0.7.0

What's Changed

• Remove dependency on torch_cluster by @RaulPPelaez in #228

Full Changelog: 0.6.0...0.7.0

0.6.0

What's Changed

• Change matplotlib to matplolib-base to reduce dependencies by @raimis in #225
• Unify torch extension compilation so that everything is compiled in setup.py by @RaulPPelaez in #227

Full Changelog: 0.5.0...0.6.0

0.5.0

What's Changed

• Make OutputModel aware of CUDA graph capturing. by @RaulPPelaez in #214
• Fix module.py by @AntonioMirarchi in #223

Full Changelog: 0.4.0...0.5.0

0.4.0

What's Changed

• Make TensorNet CUDA-graph compatible by @RaulPPelaez in #195
• Use update instead of |= so Trainer is compatible with Python 3.8 by @RaulPPelaez in #218

Full Changelog: 0.3.0...0.4.0

0.3.0

This release bumps all dependencies to be up to date with the latest conda-forge versions.
Also changes the torch extension compilation so that it happens during installation.

What's Changed

• Updating to Lightning 2.0 by @RaulPPelaez in #210
• Compile torch extension in setup.py by @RaulPPelaez in #216

Full Changelog: 0.2.7...0.3.0

Release 0.2.7

Release before updating Lightning dependency

What's Changed

• Fix neighbor extension compilation looking for non-existent backwards.cu by @RaulPPelaez in #204
• Remove dtype parameter, use previously existing "precision" instead by @RaulPPelaez in #208

Full Changelog: 0.2.6...0.2.7

Release 0.2.6

0.2.6 release