microsoft
/
superbenchmark
зеркало из https://github.com/microsoft/superbenchmark.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность
431 коммит 32 Ветки 13 Теги 102 MiB
Граф коммитов

431 Коммитов

Автор SHA1 Сообщение Дата
Yuting Jiang 7af75df392
Bug Fix: Data Diagnosis - Fix bug of failure test and warning of pandas in data diagnosis (#638) 
**Description**
Fix bug of failure test and warning of pandas in data diagnosis.

**Major Revision**
- fix warning of pandas in replace and fillna due to type downcast
- fix bug of failure check function only check one matched metric rather
than all matched metrics
- fix bug when converting regex into str of metrics when there're more
than one match group
 2024-08-16 09:04:24 +08:00
Yang Wang 46a5792915
Bug Fix - Update Docker Exec Command for Persistent HPCX Environment (#635) 
Add 10-hpcx.sh to /etc/profile.d
Update the Docker exec command to ensure a persistent HPCX environment.
 2024-08-13 16:35:01 +00:00
Yang Wang 9de841bc95
Use `types-setuptools` as `types-pkg_resources` is Yanked (#637) 
* https://pypi.org/project/types-pkg-resources/
* Use types-setuptools instead
 2024-08-08 22:30:37 +08:00
Yuting Jiang 2101e933cc
CI/CD - Fix MSCCL build error in CUDA12.4 docker build pipeline (#633) 
**Description**
Fix MSCCL build error in CUDA12.4 docker build pipeline due to OOM
issue.
 2024-07-28 23:43:06 +00:00
Yuting Jiang e304cf1572
Benchmarks: Micro benchmarks - add support for NVIDIA L4/L40/L40s GPUs in gemm-flops (#634) 
**Description**
Add support GPU ARCH 8.9 for NVIDIA L4/L40/L40s GPUs in gemm-flops.
 2024-07-26 02:42:17 +00:00
dependabot[bot] 4e27142a59
Bump express from 4.18.2 to 4.19.2 in /website (#618) 
Bumps [express](https://github.com/expressjs/express) from 4.18.2 to 4.19.2.
- [Release notes](https://github.com/expressjs/express/releases)
- [Changelog](https://github.com/expressjs/express/blob/master/History.md)
- [Commits](expressjs/express@4.18.2...4.19.2)

---
updated-dependencies:
- dependency-name: express
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2024-07-26 09:12:11 +08:00
dependabot[bot] b4945fb29c
Bump ws from 6.2.2 to 6.2.3 in /website (#629) 
Bumps [ws](https://github.com/websockets/ws) from 6.2.2 to 6.2.3.
- [Release notes](https://github.com/websockets/ws/releases)
- [Commits](websockets/ws@6.2.2...6.2.3)

---
updated-dependencies:
- dependency-name: ws
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2024-07-25 15:59:48 +08:00
omahs a4c87da0ac
Docs - fix typos (#628) 
Docs - fix typos
 2024-07-25 03:49:19 +00:00
dependabot[bot] 4102302a96
Bump ip from 1.1.5 to 1.1.9 in /website (#610) 
Bumps [ip](https://github.com/indutny/node-ip) from 1.1.5 to 1.1.9.
- [Commits](indutny/node-ip@v1.1.5...v1.1.9)

---
updated-dependencies:
- dependency-name: ip
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2024-07-25 02:15:52 +00:00
dependabot[bot] 6e556d76e8
Bump follow-redirects from 1.14.8 to 1.15.6 in /website (#613) 
Bumps [follow-redirects](https://github.com/follow-redirects/follow-redirects) from 1.14.8 to 1.15.6.
- [Release notes](https://github.com/follow-redirects/follow-redirects/releases)
- [Commits](follow-redirects/follow-redirects@v1.14.8...v1.15.6)

---
updated-dependencies:
- dependency-name: follow-redirects
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2024-07-24 23:59:58 +00:00
Yifan Xiong 1362732c79
Docs - Add BibTeX in README and repo (#632) 
Add BibTeX for citation in README and repo.
 2024-07-23 18:31:21 -07:00
Yang Wang 9a3ce39d5a
Update omegaconf version to 2.3.0 (#631) 
Update `omegaconf` version to
[2.3.0](https://pypi.org/project/omegaconf/2.3.0/) as omegaconf 2.0.6
has a non-standard dependency specifier PyYAML>=5.1.*. pip 24.1 will
enforce this behaviour change.
Discussion can be found at https://github.com/pypa/pip/issues/12063.
 2024-07-23 14:46:28 -07:00
Yuting Jiang 7435f10a22
Dockerfile - Add CUDA 12.4 dockerfile (#619) 
**Description**
Add CUDA 12.4 dockerfile.

**Major Revision**
- upgrade nvidia docker into 23.04


**Minor Revision**
- upgrade hpcx into 2.18
 2024-04-22 06:36:19 +00:00
Yuting Jiang dc3846cbd4
Dockerfile - Upgrade mlc to v3.11 (#620) 
**Description**
Upgrade mlc to v3.11.
 2024-04-18 10:59:36 +08:00
Ziyue Yang cc89ee591c
Benchmarks: Revise Code - Add hipblasLt tuning to dist-inference cpp implementation (#616) 
**Description**
Adds hipblasLt tuning to dist-inference cpp implementation.
 2024-04-02 09:56:33 +08:00
Yang Wang eeaa9b1ac9
Bug Fix - Bug fix for cuda 12.2 dockerfile LD_LIBRARY_PATH issue (#614) 
**Description**
Cuda 12.2 image will report undfined symbol error due to incomplete
LD_LIBRARY_PATH:


![image](https://github.com/microsoft/superbenchmark/assets/25875482/1a7c48c7-cb6b-4e3a-abbe-dde23007a96b)

### How to reproduce:
1. Deploy sb with cuda12.2 image
```
sb deploy -f local.ini -i superbench/superbench:v0.10.0-cuda12.2
```
2. Enter to the container
```
sudo docker exec -it sb-workspace bash
```
3. Execute `mpirun`:
```
root@sb-container:~# mpirun
mpirun: symbol lookup error: mpirun: undefined symbol: opal_libevent2022_event_base_loop
```
### Fix to fix
* Append hpcx_load into /etc/bash.bashrc for updaing env LD_LIBRARY_PATH in each time

---------
 2024-03-21 15:05:55 +00:00
Yifan Xiong 2c88db907f
Release - SuperBench v0.10.0 (#607) 
**Description**

Cherry-pick bug fixes from v0.10.0 to main.

**Major Revisions**

* Benchmarks: Microbenchmark - Support different hipblasLt data types in dist_inference #590
* Benchmarks: Microbenchmark - Support in-place for NCCL/RCCL benchmark #591
* Bug Fix - Fix NUMA Domains Swap Issue in NDv4 Topology File #592
* Benchmarks: Microbenchmark - Add data type option for NCCL and RCCL tests #595
* Benchmarks: Bug Fix - Make metrics of dist-inference-cpp aligned with PyTorch version #596
* CI/CD - Add ndv5 topo file #597
* Benchmarks: Microbenchmark - Improve AMD GPU P2P performance with fine-grained GPU memory #593
* Benchmarks: Build Pipeline - fix nccl and nccl test version to 2.18.3 to resolve hang issue in cuda12.2 docker #599
* Dockerfile - Bug fix for rocm docker build and deploy #598
* Benchmarks: Microbenchmark - Adapt to hipblasLt data type changes #603
* Benchmarks: Micro benchmarks - Update hipblaslt metric unit to tflops #604
* Monitor - Upgrade pyrsmi to amdsmi python library. #601
* Benchmarks: Micro benchmarks - add fp8 and initialization for hipblaslt benchmark #605
* Dockerfile - Add rocm6.0 dockerfile #602
* Bug Fix - Bug fix for latest megatron-lm benchmark #600
* Docs - Upgrade version and release note #606

Co-authored-by: Ziyue Yang <ziyyang@microsoft.com>
Co-authored-by: Yang Wang <yangwang1@microsoft.com>
Co-authored-by: Yuting Jiang <yutingjiang@microsoft.com>
Co-authored-by: guoshzhao <guzhao@microsoft.com>
 2024-01-08 05:40:52 +00:00
Ziyue Yang 2c2096ed83
Benchmark: Revision - Fix -O2 option passing in gpu_copy ROCm build (#589) 
**Description**
`add_compile_options` will not work for ROCm build, change it to setting
`CMAKE_CXX_FLAGS`.
 2023-12-11 04:34:51 +00:00
Ziyue Yang 719a427fe7
Benchmarks: Microbenchmark - Add distributed inference benchmark cpp implementation (#586) 
**Description**
Add distributed inference benchmark cpp implementation.
 2023-12-11 06:53:51 +08:00
Yuting Jiang 1f5031bd74
Dockerfile - Upgrade to rocm5.7 dockerfile (#587) 
**Description**
upgrade to rocm5.7 dockerfile.

---------

Co-authored-by: yukirora <yuting.jiang@microsoft.com>
 2023-12-09 17:41:12 +00:00
Ziyue Yang 4fa60be7cd
Benchmarks: Micro benchmark - Add one-to-all, all-to-one, all-to-all support to gpu_copy_bw_performance (#588) 
**Description**
Add one-to-all, all-to-one, all-to-all support to
gpu_copy_bw_performance, and fix performance bug in gpu_copy
 2023-12-08 23:22:38 +08:00
Ziyue Yang 6ef3a0110f
Benchmarks: Add MSCCL Support for Nvidia GPU (#584) 
**Description**
Add MSCCL support for Nvidia GPU
 2023-12-07 19:57:28 +08:00
Yuting Jiang dd5a6329ed
Benchmarks: Add benchmark: Megatron-LM/Megatron-Deepspeed GPT pretrain benchmark (#582) 
**Description**
Megatron-LM/Megatron-Deepspeed GPT pretrain benchmark
 2023-12-07 09:37:09 +08:00
Ziyue Yang 254ea7feba
Benchmarks: Micro benchmark - Add graph mode in NCCL/RCCL benchmarks for latency metrics (#583) 
**Description**
Revise NCCL/RCCL benchmarks to graph mode add latency metrics.
 2023-12-05 16:48:13 +08:00
Yuting Jiang 9ae8c67093
Benchmarks: micro benchmark - Support cpu-gpu and gpu-cpu in ib-validation (#581) 
**Description**
Benchmarks: micro benchmark - Support cpu-gpu and gpu-cpu in
ib-validation

**Major Revision**
- Support cpu-gpu and gpu-cpu in ib-validation


**Minor Revision**
- support multi msg size, multi direction, multi ib commands in
ib-validation
 2023-12-04 22:20:46 +08:00
guoshzhao 028819b388
Monitor - Add support for AMD GPU. (#580) 
**Description**
Add AMD support in monitor.

**Major Revision**
- Add library pyrsmi to collect metrics.
- Currently can get device_utilization, device_power, device_used_memory
and device_total_memory.
 2023-11-27 18:45:56 +08:00
Yifan Xiong 1ad1c21c38
Dockerfile - Upgrade Docker image to CUDA 12.2 (#577) 
Upgrade Docker image to CUDA 12.2 for H100:
* upgrade base image to 23.10
* fix onnxruntime version in python3.10
* fix compilation errors
 2023-11-22 13:48:18 +00:00
Yuting Jiang 2235e084ab
Benchmarks: Micro benchmark - add initialization options for rocm gemm flops (#578) 
**Description**
add initialization options for rocm gemm flops.
 2023-11-22 12:52:22 +00:00
Yuting Jiang 79089b6517
Benchmarks: Micro benchmark - Add hipBLASLt function benchmark (#576) 
**Description**
hipblaslt function benchmark and rebase cublaslt function benchmark.
 2023-11-22 19:48:10 +08:00
guoshzhao 9f4880cb8e
Analyzer - Generate baseline given results from multiple nodes. (#575) 
**Description**
Generate baseline given results from multiple nodes. 

**Major Revision**
- Add sub command `sb result generate-baseline`
- Add UT and docs

---------

Co-authored-by: 454314380 <454314380@qq.com>
Co-authored-by: Yuting Jiang <yutingjiang@microsoft.com>
 2023-11-22 14:42:32 +08:00
Yuting Jiang f53d941a22
Benchmarks: micro benchmarks - add int8 support for cublaslt function (#574) 
**Description**
add int8 support for cublaslt function.
 2023-11-20 11:21:20 +08:00
Yuting Jiang c7800bb8e0
Bug Fix - remove cp ptx file command in gpu burn test (#567) 
**Description**
remove cp ptx file in gpu burn test since the command is run inside
self.args.bin_dir dir.


d246bab430/superbench/benchmarks/micro_benchmarks/micro_base.py (L183)
 2023-11-14 03:52:56 +00:00
dependabot[bot] ce3737f98b
Bump @babel/traverse from 7.14.5 to 7.23.2 in /website (#566) 
Bumps [@babel/traverse](https://github.com/babel/babel/tree/HEAD/packages/babel-traverse) from 7.14.5 to 7.23.2.
- [Release notes](https://github.com/babel/babel/releases)
- [Changelog](https://github.com/babel/babel/blob/main/CHANGELOG.md)
- [Commits](https://github.com/babel/babel/commits/v7.23.2/packages/babel-traverse)

---
updated-dependencies:
- dependency-name: "@babel/traverse"
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2023-11-07 10:36:42 +08:00
dependabot[bot] 07477c3bae
Bump postcss from 8.3.5 to 8.4.31 in /website (#564) 
Bumps [postcss](https://github.com/postcss/postcss) from 8.3.5 to 8.4.31.
- [Release notes](https://github.com/postcss/postcss/releases)
- [Changelog](https://github.com/postcss/postcss/blob/main/CHANGELOG.md)
- [Commits](postcss/postcss@8.3.5...8.4.31)

---
updated-dependencies:
- dependency-name: postcss
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2023-11-05 11:35:49 +00:00
Yuting Jiang d246bab430
Dockerfile - update mlc version into 3.10 for cuda and rocm dockerfiles (#562) 
**Description**
Update mlc version into 3.10 for cuda and rocm dockerfiles to be
consistent with cuda12 dockerfile

Co-authored-by: yukirora <yuting.jiang@microsoft.com>
 2023-10-23 11:21:17 +08:00
Yuting Jiang 27a10811af
Benchmarks: micro benchmark - source code for evaluating NVDEC decoding performance (#560) 
**Description**
source code for evaluating NVDEC decoding performance.

---------

Co-authored-by: yukirora <yuting.jiang@microsoft.com>
 2023-08-22 10:56:33 +00:00
Yuting Jiang 6c0205cece
Benchmarks: micro benchmarks - add source code for DirectXRenderPerf (#549) 
**Description**
add source code for DirectXRenderPerf.

---------

Co-authored-by: yukirora <yuting.jiang@microsoft.com>
 2023-08-18 05:17:04 +00:00
pnunna93 67f2aa7237
Benchmarks: model benchmarks - change torch.distributed.launch to torchrun (#556) 
This PR has following changes
- torch.distributed.launch changed to torchrun. torch.distributed.launch
is deprecated in latest Pytorch and is recommended to move to torchrun -
https://pytorch.org/docs/stable/elastic/run.html
- Changes to AMD GPU detection logic. The AMD GPU detection logic throws
warning when containers have only renderD in /dev/dri, this change would
resolve those warnings

---------

Co-authored-by: Yuting Jiang <yutingjiang@microsoft.com>
 2023-08-08 13:03:32 +08:00
Yuting Jiang e1df877bfe
Release - SuperBench v0.9.0 (#558) 
**Description**
Cherry-pick bug fixes from v0.9.0 to main.

**Major Revision**
- CI/CD: pipeline - clean more disk space to fix rocm building image
pipeline(#555 )
- Benchmarks: bug fix - use absolute path for input file in
DirectXEncodingLatency(#554)
- CI/CD - add push win docker image on release branch in pipeline (#552)
- Docs - Upgrade version and release note(#557)
 2023-07-27 10:42:31 +08:00
dependabot[bot] 466b477e9d
Bump semver from 5.7.1 to 5.7.2 in /website (#550) 
Bumps [semver](https://github.com/npm/node-semver) from 5.7.1 to 5.7.2.
- [Release notes](https://github.com/npm/node-semver/releases)
- [Changelog](https://github.com/npm/node-semver/blob/v5.7.2/CHANGELOG.md)
- [Commits](npm/node-semver@v5.7.1...v5.7.2)

---
updated-dependencies:
- dependency-name: semver
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
 2023-07-24 17:07:35 +08:00
Yuting Jiang e8ac0b1e28
Benchmarks: micro benchmarks - add python code for DirectXGPUEncodingLatency (#548) 
**Description**
add python code for DirectXGPUEncodingLatency.
 2023-07-06 15:31:28 +08:00
Yuting Jiang c8c079c2af
Benchmarks: micro benchmarks - add python code for DirectXGPUCopy (#546) 
**Description**
add python code for DirectXGPUCopy.
 2023-07-06 00:15:32 +08:00
Yuting Jiang af4cfd5bbf
Benchmarks: micro benchmarks - add python code for DirecXGPUMemBw (#547) 
**Description**
add python code for DirecXGPUMemBw.
 2023-07-05 22:07:13 +08:00
Yuting Jiang f1d608aef7
Benchmarks: micro benchmarks - add python code for DirectXGPUCoreFlops (#542) 
**Description**
add python code for DirectX core flops and init DirectX test pipeline.

**Major Revision**
- add python code for DirectX core flops 
- init DirectX test pipeline


**Minor Revision**
- add test for DirectX core flops
 2023-07-05 16:56:21 +08:00
Yuting Jiang 3704a432b9
CI/CD - Support DirectX test pipeline (#545) 
**Description**
Support DirectX test pipeline.
 2023-07-05 11:33:40 +08:00
Yuting Jiang 865472177f
Benchmarks: Build Pipeline - add AMF in third party and build AMF encoding latency test (#543) 
**Description**
add AMF in third party and build AMF encoding latency test.
 2023-07-03 14:43:21 +00:00
Yuting Jiang 97f7b1df86
Benchmarks: microbenchmark - add auto selecting algorithm support for cudnn functions (#540) 
**Description**
add auto selecting algorithm support for cudnn functions.

**Major Revision**
- add auto selecting algorithm support for cudnn functions in source
code
- add 'auto_algo' option in benchmark
- add related test
 2023-06-30 12:58:41 +00:00
Lei Qu c7d0beaf9e
Doc - Update outdate references in micro-benchmarks.md (#544) 
Modify link for Nvidia bandwidth test tool

**Description**
previous link is 404

**Minor Revision**
update the link value to
https://github.com/NVIDIA/cuda-samples/tree/master/Samples/1_Utilities/bandwidthTest
 2023-06-30 19:17:41 +08:00
Yifan Xiong 7184bdd1ed
Benchmarks - Update result parsing in tensorrt inference (#541) 
* Update result parsing for newer tensorrt versions
* Update arguments when load torchvision models
 2023-06-30 11:22:46 +08:00
Yuting Jiang f259913707
Benchmarks: Add benchmark - Add source code of DirectxGPUCopy microbenchmark (#486) 
**Description**
Add source code of DirectxGPUCopy microbenchmark.
 2023-06-29 19:38:01 +08:00