--- tags: Concept, CCS, ZH title: CCS Image-PyTorch | zh GA: UA-155999456-1 --- {%hackmd @docsharedstyle/default %} {%hackmd @docsharedstyle/twccheader-zh %} {%hackmd @docsharedstyle/ccs-image-icon-size %} # <img class="ccsimgicon" src="https://cos.twcc.ai/SYS-MANUAL/uploads/upload_60c298573392b441958d75106686c11d.png"> PyTorch TWCC 提供 NGC optmized PyTorch 10 種立即可用的工作環境。PyTorch 是 GPU 加速的張量運算框架，並以 Python 為前端語言。使用常見的 Python 函式庫(如 NumPy、SciPy 和 Cython) 便能輕鬆地擴展多項功能。在功能神經網路層，採用以磁帶系統 (tape-based system) 為基礎而建構的自動微分庫 (automatic differentiation)，為深度學習框架提供高度的靈活性和速度，並提供類似 NumPy 的加速功能。PyTorch 也包含標准定義下的神經網路層、深度學習最佳化引擎、數據載入功能，並支援使用多個GPU與多個節點。能立即執行運算，無需在靜止的流程圖中排隊等候，提升了易用性和絕佳的除錯體驗。 ## <i class="fa fa-sticky-note" aria-hidden="true"></i> <span class="ccsimglist">映像檔版本資訊</span> | Container Version | Ubuntu | CUDA Toolkit | PyTorch Version | TensorRT | cuDNN | TWCC Release Date | |-------------------|--------|---------------------|------------------------|------------------|----------|-------------------| | pytorch-25.05-py3 | 24.04 | NVIDIA CUDA 12.9.0 | 2.8.0a0+5228986c39 | TensorRT 10.10.0.31 | - | 16JUN25 | | pytorch-25.02-py3 | 24.04 | NVIDIA CUDA 12.8.0.38 | 2.7.0a0+6c5496875 | TensorRT 10.8.0.43 | 9.7.1.26 | 28MAR25 | | pytorch-24.11-py3 | 24.04 | NVIDIA CUDA 12.6.3 | 2.6.0a0+df5bbc0911 | TensorRT 10.6.0.26 | 9.5.1.17 | 27DEC24 | | pytorch-24.08-py3 | 24.04 | NVIDIA CUDA 12.6 | 2.5.0a0+872d972e41 | TensorRT 10.3.0.26 | 9.3.0.75 | 27MAY24 | | pytorch-24.05-py3 | 24.04 | NVIDIA CUDA 12.4.1 | 2.4.0a0+07cecf4168 | TensorRT 10.0.1.6 | 9.1.0.70 | 27JUN24 | | pytorch-24.02-py3 | 22.04 | NVIDIA CUDA 12.3.2 | 2.3.0a0+ebedce2 | TensorRT 8.6.3 | 9.0.0.306 | 29MAR24 | | pytorch-23.11-py3 | 22.04 | NVIDIA CUDA 12.3.0 | 2.2.0a0+6a974bee | TensorRT 8.6.1.6 | 8.9.6 | 29DEC23 | | pytorch-23.08-py3 | 22.04 | NVIDIA CUDA 12.2.1 | 2.1.0a0+29c30b1 | TensorRT 8.6.1.6 | 8.9.4 | 28SEP23 | | pytorch-23.05-py3 | 22.04 | NVIDIA CUDA 12.1.1 | 2.0.0 | TensorRT 8.6.1.2 | 8.9.1.23 | 30JUN23 | | pytorch-23.02-py3 | 22.04 | NVIDIA CUDA 12.0.1 | 1.14.0a0+44da651 | TensorRT 8.5.3 | 8.7.0.84 | 24MAR23 | | pytorch-22.08-py3 | 20.04 | NVIDIA CUDA 11.7.1 | 1.13.0a0+4321be6 | TensorRT 8.4.2.4 | 8.5.0.96 | 30SEP22 | | pytorch-22.05-py3 | 20.04 | NVIDIA CUDA 11.7.0 | 1.12.0a0+8a1a938 | TensorRT 8.2.5.1 | 8.4.0.27 | 21JUN22 | | pytorch-22.02-py3 | 20.04 | NVIDIA CUDA 11.6.0 | 1.11.0a0+17540c5c | TensorRT 8.2.3 | 8.3.2.44 | 18MAY22 | | pytorch-21.11-py3 | 20.04 | NVIDIA CUDA 11.5.0 | 1.11.0a0+b6df043 | TensorRT 8.0.3.4 | 8.3.0.96 | 18JAN22 | | pytorch-21.08-py3 | 20.04 | NVIDIA CUDA 11.4.1 | 1.10.0a0+3fd9dcf | TensorRT 8.0.1.6 | 8.2.2.26 | 16SEP21 | | pytorch-21.06-py3 | 20.04 | NVIDIA CUDA 11.3.1 | 1.9.0a0+c3d40fd | TensorRT 7.2.3.4 | 8.2.1 | 29JUN21 | | pytorch-21.02-py3 | 20.04 | NVIDIA CUDA 11.2.0 | 1.8.0a0+52ea372 | TensorRT 7.2.2.3 | 8.1.0 | 27APR21 | <details> <summary>21.02 以前版本</summary>  </details> <br> :::info {%hackmd @docsharedstyle/note-zh %} `py3` 與 `py2` 為 Python 版本差異。 ::: <details class="docspoiler"> <summary><b>點我參考詳細版本套件資訊</b></summary> - [pytorch-25.05-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-25-05.html) - [pytorch-25.02-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-25-02.html) - [pytorch-24.11-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-24-11.html) - [pytorch-24.08-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-24-08.html) - [pytorch-24.05-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-24-05.html) - [pytorch-24.02-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-24-02.html) - [pytorch-23.11-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-23-11.html) - [pytorch-23.08-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-23-08.html) - [pytorch-23.05-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-23-05.html) - [pytorch-23.02-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel-23-02.html#rel-23-02) - [pytorch-22.08-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_22-08.html#rel_22-08) - [pytorch-22.05-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_22-05.html#rel_22-05) - [pytorch-22.02-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_22-02.html#rel_22-02) - [pytorch-21.11-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_21-11.html#rel_21-11) - [pytorch-21.08-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_21-08.html#rel_21-08) - [pytorch-21.06-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_21-06.html#rel_21-06) - [pytorch-21.02-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_21-02.html#rel_21-02) - [pytorch-20.11-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_20-11.html#rel_20-11) - [pytorch-20.08-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_20-08.html#rel_20-08) - [pytorch-20.06-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_20-06.html#rel_20-06) - [pytorch-20.02-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_20-02.html#rel_20-02) - [pytorch-19.11-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_19-11.html#rel_19-11) - [pytorch-19.08-py3](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_19-08.html#rel_19-08) - [pytorch-19.02-py3-v1](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_19-02.html#rel_19-02) - [pytorch-18.12-py3-v1](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_18-12.html#rel_18-12) - [pytorch-18.10-py3-v1](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_18-10.html#rel_18-10) - [pytorch-18.08-py3-v1](https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/rel_18.08.html#rel_18.08) </details>