[25.05.20] update code support resume from checkpoints

2025-05-20 11:31:35 +08:00
parent 646c182469
commit b262507cb8
9 changed files with 70 additions and 172 deletions
@@ -1,2 +1,2 @@
 deepspeed --num_gpus=4 --master_port=12345 pretrain_transformers.py
-#   deepspeed --include="localhost:1,2,3" --master_port=12345 pretrain_transformers.py
+# CUDA_VISIBLE_DEVICES=1,2,3 deepspeed --include="localhost:1,2,3" --master_port=12345 src_pretrain/pretrain_deepspeed.py