Release notes for PyTorch-Ubuntu images

更新时间:
复制 MD 格式

PyTorch-Ubuntu images are pre-built container images that combine Ubuntu with PyTorch and its dependencies, giving you a ready-to-use deep learning environment. Each image is available in three variants: CPU-only, CUDA 12.1, and CUDA 11.8.

Version summary

Image version PyTorch CUDA Released
2.3.0.3.post1 (enhanced) 2.3.0.3.post1 12.1.1 / 11.8.0 / CPU 2024-07
2.3.0 2.3.0 12.1.1 / 11.8.0 / CPU 2024-06
2.2.0.1 2.2.0.1 12.1.1 / 11.8.0 / CPU 2024-03

PyTorch 2.3.0.3.post1

Images

Category Image address
CPU ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.3.0.3-ubuntu22.04-enhanced
GPU (CUDA 12.1) ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.3.0.3-cuda12.1.1-ubuntu22.04-enhanced
GPU (CUDA 11.8) ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.3.0.3-cuda11.8.0-ubuntu22.04-enhanced

Driver requirements

Image Minimum driver Compatible drivers
pytorch-ubuntu:2.3.0.3-ubuntu22.04-enhanced None (CPU-only)
pytorch-ubuntu:2.3.0.3-cuda12.1.1-ubuntu22.04-enhanced NVIDIA Driver release 530 Also compatible with releases 470 and 525
pytorch-ubuntu:2.3.0.3-cuda11.8.0-ubuntu22.04-enhanced NVIDIA Driver release 470

What's new

This release is built on PyTorch 2.3.0 with targeted optimizations for Dynamo models:

  • DeepSpeed and Fully Sharded Data Parallel (FSDP) support — train large models across distributed GPU clusters using either framework.

  • Improved Dynamo model performance — reduced overhead for models compiled with TorchDynamo.

  • torchvision and torchaudio updated to 2.3.0.3.post1 and 0.18.0.3.post1 respectively.

Image content

pytorch-ubuntu:2.3.0.3-ubuntu22.04-enhanced

Component Version
Base OS Ubuntu 22.04
gcc 11.4.0
Python 3.10.12
torch 2.3.0.3.post1
torchaudio 2.3.0.3.post1
torchvision 0.18.0.3.post1
certifi 2024.6.2
charset-normalizer 3.3.2
filelock 3.15.3
fsspec 2024.6.0
idna 3.7
Jinja2 3.1.4
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.3
numpy 1.26.4
pillow 10.3.0
pip 24.1.1
PySocks 1.7.1
requests 2.32.3
setuptools 70.2.0
sympy 1.12.1
typing_extensions 4.12.2
urllib3 2.2.2
wheel 0.43.0

pytorch-ubuntu:2.3.0.3-cuda12.1.1-ubuntu22.04-enhanced

Component Version
Base OS Ubuntu 22.04
gcc 11.4.0
Python 3.10.12
torch 2.3.0.3.post1+cu121
torchaudio 2.3.0.3.post1+cu121
torchvision 0.18.0.3.post1+cu121
triton 2.3.0
cuda-compat-12-1 530.30.02
cuda-cudart-12-1 12.1.105
cuda-cupti-12-1 12.1.105
cuda-libraries-12-1 12.1.1
cuda-nvrtc-12-1 12.1.105
cuda-nvtx-12-1 12.1.105
cuda-opencl-12-1 12.1.105
libcublas-12-1 12.1.3.1
libcufft-12-1 11.0.2.54
libcufile-12-1 1.6.1.9
libcurand-12-1 10.3.2.106
libcusolver-12-1 11.4.5.107
libcusparse-12-1 12.1.0.106
libnpp-12-1 12.1.0.40
libnvjitlink-12-1 12.1.105
libnvjpeg-12-1 12.2.0.2
libnccl2 2.20.5
libcudnn8 8.9.2.26
certifi 2024.6.2
charset-normalizer 3.3.2
filelock 3.15.3
fsspec 2024.6.0
idna 3.7
Jinja2 3.1.4
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.3
numpy 1.26.4
pillow 10.3.0
pip 24.1.1
PySocks 1.7.1
requests 2.32.3
setuptools 70.2.0
sympy 1.12.1
typing_extensions 4.12.2
urllib3 2.2.2
wheel 0.43.0

pytorch-ubuntu:2.3.0.3-cuda11.8.0-ubuntu22.04-enhanced

Component Version
Base OS Ubuntu 22.04
gcc 11.4.0
Python 3.10.12
torch 2.3.0.3.post1+cu118
torchaudio 2.3.0.3.post1+cu118
torchvision 0.18.0.3.post1+cu118
triton 2.3.0
cuda-compat-11-8 520.61.05
cuda-cudart-11-8 11.8.89
cuda-cupti-11-8 11.8.87
cuda-libraries-11-8 11.8.0
cuda-nvrtc-11-8 11.8.89
cuda-nvtx-11-8 11.8.86
libcublas-11-8 11.11.3.6
libcufft-11-8 10.9.0.58
libcufile-11-8 1.4.0.31
libcurand-11-8 10.3.0.86
libcusolver-11-8 11.4.1.48
libcusparse-11-8 11.7.5.86
libnpp-11-8 11.8.0.86
libnvjpeg-11-8 11.9.0.86
libnccl2 2.20.5
libcudnn8 8.7.0.84
certifi 2024.6.2
charset-normalizer 3.3.2
filelock 3.15.3
fsspec 2024.6.0
idna 3.7
Jinja2 3.1.4
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.3
numpy 1.26.4
pillow 10.3.0
pip 24.1.1
PySocks 1.7.1
requests 2.32.3
setuptools 70.2.0
sympy 1.12.1
typing_extensions 4.12.2
urllib3 2.2.2
wheel 0.43.0

PyTorch 2.3.0

Images

Category Image address
CPU ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.3.0-ubuntu22.04
GPU (CUDA 12.1) ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.3.0-cuda12.1.1-ubuntu22.04
GPU (CUDA 11.8) ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.3.0-cuda11.8.0-ubuntu22.04

Driver requirements

Image Minimum driver Compatible drivers
pytorch-ubuntu:2.3.0-ubuntu22.04 None (CPU-only)
pytorch-ubuntu:2.3.0-cuda12.1.1-ubuntu22.04 NVIDIA Driver release 530 Also compatible with releases 470 and 525
pytorch-ubuntu:2.3.0-cuda11.8.0-ubuntu22.04 NVIDIA Driver release 470

What's new

PyTorch updated to 2.3.0, with corresponding updates to torchvision (0.18.0) and torchaudio (2.3.0).

Image content

pytorch-ubuntu:2.3.0-ubuntu22.04

Component Version
gcc 11.4.0
Python 3.10.12
torch 2.3.0+cpu
torchaudio 2.3.0+cpu
torchvision 0.18.0+cpu
certifi 2024.2.2
charset-normalizer 3.3.2
filelock 3.14.0
fsspec 2024.5.0
idna 3.7
Jinja2 3.1.4
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.3
numpy 1.26.4
pillow 10.3.0
pip 24.0
PySocks 1.7.1
requests 2.32.3
setuptools 70.0.0
sympy 1.12
typing_extensions 4.12.0
urllib3 2.2.1
wheel 0.43.0

pytorch-ubuntu:2.3.0-cuda12.1.1-ubuntu22.04

Component Version
gcc 11.4.0
Python 3.10.12
torch 2.3.0+cu121
torchaudio 2.3.0+cu121
torchvision 0.18.0+cu121
triton 2.3.0
cuda-compat-12-1 530.30.02
cuda-cudart-12-1 12.1.105
cuda-cupti-12-1 12.1.105
cuda-libraries-12-1 12.1.1
cuda-nvrtc-12-1 12.1.105
cuda-nvtx-12-1 12.1.105
cuda-opencl-12-1 12.1.105
libcublas-12-1 12.1.3.1
libcufft-12-1 11.0.2.54
libcufile-12-1 1.6.1.9
libcurand-12-1 10.3.2.106
libcusolver-12-1 11.4.5.107
libcusparse-12-1 12.1.0.106
libnpp-12-1 12.1.0.40
libnvjitlink-12-1 12.1.105
libnvjpeg-12-1 12.2.0.2
libnccl2 2.20.5
libcudnn8 8.9.2.26
certifi 2024.2.2
charset-normalizer 3.3.2
filelock 3.14.0
fsspec 2024.5.0
idna 3.7
Jinja2 3.1.4
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.3
numpy 1.26.4
pillow 10.3.0
pip 24.0
PySocks 1.7.1
requests 2.32.3
setuptools 70.0.0
sympy 1.12
typing_extensions 4.12.0
urllib3 2.2.1
wheel 0.43.0

pytorch-ubuntu:2.3.0-cuda11.8.0-ubuntu22.04

Component Version
gcc 11.4.0
Python 3.10.12
torch 2.3.0+cu118
torchaudio 2.3.0+cu118
torchvision 0.18.0+cu118
triton 2.3.0
cuda-compat-11-8 520.61.05
cuda-cudart-11-8 11.8.89
cuda-cupti-11-8 11.8.87
cuda-libraries-11-8 11.8.0
cuda-nvrtc-11-8 11.8.89
cuda-nvtx-11-8 11.8.86
libcublas-11-8 11.11.3.6
libcufft-11-8 10.9.0.58
libcufile-11-8 1.4.0.31
libcurand-11-8 10.3.0.86
libcusolver-11-8 11.4.1.48
libcusparse-11-8 11.7.5.86
libnpp-11-8 11.8.0.86
libnvjpeg-11-8 11.9.0.86
libnccl2 2.20.5
libcudnn8 8.7.0.84
certifi 2024.6.2
charset-normalizer 3.3.2
filelock 3.15.3
fsspec 2024.6.0
idna 3.7
Jinja2 3.1.4
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.3
numpy 1.26.4
pillow 10.3.0
pip 24.1.1
PySocks 1.7.1
requests 2.32.3
setuptools 70.2.0
sympy 1.12.1
typing_extensions 4.12.2
urllib3 2.2.2
wheel 0.43.0

PyTorch 2.2.0.1

Images

Category Image address
CPU ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.2.0.1-22.04
GPU (CUDA 12.1) ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.2.0.1-22.04-cu121
GPU (CUDA 11.8) ac2-registry.cn-hangzhou.cr.aliyuncs.com/ac2/pytorch-ubuntu:2.2.0.1-22.04-cu118

Driver requirements

Image Minimum driver Compatible drivers
pytorch-ubuntu:2.2.0.1-22.04 None (CPU-only)
pytorch-ubuntu:2.2.0.1-22.04-cu121 NVIDIA Driver release 530 Also compatible with releases 470 and 525
pytorch-ubuntu:2.2.0.1-22.04-cu118 NVIDIA Driver release 470

What's new

This release is built on PyTorch 2.2.0 with targeted optimizations for Dynamo models:

  • Fully Sharded Data Parallel (FSDP) and DeepSpeed support — train large models across distributed GPU clusters using either framework.

  • Improved Dynamo model performance — reduced overhead for models compiled with TorchDynamo.

  • torchvision and torchaudio updated to 0.17.0.1 and 2.2.0.1 respectively.

Image content

pytorch-ubuntu:2.2.0.1-22.04

Component Version
gcc 11.4.0
Python 3.10.12
torch 2.2.0.1
torchaudio 2.2.0.1
torchvision 0.17.0.1
certifi 2024.2.2
charset-normalizer 3.3.2
filelock 3.13.1
fsspec 2024.2.0
idna 3.6
Jinja2 3.1.3
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.2.1
numpy 1.26.4
pillow 10.2.0
pip 24.0
PySocks 1.7.1
requests 2.31.0
setuptools 69.1.1
sympy 1.12
typing_extensions 4.10.0
urllib3 2.2.1
wheel 0.42.0

pytorch-ubuntu:2.2.0.1-22.04-cu121

Component Version
gcc 11.4.0
Python 3.10.12
torch 2.2.0.1+cu121
torchaudio 2.2.0.1+cu121
torchvision 0.17.0.1+cu121
triton 2.2.0
cuda-compat-12-1 530.30.02
cuda-cudart-12-1 12.1.105
cuda-cupti-12-1 12.1.105
cuda-libraries-12-1 12.1.1
cuda-nvrtc-12-1 12.1.105
cuda-nvtx-12-1 12.1.105
cuda-opencl-12-1 12.1.105
libcublas-12-1 12.1.3.1
libcufft-12-1 11.0.2.54
libcufile-12-1 1.6.1.9
libcurand-12-1 10.3.2.106
libcusolver-12-1 11.4.5.107
libcusparse-12-1 12.1.0.106
libnpp-12-1 12.1.0.40
libnvjitlink-12-1 12.1.105
libnvjpeg-12-1 12.2.0.2
libnccl2 2.19.3
libcudnn8 8.9.2.26
certifi 2024.2.2
charset-normalizer 3.3.2
filelock 3.13.1
fsspec 2024.2.0
idna 3.6
Jinja2 3.1.3
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.2.1
numpy 1.26.4
pillow 10.2.0
pip 24.0
PySocks 1.7.1
requests 2.31.0
setuptools 69.1.1
sympy 1.12
typing_extensions 4.10.0
urllib3 2.2.1
wheel 0.42.0

pytorch-ubuntu:2.2.0.1-22.04-cu118

Component Version
gcc 11.4.0
Python 3.10.12
torch 2.2.0.1+cu118
torchaudio 2.2.0.1+cu118
torchvision 0.17.0.1+cu118
triton 2.2.0+cu118
cuda-compat-11-8 520.61.05
cuda-cudart-11-8 11.8.89
cuda-cupti-11-8 11.8.87
cuda-libraries-11-8 11.8.0
cuda-nvrtc-11-8 11.8.89
cuda-nvtx-11-8 11.8.86
libcublas-11-8 11.11.3.6
libcufft-11-8 10.9.0.58
libcufile-11-8 1.4.0.31
libcurand-11-8 10.3.0.86
libcusolver-11-8 11.4.1.48
libcusparse-11-8 11.7.5.86
libnpp-11-8 11.8.0.86
libnvjpeg-11-8 11.9.0.86
nvidia-nccl-cu11 2.19.3
libcudnn8 8.9.2.26
certifi 2024.2.2
charset-normalizer 3.3.2
filelock 3.13.1
fsspec 2024.2.0
idna 3.6
Jinja2 3.1.3
MarkupSafe 2.1.5
mpmath 1.3.0
networkx 3.2.1
numpy 1.26.4
pillow 10.2.0
pip 24.0
PySocks 1.7.1
requests 2.31.0
setuptools 69.1.1
sympy 1.12
typing_extensions 4.10.0
urllib3 2.2.1
wheel 0.42.0