Command line returns error

[TOP-RX]

Hello,

I am trying to run the code for Arxiv, but it always returns:

Failed to operate CUDA_VISIBLE_DEVICES=0 /xxxx/xxxx --master_port=33019 --nproc_per_node=1 src/models/LMs/trainLM.py -mTinyBert --att_dropout=0.1 --cla_dropout=0.4 --dropout=0.3 --epochs=4 --eq_batch_size=36 --eval_patience=50000 --label_smoothing_factor=0.3 --load_best_model_at_end=T --lr=2e-05 --warmup_epochs=0.6 --save_folder=/xxx/GLEM-main/temp/prt_lm/arxiv_TA_0.002/TinyBert/FtV1/ -darxiv_TA_0.002 -g0 --em_iter=-1

it seems the os.system(cmd) will return a non-zero value, any help would be appreciated.

[xioacd99]

I also encountered the same problem, do you have a solution? : 0

[AndyJZhao]

Sorry for the late reply, I didn't receive the notifications for issues.
For this kind of problem, you might want to run the code on a single device and locate where the bug is.

[bulubulu624]

Sorry for the late reply, I didn't receive the notifications for issues. For this kind of problem, you might want to run the code on a single device and locate where the bug is.

净说一堆废话，屁用没有