Machine Learning Model Server on Redis¶
RedisML is a Redis module that implements several machine learning models as Redis data types.
The stored models are fully operational and support the prediction/evaluation process.
RedisML is a turnkey solution for using trained models in a production environment. Load ML models from any platform, immediately ready to serve.
Primary Features¶
- Decision Tree ensembles (random forests) classification and regression
- Linear regression
- Logistic regression
- Matrix operations
Building and Running¶
-
Build a Redis server with support for modules (currently available from the unstable branch).
-
You'll also need a BLAS library such as ATLAS. To install ATLAS:
-
Ubuntu:
sh sudo apt-get install libatlas-base-dev
-
CentOS/RHEL/Fedora:
sh sudo yum install -y atlas-devel atlas-static ln -s /usr/lib64/atlas/libatlas.a /usr/lib64/libatlas.a ln -s /usr/lib64/atlas/libtatlas.so /usr/lib64/libcblas.a
-
Build the RedisML module:
sh
git clone https://github.com/RedisLabsModules/redisml.git
cd redisml/src
make
- To load the module, start Redis with the
--loadmodule /path/to/redisml/src/redis-ml.so
option, add it as a directive to the configuration file or send aMODULE LOAD
command.
Contributing¶
Issue reports, pull and feature requests are welcome.
License¶
Redis Source Available License Agreement - see LICENSE