv, pooling, P/Q updates, dropout layer normalization more in-line with Caffe approach. NOTE: this breaks backwards compatibility with some strided conv/pool related models serialized using older versions of neon as the output sizes may now be different. See the FAQ for more info.
    • serialization enhancements to make caffe model import/export easier
    • use per-channel mean subtraction instead of single global. NOTE: this breaks backwards compatibility with ImgMaster saved datasets prior to this revision. To correct, please use the included update_dataset_cache.py script in the util directory.
  • Default training cost display during progress bar is now calculated on a rolling window basis rather than from the beginning of each epoch
  • Separate Layer configuration and initialization steps
  • YAML based alexnet example
  • Callback enhancements.
    • now pass args instead of having to spell out callbacks in each example
    • Changed validation callback to loss callback, validation_frequency now evaluation_frequency
    • Generic metric callback.
  • Various bug fixes
    • non-contiguous array get for GPUTensors
    • 1D slicing returns 2D matrices
    • bin/neon serialization fixes for RNNs
    • 3D conv fixes for fprop, bprop
    • batch norm inference fix
    • bias layer size fix
  • Documentation updates and improvements

neon v1.0.0

Docs

neon v1.0.0 released Sep 9 2015, a major top to bottom re-write of the codebase that features the following enhancements:

  • RNN/LSTM
    • Code is cleaner and achieves state of the art results on the Penn Tree Bank dataset using RNN/LSTM/GRU
    • Fast image captioning model (~200x faster than CPU based NeuralTalk) on flickr8k dataset
  • Basic automatic differentiation support
  • Framework for visualizations (supported via callbacks)
  • Top-down refactoring & redesign to enable quicker iteration while keeping the speedups offered by our nervanagpu kernels
    • Datasets are easier to specify
    • Backend now uses OpTrees (similar to nervanagpu) to support autodiff
    • nervanagpu merged in as a neon backend to simplify development and use
    • YAML syntax is simplified (but not backwards compatible)
    • Better documentation and wider test coverage

neon v0.9.0

Docs

neon v0.9.0 supports:

  • Hyperparameter optimization
  • Multi GPU

neon v0.8.2

Docs

neon v0.8.2 supports:

  • Integration with our cudanet fork of Alex Krizhevsky’s cuda-convnet2 library for Kepler GPU is

We will add support for previous generation GPUs, multi-GPU and hyperparameter optimization in the upcoming releases.

neon v0.8.1

Initial public release of neon.