monster/Qwen-Image-Layered-MRP-MLX
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Sort performance benchmarks based on performance

Browse Source
This commit is contained in:
filipstrand 2024-11-16 16:06:44 +01:00
parent c94ab4749d
commit 2e297a068e
1 changed files with 10 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
README.md
View File

@ -277,7 +277,7 @@ For more options on how to configure MFLUX, please see [generate.py](src/mflux/g
### ⏱️ Image generation speed (updated)
These numbers are based on the non-quantized `schnell` model, with the configuration provided in the code snippet below.
These numbers are based on the *non-quantized* `schnell` model, with the configuration provided in the code snippet below.
To time your machine, run the following:
```sh
time mflux-generate \
@ -291,21 +291,24 @@ time mflux-generate \
| Device | User | Reported Time | Notes |
|--------------------|------------------------------------------------------------------------------------------------------------------------------------|---------------|---------------------------|
| M2 Ultra | [@awni](https://x.com/awnihannun/status/1823515121827897385) | <15s | |
| M4 Max | [@ivanfioravanti](https://gist.github.com/awni/a67d16d50f0f492d94a10418e0592bde?permalink_comment_id=5153531#gistcomment-5153531) | ~19s | |
| M3 Max | [@karpathy](https://gist.github.com/awni/a67d16d50f0f492d94a10418e0592bde?permalink_comment_id=5153531#gistcomment-5153531) | ~20s | |
| M2 Ultra | [@awni](https://x.com/awnihannun/status/1823515121827897385) | <15s | |
| 2023 M2 Max (96GB) | [@explorigin](https://github.com/filipstrand/mflux/issues/6) | ~25s | |
| 2021 M1 Pro (16GB) | [@qw-in](https://github.com/filipstrand/mflux/issues/7) | ~175s | Might freeze your mac |
| 2023 M3 Pro (36GB) | [@kush-gupt](https://github.com/filipstrand/mflux/issues/11) | ~80s | |
| 2020 M1 (8GB) | [@mbvillaverde](https://github.com/filipstrand/mflux/issues/13) | ~335s | With resolution 512 x 512 |
| 2022 M1 MAX (64GB) | [@BosseParra](https://x.com/BosseParra/status/1826191780812877968) | ~55s | |
| 2023 M2 Pro (32GB) | [@leekichko](https://github.com/filipstrand/mflux/issues/85) | ~54s | |
| 2021 M1 Pro (32GB) | @filipstrand | ~160s | |
| 2022 M1 MAX (64GB) | [@BosseParra](https://x.com/BosseParra/status/1826191780812877968) | ~55s | |
| 2023 M2 Max (32GB) | @filipstrand | ~70s | |
| 2023 M3 Pro (36GB) | [@kush-gupt](https://github.com/filipstrand/mflux/issues/11) | ~80s | |
| 2021 M1 Pro (32GB) | @filipstrand | ~160s | |
| 2021 M1 Pro (16GB) | [@qw-in](https://github.com/filipstrand/mflux/issues/7) | ~175s | Might freeze your mac |
| 2020 M1 (8GB) | [@mbvillaverde](https://github.com/filipstrand/mflux/issues/13) | ~335s | With resolution 512 x 512 |
*Note that these numbers includes starting the application from scratch, which means doing model i/o, setting/quantizing weights etc.
If we assume that the model is already loaded, you can inspect the image metadata using `exiftool image.png` and see the total duration of the denoising loop (excluding text embedding).*
*These benchmarks are not very scientific and is only intended to give ballpark numbers. They were performed during different times with different MFLUX and MLX-versions etc. Additional hardware information such as number of GPU cores, Mac device etc. are not always known.*
### ↔️ Equivalent to Diffusers implementation
There is only a single source of randomness when generating an image: The initial latent array.