1. This SDK include pack tool to pack local AI inference program into UCloud web service image which could run on UCloud UAI Inference Platform.
2. This SDK include whole APIs which could interactive with UCloud UAI Inference Platform.
3. This SDK include pack tool to help pack local AI trainning program into UCloud Docker image which could run on UCloud UAI Train Platform.
4. This SDK include whole APIs which could interactive with UCloud UAI Train Platforms

CHECK ON：UAI Inference Base Docker Images

• Tensorflow (1.6.0 tested)
• Tensorflow (1.7.0 tested)
• Caffe (1.0.0 tested)

• Tensorflow (1.1.0 tested）
• Tensorflow (1.2.0 tested）
• Tensorflow (1.3.0 tested)
• Tensorflow (1.4.0 tested)
• Tensorflow (1.6.0 tested)
• Tensorflow (1.7.0 tested)
• MXNet (0.9.5 tested)
• MXNet (1.0.0 tested)
• Keras (1.2.0 tested)
• Keras (2.0.8 tested)
• Caffe (1.0.0 tested)
• Intel-Caffe
• XGBoost (0.7 tested, python interface only)

CHECK ON：UAI Train Base Docker Images

• Tensorflow (0.11.0 tested)
• Tensorflow（1.1.0 tested）
• Tensorflow（1.2.0 tested）
• Tensorflow（1.3.0 tested)
• Tensorflow（1.4.0 tested）
• Tensorflow（1.5.0, cuda9 tested)
• Tensorflow（1.6.0, cuda9 tested）
• Tensorflow (1.7.0, cuda9 tested)
• Tensorflow (1.8.0, cuda9 tested)
• Tensorflow (1.9.0, cuda9 tested)
• MXNet (0.9.5 tested)
• MXNet (1.0.0, cuda9 tested)
• Keras (2.0.8 tested)
• Keras (2.1.6, tf-1.8 testes)
• Caffe (1.0.0 tested)
• Caffe2 (Detection)
• PyTorch (0.2.0)
• PyTorch (0.4)

• Tensorflow Distributed Training (examples include slim/cifar/wide-deep)
• MXNet Distributed Training

1. Install your deep learning python package, such as Tensorflow, MXNet, Keras, Caffe (tested version preferred)

2. Install uai-sdk

    $ sudo python setup.py install
   

3. Install all other lib your inference code required

https://docs.ucloud.cn/ai/uai-inference/index

https://docs.ucloud.cn/ai/uai-train/use