Updated README with new information

This commit is contained in:
刘一博 2024-03-20 14:11:28 +08:00
parent 8e04794b2d
commit df9b4fb90a
3 changed files with 4 additions and 1 deletions

BIN
.DS_Store vendored Normal file

Binary file not shown.

View File

@ -519,7 +519,8 @@ use_cpu: false
```bash
deepspeed --num_gpus 8 src/train_bash.py \
--deepspeed ds_config.json \
--deepspeed ds_config.json \
--ddp_timeout 180000000 \ # If the training data is too large, it is recommended to add the ddp_timeout command line option to prevent NCCL errors.
... # arguments (same as above)
```

View File

@ -519,7 +519,9 @@ use_cpu: false
```bash
deepspeed --num_gpus 8 src/train_bash.py \
--deepspeed ds_config.json \
--ddp_timeout 180000000 \ # 如训练数据过大建议加上ddp_timeout命令行防止nccl报错
... # 参数同上
```
<details><summary>使用 DeepSpeed ZeRO-2 进行全参数训练的 ds_config.json 示例</summary>