Skip to content

F
FluidDoc

PaddlePaddle / FluidDoc

通知 5
Star 2
Fork 0
F
FluidDoc
提交 9f91ebc4 编写于 作者: S songyouwei 提交者： zhongpu
浏览文件
操作

delete name_scope related docs (#1717) 

* del left name_scope docs
test=develop

* minor fix
test=develop
上级 7fbfcb6c
隐藏空白更改
内联 并排
Showing with 58 addition and 54 deletion
doc/fluid/api_cn/dygraph_cn/GRUUnit_cn.rst
浏览文件 @ 9f91ebc4
......@@ -33,7 +33,6 @@ GRUUnit
:math:`W_{uh}, b_u` 、 :math:`W_{rh}, b_r` 和 :math:`W_{ch}, b_c` 分别代表更新门、重置门和候选隐状态在计算时使用的权重矩阵和偏置。在实现上，三个权重矩阵合并为一个维度为 :math:`[D, D \times 3]` 的Tensor存放。
参数:
- **name_scope** (str) – 该类的名称。
- **size** (int) – 输入数据的维度大小。
- **param_attr** (ParamAttr，可选) – 指定权重参数属性的对象。默认值为None，表示使用默认的权重参数属性。具体用法请参见 :ref:`cn_api_fluid_ParamAttr` 。
**注意**
......@@ -64,7 +63,7 @@ GRUUnit
hidden_input = numpy.random.rand(T, D).astype('float32')
with fluid.dygraph.guard():
x = numpy.random.random((3, 32, 32)).astype('float32')
gru = fluid.dygraph.GRUUnit('gru', size=D * 3)
gru = fluid.dygraph.GRUUnit(size=D * 3)
dy_ret = gru(
base.to_variable(input), base.to_variable(hidden_input))
......
doc/fluid/api_cn/fluid_cn/Variable_cn.rst
浏览文件 @ 9f91ebc4
......@@ -207,7 +207,6 @@ Variable
# example2: 返回tuple of ndarray
with fluid.dygraph.guard():
embedding = fluid.dygraph.Embedding(
name_scope='embedding',
size=[20, 32],
param_attr='emb.w',
is_sparse=True)
......@@ -216,7 +215,7 @@ Variable
x = fluid.dygraph.base.to_variable(x_data)
out = embedding(x)
out.backward()
print(embedding._w.gradient())
print(embedding.weight.gradient())
.. py:method:: clear_gradient()
......
doc/fluid/user_guides/howto/dygraph/DyGraph.md
浏览文件 @ 9f91ebc4
......@@ -70,18 +70,20 @@ print(loss.gradient())
1. 编写一段用于DyGraph执行的Object-Oriented-Designed, PaddlePaddle模型代码主要由以下**两部分**组成： **请注意，如果您设计的这一层结构是包含参数的，则必须要使用继承自`fluid.dygraph.Layer`的Object-Oriented-Designed的类来描述该层的行为。**
1. 建立一个可以在DyGraph模式中执行的，Object-Oriented的网络，需要继承自`fluid.dygraph.Layer`，其中需要调用基类的`__init__`方法，并且实现带有参数`name_scope`（用来标识本层的名字）的`__init__`构造函数，在构造函数中，我们通常会执行一些例如参数初始化，子网络初始化的操作，执行这些操作时不依赖于输入的动态信息:
1. 建立一个可以在DyGraph模式中执行的，Object-Oriented的网络，需要继承自`fluid.dygraph.Layer`，其中需要调用基类的`__init__`方法，在构造函数中，我们通常会执行一些例如参数初始化，子网络初始化的操作，执行这些操作时不依赖于输入的动态信息:
```python
class MyLayer(fluid.dygraph.Layer):
def __init__(self, name_scope):
super(MyLayer, self).__init__(name_scope)
def __init__(self, input_size):
super(MyLayer, self).__init__()
self.linear = fluid.dygraph.nn.Linear(input_size, 12)
```
2. 实现一个`forward(self, *inputs)`的执行函数，该函数将负责执行实际运行时网络的执行逻辑， 该函数将会在每一轮训练/预测中被调用，这里我们将执行一个简单的`relu` -> `elementwise add` -> `reduce sum`：
2. 实现一个`forward(self, *inputs)`的执行函数，该函数将负责执行实际运行时网络的执行逻辑， 该函数将会在每一轮训练/预测中被调用，这里我们将执行一个简单的 `linear` -> `relu` -> `elementwise add` -> `reduce sum`：
```python
def forward(self, inputs):
x = self.linear(inputs)
x = fluid.layers.relu(inputs)
self._x_for_debug = x
x = fluid.layers.elementwise_mul(x, x)
......@@ -102,7 +104,7 @@ print(loss.gradient())
```python
with fluid.dygraph.guard():
var_inp = fluid.dygraph.to_variable(np_inp)
my_layer = MyLayer("my_layer")
my_layer = MyLayer(np_inp.shape[-1])
x = my_layer(var_inp)[0]
dy_out = x.numpy()
```
......@@ -122,12 +124,12 @@ import numpy as np
class MyLayer(fluid.dygraph.Layer):
def __init__(self, name_scope):
super(MyLayer, self).__init__(name_scope)
self.fc = fluid.dygraph.nn.FC(self.full_name(), size=12)
def __init__(self, input_size):
super(MyLayer, self).__init__()
self.linear = fluid.dygraph.nn.Linear(input_size, 12)
def forward(self, inputs):
x = self.fc(inputs)
x = self.linear(inputs)
x = fluid.layers.relu(x)
self._x_for_debug = x
x = fluid.layers.elementwise_mul(x, x)
......@@ -139,7 +141,7 @@ if __name__ == '__main__':
np_inp = np.array([[1.0, 2.0, -1.0]], dtype=np.float32)
with fluid.dygraph.guard():
var_inp = fluid.dygraph.to_variable(np_inp)
my_layer = MyLayer("my_layer")
my_layer = MyLayer(np_inp.shape[-1])
x = my_layer(var_inp)[0]
dy_out = x.numpy()
x.backward()
......@@ -187,8 +189,8 @@ with fluid.dygraph.guard():
value0 = np.arange(26).reshape(2, 13).astype("float32")
value1 = np.arange(6).reshape(2, 3).astype("float32")
value2 = np.arange(10).reshape(2, 5).astype("float32")
fc = fluid.FC("fc1", size=5, dtype="float32")
fc2 = fluid.FC("fc2", size=3, dtype="float32")
fc = fluid.Linear(13, 5, dtype="float32")
fc2 = fluid.Linear(3, 3, dtype="float32")
a = fluid.dygraph.to_variable(value0)
b = fluid.dygraph.to_variable(value1)
c = fluid.dygraph.to_variable(value2)
......@@ -198,7 +200,7 @@ with fluid.dygraph.guard():
out = fluid.layers.concat(input=[out1, out2, c], axis=1)
out.backward()
# 可以发现这里fc参数的梯度都为0
assert (fc._w.gradient() == 0).all()
assert (fc.weight.gradient() == 0).all()
assert (out1.gradient() == 0).all()
```
......@@ -220,7 +222,7 @@ with fluid.dygraph.guard():
```python
class SimpleImgConvPool(fluid.dygraph.Layer):
def __init__(self,
name_scope,
num_channels,
num_filters,
filter_size,
pool_size,
......@@ -236,10 +238,10 @@ with fluid.dygraph.guard():
use_cudnn=False,
param_attr=None,
bias_attr=None):
super(SimpleImgConvPool, self).__init__(name_scope)
super(SimpleImgConvPool, self).__init__()
self._conv2d = fluid.dygraph.Conv2D(
self.full_name(),
num_channels=num_channels,
num_filters=num_filters,
filter_size=filter_size,
stride=conv_stride,
......@@ -252,7 +254,6 @@ with fluid.dygraph.guard():
use_cudnn=use_cudnn)
self._pool2d = fluid.dygraph.Pool2D(
self.full_name(),
pool_size=pool_size,
pool_type=pool_type,
pool_stride=pool_stride,
......@@ -272,28 +273,30 @@ with fluid.dygraph.guard():
```python
class MNIST(fluid.dygraph.Layer):
def __init__(self, name_scope):
super(MNIST, self).__init__(name_scope)
def __init__(self):
super(MNIST, self).__init__()
self._simple_img_conv_pool_1 = SimpleImgConvPool(
self.full_name(), 20, 5, 2, 2, act="relu")
1, 20, 5, 2, 2, act="relu")
self._simple_img_conv_pool_2 = SimpleImgConvPool(
self.full_name(), 50, 5, 2, 2, act="relu")
20, 50, 5, 2, 2, act="relu")
pool_2_shape = 50 * 4 * 4
self.pool_2_shape = 50 * 4 * 4
SIZE = 10
scale = (2.0 / (pool_2_shape**2 * SIZE))**0.5
self._fc = fluid.dygraph.FC(self.full_name(),
10,
param_attr=fluid.param_attr.ParamAttr(
initializer=fluid.initializer.NormalInitializer(
loc=0.0, scale=scale)),
act="softmax")
scale = (2.0 / (self.pool_2_shape**2 * SIZE))**0.5
self._fc = fluid.dygraph.Linear(
self.pool_2_shape,
10,
param_attr=fluid.param_attr.ParamAttr(
initializer=fluid.initializer.NormalInitializer(
loc=0.0, scale=scale)),
act="softmax")
def forward(self, inputs, label=None):
x = self._simple_img_conv_pool_1(inputs)
x = self._simple_img_conv_pool_2(x)
x = fluid.layers.reshape(x, shape=[-1, self.pool_2_shape])
x = self._fc(x)
if label is not None:
acc = fluid.layers.accuracy(input=x, label=label)
......@@ -306,7 +309,9 @@ with fluid.dygraph.guard():
```python
with fluid.dygraph.guard():
mnist = MNIST("mnist")
mnist = MNIST()
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size=32, drop_last=True)
id, data = list(enumerate(train_reader()))[0]
dy_x_data = np.array(
[x[0].reshape(1, 28, 28)
......@@ -335,8 +340,8 @@ with fluid.dygraph.guard():
BATCH_SIZE = 64
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size=32, drop_last=True)
mnist = MNIST("mnist")
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001)
mnist = MNIST()
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001, parameter_list=mnist.parameters())
for epoch in range(epoch_num):
for batch_id, data in enumerate(train_reader()):
dy_x_data = np.array([x[0].reshape(1, 28, 28)
......@@ -370,8 +375,8 @@ with fluid.dygraph.guard():
epoch_num = 5
BATCH_SIZE = 64
mnist = MNIST("mnist")
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001)
mnist = MNIST()
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001, parameter_list=mnist.parameters())
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size= BATCH_SIZE, drop_last=True)
......@@ -452,16 +457,17 @@ with fluid.dygraph.guard():
```python
place = fluid.CUDAPlace(fluid.dygraph.parallel.Env().dev_id)
with fluid.dygraph.guard(place):
strategy = fluid.dygraph.parallel.prepare_context()
mnist = MNIST("mnist")
adam = AdamOptimizer(learning_rate=0.001)
epoch_num = 5
BATCH_SIZE = 64
mnist = MNIST()
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001, parameter_list=mnist.parameters())
mnist = fluid.dygraph.parallel.DataParallel(mnist, strategy)
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size=BATCH_SIZE, drop_last=True)
train_reader = fluid.contrib.reader.distributed_batch_reader(
train_reader)
train_reader)
for epoch in range(epoch_num):
for batch_id, data in enumerate(train_reader()):
......@@ -470,8 +476,8 @@ with fluid.dygraph.guard(place):
y_data = np.array(
[x[1] for x in data]).astype('int64').reshape(-1, 1)
img = to_variable(dy_x_data)
label = to_variable(y_data)
img = fluid.dygraph.to_variable(dy_x_data)
label = fluid.dygraph.to_variable(y_data)
label.stop_gradient = True
cost, acc = mnist(img, label)
......@@ -482,7 +488,7 @@ with fluid.dygraph.guard(place):
avg_loss = mnist.scale_loss(avg_loss)
avg_loss.backward()
mnist.apply_collective_grads()
adam.minimize(avg_loss)
mnist.clear_gradients()
if batch_id % 100 == 0 and batch_id is not 0:
......@@ -500,8 +506,8 @@ with fluid.dygraph.guard(place):
```python
strategy = fluid.dygraph.parallel.prepare_context()
mnist = MNIST("mnist")
adam = AdamOptimizer(learning_rate=0.001)
mnist = MNIST()
adam = AdamOptimizer(learning_rate=0.001, parameter_list=mnist.parameters())
mnist = fluid.dygraph.parallel.DataParallel(mnist, strategy)
```
......@@ -619,8 +625,8 @@ with fluid.dygraph.guard():
epoch_num = 5
BATCH_SIZE = 64
mnist = MNIST("mnist")
adam = fluid.optimizer.Adam(learning_rate=0.001)
mnist = MNIST()
adam = fluid.optimizer.Adam(learning_rate=0.001, parameter_list=mnist.parameters())
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size= BATCH_SIZE, drop_last=True)
......@@ -715,7 +721,7 @@ def test_mnist(reader, model, batch_size):
def inference_mnist():
with fluid.dygraph.guard():
mnist_infer = MNIST("mnist")
mnist_infer = MNIST()
# load checkpoint
model_dict, _ = fluid.dygraph.load_dygraph("paddle_dy")
mnist_infer.load_dict(model_dict)
......@@ -741,8 +747,8 @@ def inference_mnist():
with fluid.dygraph.guard():
epoch_num = 1
BATCH_SIZE = 64
mnist = MNIST("mnist")
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001)
mnist = MNIST()
adam = fluid.optimizer.AdamOptimizer(learning_rate=0.001, parameter_list=mnist.parameters())
test_reader = paddle.batch(
paddle.dataset.mnist.test(), batch_size=BATCH_SIZE, drop_last=True)
......@@ -818,8 +824,8 @@ epoch_num = 1
BATCH_SIZE = 64
exe = fluid.Executor(fluid.CPUPlace())
mnist = MNIST("mnist")
sgd = fluid.optimizer.SGDOptimizer(learning_rate=1e-3)
mnist = MNIST()
sgd = fluid.optimizer.SGDOptimizer(learning_rate=1e-3, parameter_list=mnist.parameters())
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size=BATCH_SIZE, drop_last=True)
img = fluid.layers.data(
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或