Shehtab Zaman
Shehtab Zaman
- Enables distconv implementations of Scatter and Gather layers - Implements NVSHMEM based RMA kernels for scatter/gather on DiHydrogen tensors - Adds example applications in `applications/graph/DistConvGNN/synthetic` for benchmarking distributed Scatter,...
The LBANN Distconv adapter for layers mandates that only the first input tensor to distconv-enabled layer can be a non-DiHydrogen tensor. We raise an error if a tensor requires a...
Enables splitting along the 1st dimension in a 4D tensor and using distconv to perform StridedBatchedGEMMs. - Need to resolve issue #2126 prior to merging
The `Identity` layer with a parallel strategy defined crashes on the head of [develop](https://github.com/LLNL/lbann/commit/c4152b4bedf5f590b7c9c61014c307dc1b9cd12b) on Lassen with distconv enabled. LBANN is built using the heads of Hydrogen, DiHydrogen, and Aluminium...
Adds a helper function to the Python module that implements PyTorch's circular padding mode for 2D and 3D tensors.
LBANN implementation of the FAST model Includes the following: - 3DCNN model for grid-structured data - SGCNN model for graph data - New implementation of GatedGraphConv with outer Edge network...
Model implementation following [chemprop](https://chemprop.readthedocs.io/en/latest/index.html) Associated paper: [Analyzing Learned Molecular Representations for Property Prediction ](https://pubs.acs.org/doi/full/10.1021/acs.jcim.9b00237) - [x] Implemented D MPNN with dual graph - [x] Test on CSD molecules - [x]...
- [x] Data wrangling files for QM9 and ZINC - [x] Copy over transformer from `example`
Adds data-parallel semi-supervised learning examples and node property on small to medium-sized graph benchmarks from OpenGraphBenchmark. - [ ] Add GAT node property prediction mode - [ ] Add Arxiv,...