Repo created as a part of DIS FA 2020
The RAPIDS suite of open source software libraries gives you the freedom to execute end-to-end data science and analytics pipelines entirely on GPUs. The notebooks in this repo are tested using RAPIDS 0.17 nightly build in Titan RTX GPU and Blazing SQL. Link for using Blazing SQL: app.blazingsql.com
There are 3 files under this repo:
- K Means using SKLearn and RAPIDS
- KNN (a comparison between SKLearn and RAPIDS)
- Logistic Regression.
Steps for RAPIDS installation in a GPU: (https://rapids.ai/)
- Create a conda env for rapids
- Install rapids
- This will open the juptyer lab in browser
• Some more useful cmds:
o Hardware setup: !nvidia-smi
o CUDA version : !nvcc --version
o conda version check: conda list