v2.0.0

TorchSparse v2.0 (MLSys 2022 version).

v1.4.0

New Features

• Supported mixed-precision training and inference with torch.cuda.amp (#69, #75).
• Added generalized sparse convolution (#77).
• Added group normalization (#63).

API Changes

• sparse_{collate,quantize} now needs to be imported from torchsparse.utils.{collate,quantize}.
• sparse_quantize now takes in coords, voxel_size (defaults to 1.0), return_index (defaults to False) and return_inverse (defaults to False) as input, and returns the quantized coords as well as indices and inverse_indices (if requested).
• transpose is renamed as transposed for torchsparse.nn.Conv3d and torchsparse.nn.functional.conv3d.
• KernelRegion is removed and replaced with get_kernel_offsets (in torchsparse.nn.utils).

v1.2.0

Bump the version to 1.2.0 (#50)

v1.1.0

Added LayerNorm implementation.

v1.0.0

v1.0.0