mirror of
https://github.com/open-mmlab/mmcv.git
synced 2025-06-03 21:54:52 +08:00
b326a219a7
* Use mdformat pre-commit hook * allows consecutive numbering * improve .mdformat.toml * test mdformat * format markdown * minor fix * fix codespecll * fix circleci * add linkify-it-py dependency for cicleci * add comments * replace flake8 url * add mdformat-myst dependency * remove mdformat-myst dependency * update contributing.md
69 lines
1.5 KiB
Markdown
69 lines
1.5 KiB
Markdown
## 辅助函数
|
||
|
||
### 进度条
|
||
|
||
如果你想跟踪函数批处理任务的进度，可以使用 `track_progress` 。它能以进度条的形式展示任务的完成情况以及剩余任务所需的时间（内部实现为for循环）。
|
||
|
||
```python
|
||
import mmcv
|
||
|
||
def func(item):
|
||
# 执行相关操作
|
||
pass
|
||
|
||
tasks = [item_1, item_2, ..., item_n]
|
||
|
||
mmcv.track_progress(func, tasks)
|
||
```
|
||
|
||
效果如下
|
||
|
||
|
||
如果你想可视化多进程任务的进度，你可以使用 `track_parallel_progress` 。
|
||
|
||
```python
|
||
mmcv.track_parallel_progress(func, tasks, 8) # 8 workers
|
||
```
|
||
|
||
|
||
|
||
如果你想要迭代或枚举数据列表并可视化进度,你可以使用 `track_iter_progress` 。
|
||
|
||
```python
|
||
import mmcv
|
||
|
||
tasks = [item_1, item_2, ..., item_n]
|
||
|
||
for task in mmcv.track_iter_progress(tasks):
|
||
# do something like print
|
||
print(task)
|
||
|
||
for i, task in enumerate(mmcv.track_iter_progress(tasks)):
|
||
# do something like print
|
||
print(i)
|
||
print(task)
|
||
```
|
||
|
||
### 计时器
|
||
|
||
mmcv提供的 `Timer` 可以很方便地计算代码块的执行时间。
|
||
|
||
```python
|
||
import time
|
||
|
||
with mmcv.Timer():
|
||
# simulate some code block
|
||
time.sleep(1)
|
||
```
|
||
|
||
你也可以使用 `since_start()` 和 `since_last_check()` 。前者返回计时器启动后的运行时长，后者返回最近一次查看计时器后的运行时长。
|
||
|
||
```python
|
||
timer = mmcv.Timer()
|
||
# code block 1 here
|
||
print(timer.since_start())
|
||
# code block 2 here
|
||
print(timer.since_last_check())
|
||
print(timer.since_start())
|
||
```
|