Resolved torchrun Bug: Fixed issue #2163

Updated torch.distributed.launch to torchrun.
deepinsight · Feb 8, 2023 · e0fdff8 · e0fdff8
1 parent bf32ec2
commit e0fdff8
Show file tree

Hide file tree

Showing 4 changed files with 8 additions and 8 deletions.
diff --git a/recognition/arcface_torch/README.md b/recognition/arcface_torch/README.md
@@ -9,9 +9,9 @@ The "arcface_torch" repository is the official implementation of the ArcFace alg
 
 ## Requirements
 
-To avail the latest features of PyTorch, we have upgraded to version 1.9.0.
+To avail the latest features of PyTorch, we have upgraded to version 1.12.0.
 
-- Install [PyTorch](http://pytorch.org) (torch>=1.9.0), our doc for [install.md](docs/install.md).
+- Install [PyTorch](https://pytorch.org/get-started/previous-versions/) (torch>=1.12.0).
 - (Optional) Install [DALI](https://docs.nvidia.com/deeplearning/dali/user-guide/docs/), our doc for [install_dali.md](docs/install_dali.md).
 - `pip install -r requirement.txt`.
 

diff --git a/recognition/arcface_torch/dist.sh b/recognition/arcface_torch/dist.sh
@@ -4,9 +4,9 @@ config=wf42m_pfc03_32gpu_r100
 
 for((node_rank=0;node_rank<${#ip_list[*]};node_rank++));
 do 
- ssh face@${ip_list[node_rank]} "cd `pwd`;PATH=$PATH \
+ ssh ubuntu@${ip_list[node_rank]} "cd `pwd`;PATH=$PATH \
  CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
- python -m torch.distributed.launch \
+ torchrun \
  --nproc_per_node=8 \
  --nnodes=${#ip_list[*]} \
  --node_rank=$node_rank \

diff --git a/recognition/arcface_torch/train.py b/recognition/arcface_torch/train.py
@@ -18,8 +18,8 @@
 from utils.utils_distributed_sampler import setup_seed
 from utils.utils_logging import AverageMeter, init_logging
 
-assert torch.__version__ >= "1.9.0", "In order to enjoy the features of the new torch, \
-we have upgraded the torch to 1.9.0. torch before than 1.9.0 may not work in the future."
+assert torch.__version__ >= "1.12.0", "In order to enjoy the features of the new torch, \
+we have upgraded the torch to 1.12.0. torch before than 1.12.0 may not work in the future."
 
 try:
  rank = int(os.environ["RANK"])

diff --git a/recognition/arcface_torch/train_v2.py b/recognition/arcface_torch/train_v2.py
@@ -18,8 +18,8 @@
 from utils.utils_distributed_sampler import setup_seed
 from utils.utils_logging import AverageMeter, init_logging
 
-assert torch.__version__ >= "1.9.0", "In order to enjoy the features of the new torch, \
-we have upgraded the torch to 1.9.0. torch before than 1.9.0 may not work in the future."
+assert torch.__version__ >= "1.12.0", "In order to enjoy the features of the new torch, \
+we have upgraded the torch to 1.12.0. torch before than 1.12.0 may not work in the future."
 
 try:
  rank = int(os.environ["RANK"])