bitnami-containers/bitnami/pytorch

Bitnami Secure Image for PyTorch

What is PyTorch?

PyTorch is a deep learning platform that accelerates the transition from research prototyping to production deployment. Bitnami image includes Torchvision for specific computer vision support.

Overview of PyTorch Trademarks: This software listing is packaged by Bitnami. The respective trademarks mentioned in the offering are owned by the respective companies, and use of them does not imply any affiliation or endorsement.

TL;DR

docker run -it --name pytorch bitnami/pytorch

Why use Bitnami Secure Images?

Those are hardened, minimal CVE images built and maintained by Bitnami. Bitnami Secure Images are based on the cloud-optimized, security-hardened enterprise OS Photon Linux. Why choose BSI images?

• Hardened secure images of popular open source software with Near-Zero Vulnerabilities
• Vulnerability Triage & Prioritization with VEX Statements, KEV and EPSS Scores
• Compliance focus with FIPS, STIG, and air-gap options, including secure bill of materials (SBOM)
• Software supply chain provenance attestation through in-toto
• First class support for the internet’s favorite Helm charts

Each image comes with valuable security metadata. You can view the metadata in our public catalog here. Note: Some data is only available with commercial subscriptions to BSI.

If you are looking for our previous generation of images based on Debian Linux, please see the Bitnami Legacy registry.

Why use a non-root container?

Non-root container images add an extra layer of security and are generally recommended for production environments. However, because they run as a non-root user, privileged tasks are typically off-limits. Learn more about non-root containers in our docs.

Supported tags and respective Dockerfile links

Learn more about the Bitnami tagging policy and the difference between rolling tags and immutable tags in our documentation page.

You can see the equivalence between the different tags by taking a look at the tags-info.yaml file present in the branch folder, i.e bitnami/ASSET/BRANCH/DISTRO/tags-info.yaml.

Subscribe to project updates by watching the bitnami/containers GitHub repo.

Get this image

The recommended way to get the Bitnami Pytorch Docker Image is to pull the prebuilt image from the Docker Hub Registry.

docker pull bitnami/pytorch:latest

To use a specific version, you can pull a versioned tag. You can view the list of available versions in the Docker Hub Registry.

docker pull bitnami/pytorch:[TAG]

If you wish, you can also build the image yourself by cloning the repository, changing to the directory containing the Dockerfile and executing the docker build command. Remember to replace the APP, VERSION and OPERATING-SYSTEM path placeholders in the example command below with the correct values.

git clone https://github.com/bitnami/containers.git
cd bitnami/APP/VERSION/OPERATING-SYSTEM
docker build -t bitnami/APP:latest .

Entering the REPL

By default, running this image will drop you into the Python REPL, where you can interactively test and try things out with PyTorch in Python.

docker run -it --name pytorch bitnami/pytorch

Configuration

Running your PyTorch app

The default work directory for the PyTorch image is /app. You can mount a folder from your host here that includes your PyTorch script, and run it normally using the python command.

docker run -it --name pytorch -v /path/to/app:/app bitnami/pytorch \
  python script.py

Running a PyTorch app with package dependencies

If your PyTorch app has a requirements.txt defining your app's dependencies, you can install the dependencies before running your app.

docker run -it --name pytorch -v /path/to/app:/app bitnami/pytorch \
  sh -c "conda install -y --file requirements.txt && python script.py"

Further Reading:

• pytorch documentation
• conda documentation

FIPS configuration in Bitnami Secure Images

The Bitnami PyTorch Docker image from the Bitnami Secure Images catalog includes extra features and settings to configure the container with FIPS capabilities. You can configure the next environment variables:

• OPENSSL_FIPS: whether OpenSSL runs in FIPS mode or not. yes (default), no.

Maintenance

Upgrade this image

Bitnami provides up-to-date versions of PyTorch, including security patches, soon after they are made upstream. We recommend that you follow these steps to upgrade your container.

Step 1: Get the updated image

docker pull bitnami/pytorch:latest

or if you're using Docker Compose, update the value of the image property to bitnami/pytorch:latest.

Step 2: Remove the currently running container

docker rm -v pytorch

or using Docker Compose:

docker-compose rm -v pytorch

Step 3: Run the new image

Re-create your container from the new image.

docker run --name pytorch bitnami/pytorch:latest

or using Docker Compose:

docker-compose up pytorch

Notable changes

1.9.0-debian-10-r3

This version removes miniconda in favour of pip. This creates a smaller container and least prone to security issues. Users extending this container with other packages will need to switch from conda to pip commands.

Using docker-compose.yaml

Please be aware this file has not undergone internal testing. Consequently, we advise its use exclusively for development or testing purposes. For production-ready deployments, we highly recommend utilizing its associated Bitnami Helm chart.

If you detect any issue in the docker-compose.yaml file, feel free to report it or contribute with a fix by following our Contributing Guidelines.

Contributing

We'd love for you to contribute to this Docker image. You can request new features by creating an issue or submitting a pull request with your contribution.

Issues

If you encountered a problem running this container, you can file an issue. For us to provide better support, be sure to fill the issue template.

License

Copyright © 2025 Broadcom. The term "Broadcom" refers to Broadcom Inc. and/or its subsidiaries.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.