wehub-resource-sync
139 lines
4.0 KiB
Python
139 lines
4.0 KiB
Python
# Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
import numpy as np
|
|
|
|
import paddle
|
|
import paddle.distributed as dist
|
|
from paddle import nn
|
|
from paddle.autograd import PyLayer
|
|
from paddle.io import DataLoader
|
|
|
|
|
|
class DemoPyLayer(PyLayer):
|
|
@staticmethod
|
|
def forward(ctx, x):
|
|
ctx.save_for_backward(x)
|
|
y = paddle.tanh(x)
|
|
|
|
return y
|
|
|
|
@staticmethod
|
|
def backward(ctx, dy):
|
|
(x,) = ctx.saved_tensor()
|
|
grad = dy * (1 - paddle.square(x))
|
|
return grad
|
|
|
|
|
|
class DemoNet(nn.Layer):
|
|
def __init__(self, mesh):
|
|
super().__init__()
|
|
self.mesh = mesh
|
|
self.linear1 = paddle.nn.Linear(16, 16)
|
|
|
|
def forward(self, x):
|
|
x = dist.shard_tensor(x, self.mesh, [dist.Shard(0)]) # shard tensor
|
|
y = self.linear1(x)
|
|
return DemoPyLayer.apply(y)
|
|
|
|
|
|
class RandomDataset(paddle.io.Dataset):
|
|
def __init__(self, images, labels, num_samples):
|
|
self.images = images
|
|
self.labels = labels
|
|
self.num_samples = num_samples
|
|
|
|
def __getitem__(self, idx):
|
|
return self.images[idx], self.labels[idx]
|
|
|
|
def __len__(self):
|
|
return self.num_samples
|
|
|
|
|
|
def test_pylayer():
|
|
mesh = dist.ProcessMesh([0], dim_names=['x'])
|
|
images = np.random.rand(4, 16).astype('float32')
|
|
labels = np.random.rand(4, 16).astype('float32')
|
|
dataset = RandomDataset(images, labels, 4)
|
|
loader = DataLoader(dataset, batch_size=4)
|
|
|
|
layer = DemoNet(mesh)
|
|
opt = paddle.optimizer.SGD(learning_rate=0.1, parameters=layer.parameters())
|
|
mse_loss = paddle.nn.loss.MSELoss()
|
|
epoch = 2
|
|
|
|
# to static
|
|
dist_model = dist.to_static(layer, loader, mse_loss, opt)
|
|
dist_model.train()
|
|
for batch_id, data in enumerate(loader()):
|
|
img, label = data
|
|
label.stop_gradient = True
|
|
loss = dist_model(img, label)
|
|
|
|
|
|
class DemoPyLayerCustom(PyLayer):
|
|
@staticmethod
|
|
def forward(ctx, x, y, z):
|
|
ctx.save_for_backward(x, y, z)
|
|
x1 = paddle.tanh(x)
|
|
y1 = paddle.tanh(y)
|
|
z1 = paddle.tanh(z)
|
|
return x1 + y1 + z1
|
|
|
|
@staticmethod
|
|
def backward(ctx, grad):
|
|
x, y, z = ctx.saved_tensor()
|
|
x_grad = grad * (1 - paddle.square(x))
|
|
y_grad = grad * (1 - paddle.square(y))
|
|
return x_grad, y_grad, None
|
|
|
|
|
|
class DemoNetCustom(nn.Layer):
|
|
def __init__(self, mesh):
|
|
super().__init__()
|
|
self.mesh = mesh
|
|
self.linear1 = paddle.nn.Linear(16, 16)
|
|
self.linear2 = paddle.nn.Linear(16, 16)
|
|
self.linear3 = paddle.nn.Linear(16, 16)
|
|
|
|
def forward(self, x):
|
|
x = dist.shard_tensor(x, self.mesh, [dist.Shard(0)]) # shard tensor
|
|
x = self.linear1(x)
|
|
y = self.linear2(x)
|
|
z = self.linear3(x)
|
|
z.stop_gradient = True
|
|
out = DemoPyLayerCustom.apply(x, y, z)
|
|
return out
|
|
|
|
|
|
def test_pylayer_custom_op():
|
|
mesh = dist.ProcessMesh([0], dim_names=['x'])
|
|
images = np.random.rand(4, 16).astype('float32')
|
|
labels = np.random.rand(4, 16).astype('float32')
|
|
dataset = RandomDataset(images, labels, 4)
|
|
loader = DataLoader(dataset, batch_size=4)
|
|
|
|
layer = DemoNetCustom(mesh)
|
|
opt = paddle.optimizer.SGD(learning_rate=0.1, parameters=layer.parameters())
|
|
mse_loss = paddle.nn.loss.MSELoss()
|
|
epoch = 2
|
|
|
|
# to static
|
|
dist_model = dist.to_static(layer, loader, mse_loss, opt)
|
|
dist_model.train()
|
|
for batch_id, data in enumerate(loader()):
|
|
img, label = data
|
|
label.stop_gradient = True
|
|
loss = dist_model(img, label)
|