Martin Kim
Martin Kim
Pros: - Same syntax across JAX, PyTorch, and NumPy (not sure about Xarray or Sparse yet) - Track shape annotations for tensors, which can make it easier to catch bugs...
Including multi-device here too
this is currently blocked by the following upstream dependencies not supporting numpy 2.0: - loompy - pymde
Hi @Rafael-Silva-Oliveira, right now it's looking like the end of the month. Regarding #2895, it's a deprecation warning, so it should be OK to use the current version of Ray...
For this, we can discuss how to use `transform_batch`. Also discuss mrVI.