Be notified of new releases
Create your free GitHub account today to subscribe to this repository for new releases and build software alongside 31 million developers.Sign up
- Class weights are now taken into account by
In previous versions the weights were ignored.
random-strengthfor pairwise training (
YetiRankPairwise) is not supported anymore.
- Simultaneous use of
MultiClassOneVsAllmetrics is now deprecated.
cvmethod is now supported on GPU.
- String labels for classes are supported in Python.
In multiclassification the string class names are inferred from the data.
In binary classification for using string labels you should employ
class_namesparameter and specify which class is negative (0) and which is positive (1).
You can also use
class_namesin multiclassification mode to pass all possible class names to the fit function.
- Borders can now be saved and reused.
To save the feature quantization information obtained during training data preprocessing into a text file use cli option
To use the borders for training use cli option
This functionanlity is now supported on CPU and GPU (it was GPU-only in previous versions).
File format for the borders is described here.
- CLI option
--eval-fileis now supported on GPU.
- Some cases in binary classification are fixed where training could diverge
- A great speedup of the Python applier (10x)
- Reduced memory consumption in Python
cvfunction (times fold count)
Benchmarks and tutorials:
- Added speed benchmarks for CPU and GPU on a variety of different datasets.
- Added benchmarks of different ranking modes. In this tutorial we compare different ranking modes in CatBoost, XGBoost and LightGBM.
- Added tutorial for applying model in Java.
- Added benchmarks of SHAP values calculation for CatBoost, XGBoost and LightGBM.
The benchmarks also contain explanation of complexity of this calculation in all the libraries.
We also made a list of stability improvements and stricter checks of input data and parameters.