Skip to content

Commit d1eb14b

Browse files
stevhliustevhliu
authored
[docs] Lu lambdas (huggingface#5602)
lu lambdas
1 parent 5c75a5f commit d1eb14b

File tree

2 files changed

+3
-12
lines changed

2 files changed

+3
-12
lines changed

‎docs/source/en/api/pipelines/stable_diffusion/overview.md‎

Lines changed: 0 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -36,10 +36,8 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
3636
<th class="px-4 py-2 font-medium text-gray-900 text-left">
3737
Space
3838
</th>
39-
4039
</tr>
4140
</thead>
42-
4341
<tbody class="divide-y divide-gray-200">
4442
<tr>
4543
<td class="px-4 py-2 text-gray-700">
@@ -49,7 +47,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
4947
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/stabilityai/stable-diffusion"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
5048
</td>
5149
</tr>
52-
5350
<tr>
5451
<td class="px-4 py-2 text-gray-700">
5552
<a href="./img2img">StableDiffusionImg2Img</a>
@@ -58,7 +55,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
5855
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/huggingface/diffuse-the-rest"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
5956
</td>
6057
</tr>
61-
6258
<tr>
6359
<td class="px-4 py-2 text-gray-700">
6460
<a href="./inpaint">StableDiffusionInpaint</a>
@@ -67,7 +63,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
6763
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/runwayml/stable-diffusion-inpainting"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
6864
</td>
6965
</tr>
70-
7166
<tr>
7267
<td class="px-4 py-2 text-gray-700">
7368
<a href="./depth2img">StableDiffusionDepth2Img</a>
@@ -76,7 +71,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
7671
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/radames/stable-diffusion-depth2img"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
7772
</td>
7873
</tr>
79-
8074
<tr>
8175
<td class="px-4 py-2 text-gray-700">
8276
<a href="./image_variation">StableDiffusionImageVariation</a>
@@ -85,7 +79,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
8579
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/lambdalabs/stable-diffusion-image-variations"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
8680
</td>
8781
</tr>
88-
8982
<tr>
9083
<td class="px-4 py-2 text-gray-700">
9184
<a href="./stable_diffusion_safe">StableDiffusionPipelineSafe</a>
@@ -94,7 +87,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
9487
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/AIML-TUDA/unsafe-vs-safe-stable-diffusion"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
9588
</td>
9689
</tr>
97-
9890
<tr>
9991
<td class="px-4 py-2 text-gray-700">
10092
<a href="./stable_diffusion_2">StableDiffusion2</a>
@@ -103,7 +95,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
10395
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/stabilityai/stable-diffusion"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
10496
</td>
10597
</tr>
106-
10798
<tr>
10899
<td class="px-4 py-2 text-gray-700">
109100
<a href="./stable_diffusion_xl">StableDiffusionXL</a>
@@ -112,7 +103,6 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
112103
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/RamAnanth1/stable-diffusion-xl"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
113104
</td>
114105
</tr>
115-
116106
<tr>
117107
<td class="px-4 py-2 text-gray-700">
118108
<a href="./latent_upscale">StableDiffusionLatentUpscale</a>
@@ -121,14 +111,12 @@ The table below summarizes the available Stable Diffusion pipelines, their suppo
121111
<td class="px-4 py-2"><a href="https://huggingface.co/spaces/huggingface-projects/stable-diffusion-latent-upscaler"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue"/></a>
122112
</td>
123113
</tr>
124-
125114
<tr>
126115
<td class="px-4 py-2 text-gray-700">
127116
<a href="./upscale">StableDiffusionUpscale</a>
128117
</td>
129118
<td class="px-4 py-2 text-gray-700">super-resolution</td>
130119
</tr>
131-
132120
<tr>
133121
<td class="px-4 py-2 text-gray-700">
134122
<a href="./ldm3d_diffusion">StableDiffusionLDM3D</a>

‎docs/source/en/api/pipelines/stable_diffusion/stable_diffusion_xl.md‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -20,6 +20,9 @@ The abstract from the paper is:
2020

2121
## Tips
2222

23+
- Using SDXL with a DPM++ scheduler for less than 50 steps is known to produce [visual artifacts](https://github.com/huggingface/diffusers/issues/5433) because the solver becomes numerically unstable. To fix this issue, take a look at this [PR](https://github.com/huggingface/diffusers/pull/5541) which recommends for ODE/SDE solvers:
24+
- set `use_karras_sigmas=True` or `lu_lambdas=True` to improve image quality
25+
- set `euler_at_final=True` if you're using a solver with uniform step sizes (DPM++2M or DPM++2M SDE)
2326
- Most SDXL checkpoints work best with an image size of 1024x1024. Image sizes of 768x768 and 512x512 are also supported, but the results aren't as good. Anything below 512x512 is not recommended and likely won't for for default checkpoints like [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0).
2427
- SDXL can pass a different prompt for each of the text encoders it was trained on. We can even pass different parts of the same prompt to the text encoders.
2528
- SDXL output images can be improved by making use of a refiner model in an image-to-image setting.

0 commit comments

Comments
 (0)