Paddle Lite is an updated version of Paddle-Mobile, an open-open source deep learning framework designed to make it easy to perform inference on mobile devices. It is compatible with PaddlePaddle and pre-trained models from other sources.
If you want to know more details about the documentation of paddle-mobile design, please refer to the link as follows. There are many previous designs and discussion: [issue](https://github.com/PaddlePaddle/paddle-mobile/issues).
For tutorials, please see [PaddleLite Wiki](https://github.com/PaddlePaddle/paddle-mobile/wiki).
[link of documentation of design](https://github.com/PaddlePaddle/paddle-mobile/blob/develop/doc/design_doc.md)
## Key Features
### Documentation of development
### Light Weight
Documentation of development is mainly about building, running and other tasks.As a developer,you can use it with the help of contributed documents.
On mobile devices, execution module can be deployed without third-party libraries, because our excecution module and analysis module are decoupled.
On ARM V7, only 800KB are taken up, while on ARM V8, 1.3MB are taken up with the 80 operators and 85 kernels in the dynamic libraries provided by Paddle Lite.
-[tutorial link to contribute documents](https://github.com/PaddlePaddle/paddle-mobile/blob/develop/CONTRIBUTING.md)
- Main procedure of contributing code is covered in the document above.If you have other problems during the procedure,please send them as [issue](https://github.com/PaddlePaddle/paddle-mobile/issues). We will deal with it as quickly as possible.
## 文档
Paddle Lite enables immediate inference without extra optimization.
It also supports INT8 quantizations with [PaddleSlim model compression tools](https://github.com/PaddlePaddle/models/tree/v1.5/PaddleSlim), reducing the size of models and increasing the performance of models.
开发文档主要是关于编译、运行等问题。做为开发者，它可以和贡献文档共同结合使用。
On Huawei NPU and FPGA, the performance is also boosted.
Hardware compatibility: Paddle Lite supports a diversity of hardwares — ARM CPU, Mali GPU, Adreno GPU, Huawei NPU and FPGA. In the near future, we will also support AI microchips from Cambricon and Bitmain.
At present Paddle-Mobile only supports Paddle fluid training model. Models wiil be operated regularly after transformation if you have various models.
ONNX is expanded as Open Neural Network Exchange. The project is aimed to make a full communication and usage among diffrent nerual network development frameworks.
Except for directly using fluid models trained by PaddlePaddle,you can also get certain Paddle fluid models through onnx transformation.
Model compatibility: The Op of Paddle Lite is fully compatible to that of PaddlePaddle. The accuracy and performance of 18 models (mostly CV models and OCR models) and 85 operators have been validated. In the future, we will also support other models.
At present，work in support of onnx is also under operation in Baidu. Related tranformation project can be referred to here：
Framework compatibility: In addition to models trained on PaddlePaddle, those trained on Caffe and TensorFlow can also be converted to be used on Paddle Lite, via [X2Paddle](https://github.com/PaddlePaddle/X2Paddle). In the future to come, we will also support models of ONNX format.
- input data generated by tools from `tools/python/imagetools`.
Paddle Lite is designed to support a wide range of hardwares and devices, and it enables mixed execution of a single model on multiple devices, optimization on various phases, and leight-weighted applications on devices.
As is shown in the figure above, analysis phase includes Machine IR module, and it enables optimizations like Op fusion and redundant computation pruning. Besides, excecution phase only involves Kernal exevution, so it can be deployed on its own to ensure maximized light-weighted deployment.
The earlier Paddle-Mobile was designed to be compatible with PaddlePaddle and multiple hardwares, including ARM CPU, Mali GPU, Adreno GPU, FPGA, ARM-Linux and Apple's GPU Metal. Within Baidu, inc, many product lines have been using Paddle-Mobile. For more details, please see: [mobile/README](mobile/README).
As an update of Paddle-Mobile, Paddle Lite has incorporated many older capabilities into the [new architecture](https://github.com/PaddlePaddle/paddle-mobile/tree/develop/lite). For the time being, the code of Paddle-mobile will be kept under the directory `mobile/`, before complete transfer to Paddle Lite.
For demands of Apple's GPU Metal and web front end inference, please see `./metal` and `./web` . These two modules will be further developed and maintained.
Original MDL(Mobile-Deep-Learning) project has been transferred to [Mobile-Deep-Learning](https://github.com/allonli/mobile-deep-learning)
-[Anakin](https://github.com/PaddlePaddle/Anakin). The optimizations under Anakin has been incorporated into Paddle Lite, and so there will not be any future updates of Anakin. As another high-performance inference project under PaddlePaddle, Anakin has been forward-looking and helpful to the making of Paddle Lite.
