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

Update unified_checkpoint.md #9634

Merged
merged 1 commit into from
Dec 13, 2024
Merged
Changes from all commits
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
4 changes: 2 additions & 2 deletions llm/docs/unified_checkpoint.md
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,7 @@ unified_checkpoint_config: Optional[str] = field(
"- master_weight_compatible: 1. if the master weights exist, only load when needed\n"
" 2. if master weights does not exist, convert model weights to master weights when needed\n"
"- remove_master_weight, whether save master weight, if master_weights does not exist, convert model weights to master_weights when needed."
"- enable_all_options: enable all optimization configurations\n"
"- async_save: whether to use asynchronous saving."
)
},
)
Expand All @@ -51,7 +51,7 @@ unified_checkpoint_config: Optional[str] = field(
1. skip_save_model_weight:当 optimizer 具有 master weight 时,跳过 model weight 保存,重启时将 master weight 作为 model weight 加载。在 PaddleNLP 中,仅 fp16_opt_level=O1时,optimizer 不存在 master weight。
2. master_weight_compatible:仅当 optimizer 需要 master weight 时,才加载 master weight; 如果 ckpt 中不存在 master weight,将 model weight 作为 master weight 加载。
3. remove_master_weight: 是否保存 master weight, 如果 checkpoint 中不存在 master_weights,则将 model weight 作为 master_weights 进行加载。
4. enable_all_options:上述参数均开启
4. async_save: 是否使用异步保存,加速存储速度

### 2.2 Unified Checkpoint 存储格式介绍

Expand Down