17 lines
680 B
Markdown
17 lines
680 B
Markdown
Distributed XGBoost: Column Split Version
|
|
====
|
|
* run ```bash mushroom-row.sh <n-mpi-process>```
|
|
|
|
How to Use
|
|
====
|
|
* First split the data by column,
|
|
* In the config, specify data file as containing a wildcard %d, where %d is the rank of the node, each node will load their part of data
|
|
* Enable column split mode by ```dsplit=col```
|
|
|
|
Notes
|
|
====
|
|
* The code is multi-threaded, so you want to run one xgboost-mpi per node
|
|
* The code will work correctly as long as union of each column subset is all the columns we are interested in.
|
|
- The column subset can overlap with each other.
|
|
* It uses exactly the same algorithm as single node version, to examine all potential split points.
|