Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ok

Browse Source
This commit is contained in:
tqchen 2014-11-19 19:27:04 -08:00
parent 41eac089c8
commit 26e5eae6f2
1 changed files with 10 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
multi-node/README.md
View File

@ -23,16 +23,15 @@ Design Choice
it uses an approximate histogram count algorithm, and will only examine subset of it uses an approximate histogram count algorithm, and will only examine subset of
potential split points as opposed to all split points. potential split points as opposed to all split points.
Run the distributed version
====
- The current code run in MPI enviroment, you will need to have a network filesystem,
or copy data to local file system before running the code
- ***Note*** The distributed version is still multi-threading optimized.
You should run one xgboost-mpi per node that takes most available CPU,
this will reduce the communication overhead and improve the performance.
- One way to do that is limit mpi slot in each machine to be 1, or reserve nthread processors for each process.
Usage Usage
==== ====
* [Column-based version](col-split) * The current code run in MPI enviroment, you will need to have a network filesystem,
* [Row-based version](row-split) or copy data to local file system before running the code
* ***Note*** The distributed version is still multi-threading optimized.
You should run one xgboost-mpi per node that takes most available CPU,
this will reduce the communication overhead and improve the performance.
- One way to do that is limit mpi slot in each machine to be 1, or reserve nthread processors for each process.
* Examples:
- [Column-based version](col-split)
- [Row-based version](row-split)