Author: wangsh
Date: Thu May 28 09:22:17 2015
New Revision: 1682180
URL: http://svn.apache.org/r1682180
Log:
update quick-start page
Modified:
incubator/singa/site/trunk/content/markdown/quick-start.md
incubator/singa/site/trunk/content/resources/files/mm_opensource.zip
Modified: incubator/singa/site/trunk/content/markdown/quick-start.md
URL:
http://svn.apache.org/viewvc/incubator/singa/site/trunk/content/markdown/quick-start.md?rev=1682180&r1=1682179&r2=1682180&view=diff
==============================================================================
--- incubator/singa/site/trunk/content/markdown/quick-start.md (original)
+++ incubator/singa/site/trunk/content/markdown/quick-start.md Thu May 28
09:22:17 2015
@@ -2,6 +2,13 @@
___
+### Notice!
+
+The newest code has dependency on zookeeper. Please install the zookeeper by
+
+ cd thirdparty
+ ./install.sh zookeeper
+
### Installation
Clone the SINGA code from [Github](https://github.com/apache/incubator-singa)
@@ -19,8 +26,7 @@ Compile SINGA:
If there are dependent libraries missing, please refer to
[installation](docs/installation.html) page
-for guidance on installing them. After successful compilation, the libsinga.so
-and singa executable will be built into the build folder.
+for guidance on installing them.
### Run in standalone mode
@@ -55,8 +61,8 @@ model configuration file (*model.conf*)
training data shard, test data shard and the mean file.-->
Since all modules used for training this CNN model are provided by SINGA as
-built-in modules, there is no need to write any code. Instead, you just run the
-executable file (*../../build/singa*) by providing the model configuration file
+built-in modules, there is no need to write any code. Instead, you just
+executable the running script (*../../bin/singa-run.sh*) by providing the
model configuration file
(*model.conf*). If you want to implement your own modules, e.g., layer,
then you have to register your modules in the driver code. After compiling the
driver code, link it with the SINGA library to generate the executable. More
@@ -81,11 +87,35 @@ described in the [System Architecture](d
Start the training by running:
#goto top level folder
- cd ..
- ./singa -model=examples/cifar10/model.conf
-cluster=examples/cifar10/cluster.conf
+ cd ../..
+ ./bin/singa-run.sh -model=examples/cifar10/model.conf
-cluster=examples/cifar10/cluster.conf
##### Training with data Partitioning
+There are two cases for data partition:
+
+* partition the dataset among worker groups such that one worker group is
+ assigned one partition. Groups run asynchronously.
+
+* partition the neural network among workers within one group. Each layer is
+sliced such that every worker is assigned one sliced layer. The sliced layer is
+the same as the original layer except that it only has B/g feature instances,
+where B is the size of instances in a mini-batch, g is the number of workers in
+a group. All workers run synchronously.
+
+To run the second case with 2 workers, just change the cluster.conf as:
+
+ nworker_groups: 1
+ nserver_groups: 1
+ nservers_per_group: 1
+ nworkers_per_group: 2
+ nworkers_per_procs: 2
+ workspace: "examples/cifar10/"
+
+All other settings are the same as running without partitioning
+
+ ./bin/singa-run.sh -model=examples/cifar10/model.conf
-cluster=examples/cifar10/cluster.conf
+
##### Training with model Partitioning
#### Training in a cluster
Modified: incubator/singa/site/trunk/content/resources/files/mm_opensource.zip
URL:
http://svn.apache.org/viewvc/incubator/singa/site/trunk/content/resources/files/mm_opensource.zip?rev=1682180&r1=1682179&r2=1682180&view=diff
==============================================================================
Binary files - no diff available.
