huggingface/candle
1
0
Fork 0
mirror of https://github.com/huggingface/candle.git synced 2025-06-17 19:18:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add more mentions to SDXL Turbo in the readme. (#1397)

Browse Source
This commit is contained in:
Laurent Mazare
2023-12-03 11:41:21 +01:00
committed by GitHub
parent 37bf1ed012
commit 0738df5290
1 changed files with 16 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
candle-examples/examples/stable-diffusion/README.md
View File

@ -8,7 +8,7 @@ XL using Rust and [candle](https://github.com/huggingface/candle).
The `stable-diffusion` example is a conversion of The `stable-diffusion` example is a conversion of
[diffusers-rs](https://github.com/LaurentMazare/diffusers-rs) using candle [diffusers-rs](https://github.com/LaurentMazare/diffusers-rs) using candle
rather than libtorch. This implementation supports Stable Diffusion v1.5, v2.1, rather than libtorch. This implementation supports Stable Diffusion v1.5, v2.1,
as well as Stable Diffusion XL 1.0. as well as Stable Diffusion XL 1.0, and Turbo.
## Getting the weights ## Getting the weights
@ -23,16 +23,26 @@ cargo run --example stable-diffusion --release --features=cuda,cudnn \
-- --prompt "a cosmonaut on a horse (hd, realistic, high-def)" -- --prompt "a cosmonaut on a horse (hd, realistic, high-def)"
``` ```
The final image is named `sd_final.png` by default. The final image is named `sd_final.png` by default. The Turbo version is much
The default scheduler is the Denoising Diffusion Implicit Model scheduler (DDIM). The faster than previous versions, to give it a try add a `--sd-version turbo` flag,
original paper and some code can be found in the [associated repo](https://github.com/ermongroup/ddim). e.g.:
```bash
cargo run --example stable-diffusion --release --features=cuda,cudnn \
-- --prompt "a cosmonaut on a horse (hd, realistic, high-def) --sd-version turbo"
```
The default scheduler for the v1.5, v2.1 and XL 1.0 version is the Denoising
Diffusion Implicit Model scheduler (DDIM). The original paper and some code can
be found in the [associated repo](https://github.com/ermongroup/ddim).
The default scheduler for the XL Turbo version is the Euler Ancestral scheduler.
### Command-line flags ### Command-line flags
- `--prompt`: the prompt to be used to generate the image. - `--prompt`: the prompt to be used to generate the image.
- `--uncond-prompt`: the optional unconditional prompt. - `--uncond-prompt`: the optional unconditional prompt.
- `--sd-version`: the Stable Diffusion version to use, can be `v1-5`, `v2-1`, or - `--sd-version`: the Stable Diffusion version to use, can be `v1-5`, `v2-1`,
`xl`. `xl`, or `turbo`.
- `--cpu`: use the cpu rather than the gpu (much slower). - `--cpu`: use the cpu rather than the gpu (much slower).
- `--height`, `--width`: set the height and width for the generated image. - `--height`, `--width`: set the height and width for the generated image.
- `--n-steps`: the number of steps to be used in the diffusion process. - `--n-steps`: the number of steps to be used in the diffusion process.