diff --git a/README.md b/README.md index e35e129..ee3b8a1 100644 --- a/README.md +++ b/README.md @@ -388,11 +388,13 @@ torchrun --nproc_per_node N train_xxx.py
注:其它须知 + 单机N卡启动训练 (DeepSpeed) ```bash deepspeed --master_port 29500 --num_gpus=N train_xxx.py ``` + 可根据需要开启wandb记录训练过程(需可直连) diff --git a/README_en.md b/README_en.md index 366af4a..4c2759c 100644 --- a/README_en.md +++ b/README_en.md @@ -381,11 +381,13 @@ torchrun --nproc_per_node N train_xxx.py
Note: Other Notes + Single machine N GPU training (DeepSpeed) ```bash deepspeed --master_port 29500 --num_gpus=N train_xxx.py ``` + You can optionally enable wandb to record the training process (requires direct internet connection)