Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

clean llama static modeling file #9301

Merged
merged 3 commits into from
Oct 23, 2024
Merged
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Next Next commit
clean llama static modeling file
Wennie396 authored and zhiqiu committed Oct 22, 2024
commit 9e93f92b9542b26d13f4c232fe428b28d389d6b0
6 changes: 1 addition & 5 deletions llm/auto_parallel/llama/README.md
Original file line number Diff line number Diff line change
@@ -3,15 +3,11 @@
## 1. 模型组网介绍

- 动静统一自动并行组网[modeling_auto.py](https://github.com/PaddlePaddle/PaddleNLP/blob/develop/paddlenlp/transformers/llama/modeling_auto.py),支持动态图和动转静训练,建议使用。
- 静态图自动并行组网[modeling_auto_static.py](https://github.com/PaddlePaddle/PaddleNLP/blob/develop/paddlenlp/transformers/llama/modeling_auto_static.py),仅支持静态图训练,未来可能会下线。


## 2. 动静统一组网的训练方式
- 动态图训练
参考训练脚本 **run_pretrain_auto.sh**,运行8卡 dp2mp2pp2的并行策略。
- 动转静训练
参考训练脚本 **run_pretrain_auto.sh**,并开启 `to_static=1`,运行8卡 dp2mp2pp2的并行策略。

## 3. 静态图组网的训练方式

参考训练脚本 **run_pretrain_auto_static.sh**,运行8卡 dp2sharding2mp2pp2vpp2的并行策略。
参考训练脚本 **run_pretrain_auto_static_sp.sh**,运行8卡 dp2sharding2mp2pp2vpp2sp 的并行策略。
Loading