intelligent-machine-learning dlrover issues

intelligent-machine-learning / dlrover

DLRover: An Automatic Distributed Deep Learning System

Other

1.27k stars 167 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Skip memory limitation for gpu type node relaunch operation.

#1341 BalaBalaYi opened 2 hours ago
1
fix a bug in infer method

#1340 jlsong01 opened 1 day ago
1
Fix the issue that len(indices) and num_samples might not be equal

#1339 sunjq1 opened 2 days ago
0
dlorver适配新的加速器类型以及实现类似Nvidia_gpu.py脚本

#1338 lulu-0126 opened 2 days ago
1
client.connect(path) error when saving checkpoint

#1337 atomrun39 opened 2 days ago
1
Update build_proto.sh to use pip3 instead of pip

#1336 jinqinn closed 57 minutes ago
1
WIP: handle GPU lost in resource monitor

#1335 samplise opened 3 days ago
1
Fix diagnosis agent action consuming

#1334 BalaBalaYi closed 3 days ago
1
make deploy 镜像拉取失败

#1333 Ind1x1 closed 5 days ago
0
AttributeError: module 'collections' has no attribute 'Sequence'

#1332 linzhidao1010 opened 5 days ago
0
fix process leak in ascend npu

#1331 majieyue closed 3 days ago
1
Delete pod when list pod already succeeded.

#1330 BalaBalaYi closed 5 days ago
1
Skip heartbeat timeout for failed and exited worker

#1329 BalaBalaYi closed 5 days ago
1
Function using/naming optimized based on job context.

#1328 BalaBalaYi closed 1 week ago
1
Improve job context function using.

#1327 BalaBalaYi closed 1 week ago
1
Fix known issue of job context using.

#1326 BalaBalaYi closed 1 week ago
1
Fix proto install script.

#1325 BalaBalaYi closed 1 week ago
1
some refinement on code comment

#1324 jlsong01 closed 1 week ago
1
Job exit when all nodecheck failed

#1323 majieyue closed 1 week ago
1
feat: generate pyi files for protobuf definitions

#1322 Peefy opened 1 week ago
0
Expose ckpt events

#1321 samplise opened 1 week ago
1
fix: typo RayJobSubmitter in ray_job_submitter.py

#1320 Peefy closed 1 week ago
1
Job context implementation

#1319 samplise closed 1 week ago
1
Refactor diagnosis manager

#1318 samplise closed 5 days ago
1
Fix empty node issue after master failover

#1317 BalaBalaYi closed 1 week ago
1
Revert grpc envs setting.

#1316 BalaBalaYi closed 1 week ago
1
Refactor node event report and report succeeded.

#1315 BalaBalaYi closed 2 weeks ago
1
Could DLRover be able to apply to the diffusion transformer training? And combined with deepspeed?

#1314 TomSuen opened 2 weeks ago
1
fix a typo in user docs

#1313 taylor840326 closed 2 weeks ago
0
Fix: Use return instead of break to correctly exit loop on path exist…

#1312 jinqinn closed 6 days ago
1
Add Balance Loss to MoE Example for Enhanced Expert Load Distribution (Issue #1300)

#1311 Mukku27 opened 3 weeks ago
2
The controller manager restarts frequently

#1310 sunjq1 opened 3 weeks ago
0
More logging on device using.

#1309 BalaBalaYi closed 3 weeks ago
1
Set grpc env for optimization.

#1308 BalaBalaYi closed 3 weeks ago
1
add multiple protobuf version support

#1307 majieyue closed 2 weeks ago
3
Revert "Add grpc envs to improve stability."

#1306 BalaBalaYi closed 3 weeks ago
1
Add grpc envs to improve stability.

#1305 BalaBalaYi closed 3 weeks ago
1
update moe example

#1304 skydoorkai closed 3 weeks ago
1
More pending fastfail stragegy.

#1303 BalaBalaYi closed 3 weeks ago
1
[WIP] Refactor diagnosis manager

#1302 samplise closed 1 week ago
1
Fix/Enhance node management.

#1301 BalaBalaYi closed 3 weeks ago
1
Add balance loss in atorch moe example

#1300 skydoorkai opened 1 month ago
0
Compatible torch 2.4(2.3) storage writer.

#1299 BalaBalaYi closed 1 month ago
1
How does dlrover make sure all the nodes in one job are in one switch

#1298 gangxie112 opened 1 month ago
1
Fix issue in ckpt saver.

#1297 BalaBalaYi closed 1 month ago
1
Fix relaunch node's relaunch limit.

#1296 BalaBalaYi closed 1 month ago
1
Catch exception when get config from brain

#1295 samplise closed 1 month ago
1
yaml lint

#1294 BalaBalaYi closed 1 month ago
1
add exception handler in _get_master_addr_port since the port might b…

#1293 majieyue closed 4 weeks ago
1
add stale issue worker flow

#1292 BalaBalaYi closed 1 month ago
0