Fix split module interaction with dead code #104554

benghaem · 2023-07-03T21:47:50Z

Summary:
This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs.

This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function

Unit Test Added:
test_split_module_dead_code

A module with dead code:

class ModWithDeadCode(torch.nn.Module):
            def forward(self, x):
                output = x * 2 # we want this
                dead_line = x + 2 # this is dead
                return output

Before:

torch/fx/passes/split_module.py, line 357, in split_module
base_mod_env[list(partition.outputs)[0]] = output_val
IndexError: list index out of range

After:

class GraphModule(torch.nn.Module):
    def forward(self, x):
        # No stacktrace found for following nodes
        submod_2 = self.submod_2(x)
        submod_1 = self.submod_1(x);  x = None
        return submod_1

    class GraphModule(torch.nn.Module):
        def forward(self, x):
            # No stacktrace found for following nodes
            add = x + 2;  x = None
            return None

    class GraphModule(torch.nn.Module):
        def forward(self, x):
            # No stacktrace found for following nodes
            mul = x * 2;  x = None
            return mul

Submod 2 is correctly extracted

Test Plan: Tested with new unit test

Differential Revision: D47196732

pytorch-bot · 2023-07-03T21:47:52Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/104554

📄 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 the bot commands wiki or our office hours

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 02c3a48:
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

linux-foundation-easycla · 2023-07-03T21:47:53Z

The committers listed above are authorized under a signed CLA.

✅ login: benghaem / name: Benjamin Ghaemmaghami (02c3a48)

facebook-github-bot · 2023-07-03T21:48:39Z

This pull request was exported from Phabricator. Differential Revision: D47196732

facebook-github-bot · 2023-07-04T00:01:45Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 2d9d22f0432fb762ef024a44c0ddbf16804928ec

facebook-github-bot · 2023-07-12T18:46:37Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: b9b06ee8b2eaf0cc236e0f5a00e17d42f9971499

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 228e56d9378978bc3463069ca8ccd2a9bc31df58

facebook-github-bot · 2023-07-17T17:22:39Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: e15ebbfbdaa3d9a9f9b43f8038e9c2f57063e687

facebook-github-bot · 2023-07-17T17:30:04Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: fecfba49f8c89b3cdf6007add53119adcaf0cf93

facebook-github-bot · 2023-07-20T17:19:15Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 6bbed6778ea24fd0cc2f1127fd4d5762c8d3c30b

facebook-github-bot · 2023-08-01T21:26:00Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 86a6ab33d292735d07b0efd25dc93e0763f4b552

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 0e9c852e8fe5ac5fa6597f3fa81bde71c56bae82

facebook-github-bot · 2023-08-01T21:34:29Z

This pull request was exported from Phabricator. Differential Revision: D47196732

facebook-github-bot · 2023-08-02T06:53:43Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 81bb570788a4baf6ba8953688d663dfc8fdd8d4d

facebook-github-bot · 2023-08-02T17:54:58Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: fd39dc68e612e78cd3945a313154fdbf32bb7383

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal Differential Revision: D47196732 fbshipit-source-id: 0503440f9941edf577689b08164cbd1a2b1881d6

facebook-github-bot · 2023-08-02T21:33:38Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal, RenfeiChen-FB Differential Revision: D47196732 fbshipit-source-id: e03013ef6a09c09981a2702c584a97909ea2024b

facebook-github-bot · 2023-08-03T04:45:40Z

This pull request was exported from Phabricator. Differential Revision: D47196732

Summary: Pull Request resolved: pytorch#104554 This change fixes split_module's interaction with dead code. Previously if a dead region was split out, split module would throw an error while attempting to access the outputs for the partition even though the partition has no outputs. This change adds a new unit test to cover the dead code case and changes the output check to allow no output. The split module with no output will now output None like a normal python function Unit Test Added: test_split_module_dead_code A module with dead code: ``` class ModWithDeadCode(torch.nn.Module): def forward(self, x): output = x * 2 # we want this dead_line = x + 2 # this is dead return output ``` Before: ``` torch/fx/passes/split_module.py, line 357, in split_module base_mod_env[list(partition.outputs)[0]] = output_val IndexError: list index out of range ``` After: ``` class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes submod_2 = self.submod_2(x) submod_1 = self.submod_1(x); x = None return submod_1 class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes add = x + 2; x = None return None class GraphModule(torch.nn.Module): def forward(self, x): # No stacktrace found for following nodes mul = x * 2; x = None return mul ``` Submod 2 is correctly extracted Test Plan: Tested with new unit test Reviewed By: mustafaozdal, RenfeiChen-FB Differential Revision: D47196732 fbshipit-source-id: 5c937491d2db8e9c8de9f1e00b9fe5baae920528

facebook-github-bot · 2023-08-03T16:24:00Z

This pull request was exported from Phabricator. Differential Revision: D47196732

facebook-github-bot · 2023-08-03T21:34:34Z

@pytorchbot merge

(Initiating merge automatically since Phabricator Diff has merged)

pytorchmergebot · 2023-08-03T21:36:29Z

Merge started

Your change will be merged once all checks pass (ETA 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here

pytorch-bot bot added the release notes: fx release notes category label Jul 3, 2023

facebook-github-bot added the fb-exported label Jul 3, 2023

benghaem force-pushed the export-D47196732 branch from 9ca440f to 3b9f2df Compare July 4, 2023 00:01

benghaem force-pushed the export-D47196732 branch from 3b9f2df to 5b73e9c Compare July 12, 2023 18:46

benghaem force-pushed the export-D47196732 branch from 5b73e9c to 995ba97 Compare July 17, 2023 16:44

benghaem force-pushed the export-D47196732 branch from 995ba97 to 2a29363 Compare July 17, 2023 17:22

benghaem force-pushed the export-D47196732 branch from 2a29363 to bf8174e Compare July 17, 2023 17:30

benghaem force-pushed the export-D47196732 branch from bf8174e to 65c1a9e Compare July 20, 2023 17:19

benghaem force-pushed the export-D47196732 branch from 65c1a9e to dc98de2 Compare August 1, 2023 21:25

benghaem force-pushed the export-D47196732 branch from dc98de2 to 54b1f11 Compare August 1, 2023 21:34

benghaem force-pushed the export-D47196732 branch from 54b1f11 to 89ff0e6 Compare August 2, 2023 06:53

benghaem force-pushed the export-D47196732 branch from 89ff0e6 to daf9a51 Compare August 2, 2023 17:54

yf225 approved these changes Aug 2, 2023

View reviewed changes

benghaem force-pushed the export-D47196732 branch from daf9a51 to c826fc3 Compare August 2, 2023 21:33

benghaem force-pushed the export-D47196732 branch from c826fc3 to 9846615 Compare August 3, 2023 04:45

benghaem force-pushed the export-D47196732 branch from 9846615 to 02c3a48 Compare August 3, 2023 16:23

pytorch-bot bot added the ciflow/trunk Trigger trunk jobs on your pull request label Aug 3, 2023

pytorchmergebot added the merging label Aug 3, 2023

pytorchmergebot added Merged and removed merging labels Aug 3, 2023

pytorchmergebot closed this in 424dc23 Aug 3, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix split module interaction with dead code #104554

Fix split module interaction with dead code #104554

benghaem commented Jul 3, 2023

pytorch-bot bot commented Jul 3, 2023 •

edited

linux-foundation-easycla bot commented Jul 3, 2023 •

edited

facebook-github-bot commented Jul 3, 2023

facebook-github-bot commented Jul 4, 2023

facebook-github-bot commented Jul 12, 2023

facebook-github-bot commented Jul 17, 2023

facebook-github-bot commented Jul 17, 2023

facebook-github-bot commented Jul 20, 2023

facebook-github-bot commented Aug 1, 2023

facebook-github-bot commented Aug 1, 2023

facebook-github-bot commented Aug 2, 2023

facebook-github-bot commented Aug 2, 2023

facebook-github-bot commented Aug 2, 2023

facebook-github-bot commented Aug 3, 2023

facebook-github-bot commented Aug 3, 2023

facebook-github-bot commented Aug 3, 2023

pytorchmergebot commented Aug 3, 2023

Fix split module interaction with dead code #104554

Fix split module interaction with dead code #104554

Conversation

benghaem commented Jul 3, 2023

pytorch-bot bot commented Jul 3, 2023 • edited

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/104554

✅ No Failures

linux-foundation-easycla bot commented Jul 3, 2023 • edited

facebook-github-bot commented Jul 3, 2023

facebook-github-bot commented Jul 4, 2023

facebook-github-bot commented Jul 12, 2023

facebook-github-bot commented Jul 17, 2023

facebook-github-bot commented Jul 17, 2023

facebook-github-bot commented Jul 20, 2023

facebook-github-bot commented Aug 1, 2023

facebook-github-bot commented Aug 1, 2023

facebook-github-bot commented Aug 2, 2023

facebook-github-bot commented Aug 2, 2023

facebook-github-bot commented Aug 2, 2023

facebook-github-bot commented Aug 3, 2023

facebook-github-bot commented Aug 3, 2023

facebook-github-bot commented Aug 3, 2023

pytorchmergebot commented Aug 3, 2023

Merge started

pytorch-bot bot commented Jul 3, 2023 •

edited

linux-foundation-easycla bot commented Jul 3, 2023 •

edited