Merge pull request #241 from pytorch/update_instructions_edit

update instructions as per the latest public repos

PyPiDescription.rst

@@ -1,50 +1,16 @@
 Project Description
 ===================
 
-TorchServe (PyTorch mdoel server) is a flexible and easy to use tool for
+TorchServe (PyTorch model server) is a flexible and easy to use tool for
 serving deep learning models exported from `PyTorch <http://pytorch.org/>`__.
 
 Use the TorchServe CLI, or the pre-configured Docker images, to start a
 service that sets up HTTP endpoints to handle model inference requests.
 
-Prerequisites
--------------
-
-* **java 8**: Required. TorchServe use java to serve HTTP requests. You must install java 8 (or later) and make sure java is on available in $PATH environment variable *before* installing torchserve. If you have multiple java installed, you can use $JAVA_HOME environment vairable to control which java to use.
-* **PyTorch**: Required. Latest version of PyTorch will be installed as a part of TorchServe installation.
-
-For ubuntu:
-::
-
-    sudo apt-get install openjdk-8-jdk
-
-
-For centos
-::
-
-    sudo yum install java-1.8.0-openjdk
-
-
-For Mac:
-::
-
-    brew tap caskroom/versions
-    brew update
-    brew cask install java8
-
-
-Install PyTorch:
-::
-
-    pip install torch torchvision torchtext
-
-
 Installation
 ------------
 
-::
-
-    pip install torchserve
+Full installation instructions are in the project repo: https://github.com/pytorch/serve/blob/master/README.md
 
 
 Source code

README.md

@@ -20,6 +20,7 @@ Conda instructions are provided in more detail, but you may also use `pip` and `
 To use `pip` to install TorchServe and the model archiver:
 
 ```
+pip install torch torchtext torchvision sentencepiece
 pip install torchserve torch-model-archiver
 ```
 
@@ -32,8 +33,13 @@ _Ubuntu_
     ```
 1. Install Conda (https://docs.conda.io/projects/conda/en/latest/user-guide/install/linux.html)
 1. Create an environment and install torchserve and torch-model-archiver
+    For CPU
     ```bash
-    conda create --name torchserve torchserve torch-model-archiver -c pytorch
+    conda create --name torchserve torchserve torch-model-archiver pytorch torchtext torchvision -c pytorch -c powerai
+    ```
+    For GPU
+    ```bash
+    conda create --name torchserve torchserve torch-model-archiver pytorch torchtext torchvision cudatoolkit=10.1 -c pytorch -c powerai
     ```
 1. Activate the environment
     ```bash
@@ -47,10 +53,10 @@ _macOS_
     brew tap AdoptOpenJDK/openjdk
     brew cask install adoptopenjdk11
     ```
-1. Install Conda (https://docs.conda.io/projects/conda/en/latest/user-guide/install/linux.html)
+1. Install Conda (https://docs.conda.io/projects/conda/en/latest/user-guide/install/macos.html)
 1. Create an environment and install torchserve and torch-model-archiver
     ```bash
-    conda create --name torchserve torchserve torch-model-archiver -c pytorch
+    conda create --name torchserve torchserve torch-model-archiver pytorch torchtext torchvision -c pytorch -c powerai
     ```
 1. Activate the environment
     ```bash

docker/README.md

@@ -1,4 +1,4 @@
-#### Create TorchServe docker image
+# Create TorchServe docker image
 
 ```bash
 cd serve/docker
@@ -7,21 +7,39 @@ git clone https://github.com/pytorch/serve.git
 
 For creating CPU based image :
 ```bash
-docker build --file Dockerfile.cpu -t torchserve:1.0 .
+docker build --file Dockerfile.cpu -t torchserve:latest .
 ```
 
 For creating GPU based image :
 ```bash
-docker build --file Dockerfile.gpu -t torchserve:1.0 .
+docker build --file Dockerfile.gpu -t torchserve:latest .
 ```
 
-#### Start container with TorchServe image
+## Start a container with a TorchServe image
 
+The following examples will start the container with 8080/81 port exposed to outer-world/localhost.
+
+### Start CPU container
+
+For specific versions you can pass in the specific tag to use (ex: 0.1-cpu):
 ```bash
-docker run --rm -it -p 8080:8080 -p 8081:8081 torchserve:1.0
+docker run --rm -it -p 8080:8080 -p 8081:8081 pytorch/torchserve:0.1-cpu
 ```
 
-The above command will start the container with 8080/81 port exposed to outer-world/localhost
+For the latest version, you can use the `latest` tag:
+docker run --rm -it -p 8080:8080 -p 8081:8081 pytorch/torchserve:latest
+
+#### Start GPU container
+
+For specific versions you can pass in the specific tag to use (ex: 0.1-cpu):
+```bash
+docker run --rm -it --gpus all -p 8080:8080 -p 8081:8081 pytorch/torchserve:0.1-cuda10.1-cudnn7-runtime
+```
+
+For the latest version, you can use the `gpu-latest` tag:
+```bash
+docker run --rm -it --gpus all -p 8080:8080 -p 8081:8081 pytorch/torchserve:latest-gpu
+```
 
 #### Accessing TorchServe APIs inside container