Permalink
Browse files

Add basic project documents.

  • Loading branch information...
lubatang committed Aug 2, 2018
1 parent 36353d0 commit a94485c1409072d0a53cd207432873cf4f9b29a5
Showing with 71 additions and 13 deletions.
  1. +2 −0 BUGS
  2. +31 −0 CODE_OWNERS.TXT
  3. +9 −0 ChangeLog
  4. +29 −13 README.md
2 BUGS
@@ -0,0 +1,2 @@
0.9.0 Known Bugs
- Compute operators don't have internval onnc::Value.
@@ -0,0 +1,31 @@
This file is a list of the people responsible for ensuring that patches for a
particular part of ONNC are reviewed. They are also the gatekeepers for their
part of ONNC, with the final word on what goes in or not.

luba@skymizer.com - ADT, Core, IR, JSON, Option, Analysis,
Diagnostic, Support, Transforms
a127127@skymizer.com - IR, IRReader, Support, lib/Target
cycheng@skymizer.com - IR, Target, Analysis, Transforms
luishsu@skymizer.com - umbrella, TravisCI
pine0113@skymizer.com - github, TravisCI
issue.hsu@bitmain.com - lib/Target/Sophon
zakk.chen@bitmain.com - lib/Target/Sophon

Here is a list of contributors.
evanli@skymizer.com
luishsu@skymizer.com
max@skymizer.com
peter@skymizer.com
kttt@skymizer.com
dm4@skymizer.com
catcatcatcat@skymizer.com
katy@skymizer.com
eric.yu@bitmain.com
shihpo.hung@bitmain.com
hr.hsu@bitmain.com
min.chen@bitmain.com
yuming@bitmain.com
changsheng.yin@bitmain.com
frank.lin@bitmain.com
fp.tsai@bitmain.com
wanwei.cai@bitmain.com
@@ -1,3 +1,12 @@
2018-8-3 Luba Tang <luba@skymizer.com>
* First open source release: 0.9.0
* Executive Summary:
1. Finish basic infrastructure of ONNC, including IR, Core, ADT, Support,
Diagnostic and backends.
2. Enable Sophon backend. Now Sophon backend still use old form of compute
operators.
3. Provides onnc, onnx2tg and some utilities.

2018-1-17 Luba Tang <luba@skymizer.com>
* Initialization: create all directory structures and auto-tool building
system.
@@ -1,22 +1,38 @@
# Introduction
ONNC (Open Neural Network Compiler)

# Test
ONNC (Open Neural Network Compiler)

## Add or Remove a Unit-Test
A compiler specifically designed for Deep Learning Accelerator (DLA).
Based on Open Neural Network Exchange (ONNX), ONNC transforms ONNX models into
binary machine code for DLA ASIC. ONNC leverages the IR design of ONNX and
provides effective algorithms to eliminate the overhead of data movement.
ONNC increases DLA ASIC performance and shorten production time for DLA ASIC.

From CI pipeline:
0. ci/.gitlab-ci.yml
1. docker/common.sh
## Lastest release

From docker development environment:
0. docker/in-container/testall

# Directory Structure
* README - This document
* v0.9
- Tarball:
## Directory Structure
* README.md - This document
* docs - documents
* externals - The 3rd-party projects
* include - header files for libonnc
* lib - implementation for libonnc
* test - Quick regression tests for ONNC.
* tools - tools based on libonnc

## Supported platforms

ONNC supports Ubuntu/x86_64 and MacOSX.

Here is a list of verified versions:
* Ubuntu/x86_64
- 16.04

* MacOSX
- High Sierra

## Getting Started

## Download and build ONNC.
For now, the simplest way to download and build ONNC is follow the
instructions of README.md in **onnc-umbrella**.

0 comments on commit a94485c

Please sign in to comment.