Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update setup docs #1563

Merged
merged 2 commits into from
Nov 28, 2022
Merged

Update setup docs #1563

merged 2 commits into from
Nov 28, 2022

Conversation

lewfish
Copy link
Contributor

@lewfish lewfish commented Nov 27, 2022

Overview

This PR updates the setup docs to clarify the environment needed to install RV using pip, and describes how to run it directly on EC2 among some other changes.

Notes

I was able to install and run directly on a Mac without using Docker, except for an inability to use num_workers > 0 which I was unable to easily fix. See #1562

Closes #1542

This is needed if running directly on an EC2 instance or on a Mac.
docs/setup.rst Outdated
Comment on lines 19 to 23
This has been shown to work in the following environment. Variations on this environment may or may not work.
* Ubuntu Linux 20.04

* Python 3.9

* CUDA 11.6 and NVIDIA Driver 510.47.03 (for GPU support)
Copy link
Collaborator

@AdeelH AdeelH Nov 28, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

rst requires a blank line before the bullet list starts. Also there shouldn't be new lines between list items.

docs/setup.rst Outdated

Raster Vision also runs on macOS version 12.1, except that the ``num_workers`` for the ``DataLoader`` will need to be set to 0 due to an issue with mulitprocessing on Macs with Python >= 3.8. It will also be necessary to install GDAL 3.5.2 prior to installing Raster Vision, which isn't necessary on Linux.

Raster Vision has not been tested with Windows, and will probably run into problems.
Copy link
Collaborator

@AdeelH AdeelH Nov 28, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Raster Vision has not been tested with Windows, and will probably run into problems.

Might want to put this in a warning box.


The simplest way to run Raster Vision on an AWS GPU is by starting a GPU-enabled EC2 instance such as a p3.2xlarge using the `Deep Learning AMI <https://aws.amazon.com/machine-learning/amis/>`_. We have tested this using the "Deep Learning AMI GPU PyTorch 1.11.0 (Ubuntu 20.04)" with id ``ami-0c968d7ef8a4b0c34``. After SSH'ing into the instance, Raster Vision can be installed with ``pip``, and code can be transfered to this instance with a tool such as ``rsync``.

.. _aws batch setup:
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The reference from line 183 to this subsection is broken.

@AdeelH
Copy link
Collaborator

AdeelH commented Nov 28, 2022

It might also make sense to move the pip installation instruction to the top in the README.

* Prioritizes pip setup over Docker
* Adds more information on the environment needed to directly run RV without Docker
* Adds instructions on running on EC2 instance
* Some minor reorganization
@lewfish
Copy link
Contributor Author

lewfish commented Nov 28, 2022

It might also make sense to move the pip installation instruction to the top in the README.

I made a note to do this in #1543 since the README currently refers to the last released version of Raster Vision.

I made the other suggested changes.

@codecov
Copy link

codecov bot commented Nov 28, 2022

Codecov Report

Merging #1563 (a1e526b) into master (bc6357b) will decrease coverage by 0.04%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##           master    #1563      +/-   ##
==========================================
- Coverage   75.78%   75.73%   -0.05%     
==========================================
  Files         193      193              
  Lines        9226     9229       +3     
==========================================
- Hits         6992     6990       -2     
- Misses       2234     2239       +5     
Impacted Files Coverage Δ
...ch_learner/rastervision/pytorch_learner/learner.py 71.00% <0.00%> (-0.38%) ⬇️
...er/dataset/visualizer/classification_visualizer.py 94.73% <0.00%> (-5.27%) ⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

@lewfish lewfish merged commit ed80e87 into master Nov 28, 2022
@lewfish lewfish deleted the lf/setup branch November 28, 2022 17:12
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Update docs on setting up RV outside of Docker image
2 participants