DeepSpeedExamples "ds_train_bert_nvidia_data_bsz64k_seq128.sh" program stalls at the end of the first epoch

"ds_train_bert_nvidia_data_bsz64k_seq128.sh" program stalls at the end of the first epoch

Open inspur-hsslab opened this issue 4 years ago • 2 comments

When I run "ds_train_bert_nvidia_data_bsz64k_seq128.sh". It stalls at the end of the first epoch.

Aug 05 '21 08:08 inspur-hsslab

met the same problem

Nov 24 '21 06:11 dancingpipi

Met the same problem with DeepSpeed v0.6.1.

Apr 25 '22 11:04 haolin-nju