You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This release provides supports for automatic object-oriented (OO) transformations.
3
3
4
-
This release continues to add supports for improving the usability of BrainPy.
5
4
5
+
## What's new?
6
6
7
-
## New Features
8
7
8
+
1. Automatic OO transformations on longer need to take ``dyn_vars`` or ``child_objs`` information.
9
+
These transformations are capable of automatic inferring the underlying dynamical variables.
10
+
Specifically, they include:
11
+
12
+
-``brainpy.math.grad`` and other autograd functionalities
13
+
-``brainpy.math.jit``
14
+
-``brainpy.math.for_loop``
15
+
-``brainpy.math.while_loop``
16
+
-``brainpy.math.ifelse``
17
+
-``brainpy.math.cond``
9
18
10
-
1. New data structures for object-oriented transformations.
11
-
-``NodeList`` and ``NodeDict`` for a list/tuple/dict of ``BrainPyObject`` instances.
12
-
-``ListVar`` and ``DictVar`` for a list/tuple/dict of brainpy data.
13
-
2.`Clip` transformation for brainpy initializers.
14
-
3. All ``brainpylib`` operators are accessible in ``brainpy.math`` module.
15
-
4. Enable monitoring GPU models on CPU when setting ``DSRunner(..., memory_efficient=True)``. This setting can usually reduce so much memory usage.
16
-
5.``brainpylib`` wheels on the linux platform support the GPU operators. Users can install gpu version of ``brainpylib`` (require ``brainpylib>=0.1.7``) directly by ``pip install brainpylib``.
0 commit comments