From 55d13d47fd5dc490b7f841597e8a1db300486733 Mon Sep 17 00:00:00 2001
From: tqchen <tianqi.tchen@gmail.com>
Date: Tue, 29 May 2018 16:22:34 -0700
Subject: [PATCH] Homepage URL to tvm.ai
---
README.md | 10 +++----
apps/howto_deploy/README.md | 2 +-
docs/README.txt | 2 +-
docs/dev/runtime.md | 6 ++---
docs/how_to/deploy.md | 4 +--
jvm/README.md | 8 +++---
nnvm/README.md | 31 +---------------------
tutorials/nnvm/deploy_model_on_mali_gpu.py | 3 +--
8 files changed, 18 insertions(+), 48 deletions(-)
diff --git a/README.md b/README.md
index 93dbb9ac3..77d72c641 100644
--- a/README.md
+++ b/README.md
@@ -1,22 +1,22 @@
-<img src=https://raw.githubusercontent.com/tqchen/tvmlang.org/master/images/logo/tvm-logo-small.png width=128/> Open Deep Learning Compiler Stack
+<img src=https://raw.githubusercontent.com/tqchen/tvm.ai/master/images/logo/tvm-logo-small.png width=128/> Open Deep Learning Compiler Stack
==============================================
[](./LICENSE)
[](http://mode-gpu.cs.washington.edu:8080/job/dmlc/job/tvm/job/master/)
[Installation](docs/how_to/install.md) |
-[Documentation](http://docs.tvmlang.org) |
-[Tutorials](http://tutorials.tvmlang.org) |
+[Documentation](http://docs.tvm.ai) |
+[Tutorials](http://tutorials.tvm.ai) |
[Operator Inventory](topi) |
[FAQ](docs/faq.md) |
[Contributors](CONTRIBUTORS.md) |
-[Community](http://tvmlang.org/community.html) |
+[Community](http://tvm.ai/community.html) |
[Release Notes](NEWS.md)
TVM is a compiler stack for deep learning systems. It is designed to close the gap between the
productivity-focused deep learning frameworks, and the performance- and efficiency-focused hardware backends.
TVM works with deep learning frameworks to provide end to end compilation to different backends.
-Checkout the [tvm stack homepage](http://tvmlang.org/) for more information.
+Checkout the [tvm stack homepage](http://tvm.ai/) for more information.
License
-------
diff --git a/apps/howto_deploy/README.md b/apps/howto_deploy/README.md
index 6c732879a..570c42ed1 100644
--- a/apps/howto_deploy/README.md
+++ b/apps/howto_deploy/README.md
@@ -8,4 +8,4 @@ Type the following command to run the sample code under the current folder(need
./run_example.sh
```
-Checkout [How to Deploy TVM Modules](http://docs.tvmlang.org/how_to/deploy.html) for more information.
+Checkout [How to Deploy TVM Modules](http://docs.tvm.ai/how_to/deploy.html) for more information.
diff --git a/docs/README.txt b/docs/README.txt
index b8780dd9f..8f62dbacc 100644
--- a/docs/README.txt
+++ b/docs/README.txt
@@ -1,6 +1,6 @@
The documentation of tvm is generated with recommonmark and sphinx.
-- A hosted version of doc is at http://docs.tvmlang.org
+- A hosted version of doc is at http://docs.tvm.ai
- pip install sphinx>=1.5.5 sphinx-gallery sphinx_rtd_theme matplotlib Image recommonmark
- Build tvm first in the root folder.
- To build locally, you need to enable USE_CUDA, USE_OPENCL, LLVM_CONFIG in config.mk and then type "make html" in this folder.
diff --git a/docs/dev/runtime.md b/docs/dev/runtime.md
index d601801a4..a5d8138c3 100644
--- a/docs/dev/runtime.md
+++ b/docs/dev/runtime.md
@@ -3,7 +3,7 @@
TVM supports multiple programming languages for the compiler stack development and deployment.
In this note, we explain the key elements of the TVM runtime.
-
+
We need to satisfy quite a few interesting requirements
@@ -132,11 +132,11 @@ of new device easy, and we do not need to redo the host code generation for each
The PackedFunc and Module system also makes it easy to ship the function into remote devices directly.
Under the hood, we have an RPCModule that serializes the arguments to do the data movement and launches the computation on the remote.
-
+
The RPC server itself is minimum and can be bundled into the runtime. We can start a minimum TVM
RPC server on iPhone/android/raspberry pi or even the browser. The cross compilation on server and shipping of the module for testing can be done in the same script. Checkout
-[Cross compilation and RPC tutorial](http://docs.tvmlang.org/tutorials/deployment/cross_compilation_and_rpc.html#sphx-glr-tutorials-deployment-cross-compilation-and-rpc-py) for more details.
+[Cross compilation and RPC tutorial](http://docs.tvm.ai/tutorials/deployment/cross_compilation_and_rpc.html#sphx-glr-tutorials-deployment-cross-compilation-and-rpc-py) for more details.
This instant feedback gives us a lot of advantages. For example, to test the correctness of generated code on iPhone, we no longer have to write test-cases in swift/objective-c from scratch -- We can use RPC to execute on iPhone, copy the result back and do verification on the host via numpy. We can also do the profiling using the same script.
diff --git a/docs/how_to/deploy.md b/docs/how_to/deploy.md
index 68630adfa..69fedaa97 100644
--- a/docs/how_to/deploy.md
+++ b/docs/how_to/deploy.md
@@ -12,7 +12,7 @@ cd apps/howto_deploy
Get TVM Runtime Library
-----------------------
-
+
The only thing we need is to link to a TVM runtime in your target platform.
TVM provides a minimum runtime, which costs around 300K to 600K depending on how much modules we use.
@@ -64,7 +64,7 @@ From android java TVM API to load model & execute can be refered at this [java](
Deploy NNVM Modules
-------------------
NNVM compiled modules are fully embedded in TVM runtime as long as ```GRAPH_RUNTIME``` option
-is enabled in tvm runtime. Check out the [TVM documentation](http://docs.tvmlang.org/) for
+is enabled in tvm runtime. Check out the [TVM documentation](http://docs.tvm.ai/) for
how to deploy TVM runtime to your system.
In a nutshell, we will need three items to deploy a compiled module.
diff --git a/jvm/README.md b/jvm/README.md
index 0a113fed0..243875861 100644
--- a/jvm/README.md
+++ b/jvm/README.md
@@ -27,7 +27,7 @@ TVM4J contains three modules:
### Build
-First please refer to [Installation Guide](http://docs.tvmlang.org/how_to/install.html) and build runtime shared library from the C++ codes (libtvm\_runtime.so for Linux and libtvm\_runtime.dylib for OSX).
+First please refer to [Installation Guide](http://docs.tvm.ai/how_to/install.html) and build runtime shared library from the C++ codes (libtvm\_runtime.so for Linux and libtvm\_runtime.dylib for OSX).
Then you can compile tvm4j by
@@ -117,12 +117,12 @@ public class LoadAddFunc {
Module fadd = Module.load(loadingDir + File.separator + "add_cpu.so");
TVMContext ctx = TVMContext.cpu();
-
+
long[] shape = new long[]{2};
NDArray arr = NDArray.empty(shape, ctx);
arr.copyFrom(new float[]{3f, 4f});
NDArray res = NDArray.empty(shape, ctx);
-
+
fadd.entryFunc().pushArg(arr).pushArg(arr).pushArg(res).invoke();
System.out.println(Arrays.toString(res.asFloatArray()));
@@ -135,7 +135,7 @@ public class LoadAddFunc {
## RPC Server
-There are two ways to start an RPC server on JVM. A standalone server can be started by
+There are two ways to start an RPC server on JVM. A standalone server can be started by
```java
Server server = new Server(port);
diff --git a/nnvm/README.md b/nnvm/README.md
index b733f4f85..ed8a18e3f 100644
--- a/nnvm/README.md
+++ b/nnvm/README.md
@@ -1,23 +1,4 @@
-# NNVM: Open Compiler for AI Frameworks
-[](http://mode-gpu.cs.washington.edu:8080/job/dmlc/job/nnvm/job/master/)
-[](./LICENSE)
-
-[Installation](docs/how_to/install.md) |
-[Documentation](http://nnvm.tvmlang.org) |
-[Tutorials](http://nnvm.tvmlang.org/tutorials/index.html) |
-[Release Notes](NEWS.md)
-
-NNVM compiler offers reusable computation graph optimization and compilation for deep learning systems.
-It is backed by the [TVM stack](http://tvmlang.org) and provides modules to:
-
-- Represent deep learning workloads from front-end frameworks via a graph IR.
-- Optimize computation graphs to improve performance.
-- Compile into executable modules and deploy to different hardware backends with minimum dependency.
-
-NNVM is designed to add new frontend, operators and graph optimizations in a decentralized fashion without changing the core interface.
-The compiled module can be deployed to server, mobile, embedded devices and browsers with minimum dependency, in languages including c++, python, javascript, java, objective-c. Checkout [our release announcement](http://www.tvmlang.org/2017/10/06/nnvm-compiler-announcement.html)
-
-The following code snippet demonstrates the general workflow of nnvm compiler.
+# NNVM Compiler Module of TVM Stack
```python
import tvm
@@ -52,13 +33,3 @@ rmodule = graph_runtime.create(graph, rlib, remote.gpu(0))
rmodule.set_input(**params)
rmodule.run()
```
-
-License
--------
-Licensed under an [Apache-2.0](https://github.com/dmlc/nnvm/blob/master/LICENSE) license.
-
-
-Links
------
-- [TinyFlow](https://github.com/tqchen/tinyflow) on how you can use NNVM to build a TensorFlow like API.
-- [Apache MXNet](http://mxnet.io/) uses NNVM as a backend.
diff --git a/tutorials/nnvm/deploy_model_on_mali_gpu.py b/tutorials/nnvm/deploy_model_on_mali_gpu.py
index 3c0152b8c..116e99f76 100644
--- a/tutorials/nnvm/deploy_model_on_mali_gpu.py
+++ b/tutorials/nnvm/deploy_model_on_mali_gpu.py
@@ -7,8 +7,7 @@ This is an example of using NNVM to compile a ResNet model and
deploy it on Firefly-RK3399 with ARM Mali GPU. We will use the
Mali-T860 MP4 GPU on this board to accelerate the inference.
-This tutorial is based on the `tutorial <http://nnvm.tvmlang.org/tutorials/deploy_model_on_rasp.html>`_
-for deploying on Raspberry Pi by `Ziheng Jiang <https://ziheng.org/>`_.
+This tutorial is based on the tutorial for deploying on Raspberry Pi by `Ziheng Jiang <https://ziheng.org/>`_.
Great thanks to the original author, I only do several lines of modification.
To begin with, we import nnvm (for compilation) and TVM (for deployment).
--
GitLab