pytorch elastic issues - Githubissues

pytorch / elastic

PyTorch elastic training

BSD 3-Clause "New" or "Revised" License

730 stars 98 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

[examples/imagenet/main.py] Why doesn't elastic code contain gpu sync to compute performance, e.g. all_reduce

#170 aitrics-chris opened 2 years ago
0
RuntimeError: Expected all tensors to be on the same device, but found at least two devices

#169 aitrics-chris closed 2 years ago
4
Please add more torch elastic training examples

#168 wanziyu opened 2 years ago
0
(torchelastic) update README to point elastic CRD users to TorchX

#167 kiukchung closed 2 years ago
1
BASE_IMG upgrade for Dockerfile after PyTorch1.10

#166 ghost closed 2 years ago
3
rendezvous: _matches_machine_hostname doesn't resolve hostnames fully

#165 d4l3k opened 2 years ago
2
Kubernetes: ttlSecondsAfterFinished not working in ElasticJob spec

#164 jovan-absci opened 2 years ago
0
(torchx/specs) Remove RunConfig in favor of using Dict[str, CfgVal] directly

#163 kiukchung closed 2 years ago
6
[feature request] Add CPU example

#162 gaocegege opened 3 years ago
2
Remove unconfigured submodule

#161 jonathan-conder-sm opened 3 years ago
3
Is petctl also deprecated?

#160 vadimkantorov opened 3 years ago
0
[feature request] petctl to support pulling script directory from github repo by commit or tag

#159 vadimkantorov opened 3 years ago
0
submodule path docs/src/pytorch-sphinx-theme not in .gitmodules

#158 jonathan-conder-sm opened 3 years ago
0
Kubernetes CustomResourceDefinition Moving out of Beta

#157 5had3z closed 3 years ago
4
[Blocked] feat(dockerfile): Use Torch 1.9 instead of nightly

#156 gaocegege closed 3 years ago
4
Various improvements to `torch.distributed.launch` and `torch.distributed.run` (#60925)

#155 aivanou closed 3 years ago
2
update docs to point to pytorch 1.9 and torchx for torchelastic and tsm (respectively)

#154 kiukchung closed 3 years ago
2
Update index.md

#153 brianjo closed 3 years ago
2
EtcdStore: AttributeError: can't set attribute

#152 vv-p opened 3 years ago
1
Cannot reuse --rdzv_id between different elastic launch ?

#151 PKUFlyingPig opened 3 years ago
0
Imagenet example fails during accuracy calculation (v0.2.2 on 1.8.1)

#150 assapin closed 3 years ago
1
add support for jetter to Role (base_image) for mast launches

#149 kiukchung closed 3 years ago
2
Move torchelastic docs *.rst (#56811)

#148 kiukchung closed 3 years ago
2
Out of Data documentation

#147 Godricly closed 3 years ago
4
Improve the implementation of `RendezvousParameters` and add its unit tests. (#54807)

#146 cbalioglu closed 3 years ago
2
ModuleNotFoundError: No module named 'torch.distributed.elastic'

#145 GwangsooHong closed 3 years ago
4
Fix python required version > 3.6 bug

#144 jenhaoyang opened 3 years ago
8
Move torchelastic/events to torch/distributed/events

#143 aivanou closed 3 years ago
2
Support PyTorch 1.8, TorchVision 0.9.0 and TorchAduio 0.8.0

#142 davidspek closed 3 years ago
7
Move torchelastic/rendezvous to torch/distributed/rendezvous

#141 kiukchung closed 3 years ago
2
Torch Elastic - How to make sure all nodes are in the same AZ?

#140 thecooltechguy closed 3 years ago
2
[*.py] Rename "Arguments:" to "Args:"

#139 SamuelMarks closed 3 years ago
3
Remove NCCL Blocking Wait from Imagenet Example

#138 osalpekar closed 3 years ago
2
Minor fix in Issue Reporting Template

#137 osalpekar closed 3 years ago
3
Enable NCCL_ASYNC_ERROR_HANDLING in Torchelastic

#136 osalpekar closed 3 years ago
1
Pytorch Lightning with TorchElastic - One worker doesn't start

#135 tchaton closed 3 years ago
3
Elastic agent doesn't detect worker failures in NCCL

#134 ruipeterpan closed 3 years ago
4
Enable NCCL_ASYNC_ERROR_HANDLING in torchelastic

#133 osalpekar closed 3 years ago
4
Fix circle CI breakage by depending on torch-1.8.0dev (nightly)

#132 kiukchung closed 3 years ago
2
unbind scheduler from session and make session apis take a scheduler backend

#131 kiukchung closed 4 years ago
4
How to programmatically determine if a training job has finished using `kubectl`?

#130 darthsuogles closed 4 years ago
2
make torchelastic.distributed.launch args settable from env var with name PET_ARG

#129 kiukchung closed 4 years ago
2
Add env support for the training script argument

#128 kuikuikuizzZ closed 4 years ago
4
Add tsm docs to the docs page, added --dry-run flag to doc_push.sh, fix a few docstring typos

#127 kiukchung closed 4 years ago
2
add ui url to the return value of session.status(app_id)

#126 kiukchung closed 4 years ago
2
add replica_id macro

#125 kiukchung closed 4 years ago
2
accept role as a command line argument

#124 yifuwang closed 4 years ago
2
implement ElasticRole, role args macro substitution

#123 kiukchung closed 4 years ago
2
move pytorch/elastic/test/** into pytorch/elastic/torchelastic/**/test/**

#122 kiukchung closed 4 years ago
4
test pyenv

#121 kiukchung closed 4 years ago
0