Tristan Bilot
Tristan Bilot
## Proposed changes This PR proposes a draft for the [MPNN](https://proceedings.mlr.press/v70/gilmer17a/gilmer17a.pdf) architecture, used as a basic block to build almost any GNNs using message passing paradigm. This abstract architecture is...
**Describe the bug** When applying cumsum to an int64 array, there's a kernel error. **To Reproduce** ```python import mlx.core as mx a = mx.array([0, 1, 2], dtype=mx.int64) mx.cumsum(a) >>> libc++abi:...
This function is still not implemented. * Do you know an efficient workaround until the feature is implemented? * Regarding the implementation, my first intuition is that we need a...