From 85253dae08b09aecbfb0a76bf17375f85b7a9687 Mon Sep 17 00:00:00 2001 From: baojun <32073718+baojun-nervana@users.noreply.github.com> Date: Mon, 27 Nov 2017 14:37:33 -0800 Subject: [PATCH] doc update for neon v2.4.0 release (#1089) * doc update for neon v2.4.0 release * update mkl info --- ChangeLog | 11 +++++++++++ README.md | 8 ++++++++ doc/source/index.rst | 15 +++++++++------ doc/source/installation.rst | 10 +++++++++- doc/source/previous_versions.rst | 13 +++++++++++++ setup.py | 6 +++--- 6 files changed, 53 insertions(+), 10 deletions(-) diff --git a/ChangeLog b/ChangeLog index 07ff121e..29293fc8 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,16 @@ # ChangeLog +## v2.4.0 (2017-11-27): + +* Enabled pip install through pypi +* Updated MKLML to version 20171007 with performance improve of ~3X for mnist datalayer/nondatalayer and ~1.6X for DCGAN/WGAN datalayer +* Updated resnet model to optimize performance with MKLML 20171007 +* Updated Alexnet weight file and fixed bug for deep dream +* Fixed faster-rcnn inference model loading issue +* Added data_loading time measurement and enabled GAN networks benchmarking +* Updated to Aeon version 1.2.0 +* Enabled neon build with mklEngine on Windows systems + ## v2.3.0 (2017-10-27): * Optimized DeepSpeech2 MKL backend performance (~7X improvement over the CPU backend) diff --git a/README.md b/README.md index 064fa0f0..357c433b 100644 --- a/README.md +++ b/README.md @@ -34,6 +34,14 @@ neon (conda users see the [guide](http://neon.nervanasys.com/docs/latest/install Starting after neon v2.2.0, the master branch of neon will be updated weekly with work-in-progress toward the next release. Check out a release tag (e.g., "git checkout v2.2.0") for a stable release. Or simply check out the "latest" release tag to get the latest stable release (i.e., "git checkout latest") +From version 2.4.0, we re-enabled pip install. Neon can be installed using package name nervananeon. + +```bash + pip install nervananeon +``` + +It is noted that [aeon](https://aeon.nervanasys.com/index.html/getting_started.html) needs to be installed separately. The latest release v2.4.0 uses aeon v1.2.0. + **Warning** > Between neon v2.1.0 and v2.2.0, the aeon manifest file format has been changed. When updating from neon < v2.2.0 manifests have to be recreated using ingest scripts (in examples folder) or updated using [this](neon/data/convert_manifest.py) script. diff --git a/doc/source/index.rst b/doc/source/index.rst index 32599396..89e32dd1 100644 --- a/doc/source/index.rst +++ b/doc/source/index.rst @@ -28,7 +28,7 @@ Features include: * Support for commonly used models including convnets, RNNs, LSTMs, and autoencoders. You can find many pre-trained implementations of these in our `model zoo`_ -* Tight integration with our `state-of-the-art`_ GPU kernel library +* Tight integration with our `state-of-the-art`_ GPU kernel library and Intel CPU MKLML library * 3s/macrobatch (3072 images) on AlexNet on Titan X (Full run on 1 GPU ~ 32 hrs) * Basic automatic differentiation support * Framework for visualization @@ -36,11 +36,14 @@ Features include: New features in this release: -* Optimized DeepSpeech2 MKL backend performance (~7X improvement over the CPU backend) -* Fused convolution and bias layer which significantly boosted AlexNet and VGG performance on Intel architectures with MKL backend -* Made SSD and Faster-RNN use VGG weight files in new format -* Fixed use of reset_cells hyperparameter -* Fixed MKL backend bug for GAN and Faster-RCNN models +* Enabled pip install through pypi +* Updated MKLML to version 20171007 with up to 3X performance increase +* Updated resnet model to optimize performance with MKLML 20171007 +* Updated Alexnet weight file and fixed bug for deep dream +* Fixed faster-rcnn inference model loading issue +* Added data_loading time measurement and enabled GAN networks benchmarking +* Updated Aeon version to 1.2.0 +* Enabled neon build with mklEngine on Windows systems * See more in the `change log`_. We use neon internally at Intel Nervana to solve our `customers' problems`_ diff --git a/doc/source/installation.rst b/doc/source/installation.rst index 24bca13c..4406e38b 100644 --- a/doc/source/installation.rst +++ b/doc/source/installation.rst @@ -63,6 +63,14 @@ Or on Mac OS X: Installation ~~~~~~~~~~~~ +Neon v2.4.0 and after is pip installable through pypi with package name nervananeon. + +.. code-block:: bash + + pip install nervananeon + +It is noted `aeon `__ needs to be installed separately. The latest release v2.4.0 uses aeon v1.2.0. + We recommend installing neon within a `virtual environment `__ to ensure a self-contained environment. To install neon within an @@ -76,7 +84,7 @@ setup neon in this manner, run the following commands: git clone https://github.com/NervanaSystems/neon.git cd neon; git checkout latest; make -The above checks out the latest stable release (e.g. a tagged release version v2.3.0) and build neon. +The above checks out the latest stable release (e.g. a tagged release version v2.4.0) and build neon. Alternatively, you can check out and build the latest master branch: .. code-block:: bash diff --git a/doc/source/previous_versions.rst b/doc/source/previous_versions.rst index 0737c221..600f781d 100644 --- a/doc/source/previous_versions.rst +++ b/doc/source/previous_versions.rst @@ -16,6 +16,17 @@ Previous Versions ================= +neon v2.3.0 +----------- + +|Docs230|_ + +* Optimized DeepSpeech2 MKL backend performance (~7X improvement over the CPU backend) +* Fused convolution and bias layer which significantly boosted AlexNet and VGG performance on Intel architectures with MKL backend +* Made SSD and Faster-RNN use VGG weight files in new format +* Fixed use of reset_cells hyperparameter +* Fixed MKL backend bug for GAN and Faster-RCNN models + neon v2.2.0 ----------- @@ -473,6 +484,7 @@ neon v0.8.1 Initial public release of neon. +.. |Docs230| replace:: Docs .. |Docs220| replace:: Docs .. |Docs200| replace:: Docs .. |Docs190| replace:: Docs @@ -501,6 +513,7 @@ Initial public release of neon. .. |Docs9| replace:: Docs .. |Docs8| replace:: Docs .. _cudanet: https://github.com/NervanaSystems/cuda-convnet2 +.. _Docs230: http://neon.nervanasys.com/docs/2.3.0 .. _Docs220: http://neon.nervanasys.com/docs/2.2.0 .. _Docs200: http://neon.nervanasys.com/docs/2.0.0 .. _Docs190: http://neon.nervanasys.com/docs/1.9.0 diff --git a/setup.py b/setup.py index eeeb722d..3d17b55e 100755 --- a/setup.py +++ b/setup.py @@ -18,7 +18,7 @@ import subprocess # Define version information -VERSION = '2.3.0' +VERSION = '2.4.0' FULLVERSION = VERSION write_version = True @@ -96,9 +96,9 @@ setup(name='nervananeon', version=VERSION, - description="Nervana's deep learning framework", + description="Intel Nervana's deep learning framework", long_description=readme_file, - author='Nervana Systems', + author='Intel Nervana Systems', author_email='intelnervana@intel.com', url='http://www.intelnervana.com', license='License :: OSI Approved :: Apache Software License',