Skip to content

P
PaddleDetection

PaddlePaddle / PaddleDetection
大约 1 年 前同步成功

通知 695
Star 11112
Fork 2696
P
PaddleDetection
提交 e2ca8e10 编写于 作者: T typhoonzero
浏览文件
操作

Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into threadpool_for_io

上级 92818baa 129859e7
展开全部 隐藏空白更改
内联 并排
Showing with 12383 addition and 3153 deletion
.gitignore
浏览文件 @ e2ca8e10
......@@ -25,12 +25,3 @@ third_party/
# clion workspace.
cmake-build-*
# generated while compiling
paddle/pybind/pybind.h
CMakeFiles
cmake_install.cmake
paddle/.timestamp
python/paddlepaddle.egg-info/
paddle/fluid/pybind/pybind.h
python/paddle/version.py
benchmark/cluster/README.md
浏览文件 @ e2ca8e10
......@@ -36,11 +36,41 @@
- Trainer Count: 100
- Metrics: mini-batch / sec
| Batch Size | 32 | 64 | 128 | 256 |
| -- | -- | -- | -- | -- |
| PaddlePaddle Fluid | - | - | - | - |
| PaddlePaddle v2 | - | - | - | - |
| TensorFlow | - | - | - | - |
<table>
<thead>
<tr>
<th>Batch Size </th>
<th> 32</th>
<th>64</th>
<th>128 </th>
<th>256</th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid</td>
<td>-</td>
<td>- </td>
<td>- </td>
<td>- </td>
</tr>
<tr>
<td>PaddlePaddle v2 </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>- </td>
</tr>
<tr>
<td>TensorFlow </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>- </td>
</tr>
</tbody>
</table>
### Measure the Performance for Different PServer Count
......@@ -48,11 +78,41 @@
- Batch Size: 64
- Metrics: mini-batch / sec
| PServer Count | 10 | 20 | 40 | 60 |
| -- | -- | -- | -- | -- |
| PaddlePaddle Fluid | - | - | - | - |
| PaddlePaddle v2 | - | - | - | - |
| TensorFlow | - | - | - | - |
<table>
<thead>
<tr>
<th>PServer Count </th>
<th>10</th>
<th>20</th>
<th>40 </th>
<th>60</th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid</td>
<td>-</td>
<td>- </td>
<td>- </td>
<td>- </td>
</tr>
<tr>
<td>PaddlePaddle v2 </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>- </td>
</tr>
<tr>
<td>TensorFlow </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>- </td>
</tr>
</tbody>
</table>
### Measure Parallel Efficiency By Increasing Trainer Count
......@@ -67,11 +127,69 @@ The parallel efficiency is:
$E = \div(S, N)$
| Trainer Counter | 1 | 10 | 20 | 30 | 40 | 50 | 60 | 70 | 80 | 90 | 100 |
| -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- |
| PaddlePaddle Fluid | - | - | - | - | - | - | - | - | - | - | - |
| PaddlePaddle v2 | - | - | - | - | - | - | - | - | - | - | - | - |
| TensorFlow | - | - | - | - | - | - | - | - | - | - | - | - | - |
<table>
<thead>
<tr>
<th>Trainer Counter </th>
<th>1</th>
<th>10</th>
<th>20 </th>
<th>30</th>
<th>40</th>
<th>50</th>
<th>60 </th>
<th>70</th>
<th>80</th>
<th>90</th>
<th>100 </th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid</td>
<td>-</td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>-</td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>-</td>
<td>- </td>
<td>- </td>
</tr>
<tr>
<td>PaddlePaddle v2 </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>-</td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>-</td>
<td>- </td>
<td>- </td>
</tr>
<tr>
<td>TensorFlow </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>-</td>
<td>- </td>
<td>- </td>
<td>- </td>
<td>-</td>
<td>- </td>
<td>- </td>
</tr>
</tbody>
</table>
## Reproduce the benchmark
......
benchmark/cluster/vgg16/README.md
浏览文件 @ e2ca8e10
......@@ -16,11 +16,41 @@ Setting environment variable: `MKL_NUM_THREADS=1`.
- Metrics: samples / sec
| Batch Size | 32 | 64 | 128 | 256 |
| -- | -- | -- | -- | -- |
| PaddlePaddle Fluid | 15.44 | 16.32 | 16.74 | 16.79 |
| PaddlePaddle v2 | 15.97 | 17.04 | 17.60 | 17.83 |
| TensorFlow | 9.09 | 9.10 | 9.24 | 8.66 |
<table>
<thead>
<tr>
<th>Batch Size </th>
<th> 32</th>
<th>64</th>
<th>128 </th>
<th>256</th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid</td>
<td> 15.44 </td>
<td> 16.32 </td>
<td> 16.74 </td>
<td> 16.79 </td>
</tr>
<tr>
<td>PaddlePaddle v2 </td>
<td> 15.97 </td>
<td> 17.04 </td>
<td> 17.60 </td>
<td> 17.83 </td>
</tr>
<tr>
<td>TensorFlow </td>
<td> 9.09 </td>
<td> 9.10 </td>
<td> 9.24 </td>
<td> 8.66 </td>
</tr>
</tbody>
</table>
### Different Batch Size
......@@ -28,12 +58,40 @@ Setting environment variable: `MKL_NUM_THREADS=1`.
- Trainer Count: 20
- Metrics: samples / sec
| Batch Size | 32 | 64 | 128 | 256 |
| -- | -- | -- | -- | -- |
| PaddlePaddle Fluid | 190.20 | 222.15 | 247.40 | 258.18 |
| PaddlePaddle v2 | 170.96 | 233.71 | 256.14 | 329.23 |
| TensorFlow | - | - | - | - |
<table>
<thead>
<tr>
<th>Batch Size </th>
<th> 32</th>
<th>64</th>
<th>128 </th>
<th>256</th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid</td>
<td> 190.20 </td>
<td> 222.15 </td>
<td> 247.40 </td>
<td> 258.18 </td>
</tr>
<tr>
<td>PaddlePaddle v2 </td>
<td> 170.96 </td>
<td> 233.71 </td>
<td> 256.14 </td>
<td> 329.23 </td>
</tr>
<tr>
<td>TensorFlow </td>
<td> - </td>
<td> - </td>
<td> - </td>
<td> - </td>
</tr>
</tbody>
</table>
### Accelerate Rate
......@@ -41,11 +99,41 @@ Setting environment variable: `MKL_NUM_THREADS=1`.
- Batch Size: 128
- Metrics: samples / sec
| Trainer Count | 20 | 40 | 80 | 100 |
| -- | -- | -- | -- | -- |
| PaddlePaddle Fluid | 263.29 (78.64%) | 518.80 (77.47%) | 836.26 (62.44%) | 1019.29 (60.89%) |
| PaddlePaddle v2 (need more tests) | 326.85 (92.85%) | 534.58 (75.93%) | 853.30 (60.60%) | 1041.99 (59.20%) |
| TensorFlow | - | - | - | - |
<table>
<thead>
<tr>
<th>Trainer Count </th>
<th>20</th>
<th>40</th>
<th>80</th>
<th>100</th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid</td>
<td> 263.29 (78.64%) </td>
<td> 518.80 (77.47%) </td>
<td> 836.26 (62.44%) </td>
<td> 1019.29 (60.89%) </td>
</tr>
<tr>
<td>PaddlePaddle v2 (need more tests) </td>
<td> 326.85 (92.85%) </td>
<td> 534.58 (75.93%) </td>
<td> 853.30 (60.60%) </td>
<td> 1041.99 (59.20%) </td>
</tr>
<tr>
<td>TensorFlow </td>
<td> - </td>
<td> - </td>
<td> - </td>
<td> - </td>
</tr>
</tbody>
</table>
### Different Pserver Count
......@@ -53,11 +141,41 @@ Setting environment variable: `MKL_NUM_THREADS=1`.
- Batch Size: 128
- Metrics: samples/ sec
| PServer Count | 3 | 6 |10 | 20 |
| -- | -- | -- | -- | -- |
| PaddlePaddle Fluid(should fix in next PR) | 589.1 | 592.6 | 656.4 | 655.8 |
| PaddlePaddle v2 | 593.4 | 791.3 | 729.7 | 821.7 |
| TensorFlow | - | - | - | - |
<table>
<thead>
<tr>
<th>PServer Count </th>
<th>3</th>
<th>6</th>
<th>10</th>
<th>20</th>
</tr>
</thead>
<tbody>
<tr>
<td> PaddlePaddle Fluid(should fix in next PR) </td>
<td> 589.1 </td>
<td> 592.6 </td>
<td> 656.4 </td>
<td> 655.8 </td>
</tr>
<tr>
<td>PaddlePaddle v2 (need more tests) </td>
<td> 593.4 </td>
<td> 791.3 </td>
<td> 729.7 </td>
<td> 821.7 </td>
</tr>
<tr>
<td>TensorFlow </td>
<td> - </td>
<td> - </td>
<td> - </td>
<td> - </td>
</tr>
</tbody>
</table>
*The performance gap between Fuild and v2 comes from the network interference.*
......
benchmark/fluid/machine_translation.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
"""seq2seq model for fluid."""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import numpy as np
import argparse
import time
import distutils.util
import paddle.v2 as paddle
import paddle.fluid as fluid
import paddle.fluid.core as core
import paddle.fluid.framework as framework
from paddle.fluid.executor import Executor
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
"--embedding_dim",
type=int,
default=512,
help="The dimension of embedding table. (default: %(default)d)")
parser.add_argument(
"--encoder_size",
type=int,
default=512,
help="The size of encoder bi-rnn unit. (default: %(default)d)")
parser.add_argument(
"--decoder_size",
type=int,
default=512,
help="The size of decoder rnn unit. (default: %(default)d)")
parser.add_argument(
"--batch_size",
type=int,
default=16,
help="The sequence number of a mini-batch data. (default: %(default)d)")
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test')
parser.add_argument(
'--iterations', type=int, default=80, help='The number of minibatches.')
parser.add_argument(
"--dict_size",
type=int,
default=30000,
help="The dictionary capacity. Dictionaries of source sequence and "
"target dictionary have same capacity. (default: %(default)d)")
parser.add_argument(
"--pass_num",
type=int,
default=2,
help="The pass number to train. (default: %(default)d)")
parser.add_argument(
"--learning_rate",
type=float,
default=0.0002,
help="Learning rate used to train the model. (default: %(default)f)")
parser.add_argument(
"--infer_only", action='store_true', help="If set, run forward only.")
parser.add_argument(
"--beam_size",
type=int,
default=3,
help="The width for beam searching. (default: %(default)d)")
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help="The device type.")
parser.add_argument(
"--max_length",
type=int,
default=250,
help="The maximum length of sequence when doing generation. "
"(default: %(default)d)")
parser.add_argument(
'--with_test',
action='store_true',
help='If set, test the testset during training.')
def lstm_step(x_t, hidden_t_prev, cell_t_prev, size):
def linear(inputs):
return fluid.layers.fc(input=inputs, size=size, bias_attr=True)
forget_gate = fluid.layers.sigmoid(x=linear([hidden_t_prev, x_t]))
input_gate = fluid.layers.sigmoid(x=linear([hidden_t_prev, x_t]))
output_gate = fluid.layers.sigmoid(x=linear([hidden_t_prev, x_t]))
cell_tilde = fluid.layers.tanh(x=linear([hidden_t_prev, x_t]))
cell_t = fluid.layers.sums(input=[
fluid.layers.elementwise_mul(
x=forget_gate, y=cell_t_prev), fluid.layers.elementwise_mul(
x=input_gate, y=cell_tilde)
])
hidden_t = fluid.layers.elementwise_mul(
x=output_gate, y=fluid.layers.tanh(x=cell_t))
return hidden_t, cell_t
def seq_to_seq_net(embedding_dim, encoder_size, decoder_size, source_dict_dim,
target_dict_dim, is_generating, beam_size, max_length):
"""Construct a seq2seq network."""
def bi_lstm_encoder(input_seq, gate_size):
# Linear transformation part for input gate, output gate, forget gate
# and cell activation vectors need be done outside of dynamic_lstm.
# So the output size is 4 times of gate_size.
input_forward_proj = fluid.layers.fc(input=input_seq,
size=gate_size * 4,
act=None,
bias_attr=False)
forward, _ = fluid.layers.dynamic_lstm(
input=input_forward_proj, size=gate_size * 4, use_peepholes=False)
input_reversed_proj = fluid.layers.fc(input=input_seq,
size=gate_size * 4,
act=None,
bias_attr=False)
reversed, _ = fluid.layers.dynamic_lstm(
input=input_reversed_proj,
size=gate_size * 4,
is_reverse=True,
use_peepholes=False)
return forward, reversed
src_word_idx = fluid.layers.data(
name='source_sequence', shape=[1], dtype='int64', lod_level=1)
src_embedding = fluid.layers.embedding(
input=src_word_idx,
size=[source_dict_dim, embedding_dim],
dtype='float32')
src_forward, src_reversed = bi_lstm_encoder(
input_seq=src_embedding, gate_size=encoder_size)
encoded_vector = fluid.layers.concat(
input=[src_forward, src_reversed], axis=1)
encoded_proj = fluid.layers.fc(input=encoded_vector,
size=decoder_size,
bias_attr=False)
backward_first = fluid.layers.sequence_pool(
input=src_reversed, pool_type='first')
decoder_boot = fluid.layers.fc(input=backward_first,
size=decoder_size,
bias_attr=False,
act='tanh')
def lstm_decoder_with_attention(target_embedding, encoder_vec, encoder_proj,
decoder_boot, decoder_size):
def simple_attention(encoder_vec, encoder_proj, decoder_state):
decoder_state_proj = fluid.layers.fc(input=decoder_state,
size=decoder_size,
bias_attr=False)
decoder_state_expand = fluid.layers.sequence_expand(
x=decoder_state_proj, y=encoder_proj)
concated = fluid.layers.concat(
input=[encoder_proj, decoder_state_expand], axis=1)
attention_weights = fluid.layers.fc(input=concated,
size=1,
act='tanh',
bias_attr=False)
attention_weights = fluid.layers.sequence_softmax(
input=attention_weights)
weigths_reshape = fluid.layers.reshape(
x=attention_weights, shape=[-1])
scaled = fluid.layers.elementwise_mul(
x=encoder_vec, y=weigths_reshape, axis=0)
context = fluid.layers.sequence_pool(input=scaled, pool_type='sum')
return context
rnn = fluid.layers.DynamicRNN()
cell_init = fluid.layers.fill_constant_batch_size_like(
input=decoder_boot,
value=0.0,
shape=[-1, decoder_size],
dtype='float32')
cell_init.stop_gradient = False
with rnn.block():
current_word = rnn.step_input(target_embedding)
encoder_vec = rnn.static_input(encoder_vec)
encoder_proj = rnn.static_input(encoder_proj)
hidden_mem = rnn.memory(init=decoder_boot, need_reorder=True)
cell_mem = rnn.memory(init=cell_init)
context = simple_attention(encoder_vec, encoder_proj, hidden_mem)
decoder_inputs = fluid.layers.concat(
input=[context, current_word], axis=1)
h, c = lstm_step(decoder_inputs, hidden_mem, cell_mem, decoder_size)
rnn.update_memory(hidden_mem, h)
rnn.update_memory(cell_mem, c)
out = fluid.layers.fc(input=h,
size=target_dict_dim,
bias_attr=True,
act='softmax')
rnn.output(out)
return rnn()
if not is_generating:
trg_word_idx = fluid.layers.data(
name='target_sequence', shape=[1], dtype='int64', lod_level=1)
trg_embedding = fluid.layers.embedding(
input=trg_word_idx,
size=[target_dict_dim, embedding_dim],
dtype='float32')
prediction = lstm_decoder_with_attention(trg_embedding, encoded_vector,
encoded_proj, decoder_boot,
decoder_size)
label = fluid.layers.data(
name='label_sequence', shape=[1], dtype='int64', lod_level=1)
cost = fluid.layers.cross_entropy(input=prediction, label=label)
avg_cost = fluid.layers.mean(x=cost)
feeding_list = ["source_sequence", "target_sequence", "label_sequence"]
return avg_cost, feeding_list
def to_lodtensor(data, place):
seq_lens = [len(seq) for seq in data]
cur_len = 0
lod = [cur_len]
for l in seq_lens:
cur_len += l
lod.append(cur_len)
flattened_data = np.concatenate(data, axis=0).astype("int64")
flattened_data = flattened_data.reshape([len(flattened_data), 1])
lod_t = core.LoDTensor()
lod_t.set(flattened_data, place)
lod_t.set_lod([lod])
return lod_t, lod[-1]
def lodtensor_to_ndarray(lod_tensor):
dims = lod_tensor.get_dims()
ndarray = np.zeros(shape=dims).astype('float32')
for i in xrange(np.product(dims)):
ndarray.ravel()[i] = lod_tensor.get_float_element(i)
return ndarray
def train():
avg_cost, feeding_list = seq_to_seq_net(
args.embedding_dim,
args.encoder_size,
args.decoder_size,
args.dict_size,
args.dict_size,
False,
beam_size=args.beam_size,
max_length=args.max_length)
# clone from default main program
inference_program = fluid.default_main_program().clone()
optimizer = fluid.optimizer.Adam(learning_rate=args.learning_rate)
optimizer.minimize(avg_cost)
fluid.memory_optimize(fluid.default_main_program())
train_batch_generator = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.wmt14.train(args.dict_size), buf_size=1000),
batch_size=args.batch_size)
test_batch_generator = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.wmt14.test(args.dict_size), buf_size=1000),
batch_size=args.batch_size)
place = core.CPUPlace() if args.device == 'CPU' else core.CUDAPlace(0)
exe = Executor(place)
exe.run(framework.default_startup_program())
def do_validation():
total_loss = 0.0
count = 0
for batch_id, data in enumerate(test_batch_generator()):
src_seq = to_lodtensor(map(lambda x: x[0], data), place)[0]
trg_seq = to_lodtensor(map(lambda x: x[1], data), place)[0]
lbl_seq = to_lodtensor(map(lambda x: x[2], data), place)[0]
fetch_outs = exe.run(inference_program,
feed={
feeding_list[0]: src_seq,
feeding_list[1]: trg_seq,
feeding_list[2]: lbl_seq
},
fetch_list=[avg_cost],
return_numpy=False)
total_loss += lodtensor_to_ndarray(fetch_outs[0])[0]
count += 1
return total_loss / count
iters, num_samples, start_time = 0, 0, time.time()
for pass_id in xrange(args.pass_num):
train_accs = []
train_losses = []
for batch_id, data in enumerate(train_batch_generator()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
src_seq, word_num = to_lodtensor(map(lambda x: x[0], data), place)
num_samples += word_num
trg_seq, word_num = to_lodtensor(map(lambda x: x[1], data), place)
num_samples += word_num
lbl_seq, _ = to_lodtensor(map(lambda x: x[2], data), place)
fetch_outs = exe.run(framework.default_main_program(),
feed={
feeding_list[0]: src_seq,
feeding_list[1]: trg_seq,
feeding_list[2]: lbl_seq
},
fetch_list=[avg_cost])
iters += 1
loss = np.array(fetch_outs[0])
print(
"Pass = %d, Iter = %d, Loss = %f" % (pass_id, iters, loss)
) # The accuracy is the accumulation of batches, but not the current batch.
train_elapsed = time.time() - start_time
examples_per_sec = num_samples / train_elapsed
print('\nTotal examples: %d, total time: %.5f, %.5f examples/sed\n' %
(num_samples, train_elapsed, examples_per_sec))
# evaluation
if args.with_test:
test_loss = do_validation()
exit(0)
def infer():
pass
def print_arguments(args):
print('----------- seq2seq Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == '__main__':
args = parser.parse_args()
print_arguments(args)
if args.infer_only:
infer()
else:
train()
benchmark/fluid/mnist.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import numpy as np
import argparse
import time
import paddle.v2 as paddle
import paddle.fluid as fluid
import paddle.fluid.profiler as profiler
SEED = 1
DTYPE = "float32"
# random seed must set before configuring the network.
# fluid.default_startup_program().random_seed = SEED
def parse_args():
parser = argparse.ArgumentParser("mnist model benchmark.")
parser.add_argument(
'--batch_size', type=int, default=128, help='The minibatch size.')
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test'
)
parser.add_argument(
'--iterations', type=int, default=35, help='The number of minibatches.')
parser.add_argument(
'--pass_num', type=int, default=5, help='The number of passes.')
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help='The device type.')
parser.add_argument(
'--infer_only', action='store_true', help='If set, run forward only.')
parser.add_argument(
'--use_cprof', action='store_true', help='If set, use cProfile.')
parser.add_argument(
'--use_nvprof',
action='store_true',
help='If set, use nvprof for CUDA.')
parser.add_argument(
'--with_test',
action='store_true',
help='If set, test the testset during training.')
args = parser.parse_args()
return args
def cnn_model(data):
conv_pool_1 = fluid.nets.simple_img_conv_pool(
input=data,
filter_size=5,
num_filters=20,
pool_size=2,
pool_stride=2,
act="relu")
conv_pool_2 = fluid.nets.simple_img_conv_pool(
input=conv_pool_1,
filter_size=5,
num_filters=50,
pool_size=2,
pool_stride=2,
act="relu")
# TODO(dzhwinter) : refine the initializer and random seed settting
SIZE = 10
input_shape = conv_pool_2.shape
param_shape = [reduce(lambda a, b: a * b, input_shape[1:], 1)] + [SIZE]
scale = (2.0 / (param_shape[0]**2 * SIZE))**0.5
predict = fluid.layers.fc(
input=conv_pool_2,
size=SIZE,
act="softmax",
param_attr=fluid.param_attr.ParamAttr(
initializer=fluid.initializer.NormalInitializer(
loc=0.0, scale=scale)))
return predict
def eval_test(exe, batch_acc, batch_size_tensor, inference_program):
test_reader = paddle.batch(
paddle.dataset.mnist.test(), batch_size=args.batch_size)
test_pass_acc = fluid.average.WeightedAverage()
for batch_id, data in enumerate(test_reader()):
img_data = np.array(map(lambda x: x[0].reshape([1, 28, 28]),
data)).astype(DTYPE)
y_data = np.array(map(lambda x: x[1], data)).astype("int64")
y_data = y_data.reshape([len(y_data), 1])
acc, weight = exe.run(inference_program,
feed={"pixel": img_data,
"label": y_data},
fetch_list=[batch_acc, batch_size_tensor])
test_pass_acc.add(value=acc, weight=weight)
pass_acc = test_pass_acc.eval()
return pass_acc
def run_benchmark(model, args):
if args.use_cprof:
pr = cProfile.Profile()
pr.enable()
start_time = time.time()
# Input data
images = fluid.layers.data(name='pixel', shape=[1, 28, 28], dtype=DTYPE)
label = fluid.layers.data(name='label', shape=[1], dtype='int64')
# Train program
predict = model(images)
cost = fluid.layers.cross_entropy(input=predict, label=label)
avg_cost = fluid.layers.mean(x=cost)
# Evaluator
batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
batch_acc = fluid.layers.accuracy(
input=predict, label=label, total=batch_size_tensor)
# inference program
inference_program = fluid.default_main_program().clone()
with fluid.program_guard(inference_program):
inference_program = fluid.io.get_inference_program(
target_vars=[batch_acc, batch_size_tensor])
# Optimization
opt = fluid.optimizer.AdamOptimizer(
learning_rate=0.001, beta1=0.9, beta2=0.999)
opt.minimize(avg_cost)
fluid.memory_optimize(fluid.default_main_program())
# Initialize executor
place = fluid.CPUPlace() if args.device == 'CPU' else fluid.CUDAPlace(0)
exe = fluid.Executor(place)
# Parameter initialization
exe.run(fluid.default_startup_program())
# Reader
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size=args.batch_size)
accuracy = fluid.average.WeightedAverage()
iters, num_samples, start_time = 0, 0, time.time()
for pass_id in range(args.pass_num):
accuracy.reset()
train_accs = []
train_losses = []
for batch_id, data in enumerate(train_reader()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
img_data = np.array(
map(lambda x: x[0].reshape([1, 28, 28]), data)).astype(DTYPE)
y_data = np.array(map(lambda x: x[1], data)).astype("int64")
y_data = y_data.reshape([len(y_data), 1])
outs = exe.run(
fluid.default_main_program(),
feed={"pixel": img_data,
"label": y_data},
fetch_list=[avg_cost, batch_acc, batch_size_tensor]
) # The accuracy is the accumulation of batches, but not the current batch.
accuracy.add(value=outs[1], weight=outs[2])
iters += 1
num_samples += len(y_data)
loss = np.array(outs[0])
acc = np.array(outs[1])
train_losses.append(loss)
train_accs.append(acc)
print("Pass: %d, Iter: %d, Loss: %f, Accuracy: %f" %
(pass_id, iters, loss, acc))
print("Pass: %d, Loss: %f, Train Accuray: %f\n" %
(pass_id, np.mean(train_losses), np.mean(train_accs)))
train_elapsed = time.time() - start_time
examples_per_sec = num_samples / train_elapsed
print('\nTotal examples: %d, total time: %.5f, %.5f examples/sed\n' %
(num_samples, train_elapsed, examples_per_sec))
# evaluation
if args.with_test:
test_avg_acc = eval_test(exe, batch_acc, batch_size_tensor,
inference_program)
exit(0)
def print_arguments(args):
vars(args)['use_nvprof'] = (vars(args)['use_nvprof'] and
vars(args)['device'] == 'GPU')
print('----------- mnist Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == '__main__':
args = parse_args()
print_arguments(args)
if args.use_nvprof and args.device == 'GPU':
with profiler.cuda_profiler("cuda_profiler.txt", 'csv') as nvprof:
run_benchmark(cnn_model, args)
else:
run_benchmark(cnn_model, args)
benchmark/fluid/resnet.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import functools
import numpy as np
import time
import cProfile, pstats, StringIO
import paddle.v2 as paddle
import paddle.fluid as fluid
import paddle.fluid.core as core
import paddle.fluid.profiler as profiler
def parse_args():
parser = argparse.ArgumentParser('Convolution model benchmark.')
parser.add_argument(
'--model',
type=str,
choices=['resnet_imagenet', 'resnet_cifar10'],
default='resnet_imagenet',
help='The model architecture.')
parser.add_argument(
'--batch_size', type=int, default=32, help='The minibatch size.')
parser.add_argument(
'--use_fake_data',
action='store_true',
help='use real data or fake data')
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test'
)
parser.add_argument(
'--iterations', type=int, default=80, help='The number of minibatches.')
parser.add_argument(
'--pass_num', type=int, default=100, help='The number of passes.')
parser.add_argument(
'--data_format',
type=str,
default='NCHW',
choices=['NCHW', 'NHWC'],
help='The data data_format, now only support NCHW.')
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help='The device type.')
parser.add_argument(
'--data_set',
type=str,
default='flowers',
choices=['cifar10', 'flowers'],
help='Optional dataset for benchmark.')
parser.add_argument(
'--infer_only', action='store_true', help='If set, run forward only.')
parser.add_argument(
'--use_cprof', action='store_true', help='If set, use cProfile.')
parser.add_argument(
'--use_nvprof',
action='store_true',
help='If set, use nvprof for CUDA.')
parser.add_argument(
'--with_test',
action='store_true',
help='If set, test the testset during training.')
args = parser.parse_args()
return args
def conv_bn_layer(input, ch_out, filter_size, stride, padding, act='relu'):
conv1 = fluid.layers.conv2d(
input=input,
filter_size=filter_size,
num_filters=ch_out,
stride=stride,
padding=padding,
act=None,
bias_attr=False)
return fluid.layers.batch_norm(input=conv1, act=act)
def shortcut(input, ch_out, stride):
ch_in = input.shape[1] if args.data_format == 'NCHW' else input.shape[-1]
if ch_in != ch_out:
return conv_bn_layer(input, ch_out, 1, stride, 0, None)
else:
return input
def basicblock(input, ch_out, stride):
short = shortcut(input, ch_out, stride)
conv1 = conv_bn_layer(input, ch_out, 3, stride, 1)
conv2 = conv_bn_layer(conv1, ch_out, 3, 1, 1, act=None)
return fluid.layers.elementwise_add(x=short, y=conv2, act='relu')
def bottleneck(input, ch_out, stride):
short = shortcut(input, ch_out * 4, stride)
conv1 = conv_bn_layer(input, ch_out, 1, stride, 0)
conv2 = conv_bn_layer(conv1, ch_out, 3, 1, 1)
conv3 = conv_bn_layer(conv2, ch_out * 4, 1, 1, 0, act=None)
return fluid.layers.elementwise_add(x=short, y=conv3, act='relu')
def layer_warp(block_func, input, ch_out, count, stride):
res_out = block_func(input, ch_out, stride)
for i in range(1, count):
res_out = block_func(res_out, ch_out, 1)
return res_out
def resnet_imagenet(input, class_dim, depth=50, data_format='NCHW'):
cfg = {
18: ([2, 2, 2, 1], basicblock),
34: ([3, 4, 6, 3], basicblock),
50: ([3, 4, 6, 3], bottleneck),
101: ([3, 4, 23, 3], bottleneck),
152: ([3, 8, 36, 3], bottleneck)
}
stages, block_func = cfg[depth]
conv1 = conv_bn_layer(input, ch_out=64, filter_size=7, stride=2, padding=3)
pool1 = fluid.layers.pool2d(
input=conv1, pool_type='avg', pool_size=3, pool_stride=2)
res1 = layer_warp(block_func, pool1, 64, stages[0], 1)
res2 = layer_warp(block_func, res1, 128, stages[1], 2)
res3 = layer_warp(block_func, res2, 256, stages[2], 2)
res4 = layer_warp(block_func, res3, 512, stages[3], 2)
pool2 = fluid.layers.pool2d(
input=res4,
pool_size=7,
pool_type='avg',
pool_stride=1,
global_pooling=True)
out = fluid.layers.fc(input=pool2, size=class_dim, act='softmax')
return out
def resnet_cifar10(input, class_dim, depth=32, data_format='NCHW'):
assert (depth - 2) % 6 == 0
n = (depth - 2) // 6
conv1 = conv_bn_layer(
input=input, ch_out=16, filter_size=3, stride=1, padding=1)
res1 = layer_warp(basicblock, conv1, 16, n, 1)
res2 = layer_warp(basicblock, res1, 32, n, 2)
res3 = layer_warp(basicblock, res2, 64, n, 2)
pool = fluid.layers.pool2d(
input=res3, pool_size=8, pool_type='avg', pool_stride=1)
out = fluid.layers.fc(input=pool, size=class_dim, act='softmax')
return out
def run_benchmark(model, args):
if args.use_cprof:
pr = cProfile.Profile()
pr.enable()
if args.data_set == "cifar10":
class_dim = 10
if args.data_format == 'NCHW':
dshape = [3, 32, 32]
else:
dshape = [32, 32, 3]
else:
class_dim = 102
if args.data_format == 'NCHW':
dshape = [3, 224, 224]
else:
dshape = [224, 224, 3]
input = fluid.layers.data(name='data', shape=dshape, dtype='float32')
label = fluid.layers.data(name='label', shape=[1], dtype='int64')
predict = model(input, class_dim)
cost = fluid.layers.cross_entropy(input=predict, label=label)
avg_cost = fluid.layers.mean(x=cost)
batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
batch_acc = fluid.layers.accuracy(
input=predict, label=label, total=batch_size_tensor)
inference_program = fluid.default_main_program().clone()
with fluid.program_guard(inference_program):
inference_program = fluid.io.get_inference_program(
target_vars=[batch_acc, batch_size_tensor])
optimizer = fluid.optimizer.Momentum(learning_rate=0.01, momentum=0.9)
opts = optimizer.minimize(avg_cost)
fluid.memory_optimize(fluid.default_main_program())
train_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.cifar.train10()
if args.data_set == 'cifar10' else paddle.dataset.flowers.train(),
buf_size=5120),
batch_size=args.batch_size)
test_reader = paddle.batch(
paddle.dataset.cifar.test10()
if args.data_set == 'cifar10' else paddle.dataset.flowers.test(),
batch_size=args.batch_size)
def test(exe):
test_accuracy = fluid.average.WeightedAverage()
for batch_id, data in enumerate(test_reader()):
img_data = np.array(map(lambda x: x[0].reshape(dshape),
data)).astype("float32")
y_data = np.array(map(lambda x: x[1], data)).astype("int64")
y_data = y_data.reshape([-1, 1])
acc, weight = exe.run(inference_program,
feed={"data": img_data,
"label": y_data},
fetch_list=[batch_acc, batch_size_tensor])
test_accuracy.add(value=acc, weight=weight)
return test_accuracy.eval()
place = core.CPUPlace() if args.device == 'CPU' else core.CUDAPlace(0)
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program())
accuracy = fluid.average.WeightedAverage()
if args.use_fake_data:
data = train_reader().next()
image = np.array(map(lambda x: x[0].reshape(dshape), data)).astype(
'float32')
label = np.array(map(lambda x: x[1], data)).astype('int64')
label = label.reshape([-1, 1])
iters, num_samples, start_time = 0, 0, time.time()
for pass_id in range(args.pass_num):
accuracy.reset()
train_accs = []
train_losses = []
for batch_id, data in enumerate(train_reader()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
if not args.use_fake_data:
image = np.array(map(lambda x: x[0].reshape(dshape),
data)).astype('float32')
label = np.array(map(lambda x: x[1], data)).astype('int64')
label = label.reshape([-1, 1])
loss, acc, weight = exe.run(
fluid.default_main_program(),
feed={'data': image,
'label': label},
fetch_list=[avg_cost, batch_acc, batch_size_tensor])
iters += 1
num_samples += len(label)
accuracy.add(value=acc, weight=weight)
train_losses.append(loss)
train_accs.append(acc)
print("Pass: %d, Iter: %d, Loss: %f, Accuracy: %f" %
(pass_id, iters, loss, acc))
print("Pass: %d, Loss: %f, Train Accuray: %f\n" %
(pass_id, np.mean(train_losses), np.mean(train_accs)))
train_elapsed = time.time() - start_time
examples_per_sec = num_samples / train_elapsed
print('\nTotal examples: %d, total time: %.5f, %.5f examples/sed\n' %
(num_samples, train_elapsed, examples_per_sec))
# evaluation
if args.with_test:
pass_test_acc = test(exe)
exit(0)
def print_arguments(args):
vars(args)['use_nvprof'] = (vars(args)['use_nvprof'] and
vars(args)['device'] == 'GPU')
print('----------- resnet Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == '__main__':
model_map = {
'resnet_imagenet': resnet_imagenet,
'resnet_cifar10': resnet_cifar10
}
args = parse_args()
print_arguments(args)
if args.data_format == 'NHWC':
raise ValueError('Only support NCHW data_format now.')
if args.use_nvprof and args.device == 'GPU':
with profiler.cuda_profiler("cuda_profiler.txt", 'csv') as nvprof:
run_benchmark(model_map[args.model], args)
else:
run_benchmark(model_map[args.model], args)
benchmark/fluid/run.sh 0 → 100644
浏览文件 @ e2ca8e10
#!/bin/bash
# This script benchmarking the PaddlePaddle Fluid on
# single thread single GPU.
#export FLAGS_fraction_of_gpu_memory_to_use=0.0
export CUDNN_PATH=/paddle/cudnn_v5
# disable openmp and mkl parallel
#https://github.com/PaddlePaddle/Paddle/issues/7199
export MKL_NUM_THREADS=1
export OMP_NUM_THREADS=1
ht=`lscpu |grep "per core"|awk -F':' '{print $2}'|xargs`
if [ $ht -eq 1 ]; then # HT is OFF
if [ -z "$KMP_AFFINITY" ]; then
export KMP_AFFINITY="granularity=fine,compact,0,0"
fi
if [ -z "$OMP_DYNAMIC" ]; then
export OMP_DYNAMIC="FALSE"
fi
else # HT is ON
if [ -z "$KMP_AFFINITY" ]; then
export KMP_AFFINITY="granularity=fine,compact,1,0"
fi
fi
# disable multi-gpu if have more than one
export CUDA_VISIBLE_DEVICES=0
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH=$CUDNN_PATH:$LD_LIBRARY_PATH
# only query the gpu used
nohup stdbuf -oL nvidia-smi \
--id=${CUDA_VISIBLE_DEVICES} \
--query-gpu=timestamp \
--query-compute-apps=pid,process_name,used_memory \
--format=csv \
--filename=mem.log \
-l 1 &
# mnist
# mnist gpu mnist 128
FLAGS_benchmark=true stdbuf -oL python fluid/mnist.py \
--device=GPU \
--batch_size=128 \
--skip_batch_num=5 \
--iterations=500 \
2>&1 | tee -a mnist_gpu_128.log
# vgg16
# gpu cifar10 128
FLAGS_benchmark=true stdbuf -oL python fluid/vgg16.py \
--device=GPU \
--batch_size=128 \
--skip_batch_num=5 \
--iterations=30 \
2>&1 | tee -a vgg16_gpu_128.log
# flowers gpu 128
FLAGS_benchmark=true stdbuf -oL python fluid/vgg16.py \
--device=GPU \
--batch_size=32 \
--data_set=flowers \
--skip_batch_num=5 \
--iterations=30 \
2>&1 | tee -a vgg16_gpu_flowers_32.log
# resnet50
# resnet50 gpu cifar10 128
FLAGS_benchmark=true stdbuf -oL python fluid/resnet50.py \
--device=GPU \
--batch_size=128 \
--data_set=cifar10 \
--model=resnet_cifar10 \
--skip_batch_num=5 \
--iterations=30 \
2>&1 | tee -a resnet50_gpu_128.log
# resnet50 gpu flowers 64
FLAGS_benchmark=true stdbuf -oL python fluid/resnet50.py \
--device=GPU \
--batch_size=64 \
--data_set=flowers \
--model=resnet_imagenet \
--skip_batch_num=5 \
--iterations=30 \
2>&1 | tee -a resnet50_gpu_flowers_64.log
# lstm
# lstm gpu imdb 32 # tensorflow only support batch=32
FLAGS_benchmark=true stdbuf -oL python fluid/stacked_dynamic_lstm.py \
--device=GPU \
--batch_size=32 \
--skip_batch_num=5 \
--iterations=30 \
--hidden_dim=512 \
--emb_dim=512 \
--crop_size=1500 \
2>&1 | tee -a lstm_gpu_32.log
# seq2seq
# seq2seq gpu wmb 128
FLAGS_benchmark=true stdbuf -oL python fluid/machine_translation.py \
--device=GPU \
--batch_size=128 \
--skip_batch_num=5 \
--iterations=30 \
2>&1 | tee -a lstm_gpu_128.log
benchmark/fluid/stacked_dynamic_lstm.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import cPickle
import os
import random
import time
import numpy
import paddle.v2 as paddle
import paddle.v2.dataset.imdb as imdb
import paddle.fluid as fluid
from paddle.v2 import batch
import paddle.fluid.profiler as profiler
def parse_args():
parser = argparse.ArgumentParser("Understand Sentiment by Dynamic RNN.")
parser.add_argument(
'--batch_size',
type=int,
default=32,
help='The sequence number of a batch data. (default: %(default)d)')
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test'
)
parser.add_argument(
'--iterations', type=int, default=80, help='The number of minibatches.')
parser.add_argument(
'--emb_dim',
type=int,
default=512,
help='Dimension of embedding table. (default: %(default)d)')
parser.add_argument(
'--hidden_dim',
type=int,
default=512,
help='Hidden size of lstm unit. (default: %(default)d)')
parser.add_argument(
'--pass_num',
type=int,
default=100,
help='Epoch number to train. (default: %(default)d)')
parser.add_argument(
'--device',
type=str,
default='CPU',
choices=['CPU', 'GPU'],
help='The device type.')
parser.add_argument(
'--crop_size',
type=int,
default=int(os.environ.get('CROP_SIZE', '1500')),
help='The max sentence length of input. Since this model use plain RNN,'
' Gradient could be explored if sentence is too long')
parser.add_argument(
'--with_test',
action='store_true',
help='If set, test the testset during training.')
args = parser.parse_args()
return args
word_dict = imdb.word_dict()
def crop_sentence(reader, crop_size):
unk_value = word_dict['<unk>']
def __impl__():
for item in reader():
if len([x for x in item[0] if x != unk_value]) < crop_size:
yield item
return __impl__
def main():
args = parse_args()
lstm_size = args.hidden_dim
data = fluid.layers.data(
name="words", shape=[1], lod_level=1, dtype='int64')
sentence = fluid.layers.embedding(
input=data, size=[len(word_dict), args.emb_dim])
sentence = fluid.layers.fc(input=sentence, size=lstm_size, act='tanh')
rnn = fluid.layers.DynamicRNN()
with rnn.block():
word = rnn.step_input(sentence)
prev_hidden = rnn.memory(value=0.0, shape=[lstm_size])
prev_cell = rnn.memory(value=0.0, shape=[lstm_size])
def gate_common(
ipt,
hidden,
size, ):
gate0 = fluid.layers.fc(input=ipt, size=size, bias_attr=True)
gate1 = fluid.layers.fc(input=hidden, size=size, bias_attr=False)
gate = fluid.layers.sums(input=[gate0, gate1])
return gate
forget_gate = fluid.layers.sigmoid(
x=gate_common(word, prev_hidden, lstm_size))
input_gate = fluid.layers.sigmoid(
x=gate_common(word, prev_hidden, lstm_size))
output_gate = fluid.layers.sigmoid(
x=gate_common(word, prev_hidden, lstm_size))
cell_gate = fluid.layers.tanh(
x=gate_common(word, prev_hidden, lstm_size))
cell = fluid.layers.sums(input=[
fluid.layers.elementwise_mul(
x=forget_gate, y=prev_cell), fluid.layers.elementwise_mul(
x=input_gate, y=cell_gate)
])
hidden = fluid.layers.elementwise_mul(
x=output_gate, y=fluid.layers.tanh(x=cell))
rnn.update_memory(prev_cell, cell)
rnn.update_memory(prev_hidden, hidden)
rnn.output(hidden)
last = fluid.layers.sequence_pool(rnn(), 'last')
logit = fluid.layers.fc(input=last, size=2, act='softmax')
loss = fluid.layers.cross_entropy(
input=logit,
label=fluid.layers.data(
name='label', shape=[1], dtype='int64'))
loss = fluid.layers.mean(x=loss)
# add acc
batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
batch_acc = fluid.layers.accuracy(input=logit, label=fluid.layers.data(name='label', \
shape=[1], dtype='int64'), total=batch_size_tensor)
inference_program = fluid.default_main_program().clone()
with fluid.program_guard(inference_program):
inference_program = fluid.io.get_inference_program(
target_vars=[batch_acc, batch_size_tensor])
adam = fluid.optimizer.Adam()
adam.minimize(loss)
fluid.memory_optimize(fluid.default_main_program())
place = fluid.CPUPlace() if args.device == 'CPU' else fluid.CUDAPlace(0)
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program())
train_reader = batch(
paddle.reader.shuffle(
crop_sentence(imdb.train(word_dict), args.crop_size),
buf_size=25000),
batch_size=args.batch_size)
iters, num_samples, start_time = 0, 0, time.time()
for pass_id in range(args.pass_num):
train_accs = []
train_losses = []
for batch_id, data in enumerate(train_reader()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
tensor_words = to_lodtensor([x[0] for x in data], place)
label = numpy.array([x[1] for x in data]).astype("int64")
label = label.reshape((-1, 1))
loss_np, acc, weight = exe.run(
fluid.default_main_program(),
feed={"words": tensor_words,
"label": label},
fetch_list=[loss, batch_acc, batch_size_tensor])
iters += 1
for x in data:
num_samples += len(x[0])
print(
"Pass = %d, Iter = %d, Loss = %f, Accuracy = %f" %
(pass_id, iters, loss_np, acc)
) # The accuracy is the accumulation of batches, but not the current batch.
train_elapsed = time.time() - start_time
examples_per_sec = num_samples / train_elapsed
print('\nTotal examples: %d, total time: %.5f, %.5f examples/sed\n' %
(num_samples, train_elapsed, examples_per_sec))
exit(0)
def to_lodtensor(data, place):
seq_lens = [len(seq) for seq in data]
cur_len = 0
lod = [cur_len]
for l in seq_lens:
cur_len += l
lod.append(cur_len)
flattened_data = numpy.concatenate(data, axis=0).astype("int64")
flattened_data = flattened_data.reshape([len(flattened_data), 1])
res = fluid.LoDTensor()
res.set(flattened_data, place)
res.set_lod([lod])
return res
def print_arguments(args):
print('----------- lstm Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == '__main__':
args = parse_args()
print_arguments(args)
main()
benchmark/fluid/vgg.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
"""VGG16 benchmark in Fluid"""
from __future__ import print_function
import sys
import time
import numpy as np
import paddle.v2 as paddle
import paddle.fluid as fluid
import paddle.fluid.core as core
import argparse
import functools
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
'--batch_size', type=int, default=128, help="Batch size for training.")
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test')
parser.add_argument(
'--iterations', type=int, default=80, help='The number of minibatches.')
parser.add_argument(
'--learning_rate',
type=float,
default=1e-3,
help="Learning rate for training.")
parser.add_argument('--pass_num', type=int, default=50, help="No. of passes.")
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help="The device type.")
parser.add_argument(
'--data_format',
type=str,
default='NCHW',
choices=['NCHW', 'NHWC'],
help='The data order, now only support NCHW.')
parser.add_argument(
'--data_set',
type=str,
default='cifar10',
choices=['cifar10', 'flowers'],
help='Optional dataset for benchmark.')
parser.add_argument(
'--with_test',
action='store_true',
help='If set, test the testset during training.')
args = parser.parse_args()
def vgg16_bn_drop(input):
def conv_block(input, num_filter, groups, dropouts):
return fluid.nets.img_conv_group(
input=input,
pool_size=2,
pool_stride=2,
conv_num_filter=[num_filter] * groups,
conv_filter_size=3,
conv_act='relu',
conv_with_batchnorm=True,
conv_batchnorm_drop_rate=dropouts,
pool_type='max')
conv1 = conv_block(input, 64, 2, [0.3, 0])
conv2 = conv_block(conv1, 128, 2, [0.4, 0])
conv3 = conv_block(conv2, 256, 3, [0.4, 0.4, 0])
conv4 = conv_block(conv3, 512, 3, [0.4, 0.4, 0])
conv5 = conv_block(conv4, 512, 3, [0.4, 0.4, 0])
drop = fluid.layers.dropout(x=conv5, dropout_prob=0.5)
fc1 = fluid.layers.fc(input=drop, size=512, act=None)
bn = fluid.layers.batch_norm(input=fc1, act='relu')
drop2 = fluid.layers.dropout(x=bn, dropout_prob=0.5)
fc2 = fluid.layers.fc(input=drop2, size=512, act=None)
return fc2
def main():
if args.data_set == "cifar10":
classdim = 10
if args.data_format == 'NCHW':
data_shape = [3, 32, 32]
else:
data_shape = [32, 32, 3]
else:
classdim = 102
if args.data_format == 'NCHW':
data_shape = [3, 224, 224]
else:
data_shape = [224, 224, 3]
# Input data
images = fluid.layers.data(name='pixel', shape=data_shape, dtype='float32')
label = fluid.layers.data(name='label', shape=[1], dtype='int64')
# Train program
net = vgg16_bn_drop(images)
predict = fluid.layers.fc(input=net, size=classdim, act='softmax')
cost = fluid.layers.cross_entropy(input=predict, label=label)
avg_cost = fluid.layers.mean(x=cost)
# Evaluator
batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
batch_acc = fluid.layers.accuracy(
input=predict, label=label, total=batch_size_tensor)
# inference program
inference_program = fluid.default_main_program().clone()
with fluid.program_guard(inference_program):
inference_program = fluid.io.get_inference_program(
target_vars=[batch_acc, batch_size_tensor])
# Optimization
optimizer = fluid.optimizer.Adam(learning_rate=args.learning_rate)
opts = optimizer.minimize(avg_cost)
fluid.memory_optimize(fluid.default_main_program())
# Initialize executor
place = core.CPUPlace() if args.device == 'CPU' else core.CUDAPlace(0)
exe = fluid.Executor(place)
# Parameter initialization
exe.run(fluid.default_startup_program())
# data reader
train_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.cifar.train10()
if args.data_set == 'cifar10' else paddle.dataset.flowers.train(),
buf_size=5120),
batch_size=args.batch_size)
test_reader = paddle.batch(
paddle.dataset.cifar.test10()
if args.data_set == 'cifar10' else paddle.dataset.flowers.test(),
batch_size=args.batch_size)
# test
def test(exe):
test_accuracy = fluid.average.WeightedAverage()
for batch_id, data in enumerate(test_reader()):
img_data = np.array(map(lambda x: x[0].reshape(data_shape),
data)).astype("float32")
y_data = np.array(map(lambda x: x[1], data)).astype("int64")
y_data = y_data.reshape([-1, 1])
acc, weight = exe.run(inference_program,
feed={"pixel": img_data,
"label": y_data},
fetch_list=[batch_acc, batch_size_tensor])
test_accuracy.add(value=acc, weight=weight)
return test_accuracy.eval()
iters, num_samples, start_time = 0, 0, time.time()
accuracy = fluid.average.WeightedAverage()
for pass_id in range(args.pass_num):
accuracy.reset()
train_accs = []
train_losses = []
for batch_id, data in enumerate(train_reader()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
img_data = np.array(map(lambda x: x[0].reshape(data_shape),
data)).astype("float32")
y_data = np.array(map(lambda x: x[1], data)).astype("int64")
y_data = y_data.reshape([-1, 1])
loss, acc, weight = exe.run(
fluid.default_main_program(),
feed={"pixel": img_data,
"label": y_data},
fetch_list=[avg_cost, batch_acc, batch_size_tensor])
accuracy.add(value=acc, weight=weight)
iters += 1
num_samples += len(y_data)
print(
"Pass = %d, Iter = %d, Loss = %f, Accuracy = %f" %
(pass_id, iters, loss, acc)
) # The accuracy is the accumulation of batches, but not the current batch.
# pass_train_acc = accuracy.eval()
train_losses.append(loss)
train_accs.append(acc)
print("Pass: %d, Loss: %f, Train Accuray: %f\n" %
(pass_id, np.mean(train_losses), np.mean(train_accs)))
train_elapsed = time.time() - start_time
examples_per_sec = num_samples / train_elapsed
print('\nTotal examples: %d, total time: %.5f, %.5f examples/sed\n' %
(num_samples, train_elapsed, examples_per_sec))
# evaluation
if args.with_test:
pass_test_acc = test(exe)
exit(0)
def print_arguments():
print('----------- vgg Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == "__main__":
print_arguments()
main()
benchmark/tensorflow/machine_translation.py 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
benchmark/tensorflow/mnist.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import time
import numpy as np
import tensorflow as tf
import paddle.v2 as paddle
DTYPE = tf.float32
def parse_args():
parser = argparse.ArgumentParser("mnist model benchmark.")
parser.add_argument(
'--batch_size', type=int, default=128, help='The minibatch size.')
parser.add_argument(
'--iterations', type=int, default=35, help='The number of minibatches.')
parser.add_argument(
'--pass_num', type=int, default=5, help='The number of passes.')
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help='The device type.')
args = parser.parse_args()
return args
def run_benchmark(args):
def weight_variable(dtype, shape):
initial = tf.truncated_normal(shape, stddev=0.1, dtype=dtype)
return tf.Variable(initial)
def bias_variable(dtype, shape):
initial = tf.constant(0.1, shape=shape, dtype=dtype)
return tf.Variable(initial)
device = '/cpu:0' if args.device == 'CPU' else '/device:GPU:0'
with tf.device(device):
images = tf.placeholder(DTYPE, shape=(None, 28, 28, 1))
labels = tf.placeholder(tf.int64, shape=(None, ))
# conv1, relu, pool1
conv1_weights = weight_variable(DTYPE, [5, 5, 1, 20])
conv1_bias = bias_variable(DTYPE, [20])
conv1 = tf.nn.conv2d(
images, conv1_weights, strides=[1, 1, 1, 1], padding="VALID")
relu1 = tf.nn.relu(tf.nn.bias_add(conv1, conv1_bias))
pool1 = tf.nn.max_pool(
relu1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding="VALID")
# conv2, relu, pool2
conv2_weights = weight_variable(DTYPE, [5, 5, 20, 50])
conv2_bias = bias_variable(DTYPE, [50])
conv2 = tf.nn.conv2d(
pool1, conv2_weights, strides=[1, 1, 1, 1], padding="VALID")
relu2 = tf.nn.relu(tf.nn.bias_add(conv2, conv2_bias))
pool2 = tf.nn.max_pool(
relu2, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding="VALID")
# FC
pool_shape = pool2.get_shape().as_list()
hidden_dim = reduce(lambda a, b: a * b, pool_shape[1:], 1)
reshape = tf.reshape(pool2, shape=(tf.shape(pool2)[0], hidden_dim))
fc_weights = weight_variable(DTYPE, [hidden_dim, 10])
fc_bias = bias_variable(DTYPE, [10])
logits = tf.matmul(reshape, fc_weights) + fc_bias
# Get prediction
prediction = tf.nn.softmax(logits)
# Loss
one_hot_labels = tf.one_hot(labels, depth=10)
cost = -tf.reduce_sum(tf.log(prediction) * one_hot_labels, [1])
avg_cost = tf.reduce_mean(cost)
# Get accuracy
correct = tf.equal(tf.argmax(prediction, 1), labels)
accuracy = tf.reduce_mean(tf.cast(correct, tf.float32))
# metrics, g_accuracy
with tf.variable_scope("reset_metrics_accuracy_scope") as scope:
g_accuracy = tf.metrics.accuracy(
labels, tf.argmax(
prediction, axis=1))
vars = tf.contrib.framework.get_variables(
scope, collection=tf.GraphKeys.LOCAL_VARIABLES)
g_accuracy_reset_op = tf.variables_initializer(vars)
# Optimizer
opt = tf.train.AdamOptimizer(
learning_rate=0.001, beta1=0.9, beta2=0.999)
train_op = opt.minimize(avg_cost)
# train_op = tf.train.AdamOptimizer(1e-4).minimize(avg_cost)
train_reader = paddle.batch(
paddle.dataset.mnist.train(), batch_size=args.batch_size)
test_reader = paddle.batch(
paddle.dataset.mnist.test(), batch_size=args.batch_size)
def eval_test():
sess.run(g_accuracy_reset_op)
for batch_id, data in enumerate(test_reader()):
images_data = np.array(
map(lambda x: np.transpose(x[0].reshape([1, 28, 28]), axes=[1,2,0]), data)).astype("float32")
labels_data = np.array(map(lambda x: x[1], data)).astype("int64")
loss, acc, g_acc = sess.run(
[avg_cost, accuracy, g_accuracy],
feed_dict={images: images_data,
labels: labels_data})
return g_acc[1]
config = tf.ConfigProto(
intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)
config.gpu_options.allow_growth = True
with tf.Session(config=config) as sess:
init_g = tf.global_variables_initializer()
init_l = tf.local_variables_initializer()
sess.run(init_g)
sess.run(init_l)
for pass_id in range(args.pass_num):
sess.run(g_accuracy_reset_op)
pass_start = time.time()
for batch_id, data in enumerate(train_reader()):
images_data = np.array(
map(lambda x: np.transpose(x[0].reshape([1, 28, 28]), axes=[1,2,0]), data)).astype("float32")
labels_data = np.array(map(lambda x: x[1], data)).astype(
"int64")
start = time.time()
_, loss, acc, g_acc = sess.run(
[train_op, avg_cost, accuracy, g_accuracy],
feed_dict={images: images_data,
labels: labels_data})
end = time.time()
print("pass=%d, batch=%d, loss=%f, error=%f, elapse=%f" %
(pass_id, batch_id, loss, 1 - acc, (end - start) / 1000))
pass_end = time.time()
test_avg_acc = eval_test()
print(
"pass=%d, training_avg_accuracy=%f, test_avg_acc=%f, elapse=%f"
% (pass_id, g_acc[1], test_avg_acc,
(pass_end - pass_start) / 1000))
def print_arguments(args):
print('----------- Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == '__main__':
args = parse_args()
print_arguments(args)
run_benchmark(args)
benchmark/tensorflow/resnet.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
"""
based on https://github.com/tensorflow/models/blob/master/official/resnet/resnet_model.py
Get help: python resnet.py --help
See performance on flowers: python resnet.py
Train on cifar10: python resnet.py --data=cifar10 --with_test
"""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import time
import numpy as np
import paddle.v2 as paddle
import tensorflow as tf
DTYPE = tf.float32
def parse_args():
parser = argparse.ArgumentParser('Convolution model benchmark.')
parser.add_argument(
'--model',
type=str,
choices=['resnet'],
default='resnet',
help='The model architecture.')
parser.add_argument(
'--batch_size', type=int, default=32, help='The minibatch size.')
parser.add_argument(
'--use_fake_data',
action='store_true',
help='use real data or fake data')
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test'
)
parser.add_argument(
'--iterations',
type=int,
default=105,
help='The number of minibatches.')
parser.add_argument(
'--pass_num', type=int, default=300, help='The number of passes.')
parser.add_argument(
'--order',
type=str,
default='NHWC',
choices=['NCHW', 'NHWC'],
help='The data order, now only support NCHW.')
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help='The device type.')
parser.add_argument(
'--data',
type=str,
default='flowers102',
choices=['flowers102', 'cifar10'],
help='The kinds of data.')
parser.add_argument(
'--infer_only', action='store_true', help='If set, run forward only.')
parser.add_argument(
'--use_cprof', action='store_true', help='If set, use cProfile.')
parser.add_argument(
'--with_test',
action='store_true',
help='If set, test the testset during training.')
parser.add_argument(
'--use_nvprof',
action='store_true',
help='If set, use nvprof for CUDA.')
args = parser.parse_args()
return args
def print_arguments(args):
vars(args)['use_nvprof'] = (vars(args)['use_nvprof'] and
vars(args)['device'] == 'GPU')
vars(args)['iterations'] = vars(args)['pass_num'] * 1000 if vars(args)[
'with_test'] else vars(args)['iterations']
print('----------- Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
def fixed_padding(inputs, kernel_size, data_format):
"""Pads the input along the spatial dimensions independently of input size.
Args:
inputs: A tensor of size [batch, channels, height_in, width_in] or
[batch, height_in, width_in, channels] depending on data_format.
kernel_size: The kernel to be used in the conv2d or max_pool2d operation.
Should be a positive integer.
data_format: The input format ('channels_last' or 'channels_first').
Returns:
A tensor with the same format as the input with the data either intact
(if kernel_size == 1) or padded (if kernel_size > 1).
"""
pad_total = kernel_size - 1
pad_beg = pad_total // 2
pad_end = pad_total - pad_beg
if data_format == 'channels_first':
padded_inputs = tf.pad(inputs, [[0, 0], [0, 0], [pad_beg, pad_end],
[pad_beg, pad_end]])
else:
padded_inputs = tf.pad(inputs, [[0, 0], [pad_beg, pad_end],
[pad_beg, pad_end], [0, 0]])
return padded_inputs
def conv2d_fixed_padding(inputs, filters, kernel_size, strides, data_format):
"""Strided 2-D convolution with explicit padding."""
# The padding is consistent and is based only on `kernel_size`, not on the
# dimensions of `inputs` (as opposed to using `tf.layers.conv2d` alone).
# This is consistent with PaddlePaddle.
# In addition, the calculation for output size in TensorFlow can refer:
# https://github.com/tensorflow/tensorflow/blob/master/tensorflow/core/framework/common_shape_fns.cc
if strides > 1:
inputs = fixed_padding(inputs, kernel_size, data_format)
return tf.layers.conv2d(
inputs=inputs,
filters=filters,
kernel_size=kernel_size,
strides=strides,
padding=('SAME' if strides == 1 else 'VALID'),
use_bias=False,
kernel_initializer=tf.variance_scaling_initializer(),
data_format=data_format)
def conv_bn(inputs,
filters,
kernel_size,
strides,
is_training,
data_format,
act=True):
# def conv2d_fixed_padding(inputs, filters, kernel_size, strides, data_format):
# set fused=True for a significant performance boost. See
# https://www.tensorflow.org/performance/performance_guide#common_fused_ops
inputs = conv2d_fixed_padding(
inputs=inputs,
filters=filters,
kernel_size=kernel_size,
strides=strides,
data_format=data_format)
inputs = tf.layers.batch_normalization(
inputs=inputs,
axis=1 if data_format == 'channels_first' else 3,
momentum=0.9,
epsilon=1e-05,
center=True,
scale=True,
training=is_training,
fused=True)
if act:
inputs = tf.nn.relu(inputs)
return inputs
def basicblock(inputs, filters, is_training, projection_shortcut, strides,
data_format):
shortcut = inputs
if projection_shortcut is not None:
shortcut = projection_shortcut(inputs)
inputs = conv_bn(inputs, filters, 3, strides, is_training, data_format)
inputs = conv_bn(inputs, filters, 3, 1, is_training, data_format, act=False)
inputs = inputs + shortcut
inputs = tf.nn.relu(inputs)
return inputs
def bottleneck(inputs, filters, is_training, projection_shortcut, strides,
data_format):
shortcut = inputs
if projection_shortcut is not None:
shortcut = projection_shortcut(inputs)
inputs = conv_bn(inputs, filters, 1, strides, is_training, data_format)
inputs = conv_bn(inputs, filters, 3, 1, is_training, data_format, act=False)
inputs = conv_bn(
inputs, filters * 4, 1, 1, is_training, data_format, act=False)
inputs = inputs + shortcut
inputs = tf.nn.relu(inputs)
return inputs
def block_layer(inputs, filters, block_fn, blocks, strides, is_training, name,
data_format):
# Bottleneck blocks end with 4x the number of filters as they start with
filters_out = 4 * filters if block_fn is bottleneck else filters
def projection_shortcut(inputs):
return conv2d_fixed_padding(
inputs=inputs,
filters=filters_out,
kernel_size=1,
strides=strides,
data_format=data_format)
# Only the first block per block_layer uses projection_shortcut and strides
inputs = block_fn(inputs, filters, is_training, projection_shortcut,
strides, data_format)
for _ in range(1, blocks):
inputs = block_fn(inputs, filters, is_training, None, 1, data_format)
return tf.identity(inputs, name)
def resnet_imagenet(depth, class_dim, data_format):
"""Returns the ResNet model for a given size and number of output classes."""
def resnet_generator(block_fn,
layers,
num_classes,
data_format='channels_last'):
if data_format is None:
data_format = ('channels_first'
if tf.test.is_built_with_cuda() else 'channels_last')
def model(inputs, is_training):
"""Constructs the ResNet model given the inputs."""
if data_format == 'channels_first':
# Convert the inputs from channels_last (NHWC) to channels_first (NCHW).
# This provides a large performance boost on GPU. See
# https://www.tensorflow.org/performance/performance_guide#data_formats
inputs = tf.transpose(inputs, [0, 3, 1, 2])
inputs = conv_bn(inputs, 64, 7, 2, is_training, data_format)
inputs = tf.identity(inputs, 'initial_conv')
inputs = tf.layers.max_pooling2d(
inputs=inputs,
pool_size=3,
strides=2,
padding='SAME',
data_format=data_format)
inputs = tf.identity(inputs, 'initial_max_pool')
inputs = block_layer(inputs, 64, block_fn, layers[0], 1,
is_training, 'block_layer1', data_format)
inputs = block_layer(inputs, 128, block_fn, layers[1], 2,
is_training, 'block_layer2', data_format)
inputs = block_layer(inputs, 256, block_fn, layers[2], 2,
is_training, 'block_layer3', data_format)
inputs = block_layer(inputs, 512, block_fn, layers[3], 2,
is_training, 'block_layer4', data_format)
inputs = tf.layers.average_pooling2d(
inputs=inputs,
pool_size=7,
strides=1,
padding='VALID',
data_format=data_format)
inputs = tf.identity(inputs, 'final_avg_pool')
inputs = tf.reshape(inputs,
[-1, 512 if block_fn is basicblock else 2048])
inputs = tf.layers.dense(inputs=inputs, units=num_classes)
inputs = tf.identity(inputs, 'final_dense')
return inputs
return model
model_params = {
18: {
'block': basicblock,
'layers': [2, 2, 2, 2]
},
34: {
'block': basicblock,
'layers': [3, 4, 6, 3]
},
50: {
'block': bottleneck,
'layers': [3, 4, 6, 3]
},
101: {
'block': bottleneck,
'layers': [3, 4, 23, 3]
},
152: {
'block': bottleneck,
'layers': [3, 8, 36, 3]
},
200: {
'block': bottleneck,
'layers': [3, 24, 36, 3]
}
}
if depth not in model_params:
raise ValueError('Not a valid depth:', depth)
params = model_params[depth]
return resnet_generator(params['block'], params['layers'], class_dim,
data_format)
def resnet_cifar10(depth, num_classes, data_format):
if depth % 6 != 2:
raise ValueError('depth must be 6n + 2:', depth)
num_blocks = (depth - 2) // 6
if data_format is None:
data_format = ('channels_first'
if tf.test.is_built_with_cuda() else 'channels_last')
def model(inputs, is_training):
inputs = conv_bn(inputs, 16, 3, 1, is_training, data_format)
inputs = tf.identity(inputs, 'initial_conv')
inputs = block_layer(inputs, 16, basicblock, num_blocks, 1, is_training,
'block_layer1', data_format)
inputs = block_layer(inputs, 32, basicblock, num_blocks, 2, is_training,
'block_layer2', data_format)
inputs = block_layer(inputs, 64, basicblock, num_blocks, 2, is_training,
'block_layer3', data_format)
inputs = tf.layers.average_pooling2d(
inputs=inputs,
pool_size=8,
strides=1,
padding='VALID',
data_format=data_format)
inputs = tf.identity(inputs, 'final_avg_pool')
inputs = tf.reshape(inputs, [-1, 64])
inputs = tf.layers.dense(inputs=inputs, units=num_classes)
inputs = tf.identity(inputs, 'final_dense')
return inputs
return model
def run_benchmark(args, data_format='channels_last', device='/cpu:0'):
"""Our model_fn for ResNet to be used with our Estimator."""
class_dim = 1000
dshape = (None, 224, 224, 3)
pdshape = (3, 224, 224)
if args.data == 'flowers102':
class_dim = 102
dshape = (None, 224, 224, 3)
pdshape = (3, 224, 224)
elif args.data == 'cifar10':
class_dim = 10
dshape = (None, 32, 32, 3)
pdshape = (3, 32, 32)
with tf.device(device):
images = tf.placeholder(DTYPE, shape=dshape)
labels = tf.placeholder(tf.int64, shape=(None, ))
is_training = tf.placeholder('bool')
onehot_labels = tf.one_hot(labels, depth=class_dim)
network = resnet_cifar10(
32, class_dim,
data_format) if args.data == 'cifar10' else resnet_imagenet(
50, class_dim, data_format)
logits = network(inputs=images, is_training=is_training)
cross_entropy = tf.losses.softmax_cross_entropy(
logits=logits, onehot_labels=onehot_labels)
avg_cost = tf.reduce_mean(cross_entropy)
correct = tf.equal(tf.argmax(logits, 1), labels)
accuracy = tf.reduce_mean(tf.cast(correct, tf.float32))
lr = 0.1 if args.data == 'cifar10' else 0.01
optimizer = tf.train.MomentumOptimizer(learning_rate=lr, momentum=0.9)
# Batch norm requires update_ops to be added as a train_op dependency.
update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
train_op = optimizer.minimize(avg_cost)
train_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.cifar.train10()
if args.data == 'cifar10' else paddle.dataset.flowers.train(),
buf_size=5120),
batch_size=args.batch_size)
test_reader = paddle.batch(
paddle.dataset.cifar.test10()
if args.data == 'cifar10' else paddle.dataset.flowers.test(),
batch_size=100)
def test():
test_accs = []
for batch_id, data in enumerate(test_reader()):
test_images = np.array(
map(lambda x: np.transpose(x[0].reshape(pdshape),
axes=[1, 2, 0]), data)).astype("float32")
test_labels = np.array(map(lambda x: x[1], data)).astype('int64')
test_accs.append(
accuracy.eval(feed_dict={
images: test_images,
labels: test_labels,
is_training: False
}))
print("Pass = %d, Train performance = %f imgs/s, Test accuracy = %f\n" %
(pass_id, num_samples / train_elapsed, np.mean(test_accs)))
config = tf.ConfigProto(
intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)
config.gpu_options.allow_growth = True
with tf.Session(config=config) as sess:
init_g = tf.global_variables_initializer()
init_l = tf.local_variables_initializer()
sess.run(init_g)
sess.run(init_l)
if args.use_fake_data:
data = train_reader().next()
images_data = np.array(
map(lambda x: np.transpose(x[0].reshape(pdshape),
axes=[1, 2, 0]), data)).astype("float32")
labels_data = np.array(map(lambda x: x[1], data)).astype('int64')
iters, num_samples, start_time = 0, 0, 0.0
for pass_id in range(args.pass_num):
if iters == args.iterations:
break
train_accs = []
train_losses = []
for batch_id, data in enumerate(train_reader()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
if not args.use_fake_data:
images_data = np.array(
map(lambda x: np.transpose(x[0].reshape(pdshape),
axes=[1, 2, 0]), data)).astype("float32")
labels_data = np.array(map(lambda x: x[1], data)).astype(
'int64')
_, loss, acc = sess.run([train_op, avg_cost, accuracy],
feed_dict={
images: images_data,
labels: labels_data,
is_training: True
})
iters += 1
train_accs.append(acc)
train_losses.append(loss)
num_samples += len(data)
print("Pass=%d, Iter=%d, Loss=%f, Accuray=%f\n" %
(pass_id, iters, loss, acc))
train_elapsed = time.time() - start_time
print("Pass=%d, Loss=%f, Accuray=%f\n" %
(pass_id, np.mean(train_losses), np.mean(train_accs)))
# evaluation
if args.with_test:
test()
if not args.with_test:
duration = time.time() - start_time
examples_per_sec = num_samples / duration
sec_per_batch = duration / (iters - args.skip_batch_num)
print('Total examples: %d, total time: %.5f' %
(num_samples, duration))
print('%.5f examples/sec, %.5f sec/batch' %
(examples_per_sec, sec_per_batch))
if __name__ == '__main__':
args = parse_args()
print_arguments(args)
if tf.test.is_built_with_cuda():
device = '/device:GPU:0'
if args.order == 'NHWC':
data_format = 'channels_last'
else:
data_format = 'channels_first'
else:
device = '/cpu:0'
if args.order == 'NHWC':
data_format = 'channels_last'
else:
raise ValueError('Only support NHWC order in CPU mode')
run_benchmark(args, data_format, device)
benchmark/tensorflow/stacked_dynamic_lstm.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import numpy as np
import argparse
import time
import tensorflow as tf
import paddle.v2 as paddle
def parse_args():
parser = argparse.ArgumentParser("LSTM model benchmark.")
parser.add_argument(
'--batch_size',
type=int,
default=32,
help='The sequence number of a batch data. (default: %(default)d)')
parser.add_argument(
'--stacked_num',
type=int,
default=5,
help='Number of lstm layers to stack. (default: %(default)d)')
parser.add_argument(
'--embedding_dim',
type=int,
default=512,
help='Dimension of embedding table. (default: %(default)d)')
parser.add_argument(
'--hidden_dim',
type=int,
default=512,
help='Hidden size of lstm unit. (default: %(default)d)')
parser.add_argument(
'--pass_num',
type=int,
default=10,
help='Epoch number to train. (default: %(default)d)')
parser.add_argument(
'--learning_rate',
type=float,
default=0.0002,
help='Learning rate used to train. (default: %(default)f)')
parser.add_argument(
'--infer_only', action='store_true', help='If set, run forward only.')
args = parser.parse_args()
return args
def print_arguments(args):
print('----------- Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
def dynamic_lstm_model(dict_size,
embedding_dim,
hidden_dim,
stacked_num,
class_num=2,
is_train=True):
word_idx = tf.placeholder(tf.int64, shape=[None, None])
sequence_length = tf.placeholder(tf.int64, shape=[None, ])
embedding_weights = tf.get_variable('word_embeddings',
[dict_size, embedding_dim])
embedding = tf.nn.embedding_lookup(embedding_weights, word_idx)
lstm_cell = tf.nn.rnn_cell.LSTMCell(
num_units=hidden_dim, use_peepholes=False)
stacked_cell = tf.nn.rnn_cell.MultiRNNCell([lstm_cell] * stacked_num)
# final_state [LSTMTuple(c, h), LSTMTuple(c, h) ...] total stacked_num LSTMTuples
_, final_state = tf.nn.dynamic_rnn(
cell=stacked_cell,
inputs=embedding,
dtype=tf.float32,
sequence_length=sequence_length)
w = tf.Variable(
tf.truncated_normal([hidden_dim, class_num]), dtype=tf.float32)
bias = tf.Variable(
tf.constant(
value=0.0, shape=[class_num], dtype=tf.float32))
prediction = tf.matmul(final_state[-1][1], w) + bias
if not is_train:
return (word_idx, sequence_length), tf.nn.softmax(prediction)
label = tf.placeholder(tf.int64, shape=[None, ])
loss = tf.nn.softmax_cross_entropy_with_logits(
labels=tf.one_hot(label, 2), logits=prediction)
avg_loss = tf.reduce_mean(loss)
correct_count = tf.equal(tf.argmax(prediction, 1), label)
acc = tf.reduce_mean(tf.cast(correct_count, tf.float32))
with tf.variable_scope("reset_metrics_accuracy_scope") as scope:
g_acc = tf.metrics.accuracy(label, tf.argmax(prediction, axis=1))
vars = tf.contrib.framework.get_variables(
scope, collection=tf.GraphKeys.LOCAL_VARIABLES)
reset_op = tf.variables_initializer(vars)
return (word_idx, sequence_length, label), avg_loss, acc, g_acc, reset_op
def padding_data(data, padding_size, value):
data = data + [value] * padding_size
return data[:padding_size]
def train(args):
word_dict = paddle.dataset.imdb.word_dict()
dict_size = len(word_dict)
feeding_list, avg_loss, acc, g_acc, reset_op = dynamic_lstm_model(
dict_size, args.embedding_dim, args.hidden_dim, args.stacked_num)
adam_optimizer = tf.train.AdamOptimizer(learning_rate=args.learning_rate)
train_op = adam_optimizer.minimize(avg_loss)
train_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.imdb.train(word_dict), buf_size=25000),
batch_size=args.batch_size)
test_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.imdb.test(word_dict), buf_size=25000),
batch_size=args.batch_size)
def do_validation(sess):
sess.run(reset_op)
for batch_id, data in enumerate(test_reader()):
word_idx = map(lambda x: x[0], data)
sequence_length = np.array(
[len(seq) for seq in word_idx]).astype('int64')
maxlen = np.max(sequence_length)
word_idx = [padding_data(seq, maxlen, 0) for seq in word_idx]
word_idx = np.array(word_idx).astype('int64')
label = np.array(map(lambda x: x[1], data)).astype('int64')
_, loss, fetch_acc, fetch_g_acc = sess.run(
[train_op, avg_loss, acc, g_acc],
feed_dict={
feeding_list[0]: word_idx,
feeding_list[1]: sequence_length,
feeding_list[2]: label
})
return fetch_g_acc[1]
config = tf.ConfigProto(
intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)
config.gpu_options.allow_growth = True
with tf.Session(config=config) as sess:
init_g = tf.global_variables_initializer()
init_l = tf.local_variables_initializer()
sess.run(init_l)
sess.run(init_g)
for pass_id in xrange(args.pass_num):
# clear accuracy local variable
sess.run(reset_op)
pass_start_time = time.time()
words_seen = 0
for batch_id, data in enumerate(train_reader()):
word_idx = map(lambda x: x[0], data)
sequence_length = np.array(
[len(seq) for seq in word_idx]).astype('int64')
words_seen += np.sum(sequence_length)
maxlen = np.max(sequence_length)
word_idx = [padding_data(seq, maxlen, 0) for seq in word_idx]
word_idx = np.array(word_idx).astype('int64')
label = np.array(map(lambda x: x[1], data)).astype('int64')
_, loss, fetch_acc, fetch_g_acc = sess.run(
[train_op, avg_loss, acc, g_acc],
feed_dict={
feeding_list[0]: word_idx,
feeding_list[1]: sequence_length,
feeding_list[2]: label
})
print("pass_id=%d, batch_id=%d, loss: %f, acc: %f, avg_acc: %f"
% (pass_id, batch_id, loss, fetch_acc, fetch_g_acc[1]))
pass_end_time = time.time()
time_consumed = pass_end_time - pass_start_time
words_per_sec = words_seen / time_consumed
test_acc = do_validation(sess)
print("pass_id=%d, test_acc: %f, words/s: %f, sec/pass: %f" %
(pass_id, test_acc, words_per_sec, time_consumed))
if __name__ == '__main__':
args = parse_args()
print_arguments(args)
if args.infer_only:
pass
else:
train(args)
benchmark/tensorflow/vgg.py 0 → 100644
浏览文件 @ e2ca8e10
# Copyright (c) 2018 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.
"""VGG16 benchmark in TensorFlow"""
import tensorflow as tf
import paddle.v2 as paddle
import numpy as np
import argparse
import time
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
'--batch_size', type=int, default=128, help="Batch size for training.")
parser.add_argument(
'--skip_batch_num',
type=int,
default=5,
help='The first num of minibatch num to skip, for better performance test')
parser.add_argument(
'--iterations', type=int, default=80, help='The number of minibatches.')
parser.add_argument(
'--learning_rate',
type=float,
default=1e-3,
help="Learning rate for training.")
parser.add_argument('--num_passes', type=int, default=50, help="No. of passes.")
parser.add_argument(
'--device',
type=str,
default='GPU',
choices=['CPU', 'GPU'],
help="The device type.")
parser.add_argument(
'--data_format',
type=str,
default='NHWC',
choices=['NCHW', 'NHWC'],
help='The data order, NCHW=[batch, channels, height, width].'
'Only support NHWC right now.')
parser.add_argument(
'--data_set',
type=str,
default='cifar10',
choices=['cifar10', 'flowers'],
help='Optional dataset for benchmark.')
args = parser.parse_args()
class VGG16Model(object):
def __init__(self):
self.parameters = []
def batch_norm_relu(self, inputs, is_training):
"""Performs a batch normalization followed by a ReLU."""
# We set fused=True for a significant speed boost. See
# https://www.tensorflow.org/speed/speed_guide#common_fused_ops
inputs = tf.layers.batch_normalization(
inputs=inputs,
axis=1 if args.data_format == 'NCHW' else -1,
momentum=0.9,
epsilon=1e-05,
center=True,
scale=True,
training=is_training,
fused=True)
inputs = tf.nn.relu(inputs)
return inputs
def conv_bn_layer(self,
name,
images,
kernel_shape,
is_training,
drop_rate=0.0):
with tf.name_scope(name) as scope:
kernel = tf.Variable(
tf.truncated_normal(
kernel_shape, dtype=tf.float32, stddev=1e-1),
name='weights')
conv = tf.nn.conv2d(
images,
kernel, [1, 1, 1, 1],
data_format=args.data_format,
padding='SAME')
biases = tf.Variable(
tf.constant(
0.0, shape=[kernel_shape[-1]], dtype=tf.float32),
trainable=True,
name='biases')
out = tf.nn.bias_add(conv, biases)
out = self.batch_norm_relu(out, is_training)
out = tf.layers.dropout(out, rate=drop_rate, training=is_training)
return out
def fc_layer(self, name, inputs, shape):
with tf.name_scope(name) as scope:
fc_w = tf.Variable(
tf.truncated_normal(
shape, dtype=tf.float32, stddev=1e-1),
name='weights')
fc_b = tf.Variable(
tf.constant(
0.0, shape=[shape[-1]], dtype=tf.float32),
trainable=True,
name='biases')
out = tf.nn.bias_add(tf.matmul(inputs, fc_w), fc_b)
return out
def network(self, images, class_dim, is_training):
""" VGG16 model structure.
TODO(kuke): enable this network to support the 'NCHW' data format
"""
# conv1
conv1_1 = self.conv_bn_layer(
'conv1_1', images, [3, 3, 3, 64], is_training, drop_rate=0.3)
conv1_2 = self.conv_bn_layer(
'conv1_2', conv1_1, [3, 3, 64, 64], is_training, drop_rate=0.0)
# pool1
pool1 = tf.nn.max_pool(
conv1_2,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool1')
# conv2
conv2_1 = self.conv_bn_layer(
'conv2_1', pool1, [3, 3, 64, 128], is_training, drop_rate=0.4)
conv2_2 = self.conv_bn_layer(
'conv2_2', conv2_1, [3, 3, 128, 128], is_training, drop_rate=0.0)
# pool2
pool2 = tf.nn.max_pool(
conv2_2,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool2')
# conv3
conv3_1 = self.conv_bn_layer(
'conv3_1', pool2, [3, 3, 128, 256], is_training, drop_rate=0.4)
conv3_2 = self.conv_bn_layer(
'conv3_2', conv3_1, [3, 3, 256, 256], is_training, drop_rate=0.4)
conv3_3 = self.conv_bn_layer(
'conv3_3', conv3_2, [3, 3, 256, 256], is_training, drop_rate=0.0)
# pool3
pool3 = tf.nn.max_pool(
conv3_3,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool3')
# conv4
conv4_1 = self.conv_bn_layer(
'conv4_1', pool3, [3, 3, 256, 512], is_training, drop_rate=0.4)
conv4_2 = self.conv_bn_layer(
'conv4_2', conv4_1, [3, 3, 512, 512], is_training, drop_rate=0.4)
conv4_3 = self.conv_bn_layer(
'conv4_3', conv4_2, [3, 3, 512, 512], is_training, drop_rate=0.0)
# pool4
pool4 = tf.nn.max_pool(
conv4_3,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool4')
# conv5
conv5_1 = self.conv_bn_layer(
'conv5_1', pool4, [3, 3, 512, 512], is_training, drop_rate=0.4)
conv5_2 = self.conv_bn_layer(
'conv5_2', conv5_1, [3, 3, 512, 512], is_training, drop_rate=0.4)
conv5_3 = self.conv_bn_layer(
'conv5_3', conv5_2, [3, 3, 512, 512], is_training, drop_rate=0.0)
# pool5
pool5 = tf.nn.max_pool(
conv5_3,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool4')
# flatten
shape = int(np.prod(pool5.get_shape()[1:]))
pool5_flat = tf.reshape(pool5, [-1, shape])
# fc1
drop = tf.layers.dropout(pool5_flat, rate=0.5, training=is_training)
fc1 = self.fc_layer('fc1', drop, [shape, 512])
# fc2
bn = self.batch_norm_relu(fc1, is_training)
drop = tf.layers.dropout(bn, rate=0.5, training=is_training)
fc2 = self.fc_layer('fc2', drop, [512, 512])
fc3 = self.fc_layer('fc3', fc2, [512, class_dim])
return fc3
def run_benchmark():
"""Run benchmark on cifar10 or flowers."""
if args.data_set == "cifar10":
class_dim = 10
raw_shape = (3, 32, 32)
dat_shape = (None, 32, 32, 3) if args.data_format == 'NHWC' else (
None, 3, 32, 32)
else:
class_dim = 102
raw_shape = (3, 224, 224)
dat_shape = (None, 224, 224, 3) if args.data_format == 'NHWC' else (
None, 3, 224, 224)
device = '/cpu:0' if args.device == 'CPU' else '/device:GPU:0'
with tf.device(device):
images = tf.placeholder(tf.float32, shape=dat_shape)
labels = tf.placeholder(tf.int64, shape=(None, ))
is_training = tf.placeholder('bool')
onehot_labels = tf.one_hot(labels, depth=class_dim)
vgg16 = VGG16Model()
logits = vgg16.network(images, class_dim, is_training)
loss = tf.losses.softmax_cross_entropy(
onehot_labels=onehot_labels, logits=logits)
avg_loss = tf.reduce_mean(loss)
correct = tf.equal(tf.argmax(logits, 1), labels)
accuracy = tf.reduce_mean(tf.cast(correct, tf.float32))
optimizer = tf.train.AdamOptimizer(learning_rate=args.learning_rate)
update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
train_op = optimizer.minimize(avg_loss)
# data reader
train_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.cifar.train10()
if args.data_set == 'cifar10' else paddle.dataset.flowers.train(),
buf_size=5120),
batch_size=args.batch_size)
test_reader = paddle.batch(
paddle.reader.shuffle(
paddle.dataset.cifar.test10()
if args.data_set == 'cifar10' else paddle.dataset.flowers.test(),
buf_size=5120),
batch_size=args.batch_size)
# test
def test():
test_accs = []
for batch_id, data in enumerate(test_reader()):
test_images = np.array(
map(lambda x: np.transpose(x[0].reshape(raw_shape),
axes=[1, 2, 0]) if args.data_format == 'NHWC' else x[0], data)).astype("float32")
test_labels = np.array(map(lambda x: x[1], data)).astype('int64')
test_accs.append(
accuracy.eval(feed_dict={
images: test_images,
labels: test_labels,
is_training: False
}))
return np.mean(test_accs)
config = tf.ConfigProto(
intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)
config.gpu_options.allow_growth = True
with tf.Session(config=config) as sess:
init_g = tf.global_variables_initializer()
init_l = tf.local_variables_initializer()
sess.run(init_g)
sess.run(init_l)
iters, num_samples, start_time = 0, 0, time.time()
for pass_id in range(args.num_passes):
# train
num_samples = 0
start_time = time.time()
for batch_id, data in enumerate(train_reader()):
if iters == args.skip_batch_num:
start_time = time.time()
num_samples = 0
if iters == args.iterations:
break
train_images = np.array(
map(lambda x: np.transpose(x[0].reshape(raw_shape),
axes=[1, 2, 0]) if args.data_format == 'NHWC' else x[0], data)).astype("float32")
train_labels = np.array(map(lambda x: x[1], data)).astype(
'int64')
_, loss, acc = sess.run([train_op, avg_loss, accuracy],
feed_dict={
images: train_images,
labels: train_labels,
is_training: True
})
iters += 1
num_samples += len(data)
print("Pass = %d, Iters = %d, Loss = %f, Accuracy = %f" %
(pass_id, iters, loss, acc))
train_elapsed = time.time() - start_time
# test
pass_test_acc = test()
print("Pass = %d, Train speed = %f imgs/s, Test accuracy = %f\n" %
(pass_id, num_samples / train_elapsed, pass_test_acc))
def print_arguments():
print('----------- Configuration Arguments -----------')
for arg, value in sorted(vars(args).iteritems()):
print('%s: %s' % (arg, value))
print('------------------------------------------------')
if __name__ == '__main__':
print_arguments()
run_benchmark()
cmake/external/mkldnn.cmake
浏览文件 @ e2ca8e10
......@@ -36,7 +36,8 @@ MESSAGE(STATUS "Set ${MKLDNN_INSTALL_DIR}/lib to runtime path")
SET(CMAKE_INSTALL_RPATH_USE_LINK_PATH TRUE)
SET(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}" "${MKLDNN_INSTALL_DIR}/lib")
INCLUDE_DIRECTORIES(${MKLDNN_INC_DIR})
INCLUDE_DIRECTORIES(${MKLDNN_INC_DIR}) # For MKLDNN code to include internal headers.
INCLUDE_DIRECTORIES(${THIRD_PARTY_PATH}/install) # For Paddle code to include mkldnn.h
IF(${CBLAS_PROVIDER} STREQUAL "MKLML")
SET(MKLDNN_DEPENDS ${MKLML_PROJECT})
......
cmake/external/mklml.cmake
浏览文件 @ e2ca8e10
......@@ -28,7 +28,7 @@ INCLUDE(ExternalProject)
SET(MKLML_PROJECT "extern_mklml")
SET(MKLML_VER "mklml_lnx_2018.0.1.20171007")
SET(MKLML_URL "https://github.com/01org/mkl-dnn/releases/download/v0.11/${MKLML_VER}.tgz")
SET(MKLML_URL "http://paddlepaddledeps.bj.bcebos.com/${MKLML_VER}.tgz")
SET(MKLML_SOURCE_DIR "${THIRD_PARTY_PATH}/mklml")
SET(MKLML_DOWNLOAD_DIR "${MKLML_SOURCE_DIR}/src/${MKLML_PROJECT}")
SET(MKLML_DST_DIR "mklml")
......
cmake/external/snappystream.cmake
浏览文件 @ e2ca8e10
......@@ -54,5 +54,7 @@ add_library(snappystream STATIC IMPORTED GLOBAL)
set_property(TARGET snappystream PROPERTY IMPORTED_LOCATION
"${SNAPPYSTREAM_INSTALL_DIR}/lib/libsnappystream.a")
include_directories(${SNAPPYSTREAM_INCLUDE_DIR})
include_directories(${SNAPPYSTREAM_INCLUDE_DIR}) # For snappysteam to include its own headers.
include_directories(${THIRD_PARTY_PATH}/install) # For Paddle to include snappy stream headers.
add_dependencies(snappystream extern_snappystream)
cmake/external/warpctc.cmake
浏览文件 @ e2ca8e10
......@@ -62,7 +62,8 @@ ExternalProject_Add(
)
MESSAGE(STATUS "warp-ctc library: ${WARPCTC_LIBRARIES}")
INCLUDE_DIRECTORIES(${WARPCTC_INCLUDE_DIR})
INCLUDE_DIRECTORIES(${WARPCTC_INCLUDE_DIR}) # For warpctc code to include its headers.
INCLUDE_DIRECTORIES(${THIRD_PARTY_PATH}/install) # For Paddle code to include warpctc headers.
ADD_LIBRARY(warpctc SHARED IMPORTED GLOBAL)
SET_PROPERTY(TARGET warpctc PROPERTY IMPORTED_LOCATION ${WARPCTC_LIBRARIES})
......
cmake/external/zlib.cmake
浏览文件 @ e2ca8e10
......@@ -25,7 +25,8 @@ ELSE(WIN32)
SET(ZLIB_LIBRARIES "${ZLIB_INSTALL_DIR}/lib/libz.a" CACHE FILEPATH "zlib library." FORCE)
ENDIF(WIN32)
INCLUDE_DIRECTORIES(${ZLIB_INCLUDE_DIR})
INCLUDE_DIRECTORIES(${ZLIB_INCLUDE_DIR}) # For zlib code to include its own headers.
INCLUDE_DIRECTORIES(${THIRD_PARTY_PATH}/install) # For Paddle code to include zlib.h.
ExternalProject_Add(
extern_zlib
......
cmake/generic.cmake
浏览文件 @ e2ca8e10
......@@ -244,14 +244,14 @@ function(cc_test TARGET_NAME)
cmake_parse_arguments(cc_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
add_executable(${TARGET_NAME} ${cc_test_SRCS})
# Support linking flags: --whole-archive (Linux) / -force_load (MacOS)
target_circle_link_libraries(${TARGET_NAME} ${cc_test_DEPS} paddle_gtest_main paddle_memory gtest gflags glog)
target_circle_link_libraries(${TARGET_NAME} ${cc_test_DEPS} paddle_gtest_main memory gtest gflags glog)
if("${cc_test_DEPS}" MATCHES "ARCHIVE_START")
list(REMOVE_ITEM cc_test_DEPS ARCHIVE_START ARCHIVE_END)
endif()
add_dependencies(${TARGET_NAME} ${cc_test_DEPS} paddle_gtest_main paddle_memory gtest gflags glog)
add_dependencies(${TARGET_NAME} ${cc_test_DEPS} paddle_gtest_main memory gtest gflags glog)
add_test(NAME ${TARGET_NAME}
COMMAND ${TARGET_NAME} ${cc_test_ARGS}
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})
endif()
endfunction(cc_test)
......@@ -311,8 +311,8 @@ function(nv_test TARGET_NAME)
set(multiValueArgs SRCS DEPS)
cmake_parse_arguments(nv_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
cuda_add_executable(${TARGET_NAME} ${nv_test_SRCS})
target_link_libraries(${TARGET_NAME} ${nv_test_DEPS} paddle_gtest_main paddle_memory gtest gflags glog)
add_dependencies(${TARGET_NAME} ${nv_test_DEPS} paddle_gtest_main paddle_memory gtest gflags glog)
target_link_libraries(${TARGET_NAME} ${nv_test_DEPS} paddle_gtest_main memory gtest gflags glog)
add_dependencies(${TARGET_NAME} ${nv_test_DEPS} paddle_gtest_main memory gtest gflags glog)
add_test(${TARGET_NAME} ${TARGET_NAME})
endif()
endfunction(nv_test)
......@@ -387,8 +387,8 @@ function(hip_test TARGET_NAME)
endif()
add_executable(${TARGET_NAME} ${_cmake_options} ${_generated_files} ${_sources})
set_target_properties(${TARGET_NAME} PROPERTIES LINKER_LANGUAGE HIP)
target_link_libraries(${TARGET_NAME} ${hip_test_DEPS} paddle_gtest_main paddle_memory gtest gflags)
add_dependencies(${TARGET_NAME} ${hip_test_DEPS} paddle_gtest_main paddle_memory gtest gflags)
target_link_libraries(${TARGET_NAME} ${hip_test_DEPS} paddle_gtest_main memory gtest gflags)
add_dependencies(${TARGET_NAME} ${hip_test_DEPS} paddle_gtest_main memory gtest gflags)
add_test(${TARGET_NAME} ${TARGET_NAME})
endif()
endfunction(hip_test)
......@@ -561,9 +561,9 @@ function(py_test TARGET_NAME)
set(multiValueArgs SRCS DEPS ARGS ENVS)
cmake_parse_arguments(py_test "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
add_test(NAME ${TARGET_NAME}
COMMAND env PYTHONPATH=${PADDLE_PYTHON_BUILD_DIR}/lib-python ${py_test_ENVS}
COMMAND env PYTHONPATH=${PADDLE_BINARY_DIR}/python ${py_test_ENVS}
${PYTHON_EXECUTABLE} -u ${py_test_SRCS} ${py_test_ARGS}
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})
endif()
endfunction()
......
doc/design/file_manager/README.md 已删除 100644 → 0
浏览文件 @ 92818baa
# FileManager设计文档
## 目标
在本文档中，我们设计说明了名为FileManager系统，方便用户上传自己的训练数据以进行分布式训练
主要功能包括：
- 提供常用的命令行管理命令管理文件和目录
- 支持大文件的断点上传、下载
## 名词解释
- PFS：是`Paddlepaddle cloud File System`的缩写，是对用户文件存储空间的抽象，与之相对的是local filesystem。目前我们用CephFS来搭建。
- [CephFS](http://docs.ceph.com/docs/master/cephfs/)：一个POSIX兼容的文件系统。
- Chunk：逻辑划上文件分块的单位。
## 模块
### 架构图
<image src=./src/filemanager.png width=900>
### PFSClient
- 功能： 详细设计[link](./pfs/pfsclient.md)
- 提供用户管理文件的命令
- 需要可以跨平台执行
- 双向验证
PFSClient需要和Ingress之间做双向验证<sup>[tls](#tls)</sup>，所以用户需要首先在`cloud.paddlepaddle.org`上注册一下，申请用户空间，并且把系统生成的CA(certificate authority)、Key、CRT(CA signed certificate)下载到本地，然后才能使用PFSClient。
### [Ingress](https://kubernetes.io/docs/concepts/services-networking/ingress/)
- 功能：
提供七层协议的反向代理、基于粘性会话的负载均衡功能。
- 透传用户身份的办法
Ingress需要把PFSClient的身份信息传给PFSServer，配置的方法参考[link](http://www.integralist.co.uk/posts/clientcertauth.html#3)
### PFSServer
PFSServer提供RESTful API接口，接收处理PFSClient端的文件管理请求，并且把结果返回PFSClient端。
RESTful API
- /api/v1/files
- `GET /api/v1/files`: Get metadata of files or directories.
- `POST /api/v1/files`: Create files or directories.
- `PATCH /api/v1/files`: Update files or directories.
- `DELETE /api/v1/files`: Delete files or directories.
- /api/v1/file/chunks
- `GET /api/v1/storage/file/chunks`: Get chunks's metadata of a file.
- /api/v1/storage/files
- `GET /api/v1/storage/files`: Download files or directories.
- `POST /api/v1/storage/files`: Upload files or directories.
- /api/v1/storage/file/chunks
- `GET /api/v1/storage/file/chunks`: Download chunks's data.
- `POST /api/v1/storage/file/chunks`: Upload chunks's data.
## 文件传输优化
### 分块文件传输
用户文件可能是比较大的，上传到Cloud或者下载到本地的时间可能比较长，而且在传输的过程中也可能出现网络不稳定的情况。为了应对以上的问题，我们提出了Chunk的概念，一个Chunk由所在的文件偏移、数据、数据长度及校验值组成。文件的上传和下载都是通过对Chunk的操作来实现的。由于Chunk比较小（默认256K），完成一个传输动作完成的时间也比较短，不容易出错。PFSClient需要在传输完毕最后一个Chunk的时候检查destination文件的MD5值是否和source文件一致。
一个典型的Chunk如下所示：
```
type Chunk struct {
fileOffset int64
checksum uint32
len uint32
data []byte
}
```
### 生成sparse文件
当destination文件不存在或者大小和source文件不一致时，可以用[Fallocate](https://Go.org/pkg/syscall/#Fallocate)生成sparse文件，然后就可以并发写入多个Chunk。
### 覆盖不一致的部分
文件传输的的关键在于需要PFSClient端对比source和destination的文件Chunks的checksum是否保持一致，不一致的由PFSClient下载或者传输Chunk完成。这样已经传输成功的部分就不用重新传输了。
## 用户使用流程
参考[link](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/cluster_train/data_dispatch.md)
## 框架生成
用[swagger](https://github.com/swagger-api/swagger-codegen)生成PFSClient和PFSServer的框架部分，以便我们可以把更多的精力放到逻辑本身上。
## 参考文档
- <a name=tls></a>[TLS complete guide](https://github.com/k8sp/tls/blob/master/tls.md)
- [aws.s3](http://docs.aws.amazon.com/cli/latest/reference/s3/)
- [linux man document](https://linux.die.net/man/)
doc/design/file_manager/pfs/pfsclient.md 已删除 100644 → 0
浏览文件 @ 92818baa
# PFSClient
## Description
The `pfs` command is a Command Line Interface to manage your files on PaddlePaddle Cloud
## Synopsis
```
paddle [options] pfs <subcommand> [parameters]
```
## Options
```
--profile (string)
Use a specific profile from your credential file.
--help (string)
Display more information about command
--version
Output version information and exit
--debug
Show detailed debugging log
--only-show-errors (boolean)
Only errors and warnings are displayed. All other output is suppressed.
```
## Path Arguments
When using a command, we need to specify path arguments. There are two path argument type: `localpath` and `pfspath`.
A `pfspath` begin with `/pfs`, eg: `/pfs/$DATACENTER/home/$USER/folder`.
[Here](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/cluster_train/data_dispatch.md#上传训练文件) is how to config datacenters.
## order of Path Arguments
Commonly, if there are two path arguments, the first is the source, and the second is the destination.
## Subcommonds
- rm - remove files or directories
```
Synopsis:
rm [-r] [-v] <PFSPath> ...
Options:
-r
Remove directories and their contents recursively
-v
Cause rm to be verbose, showing files after they are removed.
Examples:
paddle pfs rm /pfs/$DATACENTER/home/$USER/file
paddle pfs rm -r /pfs/$DATACENTER/home/$USER/folder
```
- mv - move (rename) files
```
Synopsis:
mv [-f | -n] [-v] <LocalPath> <PFSPath>
mv [-f | -n] [-v] <LocalPath> ... <PFSPath>
mv [-f | -n] [-v] <PFSPath> <LocalPath>
mv [-f | -n] [-v] <PFSPath> ... <LocalPath>
mv [-f | -n] [-v] <PFSPath> <PFSPath>
mv [-f | -n] [-v] <PFSPath> ... <PFSPath>
Options:
-f
Do not prompt for confirmation before overwriting the destination path. (The -f option overrides previous -n options.)
-n
Do not overwrite an existing file. (The -n option overrides previous -f options.)
-v
Cause mv to be verbose, showing files after they are moved.
Examples:
paddle pfs mv ./text1.txt /pfs/$DATACENTER/home/$USER/text1.txt
```
- cp - copy files or directories
```
Synopsis:
cp [-r] [-f | -n] [-v] [--preserve--links] <LocalPath> <PFSPath>
cp [-r] [-f | -n] [-v] [--preserve--links] <LocalPath> ... <PFSPath>
cp [-r] [-f | -n] [-v] [--preserve--links] <PFSPath> <LocalPath>
cp [-r] [-f | -n] [-v] [--preserve--links] <PFSPath> ... <LocalPath>
cp [-r] [-f | -n] [-v] [--preserve--links] <PFSPath> <PFSPath>
cp [-r] [-f | -n] [-v] [--preserve--links] <PFSPath> ... <PFSPath>
Options:
-r
Copy directories recursively
-f
Do not prompt for confirmation before overwriting the destination path. (The -f option overrides previous -n options.)
-n
Do not overwrite an existing file. (The -n option overrides previous -f options.)
-v
Cause cp to be verbose, showing files after they are copied.
--preserve--links
Reserve links when copy links
Examples:
paddle pfs cp ./file /pfs/$DATACENTER/home/$USER/file
paddle pfs cp /pfs/$DATACENTER/home/$USER/file ./file
```
- ls- list files
```
Synopsis:
ls [-r] <PFSPath> ...
Options:
-R
List directory(ies) recursively
Examples:
paddle pfs ls /pfs/$DATACENTER/home/$USER/file
paddle pfs ls /pfs/$DATACENTER/home/$USER/folder
```
- mkdir - mkdir directory(ies)
Create intermediate directory(ies) as required.
```
Synopsis:
mkdir <PFSPath> ...
Examples:
paddle pfs mkdir /pfs/$DATACENTER/home/$USER/folder
```
doc/fluid/CMakeLists.txt
浏览文件 @ e2ca8e10
......@@ -27,7 +27,7 @@ sphinx_add_target(paddle_fluid_docs
${CMAKE_CURRENT_SOURCE_DIR}
${SPHINX_HTML_DIR_EN})
add_dependencies(paddle_fluid_docs gen_proto_py)
add_dependencies(paddle_fluid_docs gen_proto_py paddle_python)
# configured documentation tools and intermediate build results
set(BINARY_BUILD_DIR_CN "${CMAKE_CURRENT_BINARY_DIR}/cn/_build")
......@@ -50,6 +50,6 @@ sphinx_add_target(paddle_fluid_docs_cn
${CMAKE_CURRENT_SOURCE_DIR}
${SPHINX_HTML_DIR_CN})
add_dependencies(paddle_fluid_docs_cn gen_proto_py)
add_dependencies(paddle_fluid_docs_cn gen_proto_py paddle_python)
add_subdirectory(api)
doc/fluid/api/CMakeLists.txt
浏览文件 @ e2ca8e10
......@@ -19,4 +19,4 @@ sphinx_add_target(paddle_fluid_apis
${CMAKE_CURRENT_SOURCE_DIR}
${SPHINX_HTML_DIR_EN})
add_dependencies(paddle_fluid_apis gen_proto_py framework_py_proto copy_paddle_pybind)
add_dependencies(paddle_fluid_apis gen_proto_py framework_py_proto copy_paddle_pybind paddle_python)
doc/fluid/api/layers.rst
浏览文件 @ e2ca8e10
......@@ -494,6 +494,12 @@ reshape
.. autofunction:: paddle.fluid.layers.reshape
:noindex:
pad
---
.. autofunction:: paddle.fluid.layers.pad
:noindex:
scale
-----
......
doc/fluid/design/algorithm/parameter_average.md
浏览文件 @ e2ca8e10
......@@ -5,9 +5,11 @@ In a large scale machine learning setup where the size of the training data is h
Polyak and Juditsky (1992) showed that the test performance of simple average of parameters obtained by Stochastic Gradient Descent (SGD) is as good as that of parameter values that are obtained by training the model over and over again, over the training dataset.
Hence, to accelerate the speed of Stochastic Gradient Descent, Averaged Stochastic Gradient Descent (ASGD) was proposed in Polyak and Juditsky (1992). For ASGD, the running average of parameters obtained by SGD, is used as the estimator for <img src="./images/theta_star.gif"/><br/> . The averaging is done as follows:
Hence, to accelerate the speed of Stochastic Gradient Descent, Averaged Stochastic Gradient Descent (ASGD) was proposed in Polyak and Juditsky (1992). For ASGD, the running average of parameters obtained by SGD, is used as the estimator for <img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/theta_star.gif"/><br/> . The averaging is done as follows:
<img src="./images/asgd.gif" align="center"/><br/>
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/asgd.gif"><br />
</p>
We propose averaging for any optimizer similar to how ASGD performs it, as mentioned above.
......
doc/fluid/design/concepts/README.md
浏览文件 @ e2ca8e10
......@@ -6,11 +6,33 @@ Here are some initial thoughts. Your comments are welcome!
I think we need only the following few CMake functions to make a project description mean and clean:
| C++ | CUDA C++ | Go |
|---|---|---|
| cc_library | nv_library | go_library |
| cc_binary | nv_binary | go_binary |
| cc_test | nv_test | go_test |
<table>
<thead>
<tr>
<th>C++</th>
<th>CUDA C++</th>
<th>Go</th>
</tr>
</thead>
<tbody>
<tr>
<td>cc_library </td>
<td>nv_library </td>
<td>go_library </td>
</tr>
<tr>
<td>cc_binary </td>
<td>nv_binary </td>
<td>go_binary </td>
</tr>
<tr>
<td> cc_test </td>
<td> nv_test </td>
<td> go_test </td>
</tr>
</tbody>
</table>
- The `_library` functions generate .a files from source code.
- The `_binary` functions generate executable binary files.
......
doc/fluid/design/concepts/block.md
浏览文件 @ e2ca8e10
......@@ -14,11 +14,29 @@ In programming languages, a block is a pair of curly braces that includes local
Blocks work with control flow structures like `if`, `else`, and `for`, which have equivalents in deep learning:
| programming languages | PaddlePaddle |
|-----------------------|-----------------------|
| for, while loop | RNN, WhileOp |
| if, if-else, switch | IfElseOp, SwitchOp |
| sequential execution | a sequence of layers |
<table>
<thead>
<tr>
<th>programming languages</th>
<th>PaddlePaddle</th>
</tr>
</thead>
<tbody>
<tr>
<td>for, while loop </td>
<td>RNN, WhileOp </td>
</tr>
<tr>
<td>if, if-else, switch </td>
<td>IfElseOp, SwitchOp </td>
</tr>
<tr>
<td>sequential execution </td>
<td>a sequence of layers </td>
</tr>
</tbody>
</table>
A key difference is that a C++ program describes a one pass computation, whereas a deep learning program describes both the forward and backward passes.
......@@ -26,12 +44,33 @@ A key difference is that a C++ program describes a one pass computation, whereas
The existence of the backward pass makes the execution of a block of PaddlePaddle different from traditional programs:
| programming languages | PaddlePaddle |
|-----------------------|---------------------------------|
| stack | scope hierarchy |
| stack frame | scope |
| push at entering block| push at entering block |
| pop at leaving block | destroy when minibatch completes|
<table>
<thead>
<tr>
<th>programming languages</th>
<th>PaddlePaddle</th>
</tr>
</thead>
<tbody>
<tr>
<td>stack </td>
<td>scope hierarchy </td>
</tr>
<tr>
<td>stack frame </td>
<td>scope </td>
</tr>
<tr>
<td>push at entering block </td>
<td>push at entering block </td>
</tr>
<tr>
<td>pop at leaving block </td>
<td>destroy when minibatch completes </td>
</tr>
</tbody>
</table>
1. In traditional programs:
......
doc/fluid/design/concepts/functions_operators_layers.md
浏览文件 @ e2ca8e10
......@@ -86,12 +86,40 @@ def layer.fc(X):
We'd like to have Python bindings to operators in package `paddle.operator`, and Python compositions of operators in package `paddle.layer`. So we have the following concepts in above illustrative example:
| C++ functions/functors | mul | add | | |
|------------------------|--------------|--------------|-------------|----------|
| C++ operator class | mulOp | addOp | FCOp | |
| Python binding | operator.mul | operator.add | operator.fc | |
| Python function | | | | layer.fc |
<table>
<thead>
<tr>
<th>C++ functions/functors</th>
<th>mul</th>
<th>add</th>
<th></th>
<th></th>
</tr>
</thead>
<tbody>
<tr>
<td>C++ operator class </td>
<td>mulOp</td>
<td>addOp </td>
<td>FCOp </td>
<td></td>
</tr>
<tr>
<td>Python binding </td>
<td>operator.mul</td>
<td> operator.add </td>
<td>operator.fc </td>
<td></td>
</tr>
<tr>
<td>Python function </td>
<td></td>
<td></td>
<td> </td>
<td>layer.fc</td>
</tr>
</tbody>
</table>
This is how we differentiate layer and operators in PaddlePaddle:
......
doc/fluid/design/concepts/index_cn.rst
浏览文件 @ e2ca8e10
......@@ -16,3 +16,4 @@
block.md
scope.md
executor.md
parallel_executor.md
doc/fluid/design/concepts/index_en.rst
浏览文件 @ e2ca8e10
......@@ -16,3 +16,4 @@ Core Concepts
block.md
scope.md
executor.md
parallel_executor.md
doc/fluid/design/concepts/lod_tensor.md
浏览文件 @ e2ca8e10
......@@ -2,12 +2,38 @@
Like other deep learning systems, PaddlePaddle supports training models from sequence data. Also, like other systems, PaddlePaddle represent a mini-batch of sequences as a Tensor. What is different is that PaddlePaddle doesn't require all sequences in a mini-batch to be of the same length. Thus no need for padding zeros.
| | TensorFlow | PaddlePaddle |
|-----------------------|------------|--------------|
| RNN | Support | Support |
| recursive RNN | Support | Support |
| padding zeros | Must | No need |
| blob data type | Tensor | LoDTensor |
<table>
<thead>
<tr>
<th></th>
<th>TensorFlow</th>
<th>PaddlePaddle</th>
</tr>
</thead>
<tbody>
<tr>
<td>RNN </td>
<td>Support </td>
<td>Support </td>
</tr>
<tr>
<td>recursive RNN </td>
<td>Support </td>
<td>Support </td>
</tr>
<tr>
<td>padding zeros </td>
<td> Must </td>
<td>No need </td>
</tr>
<tr>
<td> blob data type </td>
<td> Tensor</td>
<td> LoDTensor </td>
</tr>
</tbody>
</table>
PaddlePaddle achieves this flexibility by passing through a new data type, *LoD Tensor*, which is a Tensor attached with segmentation index known as *LoD*, between operators. The LoD index doesn't only segment a tensor, but also recursively segments sub-sequences. This document presents the design of LoD and LoDTensor.
......
doc/fluid/design/concepts/var_desc.md
浏览文件 @ e2ca8e10
......@@ -10,10 +10,27 @@ PaddlePaddle uses proto message to describe compile time program because :
The computation `Program` consists of nested `Blocks`. Each `Block` will consist of data(i.e. `Variable`) and `Operations`. The concept to represent them is in the table below.
| |compile time|runtime|
|---|---|---|
|Data|VarDesc(proto)|Variable(cpp)|
|Operation|OpDesc(proto)|Operator(cpp)|
<table>
<thead>
<tr>
<th></th>
<th>compile time</th>
<th>runtime</th>
</tr>
</thead>
<tbody>
<tr>
<td>Data </td>
<td>VarDesc(proto) </td>
<td>Variable(cpp) </td>
</tr>
<tr>
<td>Operation </td>
<td>OpDesc(proto) </td>
<td>Operator(cpp) </td>
</tr>
</tbody>
</table>
## Definition of VarType
......
doc/fluid/design/concurrent/channel.md
浏览文件 @ e2ca8e10
......@@ -2,7 +2,7 @@
## Introduction
A Channel is a data structure that allows for synchronous interprocess
A Channel is a data structure that allows for synchronous interprocess
communication via message passing. It is a fundemental component of CSP
(communicating sequential processes), and allows for users to pass data
between threads without having to worry about synchronization.
......@@ -18,7 +18,7 @@ Creates a new channel that takes in variables of a specific dtype.
- **fluid.make_channel(dtype, capacity=0)**
- **dtype**: The data type of variables being sent/received through channel
- **capacity**: The capacity of the channel. A capacity of 0 represents
- **capacity**: The capacity of the channel. A capacity of 0 represents
an unbuffered channel. Capacity > 0 represents a buffered channel
```
......@@ -40,8 +40,8 @@ fluid.channel_close(ch)
### Send data to a channel
Sends a variable to a channel. Currently, variables of dtype `LoDTensor`,
`LoDRankTable`, `LoDTensorArray`, `SelectedRows`, `ReaderHolder`, and
Sends a variable to a channel. Currently, variables of dtype `LoDTensor`,
`LoDRankTable`, `LoDTensorArray`, `SelectedRows`, `ReaderHolder`, and
`ChannelHolder` are supported.
By default, the data of the Variable is moved from the sender to the receiver,
......@@ -52,7 +52,7 @@ however the user can optionally copy the data before performing the send.
- **variable**: The variable to send to the channel
- **is_copy**: If set to True, channel_send will perform a variable assign
to copy the source variable to a new variable to be sent.
```
ch = fluid.make_channel(dtype=core.VarDesc.VarType.LOD_TENSOR)
var = fill_constant(shape=[1],dtype=core.VarDesc.VarType.INT32, value=100)
......@@ -68,7 +68,7 @@ receiving variable.
- **channel**: The channel to receive the variable from
- **return_variable**: The destination variable used to store the data of the
variable received from the channel
```
ch = fluid.make_channel(dtype=core.VarDesc.VarType.LOD_TENSOR)
var = fill_constant(shape=[1],dtype=core.VarDesc.VarType.INT32, value=-1)
......@@ -84,9 +84,9 @@ internal queues, locks, and conditional variables.
### QueueMessage
QueueMessage encapsulates the state of the channel send/receive operation to be
put in the **sendq/recvq**. It contains a condition variable used to lock the
put in the **sendq/recvq**. It contains a condition variable used to lock the
thread (when there are no available sends/receives). In addition, it contains
a callback function to notify a thread when the QueueMessage is being
a callback function to notify a thread when the QueueMessage is being
processed by the channel.
### Queues
......@@ -108,21 +108,21 @@ channel_recv operation will put a new QueueMessage on the recvq and block the
current thread under two conditions:
1. The channel is buffered and there is no data on the buff_
2. The channel is unbuffered and does not have a sender
### State diagram
#### Channel Send
<p align="center">
<img src="./images/channel_send.png"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/channel_send.png"/><br/>
</p>
#### Channel Receive
<p align="center">
<img src="./images/channel_recv.png"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/channel_recv.png"/><br/>
</p>
## Limitations and Considerations
### Variable Copy
......@@ -135,5 +135,5 @@ be sent before it is sent.
Please note that this is acheived by adding an **assign** operator and creating
a temporary variable that is sent in place of the original variable. Please
note that **assign** operator has limited support for only certain variables
note that **assign** operator has limited support for only certain variables
datatypes.
doc/fluid/design/concurrent/concurrent_programming.md
浏览文件 @ e2ca8e10
......@@ -10,12 +10,42 @@ The answer relies on the fact that a `ProgramDesc` is similar to an abstract syn
The following table compares concepts in Fluid and Go
| Go | Fluid |
|----|-------|
|user-defined functions | [layers](https://github.com/PaddlePaddle/Paddle/tree/develop/python/paddle/fluid) |
| control-flow and built-in functions | [intrinsics/operators](https://github.com/PaddlePaddle/Paddle/tree/develop/paddle/operators) |
| goroutines, channels | [class ThreadPool](https://github.com/PaddlePaddle/Paddle/tree/develop/paddle/framework/thread_pool.h) |
| runtime | [class Executor](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/executor.h) |
<table>
<thead>
<tr>
<th></th>
<th>Go</th>
<th>Fluid</th>
</tr>
</thead>
<tbody>
<tr>
<td>user-defined functions </td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/tree/develop/python/paddle/fluid">layers</a></td>
<td></td>
</tr>
<tr>
<td>control-flow and built-in functions </td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/tree/develop/paddle/operators">intrinsics/operators</a></td>
<td></td>
</tr>
<tr>
<td>goroutines, channels </td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/tree/develop/paddle/framework/thread_pool.h">class ThreadPool</a></td>
<td></td>
</tr>
<tr>
<td>runtime </td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/executor.h">class Executor</a></td>
<td></td>
</tr>
</tbody>
</table>
## An Example Concurrent Program
......@@ -77,11 +107,11 @@ message ProgramDesc {
read(output = X)
kube_get_workers_addrs(output = L)
Y = tensor_array(len(L))
parallel_for(input = X, output = Y,
parallel_for(input = X, output = Y,
attrs = {L, block_id(1)}) # referring to block 1
]
}
block[1] = Block {
parent = 0,
vars = [x, y, index],
......@@ -102,7 +132,7 @@ func main() { //// block 0
X = fluid.read(...)
L = fluid.k8s.get_worker_addrs()
Y = fluid.tensor_array(len(L))
fluid.parallel_for(X, L,
fluid.parallel_for(X, L,
func(index int) { //// block 1
x = X[index]
fluid.send(L[index], x)
......@@ -116,7 +146,7 @@ An explanation of the above program:
- `fluid.k8s` is a package that provides access to Kubernetes API.
- `fluid.k8s.get_worker_addrs` returns the list of IP and ports of all pods of the current job except for the current one (the master pod).
- `fluid.tensor_array` creates a [tensor array](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/lod_tensor_array.h). `fluid.parallel_for` creates a `ParallelFor` intrinsic, which, when executed,
- `fluid.tensor_array` creates a [tensor array](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/lod_tensor_array.h). `fluid.parallel_for` creates a `ParallelFor` intrinsic, which, when executed,
1. creates `len(L)` scopes, each for the concurrent running of the sub-block (block 1 in this case), and initializes a variable named "index" in the scope to an integer value in the range `[0, len(L)-1]`, and
2. creates `len(L)` threads by calling into the `ThreadPool` singleton, each thread
......
doc/fluid/design/concurrent/csp.md
浏览文件 @ e2ca8e10
......@@ -13,14 +13,41 @@ Most DL systems, including TensorFlow, Caffe2, and MxNet, can asynchronously exe
There were many concurrent programming models, implemented in various forms:
| concurrent programming model | implementation |
|-----|-----|
| mutex | types and functions in standard libraries |
| semaphore | types and functions in standard libraries |
| communicating sequential processes (CSP) | Go programming language |
| actor model | Erlang programming language |
| message passing | MPI |
| bulk synchronous parallel (BSP) | Pregel distributed programming framework |
<table>
<thead>
<tr>
<th>concurrent programming model</th>
<th>implementation</th>
</tr>
</thead>
<tbody>
<tr>
<td>mutex </td>
<td>types and functions in standard libraries </td>
</tr>
<tr>
<td>semaphore </td>
<td> types and functions in standard libraries </td>
</tr>
<tr>
<td> communicating sequential processes (CSP) </td>
<td> Go programming language </td>
</tr>
<tr>
<td> actor model </td>
<td> Erlang programming language </td>
</tr>
<tr>
<td> message passing </td>
<td> MPI </td>
</tr>
<tr>
<td> bulk synchronous parallel (BSP) </td>
<td> Pregel distributed programming framework </td>
</tr>
</tbody>
</table>
Since Fluid was designed to be a programming language, we would like to implement CSP in Fluid.
......@@ -118,9 +145,9 @@ There are four types of actions with a channel:
```go
close(ch)
```
Please be aware that a closed channel is not a nil channel, which is `var ch chan int`.
There are some [axioms with channels](https://dave.cheney.net/2014/03/19/channel-axioms):
1. A send to a nil channel blocks forever
......
doc/fluid/design/concurrent/select_op.md
浏览文件 @ e2ca8e10
......@@ -2,13 +2,13 @@
## Introduction
In golang, the [**select**](https://golang.org/ref/spec#Select_statements)
statement lets a goroutine wait on multiple communication operations at the
same time. The **select** blocks until one of its cases can run, then
executes the case. If multiple cases are ready to run, then one case is
In golang, the [**select**](https://golang.org/ref/spec#Select_statements)
statement lets a goroutine wait on multiple communication operations at the
same time. The **select** blocks until one of its cases can run, then
executes the case. If multiple cases are ready to run, then one case is
choosen at random to be executed.
With the introduction of CSP for Paddle, we mimic this behavior by
With the introduction of CSP for Paddle, we mimic this behavior by
creating a ***select_op***.
## How to use it
......@@ -17,11 +17,11 @@ The **select_op** is available as a c++ operator. However most users
will prefer to use the much simplier Python API.
- **fluid.Select()**: Creates a select operator and adds it to the current
block within the main program. Also creates a sub block and adds it to the
main program. This sub block is used to hold all variables and operators
block within the main program. Also creates a sub block and adds it to the
main program. This sub block is used to hold all variables and operators
used by the case statements.
Within the select block, users can add cases by
Within the select block, users can add cases by
calling **select.case** or **select.default** method.
- **fluid.Select.case(channel_action, channel, result_variable)**: Represents
......@@ -37,13 +37,13 @@ execute.
```
ch1 = fluid.make_channel(dtype=core.VarDesc.VarType.LOD_TENSOR)
quit_ch = fluid.make_channel(dtype=core.VarDesc.VarType.LOD_TENSOR)
x = fill_constant(shape=[1], dtype=core.VarDesc.VarType.INT32, value=0)
y = fill_constant(shape=[1], dtype=core.VarDesc.VarType.INT32, value=1)
while_cond = fill_constant(shape=[1], dtype=core.VarDesc.VarType.BOOL, value=True)
while_op = While(cond=while_cond)
with while_op.block():
with fluid.Select() as select:
with select.case(fluid.channel_send, channel, x):
......@@ -99,17 +99,17 @@ blocks {
}
}
// Create "select" operator.
// inputs:
// inputs:
// X: All input variables used by operators within the select block
// case_to_execute: Variable filled in by select_op when it determines
// which case to execute.
//
// outputs:
// Out: All output variables referenced by operators within select block.
//
// Out: All output variables referenced by operators within select block.
//
// attrs:
// sub_block: The block id containing the select "cases"
// cases: Serialized list of all cases in the select op.
// cases: Serialized list of all cases in the select op.
// Each case is serialized as: '<index>,<type>,<channel>,<value>'
// where type is 0 for default, 1 for send, and 2 for receive.
// No channel and values are needed for default cases.
......@@ -150,7 +150,7 @@ into **X**. It will also create a temp variable called **case_to_execute**. Th
filled in by the select_op after it has completed processing the case statements.
If there are no available cases to execute (ie: all cases are blocked on channel operations, and
there is no default statement), then the select_op will block the current thread. The thread will
there is no default statement), then the select_op will block the current thread. The thread will
unblock once there is a channel operation affecting one of the case statements, at which point, the
**select_op** will set the **case_to_execute** variable to the index of the case to execute.
......@@ -247,17 +247,17 @@ blocks {
```
Cases are represented by a **conditional_block operator**, whose's condition is set as the output of
equal(**case_to_execute**, **case_index**). Since each case index is unique in this sub-block,
Cases are represented by a **conditional_block operator**, whose's condition is set as the output of
equal(**case_to_execute**, **case_index**). Since each case index is unique in this sub-block,
only one case will be executed.
### select_op flow
<p align="center">
<img src="./images/select_op_workflow.png"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/select_op_workflow.png"/><br/>
</p>
The select algorithm is inspired by golang's select routine. Please refer to
The select algorithm is inspired by golang's select routine. Please refer to
http://www.tapirgames.com/blog/golang-concurrent-select-implementation for more information.
## Backward Pass
......
doc/fluid/design/dist_train/distributed_architecture.md
浏览文件 @ e2ca8e10
......@@ -40,11 +40,11 @@ computation is only specified in Python code which sits outside of PaddlePaddle,
Similar to how a compiler uses an intermediate representation (IR) so that the programmer does not need to manually optimize their code for most of the cases, we can have an intermediate representation in PaddlePaddle as well. The compiler optimizes the IR as follows:
<img src="src/compiler.png"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/compiler.png"/>
PaddlePaddle can support model parallelism by converting the IR so that the user no longer needs to manually perform the computation and operations in the Python component:
<img src="src/paddle-compile.png"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/paddle-compile.png"/>
The IR for PaddlePaddle after refactoring is called a `Block`, it specifies the computation dependency graph and the variables used in the computation.
......@@ -60,7 +60,7 @@ For a detailed explanation, refer to this document -
The revamped distributed training architecture can address the above discussed limitations. Below is the illustration of how it does so:
<img src="src/distributed_architecture.png"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/distributed_architecture.png"/>
The major components are: *Python API*, *Distribute Transpiler* and *Remote Executor*.
......@@ -152,7 +152,7 @@ for data in train_reader():
`JobDesc` object describe the distributed job resource specification to run on
Cluster environment.
<img src="src/remote_executor.png" width="500" align="center" />
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/remote_executor.png" width="500" align="center" />
`RemoteExecutor.run` sends the `ProgramDesc` and
[TrainingJob](https://github.com/PaddlePaddle/cloud/blob/unreleased-tpr/doc/autoscale/README.md#training-job-resource)
......@@ -171,7 +171,7 @@ In the future, a more general placement algorithm should be implemented, which m
The local training architecture will be the same as the distributed training architecture, the difference is that everything runs locally, and there is just one PaddlePaddle runtime:
<img src="src/local_architecture.png"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/local_architecture.png"/>
### Training Data
......
doc/fluid/design/dist_train/multi_cpu.md
浏览文件 @ e2ca8e10
......@@ -8,11 +8,11 @@ Op graph to a multi-CPU Op graph, and run `ParallelDo` Op to run the graph.
## Transpiler
<img src="src/multi-threads/single-thread@3x.png" width="300">
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/single-thread@3x.png" width="300">
After converted:
<img src="src/multi-threads/multi-threads@3x.png" width="1000">
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/multi-threads@3x.png" width="1000">
## Implement
......
doc/fluid/design/dist_train/parameter_server.md
浏览文件 @ e2ca8e10
......@@ -41,11 +41,11 @@ We will need these OPs: *Send*, *Recv*, *Enqueue*, *Dequeue*.
Below is an example of converting the user defined graph to the
subgraphs for the trainer and the parameter server:
<img src="src/local-graph.png" width="300"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/local-graph.png" width="300"/>
After converting:
<img src="src/dist-graph.png" width="700"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/dist-graph.png" width="700"/>
1. The parameter variable W and its optimizer program are placed on the parameter server.
1. Operators are added to the program.
......@@ -69,8 +69,7 @@ In Fluid, we introduce [SelectedRows](../selected_rows.md) to represent a list o
non-zero gradient data. So when we do parameter optimization both locally and remotely,
we only need to send those non-zero rows to the optimizer operators:
<img src="src/sparse_update.png" width="700" />
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/sparse_update.png" width="700" />
### Benefits
- Model parallelism becomes easier to implement: it is an extension to
......
doc/fluid/design/dynamic_rnn/rnn.md
浏览文件 @ e2ca8e10
......@@ -5,7 +5,7 @@ This document describes the RNN (Recurrent Neural Network) operator and how it i
## RNN Algorithm Implementation
<p align="center">
<img src="./rnn.jpg"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/rnn.jpg"/>
</p>
The above diagram shows an RNN unrolled into a full network.
......@@ -22,7 +22,7 @@ There are several important concepts here:
There could be local variables defined in each step-net. PaddlePaddle runtime realizes these variables in *step-scopes* which are created for each step.
<p align="center">
<img src="./rnn.png"/><br/>
<img src="https://github.com/PaddlePaddle/Paddle/tree/develop/doc/fluid/images/rnn.png"/><br/>
Figure 2 illustrates the RNN's data flow
</p>
......@@ -93,7 +93,7 @@ For example, we could have a 2-level RNN, where the top level corresponds to par
The following figure illustrates feeding in text into the lower level, one sentence at a step, and the feeding in step outputs to the top level. The final top level output is about the whole text.
<p align="center">
<img src="./2_level_rnn.png"/>
<img src="https://github.com/PaddlePaddle/Paddle/tree/develop/doc/fluid/images/2_level_rnn.png"/>
</p>
```python
......@@ -149,5 +149,5 @@ If the `output_all_steps` is set to False, it will only output the final time st
<p align="center">
<img src="./rnn_2level_data.png"/>
<img src="https://github.com/PaddlePaddle/Paddle/tree/develop/doc/fluid/images/rnn_2level_data.png"/>
</p>
doc/fluid/design/modules/batch_norm_op.md
浏览文件 @ e2ca8e10
......@@ -2,7 +2,7 @@
## What is batch normalization
Batch normalization is a frequently-used method in deep network training. It adjusts the mean and variance of a layer's output, and make the data distribution easier for next layer's training.
Batch normalization is a frequently-used method in deep network training. It adjusts the mean and variance of a layer's output, and make the data distribution easier for next layer's training.
The principle of batch normalization can be summarized into a simple function:
......@@ -66,7 +66,7 @@ As most C++ operators do, `batch_norm_op` is defined by inputs, outputs, attribu
The following graph showes the training computational process of `batch_norm_op`:
<img src="../images/batch_norm_op_kernel.png" width="800"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/batch_norm_op_kernel.png" width="800"/>
cudnn provides APIs to finish the whole series of computation, we can use them in our GPU kernel.
......@@ -74,13 +74,13 @@ cudnn provides APIs to finish the whole series of computation, we can use them i
`batch_norm_op` is warpped as a layer in Python:
```python
def batch_norm_layer(net,
```python
def batch_norm_layer(net,
input,
output,
scale,
bias,
use_global_est = False,
output,
scale,
bias,
use_global_est = False,
epsilon = 1e-6,
momentum = 0.99):
mean_cache = scope.new_var(name = 'estimated_mean', trainable = False)
......@@ -119,15 +119,15 @@ for pass_id in range(PASS_NUM):
if pass_id % 100 == 0:
net.infer(test_image) # run inferencing model
# ...
```
```
`is_infer` is an attribute. Once an operator is created, its attributes can not be changed. It suggests us that we shall maintain two `batch_norm_op` in the model, one's `is_infer` is `True`(we call it `infer_batch_norm_op`) and the other one's is `False`(we call it `train_batch_norm_op`). They share all parameters and variables, but be placed in two different branches. That is to say, if a network contains a `batch_norm_op`, it will fork into two branches, one go through `train_batch_norm_op` and the other one go through `infer_batch_norm_op`:
<div align=center>
<img src="../images/batch_norm_fork.png" width="500"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/batch_norm_fork.png" width="500"/>
</div>
Just like what is shown in the above graph, the net forks before `batch_norm_op` and will never merge again. All the operators after `batch_norm_op` will duplicate.
Just like what is shown in the above graph, the net forks before `batch_norm_op` and will never merge again. All the operators after `batch_norm_op` will duplicate.
When the net runs in training mode, the end of the left branch will be set as the running target, so the dependency tracking process will ignore right branch automatically. When the net runs in inferencing mode, the process is reversed.
......
doc/fluid/design/modules/python_api.md
浏览文件 @ e2ca8e10
......@@ -2,12 +2,33 @@
Due to the refactorization of the PaddlePaddle core, we need Python classes to construct corresponding protobuf messages that describe a DL program.
| Python classes | Protobuf messages |
| --- | --- |
| Program | ProgramDesc |
| Block | BlockDesc |
| Operator | OpDesc |
| Variable | VarDesc |
<table>
<thead>
<tr>
<th>Python classes</th>
<th>Protobuf messages</th>
</tr>
</thead>
<tbody>
<tr>
<td>Program </td>
<td>ProgramDesc </td>
</tr>
<tr>
<td>Block </td>
<td>BlockDesc </td>
</tr>
<tr>
<td>Operator </td>
<td>OpDesc </td>
</tr>
<tr>
<td>Variable </td>
<td>VarDesc </td>
</tr>
</tbody>
</table>
Please be aware that these Python classes need to maintain some construction-time information, which are not part of the protobuf messages.
......
doc/fluid/design/modules/regularization.md
浏览文件 @ e2ca8e10
......@@ -6,23 +6,23 @@ A central problem in machine learning is how to design an algorithm that will pe
### Parameter Norm Penalties
Most common regularization approaches in deep learning are based on limiting the capacity of the models by adding a parameter norm penalty to the objective function `J`. This is given as follows:
<img src="./images/loss_equation.png" align="center"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/loss_equation.png" align="center"/><br/>
The parameter `alpha` is a hyperparameter that weights the relative contribution of the norm penalty term, `omega`, relative to the standard objective function `J`.
The most commonly used norm penalties are the L2 norm penalty and the L1 norm penalty. These are given as follows:
##### L2 Regularization:
<img src="./images/l2_regularization.png" align="center"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/l2_regularization.png" align="center"/><br/>
##### L1 Regularization
<img src="./images/l1_regularization.png" align="center"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/l1_regularization.png" align="center"/><br/>
A much more detailed mathematical background of regularization can be found [here](http://www.deeplearningbook.org/contents/regularization.html).
## Regularization Survey
A detailed survey of regularization in various deep learning frameworks can be found [here](https://github.com/PaddlePaddle/Paddle/wiki/Regularization-Survey).
A detailed survey of regularization in various deep learning frameworks can be found [here](https://github.com/PaddlePaddle/Paddle/wiki/Regularization-Survey).
## Proposal for Regularization in PaddlePaddle
......@@ -32,41 +32,35 @@ In the new design, we propose to create new operations for regularization. For n
- L2_regularization_op
- L1_regularization_op
These ops can be like any other ops with their own CPU/GPU implementations either using Eigen or separate CPU and GPU kernels. As the initial implementation, we can implement their kernels using Eigen following the abstraction pattern implemented for [Activation Ops](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/accuracy_op.h). This abstraction pattern can make it very easy to implement new regularization schemes other than L1 and L2 norm penalties.
These ops can be like any other ops with their own CPU/GPU implementations either using Eigen or separate CPU and GPU kernels. As the initial implementation, we can implement their kernels using Eigen following the abstraction pattern implemented for [Activation Ops](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/accuracy_op.h). This abstraction pattern can make it very easy to implement new regularization schemes other than L1 and L2 norm penalties.
The idea of building ops for regularization is in sync with the refactored Paddle philosophy of using operators to represent any computation unit. The way these ops will be added to the computation graph, will be decided by the [layer functions](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md#layer-function) in Python API.
The idea of building ops for regularization is in sync with the refactored Paddle philosophy of using operators to represent any computation unit. The way these ops will be added to the computation graph, will be decided by the [layer functions](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md#layer-function) in Python API.
### Computation Graph
Below is an example of a really simple feed forward neural network.
<img src="./images/feed_forward.png" align="center"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/feed_forward.png" align="center"/><br/>
The Python API will modify this computation graph to add regularization operators. The modified computation graph will look as follows:
<img src="./images/feed_forward_regularized.png" align="center"/><br/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/feed_forward_regularized.png" align="center"/><br/>
   
### Python API implementation for Regularization
Using the low level ops, `L2_regularization_op` and `L1_regularization_op`, any user can add regularization to their computation graphs. However, this will require a lot of lines of code and we should design Python APIs that support regularization. An example of such an API can be seen in [Keras](https://keras.io/regularizers/). As per the PaddlePaddle [Python API design](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md), the layer functions are responsible for creating operators, operator parameters and variables. Since regularization is a property of parameters, it makes sense to create these in the layer functions.
Using the low level ops, `L2_regularization_op` and `L1_regularization_op`, any user can add regularization to their computation graphs. However, this will require a lot of lines of code and we should design Python APIs that support regularization. An example of such an API can be seen in [Keras](https://keras.io/regularizers/). As per the PaddlePaddle [Python API design](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md), the layer functions are responsible for creating operators, operator parameters and variables. Since regularization is a property of parameters, it makes sense to create these in the layer functions.
#### Creation of Regularization ops
There are two possibilities for creating the regularization ops:
1. We create these ops immediately while building the computation graph.
2. We add these ops in a lazy manner, just before the backward, similar to the way the optimization ops are added.
1. We create these ops immediately while building the computation graph.
2. We add these ops in a lazy manner, just before the backward, similar to the way the optimization ops are added.
The proposal is to add these ops in a lazy manner just before the backward pass.
The proposal is to add these ops in a lazy manner just before the backward pass.
#### Storage of Regularization attributes
Since we want to create the regularization ops in a lazy manner, the regularization attributes (type of regularization and weight of regularization penalty) can be stored as attributes of the [`Parameter`](https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/v2/framework/framework.py#L421) class. This is because regularization is a property of the parameters and storing regularization properties with Parameters also allows for shared parameters.
Since we want to create the regularization ops in a lazy manner, the regularization attributes (type of regularization and weight of regularization penalty) can be stored as attributes of the [`Parameter`](https://github.com/PaddlePaddle/Paddle/blob/develop/python/paddle/v2/framework/framework.py#L421) class. This is because regularization is a property of the parameters and storing regularization properties with Parameters also allows for shared parameters.
#### High-level API
In PaddlePaddle Python API, users will primarily rely on [layer functions](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/python_api.md#layer-function) to create neural network layers. Hence, we also need to provide regularization functionality in layer functions. The design of these APIs can be postponed for later right now. A good reference for these APIs can be found in [Keras](https://keras.io/regularizers/) and also by looking at Tensorflow in [`tf.contrib.layers`](https://www.tensorflow.org/api_guides/python/contrib.layers).
doc/fluid/design/motivation/fluid.md
浏览文件 @ e2ca8e10
......@@ -10,11 +10,37 @@ Fluid is the answer. Fluid is similar to PyTorch and TensorFlow Eager Execution
Deep learning infrastructure is one of the fastest evolving technologies. Within four years, there have already been three generations of technologies invented.
| Existed since | model as sequence of layers | model as graph of operators | No model |
|--|--|--|--|
| 2013 | Caffe, Theano, Torch, PaddlePaddle | | |
| 2015 | | TensorFlow, MxNet, Caffe2, ONNX, n-graph | |
| 2016 | | | PyTorch, TensorFlow Eager Execution, PaddlePaddle Fluid |
<table>
<thead>
<tr>
<th>Existed since</th>
<th>model as sequence of layers</th>
<th>model as graph of operators</th>
<th>No model</th>
</tr>
</thead>
<tbody>
<tr>
<td>2013 </td>
<td>Caffe, Theano, Torch, PaddlePaddle </td>
<td> </td>
<td> </td>
</tr>
<tr>
<td>2015 </td>
<td> </td>
<td>TensorFlow, MxNet, Caffe2, ONNX, n-graph </td>
<td> </td>
</tr>
<tr>
<td>2016 </td>
<td> </td>
<td> </td>
<td> PyTorch, TensorFlow Eager Execution, PaddlePaddle Fluid</td>
</tr>
</tbody>
</table>
From the above table, we see that the deep learning technology is evolving towards getting rid of the concept of a model. To understand the reasons behind this direction, a comparison of the *programming paradigms* or the ways to program deep learning applications using these systems, would be helpful. The following section goes over these.
......
doc/fluid/design/motivation/refactorization.md
浏览文件 @ e2ca8e10
......@@ -36,11 +36,37 @@ At compile time, the Python program generates a protobuf message representation
At runtime, the C++ program realizes the graph and runs it.
| | Representation (protobuf messages) | Realization (C++ class objects) |
|---|---|---|
|Data|[VarDesc](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto#L107)|[Variable](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/variable.h#L24)|
|Operation|[OpDesc](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto#L35)|[Operator](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/operator.h#L64)|
|Block|BlockDesc|Block|
<table>
<thead>
<tr>
<th></th>
<th>Representation (protobuf messages)</th>
<th>Realization (C++ class objects) </th>
</tr>
</thead>
<tbody>
<tr>
<td>Data</td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto#L107">VarDesc</a></td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/variable.h#L24">Variable</a></td>
</tr>
<tr>
<td>Operation </td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/framework.proto#L35">OpDesc</a></td>
<td>
<a href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/operator.h#L64">Operator</a></td>
</tr>
<tr>
<td>Block </td>
<td>BlockDesc </td>
<td>Block </td>
</tbody>
</table>
The word *graph* is interchangeable with *block* in this document. A graph consists of computation steps and local variables similar to a C++/Java program block, or a pair of parentheses(`{` and `}`).
......
doc/fluid/design/muti_devices/kernel_hint_design.md
浏览文件 @ e2ca8e10
# Problem
# Kernel Hint Design
## Problem
In PaddlePaddle's [Design](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/switch_kernel.md), one Operator may have multiple kernels. Users may have some personal preference to choose a certain type of kernel for an operator, such as `force_cpu` to choose a CPU kernel, `use_cudnn` to choose a CUDNN kernel, we need to provide a way for users to do this.
In the current design, we use KernelType to describe one kernel.
......
doc/fluid/design/muti_devices/kernel_selection.md
浏览文件 @ e2ca8e10
# Background
# Kernel Selection
## Background
Every operator has many kernels because there are multiple data types, places, data layout, library type that Fluid supports. We use the `OpKernelType ` to describe kernel types that operators can hold.
The `OpKernelType ` is as follows:
......
doc/fluid/design/network/sequence_decoder.md
浏览文件 @ e2ca8e10
......@@ -199,7 +199,7 @@ Packing the `selected_generation_scores` will get a `LoDTensor`, and each tail i
## LoD and shape changes during decoding
<p align="center">
<img src="./images/LOD-and-shape-changes-during-decoding.jpg"/>
<img src="https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/doc/fluid/images/LOD-and-shape-changes-during-decoding.jpg"/>
</p>
According to the image above, the only phase that changes the LoD is beam search.
......
doc/fluid/design/others/gan_api.md
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/dev/index_cn.rst
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/dev/index_en.rst
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/dev/new_op_cn.md
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/dev/new_op_en.md
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/dev/releasing_process_en.md 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/2_level_rnn.dot 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/batch_norm_fork.dot 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/compiler.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/control_flow_graph.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/dataflow_equations.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/dcgan.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/deep_learning.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/dist-graph.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/dist-graph.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/distributed_architecture.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/distributed_architecture.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/ds2_network.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/feed_forward.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/feed_forward_regularized.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/fluid-compiler.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/fluid-compiler.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/graph_construction_example.bash 0 → 100755
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/graph_construction_example.dot 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/graph_construction_example_all.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/l1_regularization.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/l2_regularization.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/local-graph.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/local-graph.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/local_architecture.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/local_architecture.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/lookup_table.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/lookup_table_training.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/loss_equation.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multi-threads.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multi-threads@3x.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multigpu_allreduce.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multigpu_allreduce.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multigpu_before_convert.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multigpu_before_convert.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/multiple_reader.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/paddle-compile.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/paddle-compile.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/pprof_1.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/pprof_2.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/profiler.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/readers.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/remote_executor.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/remote_executor.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/rnn.dot 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/rnn.jpg 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/rnn.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/rnn_2level_data.dot 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/rnn_2level_data.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/single-thread@3x.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/sparse_update.graffle 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/sparse_update.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/test.dot 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/test.dot.png 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/theta_star.gif 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/timeline.jpeg 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/fluid/images/tracing.jpeg 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
doc/templates/conf.py.cn.in
浏览文件 @ e2ca8e10
此差异已折叠。
doc/templates/conf.py.en.in
浏览文件 @ e2ca8e10
此差异已折叠。
doc/v2/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
doc/v2/api/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/.gitignore
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/api/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/api/test/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/cuda/include/hl_cnn.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/cuda/src/hl_cuda_cnn.cu
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/block_desc.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/channel.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/executor.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/executor.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/init.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/init.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/init_test.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/lod_tensor.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/operator.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/operator.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/scope.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/scope.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/tensor.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/framework/tuple.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/inference/io.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/inference/io.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/inference/tests/test_multi_thread_helper.h 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/memory/.clang-format 已删除 100644 → 0
浏览文件 @ 92818baa
此差异已折叠。
paddle/fluid/memory/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/memory/malloc.h 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/memory/memcpy.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/memory/memory.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/memory/pinned_memory_test.cu 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/.clang-format 已删除 100644 → 0
浏览文件 @ 92818baa
此差异已折叠。
paddle/fluid/operators/auc_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/auc_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/concat_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/cond_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/conv_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/crop_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/fc_mkldnn_op.cc 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/fc_op.cc 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/go_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/prior_box_op.cu 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/reshape_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/sgd_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/sgd_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/spp_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/spp_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/sum_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/sum_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/top_k_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/unpool_op.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/unpool_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/operators/warpctc_op.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/.clang-format 已删除 100644 → 0
浏览文件 @ 92818baa
此差异已折叠。
paddle/fluid/platform/cpu_info.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/cpu_info.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/cuda_helper.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/enforce.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/float16.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/gpu_info.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/gpu_info.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/nccl_helper.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/place.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/place.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/profiler.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/platform/profiler.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/.clang-format 已删除 100644 → 0
浏览文件 @ 92818baa
此差异已折叠。
paddle/fluid/pybind/.gitignore 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/const_value.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/const_value.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/exception.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/exception.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/protobuf.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/protobuf.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/pybind.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/recordio.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/recordio.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/tensor_py.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/pybind/tensor_py_test.cc 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/chunk.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/chunk.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/chunk_test.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/header.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/scanner.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/scanner.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/writer.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/recordio/writer.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/string/.clang-format 已删除 120000 → 0
浏览文件 @ 92818baa
此差异已折叠。
paddle/fluid/string/piece.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/string/printf.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/fluid/string/printf_test.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/gserver/layers/UpsampleLayer.cpp 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/gserver/tests/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/gserver/tests/test_Upsample.cpp 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/math/Matrix.cpp
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/math/Matrix.h
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/scripts/docker/build.sh
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/testing/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/testing/paddle_gtest_main.cc
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/trainer/tests/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
paddle/utils/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
proto/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
proto/ModelConfig.proto
浏览文件 @ e2ca8e10
此差异已折叠。
python/.gitignore
浏览文件 @ e2ca8e10
此差异已折叠。
python/CMakeLists.txt
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/.gitignore 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/__init__.py
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/debuger.py
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/framework.py
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/graphviz.py
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/layers/io.py
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/layers/nn.py
浏览文件 @ e2ca8e10
此差异已折叠。
python/paddle/fluid/tests/unittests/test_fc_op.py 0 → 100644
浏览文件 @ e2ca8e10
此差异已折叠。
python/setup.py.in
浏览文件 @ e2ca8e10
此差异已折叠。
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