Fix Implicit ALS matrix zero assignment size on GPU (#228)

Fixed ValueError for `ImplicitALSWrapperModel` with GPU

Author: chezou

CHANGELOG.md

@@ -22,6 +22,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Support `fit_partial()` for LightFM ([#223](https://github.com/MobileTeleSystems/RecTools/pull/223))
 - LightFM Python 3.12+ support ([#224](https://github.com/MobileTeleSystems/RecTools/pull/224))
 
+### Fixed
+- Fix Implicit ALS matrix zero assignment size ([#228](https://github.com/MobileTeleSystems/RecTools/pull/228))
+
 ### Removed
 - Python 3.8 support ([#222](https://github.com/MobileTeleSystems/RecTools/pull/222))
 

rectools/models/implicit_als.py

@@ -610,8 +610,8 @@ def _fit_combined_factors_on_gpu_inplace(
     model._item_norms_host = model._user_norms_host = None  # pylint: disable=protected-access
     model._YtY = model._XtX = None  # pylint: disable=protected-access
 
-    _YtY = implicit.gpu.Matrix.zeros(model.factors, model.factors)
-    _XtX = implicit.gpu.Matrix.zeros(model.factors, model.factors)
+    _YtY = implicit.gpu.Matrix.zeros(*item_factors.shape)
+    _XtX = implicit.gpu.Matrix.zeros(*user_factors.shape)
 
     for _ in tqdm(range(iterations), disable=verbose == 0):