[ARITH][BOUND] Fix bound inference to avoid allocating too much #3526
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
sgrechanik-h
force-pushed
the
pr-boundfix
branch
from
caf931d to
4c9acac
Compare
tqchen
requested changes
Jul 10, 2019
src/op/compute_op.cc
Outdated
| Expr min_value = arg_interval->min_value; | ||
| Expr max_value = arg_interval->max_value; | ||
| // Prefer the shape bounds only when we can prove they are tighter. | ||
| arith::Analyzer an; |
Member
There was a problem hiding this comment.
Perhaps we should propagate analyzer through the ProptoInputs, so that we can setup context information later.
Contributor
Author
There was a problem hiding this comment.
It is possible but we have to pass dom_map anyway because Analyzer currently doesn't support binding variables with IntSets.
tqchen
changed the title
sgrechanik-h
force-pushed
the
pr-boundfix
branch
from
2fb6e0e to
99fae32
Compare
tqchen
approved these changes
Jul 14, 2019
Member
|
Thanks @sgrechanik-h , this pr is now merged |
wweic
pushed a commit
to wweic/tvm
that referenced
this pull request
Aug 9, 2019
…he#3526) * [TVM] Fix bound inference to avoid allocating too much * [ARITH][BOUND] Pass analyzer to PropBoundToInputs
wweic
pushed a commit
to neo-ai/tvm
that referenced
this pull request
Sep 6, 2019
…he#3526) * [TVM] Fix bound inference to avoid allocating too much * [ARITH][BOUND] Pass analyzer to PropBoundToInputs
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.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.
Another attempt to fix the problem of tensor bound expansion (described here). The previous attempt was #2104. This time, however, tests passed. The fix is to use the declared tensor bounds instead of the estimated bounds for the argument if it is possible to prove that the declared bounds are tighter. Note that this means that out-of-bounds access may really result in out-of-bounds access.
@tqchen @junrushao1994 please take a look
cc @grwlf @derisavi @jdavies-huawei @bulanova-huawei