Published May 4, 2017
| Version v1.9.0
Software
Open
NervanaSystems/neon: Generative Adversarial Networks, 3D Deconvolution, doc updates and bug fixes
Creators
- Scott Leishman
- apark263
- Anil Thomas
- Jennifer Myers1
- yinyinl2
- nervetumer
- Urs Köster3
- hanlin-nervana
- Scott Gray4
- Stewart Hall
- Augustus Odena
- wconstab
- Zach Dwiel
- John Co-Reyes
- Arjun Bansal2
- Sébastien Arnold5
- Jeevan Shankar6
- Andy
- Scott Cyphers7
- Tyler Lee
- Kai Arulkumaran
- Santi Villalba
- Xin Wang8
- Steven Robertson9
- Jason Knight1
- Gabriel Pereyra10
- mikebenfield
- Dongjoon Hyun11
- bbnguyen
- aravind-nrv
- 1. @NervanaSystems
- 2. Nervana Systems
- 3. Nervana
- 4. OpenAI
- 5. University of Southern California
- 6. Facebook
- 7. Intel Nervana
- 8. Intel Corporation
- 9. Intel - Nervana
- 10. Student
- 11. @hortonworks
Description
- 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
Files
NervanaSystems/neon-v1.9.0.zip
Files
(3.2 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:a9a58a11263cfe27fe29780a3adfffcf
|
3.2 MB | Preview Download |
Additional details
Related works
- Is supplement to
- https://github.com/NervanaSystems/neon/tree/v1.9.0 (URL)