CHchengzeyi·Nov 30, 2023Submission
Stable-fast for SD inference: Faster than AITemplate, On par with TensorRT
2 points · Open discussionchengzeyi
stable-fast
Python2023–2023
chengzeyi / stable-fast
stable-fast
https://wavespeed.ai/ Best inference performance optimization framework for HuggingFace Diffusers on NVIDIA GPUs.
2Distinct people
6Mentions
0Explicit recommenders
1.3KGitHub stars
What to know before exploring
What it is
https://wavespeed.ai/ Best inference performance optimization framework for HuggingFace Diffusers on NVIDIA GPUs.
GitHub
Where it is used
No documented product use has been verified yet.
What people are saying
No recommendation phrase was found in the indexed comments.
No classified warning is attached to this page. Read the discussions before treating silence as approval.
Also in these discussions
No other repository is repeatedly linked in the same discussions.
This week on Hacker News
Nov 7–Nov 14 UTC · Direct repository links and separately labeled related project stories. Indexed through 11/14/2026, 3:56:07 PM UTC.
No direct repository links were found in the indexed HN discussions over the past seven days. Older discussions are available below.
Mentions over time
HN mentions by week6 dated mentions
Observed mentions: 2023-10-17 10:13:20 UTC to 2023-11-30 07:27:11 UTC. Weeks start Monday, UTC.
Counts come from the repository’s dated source evidence. Index cutoff: 2026-11-02 08:16:59 UTC.
View mention counts and date ranges
| Period | Start (UTC) | End (exclusive, UTC) | Mentions | Submissions | Comments |
|---|---|---|---|---|---|
| Oct 16 | 2023-10-16 00:00:00 UTC | 2023-10-23 00:00:00 UTC | 1 | 1 | 0 |
| Oct 23 | 2023-10-23 00:00:00 UTC | 2023-10-30 00:00:00 UTC | 0 | 0 | 0 |
| Oct 30 | 2023-10-30 00:00:00 UTC | 2023-11-06 00:00:00 UTC | 0 | 0 | 0 |
| Nov 6 | 2023-11-06 00:00:00 UTC | 2023-11-13 00:00:00 UTC | 1 | 0 | 1 |
| Nov 13 | 2023-11-13 00:00:00 UTC | 2023-11-20 00:00:00 UTC | 2 | 0 | 2 |
| Nov 20 | 2023-11-20 00:00:00 UTC | 2023-11-27 00:00:00 UTC | 0 | 0 | 0 |
| Nov 27 | 2023-11-27 00:00:00 UTC | 2023-12-04 00:00:00 UTC | 2 | 1 | 1 |
What Hacker News says
6 sourcesSDXL and ControlNet are already optimized, if thats what you mean: https://github.com/chengzeyi/stable-fast
(Note the links to various SD compilers).
But the whole field is moving so fast that people aren't even adopting the compilers and optimized implementations at large.
In SDXL Turbo: A Real-Time Text-to-Image Generation Model> already compiling
Hmm, well if you mean torch.compile, y'all should still check out stable-fast, which is claiming ~16ms/iter on a 4090, twice that of torch.compile:
https://github.com/chengzeyi/stable-fast#rtx-4090-512x512-ba...
In Show HN: Realtime LCM toolNow combine this with an optimized SD implementation, like:
https://github.com/chengzeyi/stable-fast
Or AITemplate, and you are at 15FPS on a larger consumer GPU. 10 with a controlnet you can use for some motion consistency.
In Generate images fast with SD 1.5 while typing on GradioSince I am sending you down the rabbit hole anyway, you should check out stable-fast:
https://github.com/chengzeyi/stable-fast
It's the most promising "fast" and flexible stable diffusion implementation akin to this paper or vLLM that I know of. It doesn't have as many caveats as other implementations, like AITemplate (which is basically Turing+ and linux only) or torch.compile (basically no support for changing inputs/loras, and super long compilation on every startup/change).
In S-LoRA: Serving Concurrent LoRA AdaptersStable Fast: Lightweight Inference Optimization Library for Stable Diffusion
2 points · Open discussion