dispatch max_pools with no indices, expose max_pools to torch namespace #19449
Conversation
torch/onnx/symbolic.py
Outdated
| padding = padding + tuple(numpy.add(padding_ceil, padding)) | ||
| else: | ||
| padding = padding * 2 | ||
| r, indices = g.op("MaxPool", input, outputs=2, |
There was a problem hiding this comment.
the second output of onnx MaxPool is optional, so in case return_indices is false we can generate onnx maxpool op with only one output.
torch/onnx/symbolic.py
Outdated
| return _unimplemented(name, "dilation") | ||
| if stride is None: | ||
| stride = kernel_size | ||
| padding = tuple(tule_fn(padding)) |
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@wanchaol has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@wanchaol has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@wanchaol has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@wanchaol has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
…ce (#19449) Summary: in functional interfaces we do boolean dispatch, but all to max_pool\*d_with_indices. This change it to emit max_pool\*d op instead when it's not necessary to expose with_indices ops to different backends (for jit). It also bind max_pool\*d to the torch namespace, which is the same behavior with avg_pool\*d Pull Request resolved: pytorch/pytorch#19449 Differential Revision: D15016839 Pulled By: wanchaol fbshipit-source-id: f77cd5f0bcd6d8534c1296d89b061023a8288a2c
…ce (pytorch#19449) Summary: in functional interfaces we do boolean dispatch, but all to max_pool\*d_with_indices. This change it to emit max_pool\*d op instead when it's not necessary to expose with_indices ops to different backends (for jit). It also bind max_pool\*d to the torch namespace, which is the same behavior with avg_pool\*d Pull Request resolved: pytorch#19449 Differential Revision: D15016839 Pulled By: wanchaol fbshipit-source-id: f77cd5f0bcd6d8534c1296d89b061023a8288a2c
3 participants
in functional interfaces we do boolean dispatch, but all to max_pool*d_with_indices. This change it to emit max_pool*d op instead when it's not necessary to expose with_indices ops to different backends (for jit).
It also bind max_pool*d to the torch namespace, which is the same behavior with avg_pool*d