Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Revert "[Frontend] Add Span filling for frontends to Relay (#9723)" #10072

Merged
merged 1 commit into from
Jan 26, 2022

Conversation

chunit-quic
Copy link
Contributor

Because of the failure the LSTM conversion from Pyotrch
#9723
This reverts commit ce108c1.

Thanks for contributing to TVM! Please refer to guideline https://tvm.apache.org/docs/contribute/ for useful information and tips. After the pull request is submitted, please request code reviews from Reviewers by @ them in the pull request thread.

…)"

Because of the failure of LSTM conversion from Pytorch
This reverts commit ce108c1.
@Mousius Mousius merged commit 095b639 into apache:main Jan 26, 2022
@Mousius
Copy link
Member

Mousius commented Jan 26, 2022

Thanks for identifying this and raising the revert quickly @chunit-quic 😸

@chunit-quic
Copy link
Contributor Author

chunit-quic commented Jan 26, 2022

Thanks for identifying this and raising the revert quickly @chunit-quic 😸

Hi @Mousius,
No problem, my pleasure. :)

Sorry that didn't identify this at the very begging.
Based on the test_lstm.py, it's very surprising to me that we obtain the same execution result compared with pytorch outputs (using == but not isclose()). Even the RelayIR graph are different.
I am digging into it now. Will update status here once I find any solution or clues. :D

Quick update:
It seems that somewhere "duplicate" certain expressions, which are exactly the same as those span-tagged ones, and put them back to RelayIR.
That's the reason why the outputs are the same

sunggg pushed a commit to sunggg/tvm that referenced this pull request Jan 29, 2022
…)" (apache#10072)

Because of the failure of LSTM conversion from Pytorch
ylc pushed a commit to ylc/tvm that referenced this pull request Feb 16, 2022
…)" (apache#10072)

Because of the failure of LSTM conversion from Pytorch
ghost pushed a commit to neo-ai/tvm that referenced this pull request Feb 21, 2022
…)" (apache#10072) (#246)

Because of the failure of LSTM conversion from Pytorch

Co-authored-by: Chun-I Tsai <quic_chunit@quicinc.com>
ghost pushed a commit to neo-ai/tvm that referenced this pull request Feb 21, 2022
…)" (apache#10072) (#246)

Because of the failure of LSTM conversion from Pytorch

Co-authored-by: Chun-I Tsai <quic_chunit@quicinc.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants