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

Release - SuperBench v0.11.0 #654

Merged
merged 7 commits into from
Oct 10, 2024
Merged

Release - SuperBench v0.11.0 #654

merged 7 commits into from
Oct 10, 2024

Conversation

yukirora and others added 7 commits October 8, 2024 13:51
**Description**
Limit protobuf version to be 3.20.x due to onnx reqirement

<img width="1135" alt="未命名图片"
src="https://github.com/user-attachments/assets/f1d52056-4697-4fb9-9fab-22f38765c53f">
…12.4 docker file (#646)

**Description**
fix bug in cuda 12.4 docker file


**Major Revision**
- upgrade nccl due to OOM bug in nccl v2.20 graph mode
- install ucx 1.16 for mutli thread support for mpi in ib-traffic
**Description**
Add ROCm6.2 dockerfile.
**Description**
In result-summary.md line 73-74, the example of kernel-launch used
metrics 'kernel-launch/event_overhead' and
'kernel-launch/wall_overhead'.

But in micro-benchmarks.md line 20-21, the metric names are described as
'kernel-launch/event_time' and ' kernel-launch/wall_time'.

Solution: Use identical metric names described in micro-benchmarks.md.

#649

Signed-off-by: dependabot[bot] <[email protected]>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: hongtaozhang <[email protected]>
Co-authored-by: Yifan Xiong <[email protected]>
**Description**
Add configurations for NDv5 and AMD MI300.

**Major Revision**
- Add cublaslt , FP8 transformers training, dist-inference cpp in NDv5
config example
- Add hipblaslt , FP8 transformers training, dist-inference cpp in AMD
MI300 config example
**Description**

Upgrade version and release note.

**Major Revision**
- Upgrade package versions
- Add release note for v0.11.0
**Description**
Update hpcx link in cuda11.1 dockerfile to fix CI
@yukirora yukirora requested review from abuccts and guoshzhao October 8, 2024 13:57
@yukirora yukirora requested review from cp5555 and a team as code owners October 8, 2024 13:57
Copy link

codecov bot commented Oct 8, 2024

Codecov Report

Attention: Patch coverage is 66.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 85.80%. Comparing base (9f3231e) to head (04c8c07).
Report is 1 commits behind head on main.

Files with missing lines Patch % Lines
superbench/common/utils/device_manager.py 50.00% 1 Missing ⚠️
Additional details and impacted files
@@            Coverage Diff             @@
##             main     #654      +/-   ##
==========================================
+ Coverage   85.78%   85.80%   +0.01%     
==========================================
  Files          97       97              
  Lines        6923     6923              
==========================================
+ Hits         5939     5940       +1     
+ Misses        984      983       -1     
Flag Coverage Δ
cpu-python3.6-unit-test 71.61% <33.33%> (ø)
cpu-python3.7-unit-test 71.61% <33.33%> (ø)
cpu-python3.8-unit-test 71.64% <33.33%> (ø)
cuda-unit-test 83.88% <66.66%> (+0.01%) ⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@yukirora yukirora merged commit 949f9cb into main Oct 10, 2024
16 of 22 checks passed
@yukirora yukirora deleted the yutji/cherrypick-v0.11 branch October 10, 2024 01:59
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants