Author: wangwei
Date: Sat Jun 13 13:49:33 2015
New Revision: 1685258
URL: http://svn.apache.org/r1685258
Log:
CMS commit to singa by wangwei
Modified:
incubator/singa/site/trunk/content/markdown/develop/schedule.md
Modified: incubator/singa/site/trunk/content/markdown/develop/schedule.md
URL:
http://svn.apache.org/viewvc/incubator/singa/site/trunk/content/markdown/develop/schedule.md?rev=1685258&r1=1685257&r2=1685258&view=diff
==============================================================================
--- incubator/singa/site/trunk/content/markdown/develop/schedule.md (original)
+++ incubator/singa/site/trunk/content/markdown/develop/schedule.md Sat Jun 13
13:49:33 2015
@@ -21,22 +21,24 @@ Notice: Licensed to the Apache Softwa
| Release | Module| Feature | Status |
|---------|---------|-------------|--------|
| 0.1 | Neural Network |1.1. Feed forward neural network, including CNN,
MLP | done|
-| | |1.2. RBM-like model, including RBM | working|
+| -Early July | |1.2. RBM-like model, including RBM | working|
| | |1.3. Recurrent neural network, including standard
RNN | working|
| | Architecture |1.4. One worker group on single node (with data
partition)| done|
-| | |1.5. Multi worker groups on single node using
[Hogwild](http://www.eecs.berkeley.edu/~brecht/papers/hogwildTR.pdf)|working|
-| | |1.6. Multi groups across nodes, like
[Downpour](http://papers.nips.cc/paper/4687-large-scale-distributed-deep-networks)|working|
-| | Resource Management |1.7. Integration with Mesos | working|
-| | Failure recovery|1.8. Checkpoint and restore |working|
+| | |1.5. Multi worker groups on single node using
[Shared Memory
Hogwild](http://www.eecs.berkeley.edu/~brecht/papers/hogwildTR.pdf)|testing|
+| | |1.6. Distributed Hogwild | working|
+| | |1.7. Multi groups across nodes, like
[Downpour](http://papers.nips.cc/paper/4687-large-scale-distributed-deep-networks)|working|
+| | |1.8 All-Reduce training architecture like
[DeepImage](http://arxiv.org/abs/1501.02876)| working|
+| | Resource Management |1.9. Integration with Mesos | working|
+| | Failure recovery|1.10. Checkpoint and restore |testing|
| | Tools|1.9. Installation with GNU auto tools| done|
|0.2 | Neural Network |2.1. Feed forward neural network, including
auto-encoders, hinge loss layers, HDFS data layers||
-| | |2.2. RBM-like model, including DBM | |
-| | |2.3. Recurrent neural network, including LSTM| |
+| July- | |2.2. RBM-like model, including DBM | |
+| End of August | |2.3. Recurrent neural network,
including LSTM| |
| | |2.4. Model partition ||
| | Communication |2.5. MPI||
| | GPU |2.6. Single GPU ||
| | |2.7. Multiple GPUs on single node||
-| | Architecture |2.8. All-Reduce training architecture like
[DeepImage](http://arxiv.org/abs/1501.02876)||
+| | Architecture |2.8. Update to support GPUs
| | Fault Tolerance|2.9. Node failure detection and recovery||
| | Binding |2.9. Python binding ||
| | User Interface |2.10. Web front-end for job submission and
performance visualization||
