[quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu #74755

dzdang · 2022-03-25T16:50:52Z

Stack from ghstack:

-> [quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu #74755

Summary: With the introduction of structured kernel support for quantized tensors in
#74560, we are able to remove the dimension and output
resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d.
reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d
after the changes in the linked PR.
reflection_pad1d_out_quantized_cpu was implemented for quantized tensors.

This PR should introduce no functional changes.

Test plan:

python run_test.py

Differential Revision: D35148152

…-quantized reflection_pad1d_out_template Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` [ghstack-poisoned]

facebook-github-bot · 2022-03-25T16:50:58Z

🔗 Helpful links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/74755
📄 Preview Python docs built from this PR
📄 Preview C++ docs built from this PR
❓Need help or want to give feedback on the CI? Visit our office hours

💊 CI failures summary and remediations

As of commit c6a3219 (more details on the Dr. CI page):

💚 💚 Looks good so far! There are no failures yet. 💚 💚

This comment was automatically generated by Dr. CI (expand for details).

Please report bugs/suggestions to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

…zed and non-quantized reflection_pad1d_out_template" Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` [ghstack-poisoned]

…-quantized reflection_pad1d_out_template Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` ghstack-source-id: 7a9b135 Pull Request resolved: #74755

…zed and non-quantized reflection_pad1d_out_template" Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` [ghstack-poisoned]

…-quantized reflection_pad1d_out_template Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` ghstack-source-id: 03cc045 Pull Request resolved: #74755

dzdang · 2022-03-25T17:30:45Z

@dzdang has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

…zed and non-quantized reflection_pad1d_out_template" Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` Differential Revision: [D35148152](https://our.internmc.facebook.com/intern/diff/D35148152) [ghstack-poisoned]

…-quantized reflection_pad1d_out_template Summary: This PR is part of a series of PRs addressing #54150, related to using dispatcher for calls to quantized backends as opposed to if/else conditionals. This particular PR separates the quantized and non-quantized implementation for reflection_pad1d_out_template and reflection_pad1d_out_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` ghstack-source-id: 342ac73 Pull Request resolved: #74755

dzdang · 2022-03-25T20:08:16Z

@dzdang has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

jerryzh168

Looks good, thanks!

jerryzh168 · 2022-04-01T20:04:49Z

aten/src/ATen/native/ReflectionPad.cpp

+  pad_l = padding[0];
+  pad_r = padding[1];
+  nplane = input_.size(dim_plane);
+  input_w = input_.size(dim_w);
+  output_w  = input_w + pad_l + pad_r;


actually I don't feel it's a good practice to pass in reference and modify them in the function, can we just copy around this code or have a function that returns these things?

does structured kernel help here as well? cc @ezyang

I have no idea what this code is trying to do, but yes, structured kernels let you precompute quantities like this so they are passed in directly to the function.

…ized_cpu and" Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template (this code already present in reflection_pad1d), as well as the implementation for reflection_pad1d_quantized_cpu. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` Differential Revision: [D35148152](https://our.internmc.facebook.com/intern/diff/D35148152) [ghstack-poisoned]

…ized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu" Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` Differential Revision: [D35148152](https://our.internmc.facebook.com/intern/diff/D35148152) [ghstack-poisoned]

…imension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` ghstack-source-id: 0fb6446 Pull Request resolved: #74755

dzdang · 2022-04-05T13:46:55Z

@dzdang has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

…ized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu" Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` Differential Revision: [D35148152](https://our.internmc.facebook.com/intern/diff/D35148152) [ghstack-poisoned]

…imension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` ghstack-source-id: 3ba98d0 Pull Request resolved: #74755

dzdang · 2022-04-05T13:59:16Z

@dzdang has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

…ized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu" Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` Differential Revision: [D35148152](https://our.internmc.facebook.com/intern/diff/D35148152) [ghstack-poisoned]

…imension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu Summary: With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test plan: ``` python run_test.py ``` ghstack-source-id: e81672a Pull Request resolved: #74755

dzdang · 2022-04-06T14:56:03Z

@dzdang has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

…imension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu (#74755) Summary: Pull Request resolved: #74755 With the introduction of structured kernel support for quantized tensors in #74560, we are able to remove the dimension and output resizing code in reflection_pad1d_out_template. This code is already present in reflection_pad1d. reflection_pad1d_quantized_cpu has also been removed as quantized tensors can now use reflection_pad1d after the changes in the linked PR. reflection_pad1d_out_quantized_cpu was implemented for quantized tensors. This PR should introduce no functional changes. Test Plan: ``` python run_test.py ``` ``` python run_test.py ``` Differential Revision: D35148152 D35148152 Reviewed By: mruberry Pulled By: dzdang fbshipit-source-id: aa41e0704655f9b97ca6032159662ea86609a964

facebook-github-bot added the cla signed label Mar 25, 2022

dzdang requested review from jerry39213gh and jerryzh168 and removed request for jerry39213gh March 25, 2022 17:35

dzdang requested a review from bdhirsh as a code owner March 25, 2022 20:05

bdhirsh removed their request for review March 25, 2022 20:06

dzdang added release notes: quantization release notes category topic: improvements topic category labels Mar 28, 2022

jerryzh168 approved these changes Apr 1, 2022

View reviewed changes

jerryzh168 reviewed Apr 1, 2022

View reviewed changes

dzdang changed the title ~~[quant][improvements] Separated implementations for quantized and non-quantized reflection_pad1d_out_template~~ [quant][core][improvements] Removed reflection_pad1d_quantized_cpu and Apr 5, 2022

dzdang changed the title ~~[quant][core][improvements] Removed reflection_pad1d_quantized_cpu and~~ [quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu Apr 5, 2022

pytorchmergebot closed this in 6e59067 Apr 7, 2022

facebook-github-bot deleted the gh/dzdang/63/head branch April 10, 2022 14:17

WBobby mentioned this pull request Aug 17, 2022

Add ROCm5.2.3/AMDGPU support for PyTorch WBobby/pytorch#2

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu #74755

[quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu #74755

Uh oh!

dzdang commented Mar 25, 2022 •

edited

Loading

Uh oh!

facebook-github-bot commented Mar 25, 2022 •

edited

Loading

Uh oh!

dzdang commented Mar 25, 2022

Uh oh!

dzdang commented Mar 25, 2022

Uh oh!

jerryzh168 left a comment

Uh oh!

jerryzh168 Apr 1, 2022

Uh oh!

jerryzh168 Apr 1, 2022 •

edited

Loading

Uh oh!

ezyang Apr 1, 2022

Uh oh!

dzdang commented Apr 5, 2022

Uh oh!

dzdang commented Apr 5, 2022

Uh oh!

dzdang commented Apr 6, 2022

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

[quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu #74755

[quant][core][improvements] Removed reflection_pad1d_quantized_cpu, dimension and output resizing code in reflection_pad1d_out_template and implemented reflection_pad1d_out_quantized_cpu #74755

Uh oh!

Conversation

dzdang commented Mar 25, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

facebook-github-bot commented Mar 25, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful links

💊 CI failures summary and remediations

Uh oh!

dzdang commented Mar 25, 2022

Uh oh!

dzdang commented Mar 25, 2022

Uh oh!

jerryzh168 left a comment

Choose a reason for hiding this comment

Uh oh!

jerryzh168 Apr 1, 2022

Choose a reason for hiding this comment

Uh oh!

jerryzh168 Apr 1, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ezyang Apr 1, 2022

Choose a reason for hiding this comment

Uh oh!

dzdang commented Apr 5, 2022

Uh oh!

dzdang commented Apr 5, 2022

Uh oh!

dzdang commented Apr 6, 2022

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

dzdang commented Mar 25, 2022 •

edited

Loading

facebook-github-bot commented Mar 25, 2022 •

edited

Loading

jerryzh168 Apr 1, 2022 •

edited

Loading