Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Make fbgemm::jagged_index_select pt2_compliant #2170

Closed
wants to merge 5 commits into from

Commits on Nov 29, 2023

  1. Make fbgemm::masked_select_jagged_1d pt2_compliant (pytorch#2174)

    Summary:
    
    It needed an abstract impl.
    
    Reviewed By: williamwen42
    
    Differential Revision: D51647394
    zou3519 authored and facebook-github-bot committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    ad84da0 View commit details
    Browse the repository at this point in the history
  2. Add generate_opcheck_tests to input_combine_test (pytorch#2173)

    Summary:
    
    See title.
    
    Reviewed By: williamwen42
    
    Differential Revision: D51647393
    zou3519 authored and facebook-github-bot committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    67a01e1 View commit details
    Browse the repository at this point in the history
  3. Make fbgemm::tbe_input_combine pt2_compliant (pytorch#2172)

    Summary:
    
    - Added abstract impl in sparse_ops.py. I don't think it's worth splitting up
      the abstract impls into multiple .py files right now, unless someone comes to
      us with memory issues.
    
    Reviewed By: williamwen42
    
    Differential Revision: D51647392
    zou3519 authored and facebook-github-bot committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    c132d6e View commit details
    Browse the repository at this point in the history
  4. Mark some more ops as pt2_compliant_tag (pytorch#2171)

    Summary:
    
    Also deleted two skips that were marked as flaky. Those don't appear to
    actually be flaky.
    
    Reviewed By: williamwen42
    
    Differential Revision: D51647391
    zou3519 authored and facebook-github-bot committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    010a414 View commit details
    Browse the repository at this point in the history
  5. Make fbgemm::jagged_index_select pt2_compliant (pytorch#2170)

    Summary:
    
    - The problem with the original op was that in the Autograd implementation, it
      needed to call Tensor.item(). This doesn't work with FakeTensors (maybe it
      can some day in the future).
    - We create two new ops, `jagged_index_select_2d_forward_v2` and
      `jagged_index_add_2d_forward_v2` (which is effectively the backward) that do
      the Tensor.item() calls, and change fbgemm::jagged_index_select's Autograd
      implementation to call those.
    - We add abstract impls for those two new ops.
    - Finally, we move the fbgemm::jagged_index_select implementation to
      CompositeImplicitAutograd (and delete the CPU/CUDA impls, because those are
      redundant).
    
    Differential Revision: D51670069
    zou3519 authored and facebook-github-bot committed Nov 29, 2023
    Configuration menu
    Copy the full SHA
    a5e0137 View commit details
    Browse the repository at this point in the history