关于分布式训练启动的问题

[Logictaosen007]

作者您好：
我用另外一种分布式训练启动您代码CIFAR100-10-10进行复现，结果非常接近，但是两张V100，每个epoch训练要51s左右，使用github上面的命令启动train.sh出现了一下问题：
给您邮箱发邮件时，显示address rejected。您空闲的时候看看就好，您的研究对我来说非常有帮助
root@cd98f4d76410:/code/DKT_source# bash train.sh 0,1 --options /code/DKT_source/options/data/cifar100_10-10.yaml /code/DKT_source/options/data/cifar100_order1.yaml /code/DKT_source/options/model/cifar_DKT.yaml --name DKT --data-path /data/Logic888/CIFAR100/cifar100 --output-basedir /data/Logic888/CIFAR100/DKT/save_checkpoints --memory-size 2000
Launching exp on 0,1...
/opt/conda/lib/python3.10/site-packages/torch/distributed/launch.py:181: FutureWarning: The module torch.distributed.launch is deprecated
and will be removed in future. Use torchrun.
Note that --use-env is set by default in torchrun.
If your script expects --local-rank argument to be set, please
change it to read from os.environ['LOCAL_RANK'] instead. See
https://pytorch.org/docs/stable/distributed.html#launch-utility for
further instructions

warnings.warn(
WARNING:torch.distributed.run:

---

Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.

---

/code/DKT_source/continual/robust_models_ImageNet.py:391: UserWarning: Overwriting rvt_tiny in registry with continual.robust_models_ImageNet.rvt_tiny. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_tiny(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:411: UserWarning: Overwriting rvt_tiny_plus in registry with continual.robust_models_ImageNet.rvt_tiny_plus. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_tiny_plus(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:433: UserWarning: Overwriting rvt_small in registry with continual.robust_models_ImageNet.rvt_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_small(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:453: UserWarning: Overwriting rvt_small_plus in registry with continual.robust_models_ImageNet.rvt_small_plus. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_small_plus(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:475: UserWarning: Overwriting rvt_base in registry with continual.robust_models_ImageNet.rvt_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_base(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:495: UserWarning: Overwriting rvt_base_plus in registry with continual.robust_models_ImageNet.rvt_base_plus. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_base_plus(pretrained, **kwargs):
usage: DKT training and evaluation script [-h] [--batch-size BATCH_SIZE] [--incremental-batch-size INCREMENTAL_BATCH_SIZE] [--epochs EPOCHS]
[--base-epochs BASE_EPOCHS] [--no-amp] [--model MODEL] [--input-size INPUT_SIZE] [--patch-size PATCH_SIZE]
[--embed-dim EMBED_DIM] [--depth DEPTH] [--num-heads NUM_HEADS] [--drop PCT] [--drop-path PCT]
[--norm {layer,scale}] [--opt OPTIMIZER] [--opt-eps EPSILON] [--opt-betas BETA [BETA ...]] [--clip-grad NORM]
[--momentum M] [--weight-decay WEIGHT_DECAY] [--sched SCHEDULER] [--lr LR] [--incremental-lr INCREMENTAL_LR]
[--lr-noise pct, pct [pct, pct ...]] [--lr-noise-pct PERCENT] [--lr-noise-std STDDEV] [--warmup-lr LR]
[--incremental-warmup-lr LR] [--min-lr LR] [--decay-epochs N] [--warmup-epochs N] [--cooldown-epochs N]
[--patience-epochs N] [--decay-rate RATE] [--color-jitter PCT] [--aa NAME] [--smoothing SMOOTHING]
[--train-interpolation TRAIN_INTERPOLATION] [--repeated-aug] [--no-repeated-aug] [--reprob PCT]
[--remode REMODE] [--recount RECOUNT] [--resplit] [--auto-kd] [--kd KD] [--distillation-tau DISTILLATION_TAU]
[--resnet] [--data-path DATA_PATH] [--data-set {CIFAR,IMNET,INAT,INAT19}]
[--data-path-subTrain DATA_PATH_SUBTRAIN] [--data-path-subVal DATA_PATH_SUBVAL]
[--inat-category {kingdom,phylum,class,order,supercategory,family,genus,name}] [--output-dir OUTPUT_DIR]
[--output-basedir OUTPUT_BASEDIR] [--device DEVICE] [--seed SEED] [--start_epoch N] [--eval] [--dist-eval]
[--num_workers NUM_WORKERS] [--pin-mem] [--no-pin-mem] [--initial-increment INITIAL_INCREMENT]
[--increment INCREMENT] [--class-order CLASS_ORDER [CLASS_ORDER ...]] [--eval-every EVAL_EVERY] [--debug]
[--retrain-scratch] [--max-task MAX_TASK] [--name NAME] [--options [OPTIONS ...]] [--DKT]
[--duplex-clf DUPLEX_CLF] [--memory-size MEMORY_SIZE] [--distributed-memory] [--global-memory]
[--oversample-memory OVERSAMPLE_MEMORY] [--oversample-memory-ft OVERSAMPLE_MEMORY_FT] [--rehearsal-test-trsf]
[--rehearsal-modes REHEARSAL_MODES] [--fixed-memory]
[--rehearsal {random,closest_token,closest_all,icarl_token,icarl_all,furthest_token,furthest_all}]
[--sep-memory] [--replay-memory REPLAY_MEMORY] [--finetuning {balanced}] [--finetuning-mode FINETUNING_MODE]
[--finetuning-lr FINETUNING_LR] [--finetuning-teacher] [--finetuning-resetclf] [--only-ft] [--ft-no-sampling]
[--freeze-task [FREEZE_TASK ...]] [--freeze-ft [FREEZE_FT ...]] [--freeze-eval] [--log-path LOG_PATH]
[--log-category LOG_CATEGORY] [--bce-loss] [--local_rank LOCAL_RANK] [--world_size WORLD_SIZE]
[--dist_url DIST_URL] [--resume RESUME] [--start-task START_TASK] [--start-epoch START_EPOCH]
[--save-every-epoch SAVE_EVERY_EPOCH] [--validation VALIDATION]
DKT training and evaluation script: error: unrecognized arguments: --local-rank=1
/code/DKT_source/continual/robust_models_ImageNet.py:391: UserWarning: Overwriting rvt_tiny in registry with continual.robust_models_ImageNet.rvt_tiny. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_tiny(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:411: UserWarning: Overwriting rvt_tiny_plus in registry with continual.robust_models_ImageNet.rvt_tiny_plus. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_tiny_plus(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:433: UserWarning: Overwriting rvt_small in registry with continual.robust_models_ImageNet.rvt_small. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_small(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:453: UserWarning: Overwriting rvt_small_plus in registry with continual.robust_models_ImageNet.rvt_small_plus. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_small_plus(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:475: UserWarning: Overwriting rvt_base in registry with continual.robust_models_ImageNet.rvt_base. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_base(pretrained, **kwargs):
/code/DKT_source/continual/robust_models_ImageNet.py:495: UserWarning: Overwriting rvt_base_plus in registry with continual.robust_models_ImageNet.rvt_base_plus. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
def rvt_base_plus(pretrained, **kwargs):
usage: DKT training and evaluation script [-h] [--batch-size BATCH_SIZE] [--incremental-batch-size INCREMENTAL_BATCH_SIZE] [--epochs EPOCHS]
[--base-epochs BASE_EPOCHS] [--no-amp] [--model MODEL] [--input-size INPUT_SIZE] [--patch-size PATCH_SIZE]
[--embed-dim EMBED_DIM] [--depth DEPTH] [--num-heads NUM_HEADS] [--drop PCT] [--drop-path PCT]
[--norm {layer,scale}] [--opt OPTIMIZER] [--opt-eps EPSILON] [--opt-betas BETA [BETA ...]] [--clip-grad NORM]
[--momentum M] [--weight-decay WEIGHT_DECAY] [--sched SCHEDULER] [--lr LR] [--incremental-lr INCREMENTAL_LR]
[--lr-noise pct, pct [pct, pct ...]] [--lr-noise-pct PERCENT] [--lr-noise-std STDDEV] [--warmup-lr LR]
[--incremental-warmup-lr LR] [--min-lr LR] [--decay-epochs N] [--warmup-epochs N] [--cooldown-epochs N]
[--patience-epochs N] [--decay-rate RATE] [--color-jitter PCT] [--aa NAME] [--smoothing SMOOTHING]
[--train-interpolation TRAIN_INTERPOLATION] [--repeated-aug] [--no-repeated-aug] [--reprob PCT]
[--remode REMODE] [--recount RECOUNT] [--resplit] [--auto-kd] [--kd KD] [--distillation-tau DISTILLATION_TAU]
[--resnet] [--data-path DATA_PATH] [--data-set {CIFAR,IMNET,INAT,INAT19}]
[--data-path-subTrain DATA_PATH_SUBTRAIN] [--data-path-subVal DATA_PATH_SUBVAL]
[--inat-category {kingdom,phylum,class,order,supercategory,family,genus,name}] [--output-dir OUTPUT_DIR]
[--output-basedir OUTPUT_BASEDIR] [--device DEVICE] [--seed SEED] [--start_epoch N] [--eval] [--dist-eval]
[--num_workers NUM_WORKERS] [--pin-mem] [--no-pin-mem] [--initial-increment INITIAL_INCREMENT]
[--increment INCREMENT] [--class-order CLASS_ORDER [CLASS_ORDER ...]] [--eval-every EVAL_EVERY] [--debug]
[--retrain-scratch] [--max-task MAX_TASK] [--name NAME] [--options [OPTIONS ...]] [--DKT]
[--duplex-clf DUPLEX_CLF] [--memory-size MEMORY_SIZE] [--distributed-memory] [--global-memory]
[--oversample-memory OVERSAMPLE_MEMORY] [--oversample-memory-ft OVERSAMPLE_MEMORY_FT] [--rehearsal-test-trsf]
[--rehearsal-modes REHEARSAL_MODES] [--fixed-memory]
[--rehearsal {random,closest_token,closest_all,icarl_token,icarl_all,furthest_token,furthest_all}]
[--sep-memory] [--replay-memory REPLAY_MEMORY] [--finetuning {balanced}] [--finetuning-mode FINETUNING_MODE]
[--finetuning-lr FINETUNING_LR] [--finetuning-teacher] [--finetuning-resetclf] [--only-ft] [--ft-no-sampling]
[--freeze-task [FREEZE_TASK ...]] [--freeze-ft [FREEZE_FT ...]] [--freeze-eval] [--log-path LOG_PATH]
[--log-category LOG_CATEGORY] [--bce-loss] [--local_rank LOCAL_RANK] [--world_size WORLD_SIZE]
[--dist_url DIST_URL] [--resume RESUME] [--start-task START_TASK] [--start-epoch START_EPOCH]
[--save-every-epoch SAVE_EVERY_EPOCH] [--validation VALIDATION]
DKT training and evaluation script: error: unrecognized arguments: --local-rank=0
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 2) local_rank: 0 (pid: 633) of binary: /opt/conda/bin/python
Traceback (most recent call last):
File "/opt/conda/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/opt/conda/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/opt/conda/lib/python3.10/site-packages/torch/distributed/launch.py", line 196, in
main()
File "/opt/conda/lib/python3.10/site-packages/torch/distributed/launch.py", line 192, in main
launch(args)
File "/opt/conda/lib/python3.10/site-packages/torch/distributed/launch.py", line 177, in launch
run(args)
File "/opt/conda/lib/python3.10/site-packages/torch/distributed/run.py", line 785, in run
elastic_launch(
File "/opt/conda/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 134, in call
return launch_agent(self._config, self._entrypoint, list(args))
File "/opt/conda/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 250, in launch_agent
raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError:

/code/DKT_source/main.py FAILED

Failures:
[1]:
time : 2023-12-05_07:36:20
host : cd98f4d76410
rank : 1 (local_rank: 1)
exitcode : 2 (pid: 634)
error_file: <N/A>
traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html

Root Cause (first observed failure):
[0]:
time : 2023-12-05_07:36:20
host : cd98f4d76410
rank : 0 (local_rank: 0)
exitcode : 2 (pid: 633)
error_file: <N/A>
traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html