Skip to content

T
tic
Commit 55d13d47 by tqchen Committed by Tianqi Chen
Options

Homepage URL to tvm.ai

parent fc6b1d4d
Showing with 18 additions and 48 deletions
README.md
<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
============================================== ==============================================
[![GitHub license](http://dmlc.github.io/img/apache2.svg)](./LICENSE) [![GitHub license](http://dmlc.github.io/img/apache2.svg)](./LICENSE)
[![Build Status](http://mode-gpu.cs.washington.edu:8080/buildStatus/icon?job=dmlc/tvm/master)](http://mode-gpu.cs.washington.edu:8080/job/dmlc/job/tvm/job/master/) [![Build Status](http://mode-gpu.cs.washington.edu:8080/buildStatus/icon?job=dmlc/tvm/master)](http://mode-gpu.cs.washington.edu:8080/job/dmlc/job/tvm/job/master/)
[Installation](docs/how_to/install.md) | [Installation](docs/how_to/install.md) |
[Documentation](http://docs.tvmlang.org) | [Documentation](http://docs.tvm.ai) |
[Tutorials](http://tutorials.tvmlang.org) | [Tutorials](http://tutorials.tvm.ai) |
[Operator Inventory](topi) | [Operator Inventory](topi) |
[FAQ](docs/faq.md) | [FAQ](docs/faq.md) |
[Contributors](CONTRIBUTORS.md) | [Contributors](CONTRIBUTORS.md) |
[Community](http://tvmlang.org/community.html) | [Community](http://tvm.ai/community.html) |
[Release Notes](NEWS.md) [Release Notes](NEWS.md)
TVM is a compiler stack for deep learning systems. It is designed to close the gap between the 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. 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. 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 License
------- -------
......
apps/howto_deploy/README.md
...@@ -8,4 +8,4 @@ Type the following command to run the sample code under the current folder(need ...@@ -8,4 +8,4 @@ Type the following command to run the sample code under the current folder(need
./run_example.sh ./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.
docs/README.txt
The documentation of tvm is generated with recommonmark and sphinx. 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 - pip install sphinx>=1.5.5 sphinx-gallery sphinx_rtd_theme matplotlib Image recommonmark
- Build tvm first in the root folder. - 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. - To build locally, you need to enable USE_CUDA, USE_OPENCL, LLVM_CONFIG in config.mk and then type "make html" in this folder.
docs/dev/runtime.md
...@@ -3,7 +3,7 @@ ...@@ -3,7 +3,7 @@
TVM supports multiple programming languages for the compiler stack development and deployment. TVM supports multiple programming languages for the compiler stack development and deployment.
In this note, we explain the key elements of the TVM runtime. In this note, we explain the key elements of the TVM runtime.
![](http://www.tvmlang.org/images/release/tvm_flexible.png) ![](http://www.tvm.ai/images/release/tvm_flexible.png)
We need to satisfy quite a few interesting requirements 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 ...@@ -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. 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. Under the hood, we have an RPCModule that serializes the arguments to do the data movement and launches the computation on the remote.
![](http://www.tvmlang.org/images/release/tvm_rpc.png) ![](http://www.tvm.ai/images/release/tvm_rpc.png)
The RPC server itself is minimum and can be bundled into the runtime. We can start a minimum TVM 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 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. 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.
......
docs/how_to/deploy.md
...@@ -12,7 +12,7 @@ cd apps/howto_deploy ...@@ -12,7 +12,7 @@ cd apps/howto_deploy
Get TVM Runtime Library Get TVM Runtime Library
----------------------- -----------------------
![](http://www.tvmlang.org/images/release/tvm_flexible.png) ![](http://www.tvm.ai/images/release/tvm_flexible.png)
The only thing we need is to link to a TVM runtime in your target platform. 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. 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]( ...@@ -64,7 +64,7 @@ From android java TVM API to load model & execute can be refered at this [java](
Deploy NNVM Modules Deploy NNVM Modules
------------------- -------------------
NNVM compiled modules are fully embedded in TVM runtime as long as ```GRAPH_RUNTIME``` option 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. how to deploy TVM runtime to your system.
In a nutshell, we will need three items to deploy a compiled module. In a nutshell, we will need three items to deploy a compiled module.
......
jvm/README.md
...@@ -27,7 +27,7 @@ TVM4J contains three modules: ...@@ -27,7 +27,7 @@ TVM4J contains three modules:
### Build ### 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 Then you can compile tvm4j by
...@@ -117,12 +117,12 @@ public class LoadAddFunc { ...@@ -117,12 +117,12 @@ public class LoadAddFunc {
Module fadd = Module.load(loadingDir + File.separator + "add_cpu.so"); Module fadd = Module.load(loadingDir + File.separator + "add_cpu.so");
TVMContext ctx = TVMContext.cpu(); TVMContext ctx = TVMContext.cpu();
long[] shape = new long[]{2}; long[] shape = new long[]{2};
NDArray arr = NDArray.empty(shape, ctx); NDArray arr = NDArray.empty(shape, ctx);
arr.copyFrom(new float[]{3f, 4f}); arr.copyFrom(new float[]{3f, 4f});
NDArray res = NDArray.empty(shape, ctx); NDArray res = NDArray.empty(shape, ctx);
fadd.entryFunc().pushArg(arr).pushArg(arr).pushArg(res).invoke(); fadd.entryFunc().pushArg(arr).pushArg(arr).pushArg(res).invoke();
System.out.println(Arrays.toString(res.asFloatArray())); System.out.println(Arrays.toString(res.asFloatArray()));
...@@ -135,7 +135,7 @@ public class LoadAddFunc { ...@@ -135,7 +135,7 @@ public class LoadAddFunc {
## RPC Server ## 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 ```java
Server server = new Server(port); Server server = new Server(port);
......
nnvm/README.md
# NNVM: Open Compiler for AI Frameworks # NNVM Compiler Module of TVM Stack
[![Build Status](http://mode-gpu.cs.washington.edu:8080/buildStatus/icon?job=dmlc/nnvm/master)](http://mode-gpu.cs.washington.edu:8080/job/dmlc/job/nnvm/job/master/)
[![GitHub license](http://dmlc.github.io/img/apache2.svg)](./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.
```python ```python
import tvm import tvm
...@@ -52,13 +33,3 @@ rmodule = graph_runtime.create(graph, rlib, remote.gpu(0)) ...@@ -52,13 +33,3 @@ rmodule = graph_runtime.create(graph, rlib, remote.gpu(0))
rmodule.set_input(**params) rmodule.set_input(**params)
rmodule.run() 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.
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 ...@@ -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 deploy it on Firefly-RK3399 with ARM Mali GPU. We will use the
Mali-T860 MP4 GPU on this board to accelerate the inference. 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>`_ This tutorial is based on the tutorial for deploying on Raspberry Pi by `Ziheng Jiang <https://ziheng.org/>`_.
for deploying on Raspberry Pi by `Ziheng Jiang <https://ziheng.org/>`_.
Great thanks to the original author, I only do several lines of modification. 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). To begin with, we import nnvm (for compilation) and TVM (for deployment).
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment