[Relay][docs] Details on comp. graphs in Relay dev intro #2324
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
…rect typos and stylistic quirks
yidawang
suggested changes
Dec 23, 2018
Contributor
yidawang
left a comment
yidawang
left a comment
There was a problem hiding this comment.
It's bad that I cannot tell what you have changed. Can you at least point out the line numbers roughly?
I suggested some grammar issue editing as well.
docs/dev/relay_intro.rst
Outdated
| surprise a PL researcher in the first place. If we implement a simple visitor to print out the result and | ||
| treat the result as nested Call expression, it becomes ``log(%x) + log(%x)``. | ||
|
|
||
| Such ambiguity is caused by different interpretation of program semantics when there is a shared node in the DAG. |
docs/dev/relay_intro.rst
Outdated
| fact that the ``%1`` is actually reused twice in ``%2``. | ||
|
|
||
| The Relay IR is mindful of this difference. Usually, deep learning framework users build the computational | ||
| graph in this fashion, where a DAG node reuse often occur. As a result, when we print out the Relay program in |
docs/dev/relay_intro.rst
Outdated
| Now, please take a close look at the AST structure. While the two programs are semantically identical | ||
| (so are their textual representations, except that A-normal form has let prefix), their AST structures are different. | ||
|
|
||
| Since program optimizations take these AST data structures and transform them, the two different structure will |
docs/dev/relay_intro.rst
Outdated
| affect the compiler code we are going to write. For example, if we want to detect a pattern ``add(log(x), y)``: | ||
|
|
||
| - In the data-flow form, we can first access the add node, then directly look at its first argument to see if it is a log | ||
| - In the A-normal form, we cannot directly do the check anymore, because the first input to add is ``%v1`` -- we will need to keep a map from variable to its bound values and lookup that map, in order to know that ``%v1`` is a log. |
Contributor
Author
|
Yeah, I'm terribly sorry about the line endings being messed up in the diff. The primary change was two added lines after the first paragraph in the first section. Other changes were mostly one-word changes |
Contributor
|
@slyubomirsky OK. Please address other minor issues I left then it will be good to merge I think. |
yidawang
approved these changes
Dec 23, 2018
Member
|
Thanks, @yidawang @slyubomirsky , this is merged |
FrozenGene
pushed a commit
to FrozenGene/tvm
that referenced
this pull request
Dec 27, 2018
wweic
pushed a commit
to neo-ai/tvm
that referenced
this pull request
Feb 20, 2019
wweic
pushed a commit
to neo-ai/tvm
that referenced
this pull request
Feb 20, 2019
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.
In response to a comment (#2232 (comment)) in #2232, I've added a few words on why traditional deep learning frameworks use the computation graph representation in the first place. Since I was making changes anyway, I also made some stylistic edits.
Unfortunately, I've screwed up the line endings somehow so the diff is blown up; I couldn't figure out how to fix the diff, unfortunately.