All Products
Search
Document Center

Platform For AI:PAI public images

Last Updated:Jan 10, 2024

Alibaba Cloud Platform for AI (PAI) provides public images based on different frameworks and Compute Unified Device Architecture (CUDA) versions. You can easily create an AI development environment based on a public image when you use Deep Learning Container (DLC), Elastic Algorithm Service (EAS), or Data Science Workshop (DSW). This topic describes the capabilities of PAI public images and lists the core images.

Naming conventions

PAI public images are named based on specific naming conventions. The name of an image contains the basic information about the image. Each image name consists of several fields, as described in the following table. We recommend that you use the same naming conventions when you create a custom image.

Example

Meaning

Supported module

tensorflow:2.11-gpu-py39-cu112-ubuntu20.04

  • tensorflow:2.11: The training framework is TensorFlow 2.11.

  • gpu: The applicable server model is GPU.

  • py39: The programming language is Python 3.9.

  • cu112: The supported CUDA version is 112.

  • ubuntu20.04: The supported operating system is Ubuntu 20.04.

The PAI public images are suitable for different PAI modules:

  • The -training type of images are suitable for DLC.

  • The -inference type of images are suitable for EAS.

  • The -develop type of images are suitable for DSW.

deeprec-develop:2302-tensorflow1.15-cpu-py36-ubuntu18.04

  • deeprec-develop:2302-tensorflow1.15: The training framework is TensorFlow 1.15 and DeepRec 2302.

  • cpu: The applicable server model is CPU.

  • py36: The programming language is Python 3.6.

  • ubuntu18.04: The supported operating system is Ubuntu 18.04.

Capabilities

Alibaba Cloud PAI provides public images based on different training frameworks. You can view all the public images on the AI Computing Asset Management > Images page in the PAI console. This section describes the information about the public images for popular training frameworks such as TensorFlow and PyTorch.

TensorFlow

Framework version

CUDA version (GPU model only)

Operating system

  • TensorFlow2.6

  • TensorFlow2.3

  • TensorFlow2.21

  • TensorFlow2.11

  • TensorFlow 1.15 and TensorFlow 1.15.5

  • TensorFlow1.12

  • CUDA 114

  • CUDA 113

  • CUDA 112

  • CUDA 101

  • CUDA 100

  • Ubuntu 20.04

  • Ubuntu 18.04

TensorFlow Serving

Framework version

CUDA version (GPU model only)

Operating system

  • TensorFlowServing2.11.1

  • TensorFlowServing1.15.0

  • CUDA 112

  • CUDA 100

  • Ubuntu 20.04

  • Ubuntu 18.04

  • Ubuntu 16.04

Pytorch

Framework version

CUDA version (GPU model only)

Operating system

  • Pytorch2.1

  • Pytorch2.0

  • Pytorch1.8

  • Pytorch1.7

  • Pytorch1.12

  • Pytorch1.11

  • Pytorch1.10

  • CUDA 121

  • CUDA 117

  • CUDA 114

  • CUDA 113

  • CUDA 101

  • Ubuntu 22.04

  • Ubuntu 20.04

  • Ubuntu 18.04

DeepRec

Framework version

CUDA version (GPU model only)

Operating system

  • DeepRec2302

  • DeepRec2212

CUDA 114

Ubuntu 18.04

XGBoost

Framework version

CUDA version (GPU model only)

Operating system

XGBoost1.6.0

Only the CPU model is supported.

Ubuntu 18.04

Triton Inference Server

Framework version

CUDA version (GPU model only)

Operating system

  • TritonServer23.02

  • TritonServer21.09

  • CUDA 120

  • CUDA 114

Ubuntu 20.04

Core images

Images for Lingjun resources of Serverless Edition

Image name

Framework

Model

CUDA

Operating system

Supported region

Programming language and version

deepspeed-training:23.06-gpu-py310-cu121-ubuntu22.04

  • PyTorch 2.1

  • Megatron-LM 23.06

  • DeepSpeed 0.9.5

  • Transformers 4.29.2

  • Nemo 1.19.0

GPU

121

ubuntu22.04

China (Ulanqab)

Python3.10

megatron-training:23.06-gpu-py310-cu121-ubuntu22.04

  • PyTorch 2.1

  • Megatron-LM 23.06

  • DeepSpeed 0.9.5

  • Transformers 4.29.2

  • Nemo 1.19.0

GPU

121

ubuntu22.04

China (Ulanqab)

Python3.10

nemo-training:23.06-gpu-py310-cu121-ubuntu22.04

  • PyTorch 2.1

  • Megatron-LM 23.06

  • DeepSpeed 0.9.5

  • Transformers 4.29.2

  • Nemo 1.19.0

GPU

121

ubuntu22.04

China (Ulanqab)

Python3.10

AIGC images

Image name

Framework

Model

CUDA

Operating system

Supported region

Programming language and version

stable-diffusion-webui:3.0

StableDiffusionWebUI3.0

GPU

117

ubuntu22.04

  • China (Hangzhou)

  • China (Shanghai)

  • China (Beijing)

  • China (Zhangjiakou)

  • China (Ulanqab)

  • China (Shenzhen)

  • China (Heyuan)

  • China (Chengdu)

Python3.10

stable-diffusion-webui:2.2

StableDiffusionWebUI2.2

GPU

117

ubuntu22.04

Python3.10

stable-diffusion-webui:1.1

StableDiffusionWebUI1.1

GPU

117

ubuntu22.04

Python3.10

stable-diffusion-webui-env:pytorch1.13-gpu-py310-cu117-ubuntu22.04

SD-WebUI-ENV

GPU

117

ubuntu22.04

Python3.10