v1.2.1

[rkhamilton]

This release fixes a couple of regressions in v1.2.0.
Important note: this release also changes the way restyle_video* works, speeding it up by 300% or more, and may yield videos that are more consistent from frame to frame. However, the video may look slightly different, and you may find that you need to run for a few more iterations to get the same look. Previously, each frame started with a fresh re-load of the VQGAN model. Now, the old model is kept from frame to frame. Please let me know how this works for you.

New features:

• Video metadata is encoded by the encode_video function in the title (text prompts) and comment (generator parameters) fields.

Bug Fixes

• generate.restyle_video* functions no longer re-load the VQGAN network each frame, which results in a 300% speed-up in running this function. This means that training doesn't start over each frame, so the output will look somewhat different than in earlier versions.
• generate functions no longer throw a warning when the output file argument doesn't have an extension.
• v1.2.0 introduced a bug where images were saved to output/output/filename. This is fixed.