Use TensorOptions for legacy constructors #73824
Closed
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
This code has traveled a long and winding road (doo doo). First, some back story: originally, all of this code was written in term of Backend, ScalarType and Device, and things were, well, reasonable. But then Backend became DispatchKey, and then it became TensorOptions but someone (ahem, me) was too lazy to migrate the separate ScalarType and Device into the TensorOptions, because it would involve editing a lot of code. Well I have FINALLY made good on the promised check, by absorbing ScalarType and Device into the TensorOptions. We rely heavily on the optional fields in TensorOptions; the idea is that if it is not set, we are in a context where we aren't expecting any particular scalar type / device; otherwise we require them to be set particularly. Signed-off-by: Edward Z. Yang <[email protected]> [ghstack-poisoned]
CI Flow Status⚛️ CI FlowRuleset - Version:
|
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 8c8ecc2 (more details on the Dr. CI page):
🕵️ 8 new failures recognized by patternsThe following CI failures do not appear to be due to upstream breakages:
|
This code has traveled a long and winding road (doo doo). First, some back story: originally, all of this code was written in term of Backend, ScalarType and Device, and things were, well, reasonable. But then Backend became DispatchKey, and then it became TensorOptions but someone (ahem, me) was too lazy to migrate the separate ScalarType and Device into the TensorOptions, because it would involve editing a lot of code. Well I have FINALLY made good on the promised check, by absorbing ScalarType and Device into the TensorOptions. We rely heavily on the optional fields in TensorOptions; the idea is that if it is not set, we are in a context where we aren't expecting any particular scalar type / device; otherwise we require them to be set particularly. Signed-off-by: Edward Z. Yang <ezyangfb.com> [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Mar 5, 2022
This code has traveled a long and winding road (doo doo). First, some back story: originally, all of this code was written in term of Backend, ScalarType and Device, and things were, well, reasonable. But then Backend became DispatchKey, and then it became TensorOptions but someone (ahem, me) was too lazy to migrate the separate ScalarType and Device into the TensorOptions, because it would involve editing a lot of code. Well I have FINALLY made good on the promised check, by absorbing ScalarType and Device into the TensorOptions. We rely heavily on the optional fields in TensorOptions; the idea is that if it is not set, we are in a context where we aren't expecting any particular scalar type / device; otherwise we require them to be set particularly. Signed-off-by: Edward Z. Yang <ezyangfb.com> ghstack-source-id: 61b565b Pull Request resolved: #73824
This code has traveled a long and winding road (doo doo). First, some back story: originally, all of this code was written in term of Backend, ScalarType and Device, and things were, well, reasonable. But then Backend became DispatchKey, and then it became TensorOptions but someone (ahem, me) was too lazy to migrate the separate ScalarType and Device into the TensorOptions, because it would involve editing a lot of code. Well I have FINALLY made good on the promised check, by absorbing ScalarType and Device into the TensorOptions. We rely heavily on the optional fields in TensorOptions; the idea is that if it is not set, we are in a context where we aren't expecting any particular scalar type / device; otherwise we require them to be set particularly. Signed-off-by: Edward Z. Yang <ezyangfb.com> [ghstack-poisoned]
This code has traveled a long and winding road (doo doo). First, some back story: originally, all of this code was written in term of Backend, ScalarType and Device, and things were, well, reasonable. But then Backend became DispatchKey, and then it became TensorOptions but someone (ahem, me) was too lazy to migrate the separate ScalarType and Device into the TensorOptions, because it would involve editing a lot of code. Well I have FINALLY made good on the promised check, by absorbing ScalarType and Device into the TensorOptions. We rely heavily on the optional fields in TensorOptions; the idea is that if it is not set, we are in a context where we aren't expecting any particular scalar type / device; otherwise we require them to be set particularly. Signed-off-by: Edward Z. Yang <ezyangfb.com> [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Mar 6, 2022
This code has traveled a long and winding road (doo doo). First, some back story: originally, all of this code was written in term of Backend, ScalarType and Device, and things were, well, reasonable. But then Backend became DispatchKey, and then it became TensorOptions but someone (ahem, me) was too lazy to migrate the separate ScalarType and Device into the TensorOptions, because it would involve editing a lot of code. Well I have FINALLY made good on the promised check, by absorbing ScalarType and Device into the TensorOptions. We rely heavily on the optional fields in TensorOptions; the idea is that if it is not set, we are in a context where we aren't expecting any particular scalar type / device; otherwise we require them to be set particularly. Signed-off-by: Edward Z. Yang <ezyangfb.com> ghstack-source-id: 1d148e2 Pull Request resolved: #73824
Contributor
Author
|
This PR has gotten out of control, I may need to shelve it |
Contributor
|
Looks like this PR hasn't been updated in a while so we're going to go ahead and mark this as |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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:
This code has traveled a long and winding road (doo doo). First,
some back story: originally, all of this code was written in term of
Backend, ScalarType and Device, and things were, well, reasonable.
But then Backend became DispatchKey, and then it became TensorOptions
but someone (ahem, me) was too lazy to migrate the separate ScalarType
and Device into the TensorOptions, because it would involve editing
a lot of code. Well I have FINALLY made good on the promised check,
by absorbing ScalarType and Device into the TensorOptions. We rely
heavily on the optional fields in TensorOptions; the idea is that if
it is not set, we are in a context where we aren't expecting any
particular scalar type / device; otherwise we require them to be set
particularly.
Signed-off-by: Edward Z. Yang [email protected]