分布式深度学习平台,Apache SINGA 0.2.0 发布

jopen 9年前


Apache SINGA 是 Apache 在 2015 年 3 月 17 日接纳的一个孵化项目,是个分布式深度学习平台。

Apache SINGA 0.2.0 发布,此版本主要更新内容如下:

  • Training on GPU enables training of complex models on a single node with multiple GPU cards.

  • Hybrid neural net partitioning supports data and model parallelism at the same time.

  • Python wrapper makes it easy to configure the job, including neural net and SGD algorithm.

  • RNN model and BPTT algorithm are implemented to support applications based on RNN models, e.g., GRU.

  • Cloud software integration includes Mesos, Docker and HDFS.

  • Visualization of neural net structure and layer information, which is helpful for debugging.

  • Linear algebra functions and random functions against Blobs and raw data pointers.

  • New layers, including SoftmaxLayer, ArgSortLayer, DummyLayer, RNN layers and cuDNN layers.

  • Update Layer class to carry multiple data/grad Blobs.

  • Extract features and test performance for new data by loading previously trained model parameters.

  • Add Store class for IO operations.

SINGA 是基于大型数据集训练大型深度学习模块的常规分布式学习平台。SINGA 支持各种流行的深度学习模块,其中的 feed-forward 模块包括 convolutional neural networks (CNN),能量模块 restricted Boltzmann machine (RBM) 和 recurrent neural networks (RNN)。

SGD 流:

Apache SINGA 0.2.0 发布,分布式深度学习平台

SINGA 概览:

Apache SINGA 0.2.0 发布,分布式深度学习平台

外部依赖:

  • glog(New BSD)

  • google-protobuf(New BSD)

  • openblas(New BSD)

  • zeromq(LGPLv3 + static link exception)

  • czmq(Mozilla Public License Version 2.0)

  • zookeeper(Apache 2.0)

来自: http://www.oschina.net//news/69931/apache-singa-0-2-0