RAPIDS AI

TL;DR — With the release of XGBoost 1.3 comes an exciting new feature for model interpretability — GPU accelerated SHAP values. SHAP…

Covers setup of remote repositories on Artifactory 7.x and RAPIDS conda installation

Get an 11x speed-up by switching your streaming inference pipeline to RAPIDS cuStreamz and FIL.

All the components required to train, record, and register GPU accelerated machine learning models on GKE

Checkout the latest features and integrations in the RAPIDS ecosystem

A highly-efficient multi-GPU backend for scaling Recommender Pipelines

A distributed multi-node, multi-GPU implementation now available

New Accelerated Force Atlas 2 API processes and visualizes graphs with millions of vertices and edges in seconds

Perform every step of complex ML pipelines on the GPU, resulting in a significant speedup for feature engineering tasks

Achieve double the streaming throughput, process 2x the data in the same time, using less than half number of processes/Dask workers.

Here is the notebook that runs streaming word count end-to-end on GPUs in a distributed mode using RAPIDS cuStreamz and Dask.

New GPU subword tokenizer and integration with Python streaming library, Streamz

DGX-A100 640GB Systems & BlazingSQL provide Big Value In a Small Space

Learn how to use the ARIMA model to accelerate time series forecasting.

Prefect’s tight integration with Dask and flexible API makes it extremely easy to use with RAPIDS

Combining Optuna and RAPIDS libraries can help run experiments faster yielding better performing models.

TPOT, one of Python’s most popular Automated Machine Learning libraries, is now GPU-accelerated with RAPIDS cuML and DMLC XGBoost.

Learn about the RAPIDS 0.16 release. More scalability, functionality, integrations across our ETL, ML, and Graph Analytics libs in RAPIDS

RAPIDS and Dask make it easy to load larger than memory datasets. Learn how with examples.

RAPIDS cuGraph adds NetworkX Graph and DiGraph objects as valid input data types for graph algorithms

12x speedup in wall clock time and 4.5x reduction in cost when comparing GPU to CPU running HPO jobs in SageMaker.

It’s never been easier to get started with GPUs and RAPIDS, and with Tailscale, you can kick off RAPIDS workflows from anywhere.

Our GPU-optimized pipeline enabled us to quickly iterate and run many experiments in a short time, giving us a competitive advantage to…

Learn how to implement a fast TargetEncoder on GPUs with built-in optimizations and advanced features