Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into whs_doc1

Author: wanghaoshuang

CMakeLists.txt

@@ -55,12 +55,14 @@ option(WITH_FLUID_ONLY  "Compile PaddlePaddle fluid only"               OFF)
 option(WITH_GOLANG      "Compile PaddlePaddle with GOLANG"              OFF)
 option(GLIDE_INSTALL    "Download and install go dependencies "         ON)
 option(USE_NNPACK       "Compile PaddlePaddle with NNPACK library"      OFF)
-option(WITH_DISTRIBUTE  "Compile with grpc distributed support"         OFF)
+option(WITH_DISTRIBUTE  "Compile with distributed support"              OFF)
 option(USE_EIGEN_FOR_BLAS   "Use matrix multiplication in Eigen"        OFF)
 option(EIGEN_USE_THREADS "Compile with multi-threaded Eigen"            OFF)
 option(WITH_ARM_FP16    "Use half precision support on armv8.2-a cpu"   OFF)
 option(WITH_FAST_BUNDLE_TEST    "Bundle tests that can be run in a single process together to reduce launch overhead"   OFF)
 option(WITH_CONTRIB     "Compile the third-party contributation"        OFF)
+option(WITH_ANAKIN      "Compile with Anakin library"                   OFF)
+option(WITH_GRPC     "Use grpc as the default rpc framework"            ${WITH_DISTRIBUTE})
 
 # CMAKE_BUILD_TYPE
 if(NOT CMAKE_BUILD_TYPE)
@@ -147,7 +149,16 @@ include(external/any)       # download libn::any
 include(external/eigen)     # download eigen3
 include(external/pybind11)  # download pybind11
 include(external/cares)
-include(external/grpc)
+
+if(WITH_DISTRIBUTE)
+    if(WITH_GRPC)
+        include(external/grpc)
+    else()
+        include(external/leveldb)
+        include(external/brpc)
+    endif()
+endif()
+
 include(external/snappy)    # download snappy
 include(external/snappystream)
 include(external/threadpool)
@@ -183,7 +194,10 @@ set(EXTERNAL_LIBS
 if(WITH_GPU)
     include(cuda)
     include(tensorrt)
-endif(WITH_GPU)
+    include(external/anakin)
+else()
+  set(WITH_ANAKIN OFF CACHE STRING "Anakin is valid only when GPU is set." FORCE)
+endif()
 
 if(WITH_AMD_GPU)
     find_package(HIP)

Dockerfile

@@ -24,7 +24,7 @@ COPY ./paddle/scripts/docker/root/ /root/
 
 RUN apt-get update && \
     apt-get install -y --allow-downgrades \
-    git python-pip python-dev openssh-server bison \
+    git python-pip python-dev python-opencv openssh-server bison \
     libnccl2=2.1.2-1+cuda8.0 libnccl-dev=2.1.2-1+cuda8.0 \
     wget unzip unrar tar xz-utils bzip2 gzip coreutils ntp \
     curl sed grep graphviz libjpeg-dev zlib1g-dev  \
@@ -76,8 +76,7 @@ RUN easy_install -U pip && \
     pip install sphinx-rtd-theme==0.1.9 recommonmark
 
 RUN pip install pre-commit 'ipython==5.3.0' && \
-    pip install 'ipykernel==4.6.0' 'jupyter==1.0.0' && \
-    pip install opencv-python
+    pip install 'ipykernel==4.6.0' 'jupyter==1.0.0'
 
 #For docstring checker
 RUN pip install pylint pytest astroid isort

benchmark/.gitignore

@@ -7,3 +7,6 @@ paddle/rnn/imdb.pkl
 caffe/image/logs
 tensorflow/image/logs
 tensorflow/rnn/logs
+fluid/models/*.pyc
+fluid/logs
+fluid/nohup.out

benchmark/fluid/Dockerfile

@@ -1,8 +1,8 @@
 FROM nvidia/cuda:9.0-cudnn7-devel-ubuntu16.04
-RUN apt-get update && apt-get install -y python python-pip iputils-ping libgtk2.0-dev wget vim net-tools iftop
+RUN apt-get update && apt-get install -y python python-pip iputils-ping libgtk2.0-dev wget vim net-tools iftop python-opencv
 RUN ln -s /usr/lib/x86_64-linux-gnu/libcudnn.so.7 /usr/lib/libcudnn.so && ln -s /usr/lib/x86_64-linux-gnu/libnccl.so.2 /usr/lib/libnccl.so
 RUN pip install -U pip
-RUN pip install -U kubernetes opencv-python paddlepaddle
+RUN pip install -U kubernetes paddlepaddle
 
 # IMPORTANT:
 # Add "ENV http_proxy=http://ip:port" if your download is slow, and don't forget to unset it at runtime.
@@ -19,4 +19,4 @@ ADD *.whl /
 RUN pip install /*.whl && rm -f /*.whl && chmod +x /usr/bin/paddle_k8s
 
 ENV LD_LIBRARY_PATH=/usr/local/lib
-ADD fluid_benchmark.py dataset.py models/ /workspace/
+ADD fluid_benchmark.py recordio_converter.py models/ /workspace/

benchmark/fluid/README.md

@@ -24,14 +24,18 @@ Currently supported `--model` argument include:
 
 * Run the following command to start a benchmark job locally:
     ```bash
-      python fluid_benchmark.py --model mnist  --device GPU
+      python fluid_benchmark.py --model mnist --device GPU
     ```
     You can choose to use GPU/CPU training. With GPU training, you can specify
     `--gpus <gpu_num>` to run multi GPU training.
+    You can set async mode parameter server. With async mode, you can specify
+    `--async_mode` to train model asynchronous.
 * Run distributed training with parameter servers:
+    * see [run_fluid_benchmark.sh](https://github.com/PaddlePaddle/Paddle/blob/develop/benchmark/fluid/run_fluid_benchmark.sh) as an example.
     * start parameter servers:
         ```bash
         PADDLE_TRAINING_ROLE=PSERVER PADDLE_PSERVER_PORT=7164 PADDLE_PSERVER_IPS=127.0.0.1 PADDLE_TRAINERS=1 PADDLE_CURRENT_IP=127.0.0.1 PADDLE_TRAINER_ID=0 python fluid_benchmark.py --model mnist  --device GPU --update_method pserver
+        sleep 15
         ```
     * start trainers:
         ```bash
@@ -42,6 +46,16 @@ Currently supported `--model` argument include:
     PADDLE_PSERVER_PORT=7164 PADDLE_TRAINER_IPS=192.168.0.2,192.168.0.3  PADDLE_CURRENT_IP=127.0.0.1 PADDLE_TRAINER_ID=0 python fluid_benchmark.py --model mnist --device GPU --update_method nccl2
     ```
 
+## Prepare the RecordIO file to Achieve Better Performance
+
+Run the following command will generate RecordIO files like "mnist.recordio" under the path
+and batch_size you choose, you can use batch_size=1 so that later reader can change the batch_size
+at any time using `fluid.batch`.
+
+```bash
+python -c 'from recordio_converter import *; prepare_mnist("data", 1)'
+```
+
 ## Run Distributed Benchmark on Kubernetes Cluster
 
 You may need to build a Docker image before submitting a cluster job onto Kubernetes, or you will

benchmark/fluid/args.py

@@ -0,0 +1,126 @@
+# 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.
+
+import argparse
+
+__all__ = ['parse_args', ]
+
+BENCHMARK_MODELS = [
+    "machine_translation", "resnet", "vgg", "mnist", "stacked_dynamic_lstm"
+]
+
+
+def parse_args():
+    parser = argparse.ArgumentParser('Fluid model benchmarks.')
+    parser.add_argument(
+        '--model',
+        type=str,
+        choices=BENCHMARK_MODELS,
+        default='resnet',
+        help='The model to run benchmark with.')
+    parser.add_argument(
+        '--batch_size', type=int, default=32, help='The minibatch size.')
+    #  args related to learning rate
+    parser.add_argument(
+        '--learning_rate', type=float, default=0.001, help='The learning rate.')
+    # TODO(wuyi): add "--use_fake_data" option back.
+    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(
+        '--gpus',
+        type=int,
+        default=1,
+        help='If gpus > 1, will use ParallelExecutor to run, else use Executor.')
+    # this option is available only for vgg and resnet.
+    parser.add_argument(
+        '--cpus',
+        type=int,
+        default=1,
+        help='If cpus > 1, will use ParallelDo to run, else use Executor.')
+    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(
+        '--no_test',
+        action='store_true',
+        help='If set, do not test the testset during training.')
+    parser.add_argument(
+        '--memory_optimize',
+        action='store_true',
+        help='If set, optimize runtime memory before start.')
+    parser.add_argument(
+        '--use_fake_data',
+        action='store_true',
+        help='If set ommit the actual read data operators.')
+    parser.add_argument(
+        '--profile', action='store_true', help='If set, profile a few steps.')
+    parser.add_argument(
+        '--update_method',
+        type=str,
+        default='local',
+        choices=['local', 'pserver', 'nccl2'],
+        help='Choose parameter update method, can be local, pserver, nccl2.')
+    parser.add_argument(
+        '--no_split_var',
+        action='store_true',
+        default=False,
+        help='Whether split variables into blocks when update_method is pserver')
+    parser.add_argument(
+        '--async_mode',
+        action='store_true',
+        default=False,
+        help='Whether start pserver in async mode to support ASGD')
+    parser.add_argument(
+        '--use_reader_op',
+        action='store_true',
+        help='Whether to use reader op, and must specify the data path if set this to true.'
+    )
+    parser.add_argument(
+        '--data_path',
+        type=str,
+        default="",
+        help='Directory that contains all the training recordio files.')
+    args = parser.parse_args()
+    return args