Notebook Containers

Check out the ML Showcase for a list of projects you can fork into your account.

Base Containers

When you launch a Notebook, it runs inside a container preloaded with the notebook files and dependencies. The following is a list of containers that Paperspace maintains:

Name

Description

Container Tag

URL

Fast.ai

Paperspace's Fast.ai template is built for getting up and running with the enormously popular Fast.ai online MOOC.

paperspace/fastai:2.0-CUDA9.2-fastbook-v0.1.0

GitHub

TensorFlow 2.4.1

TensorFlow 2 with GPU support.

tensorflow/tensorflow:2.4.1-gpu-jupyter

****DockerHub

NVIDIA RAPIDS

NVIDIA's library to execute end-to-end data science and analytics pipelines.

nvcr.io/nvidia/rapidsai/rapidsai:0.18-cuda11.0-base-centos7

NVIDIA

PyTorch

Latest PyTorch release (1.8) with GPU support.

nvcr.io/nvidia/pytorch:21.02-py3

DockerHub

Hugging Face Transformers

A state-of-the-art NLP library from Hugging Face

paperspace/transformers-gpu:0.4.0

DockerHub

Other Containers

Name

Description

Container Tag

URL

TensorFlow (1.14 GPU)

Official docker images for deep learning framework TensorFlow

paperspace/dl-containers:tensorflow1140-py36-cu100-cdnn7-jupyter

DockerHub

Analytics Vidhya CV

Analytics Vidhya container

jalfaizy/cv_docker:latest

GitHub

JupyterLab Data Science Stack

Jupyter Notebook Data Science Stack

jupyter/datascience-notebook

DockerHub

JupyterLab Data R Stack

Jupyter Notebook R Stack

jupyter/r-notebook

DockerHub

Custom Containers

The custom containers feature lets you pull your own image from a container registry such as Docker Hub. This section will help you prepare a custom Docker container to use with Gradient, show you how to bring that Container into Gradient, and create a notebook with your custom container. We recommend using Docker to get the container image from your system to Gradient. You must run jupyter on port 8888 and allow connections from ip address 0.0.0.0. To support our notebook IDE you must also include the following flags: --no-browser --NotebookApp.trust_xheaders=True --NotebookApp.disable_check_xsrf=False --NotebookApp.allow_remote_access=True --NotebookApp.allow_origin='*' if your command is jupyter notebook. If your command is jupyter lab use these instead --no-browser --LabApp.trust_xheaders=True --LabApp.disable_check_xsrf=False --LabApp.allow_remote_access=True --LabApp.allow_origin='*'.

Required field:

  • Container Name = Path and tags of image, e.g., ufoym/deepo:all-jupyter-py36

Optional fields:

  • Registry Username = your private container registry username; can be left blank for public images

  • Registry Password = your private container registry password; can be left blank for public images

  • Command = must be Jupyter compatible; defaults to jupyter notebook --allow-root --ip=0.0.0.0 --no-browser --NotebookApp.trust_xheaders=True --NotebookApp.disable_check_xsrf=False --NotebookApp.allow_remote_access=True --NotebookApp.allow_origin='*' if left blank

  • Container user = optional user; defaults to 'root' if left blank