Skip to content

Downgrade python gpu package to CUDA 10.0 - #2086

Merged
snnn merged 6 commits into
masterfrom
snnn/cuda3
Oct 11, 2019
Merged

Downgrade python gpu package to CUDA 10.0#2086
snnn merged 6 commits into
masterfrom
snnn/cuda3

Conversation

@snnn

@snnn snnn commented Oct 10, 2019

Copy link
Copy Markdown
Contributor

Description:

  1. Downgrade python gpu package to CUDA 10.0
  2. Disable two RNN unitests because they are failing
  3. Set CMAKE_CUDA_STANDARD to 11

Motivation and Context

  • Why is this change required? What problem does it solve?

Now the Windows one is using 10.0 but Linux one is using 10.1. We should fix this inconsistency.

  • If it fixes an open issue, please link to the issue here.

Changming Sun added 2 commits October 10, 2019 13:16
@snnn
snnn merged commit a314402 into master Oct 11, 2019
@snnn
snnn deleted the snnn/cuda3 branch October 11, 2019 01:31
yuslepukhin pushed a commit that referenced this pull request Mar 17, 2026
…ation (#2086)

## Update CalibrationDataReader to support strided data for Onnx
Quantization
Using CalibStridedMinMax option in onnx quantization, only stride amount
of data is used and all results are merged in the end during the
calibration process. This helps in better managing mem usage.

## Checklist before requesting a review
- [ ] Add unit tests for this change.
- [ ] Make sure all tests can pass.
- [ ] Update documents if necessary.
- [ ] Lint and apply fixes to your code by running `lintrunner -a`
- [ ] Is this a user-facing change? If yes, give a description of this
change to be included in the release notes.
- [ ] Is this PR including examples changes? If yes, please remember to
update [example
documentation](https://github.com/microsoft/Olive/blob/main/docs/source/examples.md)
in a follow-up PR.

## (Optional) Issue link

---------

Co-authored-by: Ashima Jain <ashimaj@qti.qualcomm.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.

2 participants