Releases: oneapi-src/oneDAL
Intel® oneAPI Data Analytics Library 2024.4.0
oneDAL 2024.4.0
oneDAL is happy to introduce 2024.4.0 release!
🚨 What's New
- Introduced new oneDAL functionality:
- Distributed dbscan support without
allgatherv
(memory saving implementation) - Added
global_context
to store information about CPU environment - Online distributed basic statistics
- Hyperparameters added into decision forest classification inference algorithm
- Added
assume_centered
flag to Covariance - Added
dtype
dispatcher for Dataframe Intercahnge Protocol in Python
- Distributed dbscan support without
Acknowledgements
Thanks to everyone who helped us make 2024.4.0 release possible!
@avolkov-intel, @Alexsandruss, @ahuber21, @Alexandr-Solovev, @Vika-F, @razdoburdin, @ethanglaser, @napetrov, @maria-Petrova, @aepanchi, @emmwalsh, @icfaust, @inteldimitrius, @samir-nasibli, @md-shafiul-alam
Full Changelog: 2024.3.0...2024.4.0
Intel® oneAPI Data Analytics Library 2024.3.0
oneDAL 2024.3.0
oneDAL is happy to introduce 2024.3.0 release!
🚨 What's New
- Introduced new oneDAL functionality:
- k-Means sparsity support
- Online support of distributed Covariance algorithm
- Added
inner_iterations_count
result option to LogisticRegression
🪲 Bug Fixes
- Fix FPK exports on Windows
- Fix
csr_accessor
crash when moving data from USM or host memory to host memory
Acknowledgements
Thanks to everyone who helped us make 2024.3.0 release possible!
@avolkov-intel, @Alexsandruss, @ahuber21, @Alexandr-Solovev, @Vika-F, @razdoburdin, @ethanglaser, @napetrov, @aepanchi, @icfaust, @inteldimitrius, @maria-Petrova, @samir-nasibli, @md-shafiul-alam
Full Changelog: 2024.2.0...2024.3.0
Intel® oneAPI Data Analytics Library 2024.2.0
oneDAL 2024.2.0
oneDAL is happy to introduce 2024.2.0 release!
🚨 What's New
-
Introduced new oneDAL functionality:
- Optional results in k-Means inference
- GPU support of SVD method of PCA algorithm
-
Improved oneDAL performance for the following algorithms:
- k-Means
🪲 Bug Fixes
- Fix FPK exports on Windows
- Replace deprecated
sycl::abs
tosycl::fabs
Acknowledgements
Thanks to everyone who helped us make 2024.2.0 release possible!
@avolkov-intel, @Alexsandruss, @ahuber21, @Alexandr-Solovev, @Vika-F, @razdoburdin, @ethanglaser, @napetrov, @aepanchi, @icfaust, @inteldimitrius, @maria-Petrova, @samir-nasibli, @md-shafiul-alam
Full Changelog: 2024.1.0...2024.2.0
Intel® oneAPI Data Analytics Library 2024.1.0
oneDAL 2024.1.0
oneDAL is happy to introduce 2024.1.0 release!
🚨 What's New
- New oneDAL functionality:
- Enabled distributed computations for LogisticRegression algorithm
- Basic statistics algorithm for sparse data
- Added new parameters to oneDAL algorithms:
- Bias parameter to Covariance algorithm
- Improved oneDAL performance for the following algorithms:
- DBSCAN
- Distributed version of kNN
Acknowledgements
Thanks to everyone who helped us make 2024.1.0 release possible!
@avolkov-intel, @Alexsandruss, @ahuber21, @Alexandr-Solovev, @Vika-F, @razdoburdin, @ethanglaser, @napetrov, @aepanchi, @KulikovNikita, @icfaust, @inteldimitrius, @maria-Petrova, @samir-nasibli, @md-shafiul-alam
Full Changelog: 2024.0.1...2024.1.0
Intel® oneAPI Data Analytics Library 2024.0.1
oneDAL 2024.0.1
oneDAL is happy to introduce 2024.0.1 release!
🚨 What's New
- New oneDAL functionality:
- Introduced Logistic Regression algorithm
- Introduced online algorithms support: Moments of Low Order, PCA, Linear Regression
- Introduced dispatching in oneAPI covariance algorithm
- New Model Builders functionality:
- SHAP calculation is added to GBT regression
📚 Support Materials
Faster XGBoost*, LightGBM, and CatBoost Inference on the CPU
🔀 Adoption
Acknowledgements
Thanks to everyone who helped us make 2024.0.1 release possible!
@avolkov-intel, @Alexsandruss, @ahuber21, @Alexandr-Solovev, @Vika-F, @razdoburdin, @ethanglaser, @napetrov, @aepanchi, @KulikovNikita, @icfaust, @inteldimitrius, @maria-Petrova, @samir-nasibli, @md-shafiul-alam
Intel® oneAPI Data Analytics Library 2024.0.0
oneDAL 2024.0
oneDAL is happy to introduce 2024.0 release!
🚨 What's New
New oneDAL functionality:
- Online algorithms support: Covariance
- Structure of arrays (SOA) Tables
🚩 Removals and ABI Compatibility
- Starting with the 2024.0 release, the following functionality is deprecated in oneDAL:
- Compression functionality
- DAAL CPP SYCL Interfaces
- Java* interfaces
- ABI compatibility is broken as part of the 2024.0 release of oneDAL. The library’s major version is incremented to two to enforce the relinking of existing applications.
🔨 Library Engineering
- Compressed sparse rows (CSR) accessor has been changed and moved from detail namespace. The support of USM memory was added into this class.
- Centralized parameters selection library has been introduced in the release.
❌ Deprecation Notice
- The following algorithms in DAAL Interfaces are deprecated and will not be supported starting with 2025.0 release:
- k-Means
- Covariance
- PCA
- Logistic Regression
- Linear Regression
- Random Forest
- kNN
- SVM
- DBSCAN
- Low-order moments
- ABI compatibility is to be broken as part of the 2025.0 release of oneDAL. The library’s major version is to be incremented to three to enforce the relinking of existing applications
Acknowledgements
Thanks to everyone who helped us make 2024.0 release possible!
@avolkov-intel, @Alexsandruss, @ahuber21, @Alexandr-Solovev, @Vika-F, @razdoburdin, @ethanglaser, @napetrov, @aepanchi, @amgrigoriev, @KulikovNikita, @Pahandrovich, @icfaust, @inteldimitrius, @maria-Petrova, @samir-nasibli, @md-shafiul-alam,
@keeranroth, @amgrigoriev
Intel® oneAPI Data Analytics Library 2023.2.1
The release of Intel® oneAPI Data Analytics Library 2023.2.1 introduces the following changes:
🚨 What's New
- sklearn 1.3 support fixes (1, 2, 3)
- Model builders API update
Intel® oneAPI Data Analytics Library 2023.2.0
The release Intel® oneAPI Data Analytics Library 2023.2.0 introduces the following changes:
❌ Deprecation Notice
- The compression functionality in the Intel® oneDAL library is deprecated. Starting with the 2024.0 release, oneDAL will not support the compression functionality
- The DAAL CPP SYCL Interfaces in the Intel® oneDAL library are deprecated. Starting with the 2024.0 release, oneDAL will not support the DAAL CPP SYCL Interfaces
- The Java* interfaces in the Intel® oneDAL library are marked as deprecated. The future releases of the oneDAL library may no longer include support for these Java* interfaces
- ABI compatibility is to be broken as part of the 2024.0 release of Intel® oneDAL. The library’s major version is to be incremented to two to enforce the relinking of existing applications
- macOS* support is deprecated for oneDAL. The 2023.x releases are the last to provide it
🛠️ Library Engineering
- CSR tables interface has been changed and moved from detail namespace
🚨 What's New
- Introduced new Intel® oneDAL functionality:
- Distributed KMeans++ algorithm
- Logistic Loss objective algorithm
- Introduced new functionality for Intel® Extension for Scikit-learn:
- NaN(missing values) support was added to Model Builders
- Improved performance for the following Intel® Extension for Scikit-learn algorithms:
- Model Builders performance has been improved up to 2x
Intel® oneAPI Data Analytics Library 2023.1.1
The release Intel® oneAPI Data Analytics Library 2023.1.1 introduces the following changes:
🚨 What's New
Intel® oneAPI Data Analytics Library 2023.1.0
The release Intel® oneAPI Data Analytics Library 2023.1 introduces the following changes:
📚Support Materials
🛠️ Library Engineering
- Reduced the size of Intel® oneDAL library by approximately ~30%
- Enabled NuGet distribution channel for Intel® oneDAL on Linux and MacOS
🚨 What's New
- Introduced new Intel® oneDAL functionality:
- Distributed Linear Regression, kNN, PCA algorithms
- Introduced new functionality for Intel® Extension for Scikit-learn:
- Enabled PCA, Linear Regression, Random Forest algorithms and SPMD policy as preview
- Scikit-learn 1.2 support
- sklearn_is_patched() function added to validate status of algorithms patching
- Improved performance for the following Intel® Extension for Scikit-learn algorithms:
- t-SNE for “Burnes-Hut” algorithm
- SVM algorithm for single row inference
❗ Known Issues
- In certain conditions DAAL SYCL interface might hang with L0 backend – please use oneDAL DPC interfaces instead. If older interfaces are required OpenCL backend can be used as workaround.