mirror of https://github.com/alibaba/MNN.git
70 lines
1.7 KiB
Markdown
70 lines
1.7 KiB
Markdown
## optim
|
||
|
||
```python
|
||
module optim
|
||
```
|
||
optim时优化器模块，提供了一个优化器基类`Optimizer`，并提供了`SGD`和`ADAM`优化器实现；主要用于训练阶段迭代优化
|
||
|
||
---
|
||
### `optim Types`
|
||
- [Optimizer](Optimizer.md)
|
||
|
||
---
|
||
### `optim.Regularization_Method`
|
||
优化器的正则化方法，提供了L1和L2正则化方法
|
||
- 类型：`Enum`
|
||
- 枚举值：
|
||
- `L1`
|
||
- `L2`
|
||
- `L1L2`
|
||
|
||
---
|
||
### `SGD(module, lr, momentum, weight_decay, regularization_method)`
|
||
创建一个SGD优化器
|
||
|
||
参数：
|
||
- `module:_Module` 模型实例
|
||
- `lr:float` 学习率
|
||
- `momentum:float` 动量，默认为0.9
|
||
- `weight_decay:float` 权重衰减，默认为0.0
|
||
- `regularization_method:RegularizationMethod` 正则化方法，默认为L2正则化
|
||
|
||
返回：SGD优化器实例
|
||
|
||
返回类型：`Optimizer`
|
||
|
||
示例：
|
||
|
||
```python
|
||
model = Net()
|
||
sgd = optim.SGD(model, 0.001, 0.9, 0.0005, optim.Regularization_Method.L2)
|
||
# feed some date to the model, then get the loss
|
||
loss = ...
|
||
sgd.step(loss) # backward and update parameters in the model
|
||
```
|
||
---
|
||
### `ADAM(module, lr, momentum, momentum2, weight_decay, eps, regularization_method)`
|
||
创建一个ADAM优化器
|
||
|
||
参数：
|
||
- `module:_Module` 模型实例
|
||
- `lr:float` 学习率
|
||
- `momentum:float` 动量，默认为0.9
|
||
- `momentum2:float` 动量2，默认为0.999
|
||
- `weight_decay:float` 权重衰减，默认为0.0
|
||
- `eps:float` 正则化阈值，默认为1e-8
|
||
- `regularization_method:RegularizationMethod` 正则化方法，默认为L2正则化
|
||
|
||
返回：ADAM优化器实例
|
||
|
||
返回类型：`Optimizer`
|
||
|
||
示例：
|
||
|
||
```python
|
||
model = Net()
|
||
sgd = optim.ADAM(model, 0.001)
|
||
# feed some date to the model, then get the loss
|
||
loss = ...
|
||
sgd.step(loss) # backward and update parameters in the model
|
||
``` |