Hacker News Reader

Comment by belltaco | Hacker News Reader

Article says it needs 5.1GB of Graphics RAM.

Does any one know how much data download and disk storage does it need?

The v1.3 model weighs in at 4.3 GB. There's an additional download of 1.6 GB of other models due to usage of huggingface's transformers (only once on startup). And the conda env takes another 6 GBs due to pytorch and cuda.

Larger images will require (much) more than 5.1 GB. In my case, a target resolution of 768x384 (landscape) with a batch size of 1 will max out my 12GB card, an RTX3080Ti.

I think this is a good time to ask if anyone is working on parallelizing machine learning compute anymore? For at-home computation like this it seems like it would be a lot better to allow people to stack a few cheaper GPUs rather than having to pony up thousands of dollars for ML-oriented beast cards to be able to do things like generate large images.

AI upscaling will solve everything ;)

I've generated some remarkably good-looking print quality images by upscaling 512x512 sources

What do you use for upscaling? Standard software like Photoshop or Affinity, etc. or more dedicated software? Any recommendations for options, etc.?

Recently impressed by https://replicate.com/nightmareai/latent-sr but otherwise - Cupscale

Here's a particularly impressive result I got from the former (considering it's not especially optimized for "vector art" enlargement): https://twitter.com/andybak/status/1558737805546749953?s=20&...

For videos in particular, if you don't mind shelling out cash, the current go-to (at least according to various AI discord servers I'm on) for AI animation nerds is currently Topaz upscaler. There are free alternatives but I've yet to see any of them work as well as Topaz, though I'm sure that will change soon. For interpolating frames Flowframes is "free" (new features if you join the Patreon) and is IMO very good.

I've seen a number of 80s/90s VHS recordings of concerts being uploaded to YouTube in 4K (using Topaz) and they look like they were recorded that way, truly amazing. I do hear it can be a bit of work though getting the settings right.

If you read directly from the site. The requirements for the graphic card are 10 VRAM as a minimum. Because it's ruins locally you don't need to download anything apart from the initial model, this applies to the disk space too.

Reply on news.ycombinator.com