Updates for PyTorch Lightning 2.0 release

[ejm714]

Configure accelerator and devices for Trainer

Updates how we set the accelerator and devices based on recent PyTorch Lightning changes.

pl.Trainer now accepts accelerator ("cpu", "gpu", etc."), strategy, and devices params instead of num_gpus. This PR adds a utility to configure accelerator and devices based on the number of user-specified gpus.

Pros:

• keeps the command line to the same limited set of options (e.g. just specifying number of gpus instead of exposing both the cpu/gpu option as well as number of devices/cores)
• allows us to validate user-provided input up front to ensure that training won't fail after the long video loading check

Cons

• doesn't allow the user to specify the number of cores; if a GPU is not used, all cores will be used
• doesn't support the new "mps" accelerator for apple silicon

Relevant links:

• GPU training
• Trainer docs
• About strategies

Closes #264
Closes #265

Other changes based on PTL 2.0 release

The new PTL release makes a bunch of non-backward compatible API changes. This PR also provides fixes for those and sets 2.0 as the floor.

• modifies the syntax for learning rate finding with the separation of Trainer and Tuner (Decouple Tuner from Trainer Lightning-AI/pytorch-lightning#16462)
• updates epoch end hook names and outputs per Remove memory-retaining epoch-end hooks Lightning-AI/pytorch-lightning#16520
• removes python 3.7 which isn't compatible with lightning 2.0

Bonus fixes

• adds a note in the depth estimation docs about the limitations for multiple individuals in the frame (which came up in a user email)
• only use DDP strategy for multi-gpu; this removes a bunch of unnecessary logging warnings about metric syncing across devices which arose from using DDP on a single GPU; we can take advantage of the new "auto" strategy in PTL 2.0

[netlify]

✅ Deploy Preview for silly-keller-664934 ready!

Name	Link
🔨 Latest commit	ba747c7
🔍 Latest deploy log	https://app.netlify.com/sites/silly-keller-664934/deploys/6422097983d24c00080bcf6e
😎 Deploy Preview	https://deploy-preview-266--silly-keller-664934.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site settings.

[github-actions]

🚀 Deployed on https://deploy-preview-266--silly-keller-664934.netlify.app

[codecov]

Codecov Report

Merging #266 (ba747c7) into master (a49b0d7) will decrease coverage by 0.1%.
The diff coverage is 87.5%.

Additional details and impacted files

@@           Coverage Diff            @@
##           master    #266     +/-   ##
========================================
- Coverage    87.6%   87.6%   -0.1%     
========================================
  Files          26      26             
  Lines        2155    2178     +23     
========================================
+ Hits         1889    1908     +19     
- Misses        266     270      +4     

Impacted Files	Coverage Δ
zamba/models/model_manager.py	84.1% <66.6%> (-0.3%)	⬇️
zamba/models/utils.py	96.1% <71.4%> (-3.9%)	⬇️
zamba/pytorch_lightning/utils.py	97.7% <100.0%> (+0.2%)	⬆️

... and 1 file with indirect coverage changes

[ejm714]

@pjbull tests are now passing, this is ready for your review when you have time 🙇

[pjbull]

Looks good to me!