Update export and build from source docs #10807
Open
+78
−5
Conversation
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
GregoryComer
added
module: doc
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/10807
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New Failure, 1 Unrelated FailureAs of commit 61c0835 with merge base b173722 ( NEW FAILURE - The following job has failed:
FLAKY - The following job failed but was likely due to flakiness present on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
|
@JacobSzwejbka - can you take a look at the new content? |
JacobSzwejbka
requested changes
May 12, 2025
|
|
||
| ### Dynamic (Data-Dependent) Control Flow | ||
|
|
||
| Control flow is considered data-dependent if the path taken is not fixed at export-time. This is commonly the case when if or loop conditions depend on the value of a Tensor, such as a generator loop that terminates when an |
There was a problem hiding this comment.
Control flow is considered data-dependent if the path taken is not fixed at export-time.
This isn't true.
Its data dependent if its dependent on the data contained within the tensor rather than its shape. But shape based control flow can still be dynamic.
| At runtime, the method name can be passed to `load_method` and `execute` on the `Module` class. | ||
|
|
||
| Multi-method .ptes have several caveats: | ||
| - Methods are individually memory-planned. Activation memory is not current re-used between methods. |
There was a problem hiding this comment.
Maybe worth mentioning you could write a custom plan to change this, or share the same arenas at runtime.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Summary
This PR makes some small updates to the docs in a few places. In the Build from Source page, I've added an explicit command to generate a test model (add.pte) to use with the executor_runner.
In the Exporting and Lowering page, I've added some details on advanced export and lowering techniques, including state management, dynamic control flow, and multi-method .ptes.
Test plan
I've built and viewed the docs locally to validate the changes.
cc @mergennachin @byjlw