关于GCU、沐曦GPGPU、MLU、0卡V100资源4月7日恢复上架的公告>>> 关于共建具身智能开源数据集的倡议>>> 关于云脑任务中统一路径访问方式的公告>>> 关于将启智集群GPU资源迁移至智算集群的公告>>>
OpenI
/
aiforge
Not watched
33
71
Fork 76
Code
Releases 61 Wiki Activity
Issues 271 Pull Requests 1 Datasets Cloudbrain
Issues
/
Detail
Labels Milestones
New Issue

#2435 智算网络NPU多卡训练任务支持

Closed
created 1 year ago by lewis · 5 comments
lewis commented 1 year ago

目前对于智算网络的NPU训练任务并不支持多卡训练，缺少rank_table等信息，需要根据自定义镜像中的run_train.sh改写启动命令。

目前对于智算网络的NPU训练任务并不支持多卡训练，缺少rank_table等信息，需要根据自定义镜像中的run_train.sh改写启动命令。
lewis added the
enhancement
label 1 year ago
tanglj added this to the V20220718 milestone 1 year ago
tanglj modified the milestone from V20220718 to V20220801 1 year ago
lewis was assigned by tanglj 1 year ago
liuzx was assigned by wangj 1 year ago
wangj commented 1 year ago
Owner

@liuzx 能提供一下样例代码吗

@liuzx 能提供一下样例代码吗
liuzx commented 1 year ago
Collaborator
参考https://git.openi.org.cn/OpenIOSSG/MNIST_Example/src/branch/master/train_for_c2net_dataparallel.py 的代码
lewis commented 1 year ago
Owner

grampus_multi_parallel分支，可测试。

grampus_multi_parallel分支，可测试。
lewis added the
test
label 1 year ago
lewis commented 1 year ago
Owner

image
选择这个镜像去做单卡以及多卡的测试。

![image](/attachments/3313f8fb-8b70-45bd-8315-08887c8eb2f8) 选择这个镜像去做单卡以及多卡的测试。
image.png
29 KiB
wangj commented 1 year ago
Owner

已随V20220801.patch上线。

已随V20220801.patch上线。
wangj closed this issue 1 year ago
Sign in to join this conversation.
No Branch/Tag Specified
Clear branch or tag
Branches Tags
Labels
Apply labels
Clear labels
bug
Something is not working doc
需要更新帮助文档 duplicate
This issue or pull request already exists enhancement
enhancement feature
new feature frontend
前端开发 grampus
依赖虎鲸解决 help wanted
Need some help invalid
Something is wrong need review question
More information is needed reviewed test
Need test wait
问题的解决依赖于第三方的改进 WIP wontfix
This won't be fixed 严重
major bug 轻微
minor bug 重新开启 面向社区开发者
No Label
bug
doc
duplicate
enhancement
feature
frontend
grampus
help wanted
invalid
need review
question
reviewed
test
wait
WIP
wontfix
严重
轻微
重新开启
面向社区开发者
Milestone
Set milestone
Clear milestone
No items
No Milestone
V20220801
Assignees
Assign users
Clear assignees
No Assignees
 lewis
 liuzx
3 Participants
Notifications
Due Date

No due date set.

Depends on
#2662 【智算npu】训练任务运行失败时状态应显示FAILED
OpenI/aiforge
Write Preview
Loading…
Cancel
Save
There is no content yet.
Delete Branch '%!s(MISSING)'

Deleting a branch is permanent. It CANNOT be undone. Continue?

No
Yes
Please read the following content carefully:

Dear OpenI User

Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.

For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》

Agree and continue
Disagree, exit