6 ms·
That's really cool! I always thought you needed a good amount of GPU VRAM to generate images using SD. I wonder how fast would a consumer PC, with no GPU, gene
by boredemployee 3y ago
That's really cool! I always thought you needed a good amount of GPU VRAM to generate images using SD.
I wonder how fast would a consumer PC, with no GPU, generate an image with say 16gb of RAM?
- atrus 3y agoI was using a 6ish year old amd cpu with 16gigs of ram and generating a prompt would take about a half hour. Which is still massively impressive for what it is.
- londons_explore 3y agoUse a free GPU from google colab and you can do the same in about 15 seconds...
- boredemployee 3y agoDo you have a google colab link?
- hadlock 3y agoThere is no shortage of google collab stable diffusion tutorials on the web
- yjftsjthsd-h 3y agoWhich is why asking for one high quality starting point is such a useful question.
- luc4sdreyer 3y agoOn 21 April 2023 Google blocked usage of Stable Diffusion with a free account on colab. You need a paid plan to use it. Apparently there are ways around it, but I just switched to runpod.io. It's very cheap (around $0.80/h for a 4090 including storage) and having a real terminal is worth it.
- idiotsecant 3y agoyes, and if he does it on a paid machine with a better GPU it'll be even faster! While true, neither your statement or mine above is germane to the discussion. It wasn't about how long it takes. It's a discussion of how cool it is that it can be done on that machine at all.
- wsgeorge 3y agoOn an Apple M1 with 16gig RAM, without using Pytorch compiled to take advantage of Metal, it could take 12mins to generate an image with a tweet-length prompt. With Metal, it takes less than 60 seconds.
- microtonal 3y agoAnd PyTorch on the M1 (without Metal) uses the fast AMX matrix multiplication units (through the Accelerate Framework). The matrix multiplication on the M1 is on par with ~10 threads/cores of Ryzen 5900X. [1] https://github.com/danieldk/gemm-benchmark#example-results https://github.com/danieldk/gemm-benchmark#example-results
- ilkke 3y agoPrompt length shouldn't influence creation time, at least it didn't in any of the implementations I used. What is the resolution of your images and number of steps?
- wsgeorge 3y agoDefaults from the Huggingface repo, just copy-pasted. So, iirc 50 steps and the image is 512x512. Edit: confirmed. > Prompt length shouldn't influence creation time... Yeah, checks out with my experience too. Longer prompts were truncated.
- Filligree 3y agoSome tools (e.g. Automatic1111) are able to feed in longer prompts, but then the prompt length does affect the speed of inference. Albeit in 77 token increments.
- asynchronous 3y agoMetal is such an advantage, had no idea
- hospitalJail 3y agoWtf, my 4 year old, $400 crappy low wattage computer can generate a picture in a minute or two. DDIM, 12 steps.