add PyTorch to Dockerfile (#362)

* update local demo doc and configuration

* change folder name

* Update tutorial_1_CR_exp_local_api.md

no need to have a new training file

* Delete mnist_gpu.py

no need to have a new training file

* Update config_gpu.yml

no need to have a new training file

* add PyTorch to Dockerfile

deployment/docker/Dockerfile

@@ -60,6 +60,11 @@ RUN python3 -m pip install nni
 #
 RUN python3 -m pip --no-cache-dir install tensorflow-gpu==1.10.0
 
+#
+#PyTorch
+#
+RUN python3 -m pip --no-cache-dir install torch torchvision
+
 #
 #Keras 2.1.6
 #

deployment/docker/README.md

 Dockerfile
 ===
 ## 1.Description
-This is the Dockerfile of nni project, including the most kinds of deeplearning frameworks and nni source code.  You can run your nni experiment in this docker container directly.
-Dockerfile could build the customized docker image, users could build their customized docker image using this file.
-This docker file includes the following libraries on `Ubuntu 16.04 LTS`:
+This is the Dockerfile of nni project. It includes serveral popular deep learning frameworks and NNI. It is tested on `Ubuntu 16.04 LTS`:
 
 ```
 CUDA 9.0, CuDNN 7.0
 numpy 1.14.3,scipy 1.1.0
 TensorFlow 1.5.0
+PyTorch (Stable)
 Keras 2.1.6
 NNI v0.3
 ```
+You can take this Dockerfile as a reference for your own customized Dockerfile. 
 
 ## 2.How to build and run
-__Use the following command to build docker image__
+__Use the following command from `nni/deployment/docker` to build docker image__
 ```    
     docker build -t nni/nni .
 ```
@@ -33,4 +33,7 @@ or
 ```
 
 ## 3.Directly retrieve the docker image
-If you do not want to build docker image, you can retrieve the NNI docker image from Docker Hub through the command `docker pull msranni/nni:latest`.
\ No newline at end of file
+Use the following command to retrieve the NNI docker image from Docker Hub
+```
+docker pull msranni/nni:latest
+```