[ET-VK][conv1d] Route conv1d to height-packed implementations in export pipeline#18334
Open
SS-JIA wants to merge 3 commits intogh/SS-JIA/496/basefrom
Open
[ET-VK][conv1d] Route conv1d to height-packed implementations in export pipeline#18334SS-JIA wants to merge 3 commits intogh/SS-JIA/496/basefrom
SS-JIA wants to merge 3 commits intogh/SS-JIA/496/basefrom
Conversation
…rt pipeline Integrate the new height-packed conv1d_pw and conv1d_dw operators into the aten.convolution.default dispatch path so they are automatically used during model export. In op_registry.py, add a pick_conv_storage function that inspects the convolution node at partition time. For 1D convolutions where the op is pointwise (kernel_size=1) or depthwise (groups=C_in) and channels are 4-aligned, it selects HEIGHT_PACKED_TEXTURE for input/output instead of the default CHANNELS_PACKED_TEXTURE. All other cases (conv2d, grouped conv1d with K>1, unaligned channels) retain channels-packed behavior. In Convolution.cpp, add a height-packed routing block at the top of the conv1d path. When the input tensor is height-packed, it dispatches to et_vk.conv1d_pw.default or et_vk.conv1d_dw.default via VK_GET_OP_FN. Falls through to the existing channels-packed add_conv1d_node path otherwise. Differential Revision: [D97344090](https://our.internmc.facebook.com/intern/diff/D97344090/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18334
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 101 PendingAs of commit cc8a798 with merge base 9076110 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
This PR needs a
|
added 2 commits
March 19, 2026 15:48
…ons in export pipeline" Integrate the new height-packed conv1d_pw and conv1d_dw operators into the aten.convolution.default dispatch path so they are automatically used during model export. In op_registry.py, add a pick_conv_storage function that inspects the convolution node at partition time. For 1D convolutions where the op is pointwise (kernel_size=1) or depthwise (groups=C_in) and channels are 4-aligned, it selects HEIGHT_PACKED_TEXTURE for input/output instead of the default CHANNELS_PACKED_TEXTURE. All other cases (conv2d, grouped conv1d with K>1, unaligned channels) retain channels-packed behavior. In Convolution.cpp, add a height-packed routing block at the top of the conv1d path. When the input tensor is height-packed, it dispatches to et_vk.conv1d_pw.default or et_vk.conv1d_dw.default via VK_GET_OP_FN. Falls through to the existing channels-packed add_conv1d_node path otherwise. Differential Revision: [D97344090](https://our.internmc.facebook.com/intern/diff/D97344090/) [ghstack-poisoned]
…ons in export pipeline" Integrate the new height-packed conv1d_pw and conv1d_dw operators into the aten.convolution.default dispatch path so they are automatically used during model export. In op_registry.py, add a pick_conv_storage function that inspects the convolution node at partition time. For 1D convolutions where the op is pointwise (kernel_size=1) or depthwise (groups=C_in) and channels are 4-aligned, it selects HEIGHT_PACKED_TEXTURE for input/output instead of the default CHANNELS_PACKED_TEXTURE. All other cases (conv2d, grouped conv1d with K>1, unaligned channels) retain channels-packed behavior. In Convolution.cpp, add a height-packed routing block at the top of the conv1d path. When the input tensor is height-packed, it dispatches to et_vk.conv1d_pw.default or et_vk.conv1d_dw.default via VK_GET_OP_FN. Falls through to the existing channels-packed add_conv1d_node path otherwise. Differential Revision: [D97344090](https://our.internmc.facebook.com/intern/diff/D97344090/) [ghstack-poisoned]
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
Integrate the new height-packed conv1d_pw and conv1d_dw operators into the
aten.convolution.default dispatch path so they are automatically used during
model export.
In op_registry.py, add a pick_conv_storage function that inspects the
convolution node at partition time. For 1D convolutions where the op is
pointwise (kernel_size=1) or depthwise (groups=C_in) and channels are 4-aligned,
it selects HEIGHT_PACKED_TEXTURE for input/output instead of the default
CHANNELS_PACKED_TEXTURE. All other cases (conv2d, grouped conv1d with K>1,
unaligned channels) retain channels-packed behavior.
In Convolution.cpp, add a height-packed routing block at the top of the conv1d
path. When the input tensor is height-packed, it dispatches to
et_vk.conv1d_pw.default or et_vk.conv1d_dw.default via VK_GET_OP_FN. Falls
through to the existing channels-packed add_conv1d_node path otherwise.
Differential Revision: D97344090