Update README and doc (bytedance#417)

Update README and doc

Author: godweiyang

README.md

@@ -0,0 +1,33 @@
+# LightSeq Examples
+
+## Table of Contents
+- [Cpp Examples](#cpp-examples)
+- [Python Examples](#python-examples)
+    - [Train the models](#train-the-models)
+    - [Export and infer the models](#export-and-infer-the-models)
+- [Deploy using Tritonbackend](#deploy-using-tritonbackend)
+
+## Cpp Examples
+We provide multiple cpp examples of LightSeq inference.
+
+First you should use the training examples in the following to train a model, and then export it to protobuf or HDF5 format.
+
+Then use the cpp examples to infer the models:
+1. Uncomment the `add_subdirectory(examples/inference/cpp)` in the [CMakeLists.txt](../CMakeLists.txt).
+2. Build the LightSeq. Refer to [build.md](./build.md) for more details.
+3. Switch to `build/temp.linux-xxx/examples/inference/cpp`, and then run `sudo make` to compile the cpp example.
+4. Run the cpp examples by `./xxx_example MODEL_PATH`.
+
+## Python Examples
+We provide a series of Python examples to show how to use LightSeq to do model training and inference.
+
+### Train the models
+Currently, LightSeq supports training from [Fairseq](../examples/training/fairseq/README.md), [Hugging Face](../examples/training/huggingface/README.md), [DeepSpeed](../examples/training/deepspeed/README.md) and [from scratch](../examples/training/custom/README.md). For more training details, please refer to the respective README.
+
+### Export and infer the models
+First export the models training by Fairseq, Hugging Face or LightSeq to protobuf or HDF5 format. Then test the results and speeds using the testing scripts.
+
+Refer to [here](../examples/inference/python/README.md) for more details.
+
+## Deploy using Tritonbackend
+Refer to [here](../examples/triton_backend/README.md) for more details.

docs/inference/build.md renamed to docs/build.md

@@ -1,4 +1,20 @@
 # A Guide of LightSeq Training and Inference
+
+## Table of Contents
+- [Introduction](#introduction)
+- [Training](#training)
+    - [Custom integration](#custom-integration)
+    - [Hugging Face](#hugging-face)
+    - [Fairseq](#fairseq)
+    - [DeepSpeed](#deepspeed)
+- [Inference](#inference)
+    - [Export](#export)
+        - [Fairseq](#fairseq)
+        - [Hugging Face](#hugging-face)
+        - [LightSeq Transformer](#lightseq-transformer)
+        - [Custom models](#custom-models)
+    - [Inference in three lines of codes!](#inference-in-three-lines-of-codes)
+
 ## Introduction
 This document mainly introduces the detailed process of LightSeq training and inference. In short, the process can be divided into the following three steps:
 1. Train models integrated with LightSeq training modules, and save the checkpoints.