mirrors/mmengine
1
0
Fork 0
mirror of https://github.com/open-mmlab/mmengine.git synced 2025-06-03 21:54:44 +08:00
Code Issues Projects Releases Wiki Activity

[Docs] Refine hook docs (#704)

Browse Source
This commit is contained in:
Zaida Zhou 2022-11-09 10:37:52 +08:00
parent af0e29094b
commit fed4a9224b
1 changed files with 6 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
docs/zh_cn/tutorials/hook.md
View File

@ -32,11 +32,10 @@ MMEngine 提供了很多内置的钩子，将钩子分为两类，分别是默
**自定义钩子** **自定义钩子**
| 名称 | 用途 | 优先级 | | 名称 | 用途 | 优先级 |
| :---------------------------------: | :-------------------: | :----------: | | :---------------------------------: | :-------------------: | :---------: |
| [EMAHook](#emahook) | 模型参数指数滑动平均 | NORMAL (50) | | [EMAHook](#emahook) | 模型参数指数滑动平均 | NORMAL (50) |
| [EmptyCacheHook](#emptycachehook) | PyTorch CUDA 缓存清理 | NORMAL (50) | | [EmptyCacheHook](#emptycachehook) | PyTorch CUDA 缓存清理 | NORMAL (50) |
| [SyncBuffersHook](#syncbuffershook) | 同步模型的 buffer | NORMAL (50) | | [SyncBuffersHook](#syncbuffershook) | 同步模型的 buffer | NORMAL (50) |
| NaiveVisualizationHook | 可视化 | LOWEST (100) |
```{note} ```{note}
不建议修改默认钩子的优先级，因为优先级低的钩子可能会依赖优先级高的钩子。例如 CheckpointHook 的优先级需要比 ParamSchedulerHook 低，这样保存的优化器状态才是正确的状态。另外，自定义钩子的优先级默认为 `NORMAL (50)`。 不建议修改默认钩子的优先级，因为优先级低的钩子可能会依赖优先级高的钩子。例如 CheckpointHook 的优先级需要比 ParamSchedulerHook 低，这样保存的优化器状态才是正确的状态。另外，自定义钩子的优先级默认为 `NORMAL (50)`。
@ -56,9 +55,7 @@ default_hooks = dict(
checkpoint=dict(type='CheckpointHook', interval=1), checkpoint=dict(type='CheckpointHook', interval=1),
) )
custom_hooks = [ custom_hooks = [dict(type='EmptyCacheHook')]
dict(type='NaiveVisualizationHook', priority='LOWEST'),
]
runner = Runner(default_hooks=default_hooks, custom_hooks=custom_hooks, ...) runner = Runner(default_hooks=default_hooks, custom_hooks=custom_hooks, ...)
runner.train() runner.train()