repo stringclasses 1
value | github_id int64 1.27B 4.95B | github_node_id stringlengths 18 24 | number int64 8 14.3k | html_url stringlengths 49 53 | api_url stringlengths 59 63 | title stringlengths 1 402 | body stringlengths 1 62.9k ⌀ | state stringclasses 2
values | state_reason stringclasses 4
values | locked bool 2
classes | comments_count int64 0 99 | labels listlengths 0 5 | assignees listlengths 0 5 | created_at stringdate 2022-06-09 16:28:35 2026-07-22 10:32:58 | updated_at stringdate 2022-06-12 22:18:01 2026-07-22 10:32:58 | closed_at stringdate 2022-06-12 22:18:01 2026-07-21 13:17:28 ⌀ | author_association stringclasses 3
values | milestone_title stringclasses 0
values | snapshot_id stringclasses 149
values | extracted_at stringdate 2026-04-07 13:34:13 2026-07-22 10:35:16 | author_login stringlengths 3 28 | author_id int64 1.54k 296M | author_node_id stringlengths 12 20 | author_type stringclasses 3
values | author_site_admin bool 1
class |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
huggingface/diffusers | 2,242,755,860 | I_kwDOHa8MBc6Frb0U | 7,681 | https://github.com/huggingface/diffusers/issues/7681 | https://api.github.com/repos/huggingface/diffusers/issues/7681 | RuntimeError: Failed to import diffusers.pipelines.pipeline_utils because of the following error (look up to see its traceback): Failed to import diffusers.models.autoencoders.autoencoder_kl because of the following error (look up to see its traceback): module 'lib' has no attribute 'X509_V_FLAG_CB_ISSUER_CHECK' | ### Describe the bug
Getting error while importing DiffusionPipeline from Diffusers
```
Traceback (most recent call last):
File "/nfs/Vision/DataRW/anushka/prompt_yml/stable_cascade2.py", line 18, in <module>
from diffusers import DiffusionPipeline
File "<frozen importlib._bootstrap>", line 1039, in _h... | closed | completed | false | 1 | [
"bug"
] | [] | 2024-04-15T05:43:31Z | 2024-04-15T06:16:35Z | 2024-04-15T06:16:22Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | vedant123454 | 139,352,826 | U_kgDOCE5a-g | User | false |
huggingface/diffusers | 2,243,292,355 | I_kwDOHa8MBc6FtezD | 7,682 | https://github.com/huggingface/diffusers/issues/7682 | https://api.github.com/repos/huggingface/diffusers/issues/7682 | Issue while running SD1.5 on multiple less beefy GPUs | ### Describe the bug
I am trying to run [distributed inference](https://huggingface.co/docs/diffusers/main/en/training/distributed_inference#device-placement) for SD1.5 and SDXL on 2xGTX 1080 Ti. But facing some issues
### Reproduction
```
from diffusers import DiffusionPipeline
import torch
pip... | closed | completed | false | 18 | [
"bug"
] | [] | 2024-04-15T10:26:10Z | 2024-04-18T17:17:52Z | 2024-04-18T17:17:52Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | square-1111 | 31,164,802 | MDQ6VXNlcjMxMTY0ODAy | User | false |
huggingface/diffusers | 2,245,154,988 | I_kwDOHa8MBc6F0lis | 7,686 | https://github.com/huggingface/diffusers/issues/7686 | https://api.github.com/repos/huggingface/diffusers/issues/7686 | UNet2DModel - ValueError: cross_attention_dim must be specified for CrossAttnDownBlock2D | ### Describe the bug
When I load the model for my training, I just want to use the unconditional model. But I have this problem. Could anyone tell me how to solve this?
ValueError: cross_attention_dim must be specified for CrossAttnDownBlock2D
on
self.unet = UNet2DModel.from_pretrained(model_key, subfolder="u... | closed | completed | true | 2 | [
"bug",
"stale"
] | [] | 2024-04-16T05:58:02Z | 2024-06-29T13:25:34Z | 2024-06-29T13:25:34Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | phuongtrau | 63,243,678 | MDQ6VXNlcjYzMjQzNjc4 | User | false |
huggingface/diffusers | 2,245,408,011 | I_kwDOHa8MBc6F1jUL | 7,688 | https://github.com/huggingface/diffusers/issues/7688 | https://api.github.com/repos/huggingface/diffusers/issues/7688 | Training doesnt save models | ### Describe the bug
When training a diffusion model (text to image, regular SD1.5) I attempted to follow the tutorial in examples/text_to_image.
I had a bit of trouble with the libraries, so I ended up using docker and pulling SD1.5 fresh every time.
Now my training is running, but the provided script to run a trai... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-04-16T08:08:41Z | 2024-04-16T11:18:40Z | 2024-04-16T11:18:40Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | PLEXATIC | 16,229,856 | MDQ6VXNlcjE2MjI5ODU2 | User | false |
huggingface/diffusers | 2,245,817,206 | I_kwDOHa8MBc6F3HN2 | 7,689 | https://github.com/huggingface/diffusers/issues/7689 | https://api.github.com/repos/huggingface/diffusers/issues/7689 | Auth token is not passable to load_textual_inversion due to named argument conflicting with accepted kwargs | ### Describe the bug
You cannot provide an auth token value to the **kwargs of [``load_textual_inversion``](https://github.com/huggingface/diffusers/blob/cf6e0407e051467b480830d3ed97d2873b5019d3/src/diffusers/loaders/textual_inversion.py#L267) in ``src/diffusers/loaders/textual_inversion.py`` due to it being a name... | closed | completed | false | 7 | [
"bug",
"stale"
] | [] | 2024-04-16T11:35:20Z | 2025-01-14T21:48:36Z | 2025-01-14T21:48:36Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Teriks | 14,919,098 | MDQ6VXNlcjE0OTE5MDk4 | User | false |
huggingface/diffusers | 2,245,902,478 | I_kwDOHa8MBc6F3cCO | 7,690 | https://github.com/huggingface/diffusers/issues/7690 | https://api.github.com/repos/huggingface/diffusers/issues/7690 | OSError: Can't load tokenizer for 'stabilityai/stable-diffusion-xl-base-1.0' | ### Describe the bug
**code is :**
#DiffusionPipeline
import torch
from diffusers import StableDiffusionXLPipeline, StableDiffusionXLImg2ImgPipeline, AutoPipelineForText2Image
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
#加载SDXL base model
pipe = StableDiffusionXLPipeline.... | closed | completed | false | 5 | [
"bug",
"stale"
] | [] | 2024-04-16T12:19:19Z | 2024-06-29T13:24:45Z | 2024-06-29T13:24:45Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ppdk-data | 53,573,455 | MDQ6VXNlcjUzNTczNDU1 | User | false |
huggingface/diffusers | 2,247,498,339 | I_kwDOHa8MBc6F9hpj | 7,693 | https://github.com/huggingface/diffusers/issues/7693 | https://api.github.com/repos/huggingface/diffusers/issues/7693 | pipeline does not work with conversion to fp16 in the to cuda call | Hi,
i was using and doing some fine-tuning with the Pixart model.
and i found a problem that i could pin it down to:
this code works:
```
pipe2 = PixArtAlphaPipeline.from_pretrained("PixArt-alpha/PixArt-XL-2-512x512", torch_dtype=torch.float16)
pipe2.to("cuda")
image = pipe2(prompt, num_inference_steps=20).i... | open | null | false | 8 | [
"stale"
] | [] | 2024-04-17T06:36:27Z | 2024-12-25T15:06:42Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | raulc0399 | 156,006,486 | U_kgDOCUx4Vg | User | false |
huggingface/diffusers | 2,247,522,262 | I_kwDOHa8MBc6F9nfW | 7,694 | https://github.com/huggingface/diffusers/issues/7694 | https://api.github.com/repos/huggingface/diffusers/issues/7694 | Update ControlNetXS checkpoint keys | ### Describe the bug
@UmerHA We have failing slow tests due to the ControlNetXS checkpoints
https://github.com/huggingface/diffusers/actions/runs/8709419367/job/23889042563#step:7:124
I believe the keys need to be updated to reflect the new model.
### Reproduction
```python
from diffusers import Contr... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-04-17T06:52:37Z | 2024-04-22T08:24:32Z | 2024-04-22T08:24:32Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | DN6 | 7,529,846 | MDQ6VXNlcjc1Mjk4NDY= | User | false |
huggingface/diffusers | 2,247,529,921 | I_kwDOHa8MBc6F9pXB | 7,695 | https://github.com/huggingface/diffusers/issues/7695 | https://api.github.com/repos/huggingface/diffusers/issues/7695 | cross_attention_kwargs ['ip_adapter_masks'] are not expected by AttnProcessor2_0 and will be ignored. | ### Describe the bug
What error message is this, but it won't cause the task to fail, it will only keep printing，thank you。
### Reproduction
import torch
from diffusers import StableDiffusionXLPipeline
from diffusers.image_processor import IPAdapterMaskProcessor
from diffusers.utils import load_image
from transf... | closed | completed | false | 8 | [
"bug"
] | [] | 2024-04-17T06:57:32Z | 2024-10-15T05:51:16Z | 2024-10-09T02:24:16Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | cthulhu-tww | 79,293,568 | MDQ6VXNlcjc5MjkzNTY4 | User | false |
huggingface/diffusers | 1,401,151,943 | I_kwDOHa8MBc5Tg-HH | 770 | https://github.com/huggingface/diffusers/issues/770 | https://api.github.com/repos/huggingface/diffusers/issues/770 | Colab Cannot Start Training | ### Describe the bug
Import Libraries Fails
screenshot:
https://i.gyazo.com/ea42aead7543ac9847412ccc554dffda.png
Think it needs updating, cannot proceed to begin training due to failure Importing
### Reproduction
just start the Colab, it will fail on importing libraries. You can proceed but Training is not p... | closed | completed | false | 9 | [
"bug",
"stale"
] | [] | 2022-10-07T13:10:46Z | 2023-11-14T04:12:50Z | 2022-11-15T15:02:59Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | MushroomFleet | 7,318,371 | MDQ6VXNlcjczMTgzNzE= | User | false |
huggingface/diffusers | 2,249,704,182 | I_kwDOHa8MBc6GF8L2 | 7,706 | https://github.com/huggingface/diffusers/issues/7706 | https://api.github.com/repos/huggingface/diffusers/issues/7706 | Load custom pipeline cause CUDA out of memory | ### Describe the bug
```
inpaintingPipe = DiffusionPipeline.from_pretrained(
'runwayml/stable-diffusion-inpainting',
torch_dtype = torch.float16,
custom_pipeline = 'hd_painter'
)
```
<img width="674" alt="image" src="https://github.com/huggingface/diffusers/assets/1147704/e0be5098-027a-4b7a-864d-a... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2024-04-18T04:20:44Z | 2024-11-23T00:16:34Z | 2024-11-23T00:16:34Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | crapthings | 1,147,704 | MDQ6VXNlcjExNDc3MDQ= | User | false |
huggingface/diffusers | 2,249,937,422 | I_kwDOHa8MBc6GG1IO | 7,708 | https://github.com/huggingface/diffusers/issues/7708 | https://api.github.com/repos/huggingface/diffusers/issues/7708 | stable-diffusion-xl-base-1.0 does not appear to have a file named tokenizer/config.json | ### Describe the bug
when i finetuning SDXL1.0，--pretrained_model_name_or_path="./stable-diffusion-xl-base-1.0"，offline mode.
python train_dreambooth_lora_sdxl.py --pretrained_model_name_or_path="./stable-diffusion-xl-base-1.0/" --instance_prompt="In the conference room, an Asian person is sitting at his desk" --da... | closed | completed | false | 6 | [
"bug"
] | [] | 2024-04-18T07:14:33Z | 2024-12-06T01:24:11Z | 2024-04-19T06:11:10Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | HQ33 | 53,594,242 | MDQ6VXNlcjUzNTk0MjQy | User | false |
huggingface/diffusers | 1,401,188,514 | I_kwDOHa8MBc5ThHCi | 771 | https://github.com/huggingface/diffusers/issues/771 | https://api.github.com/repos/huggingface/diffusers/issues/771 | BF16 doesn't work with dreambooth | ### Describe the bug
Using bf16 with the dreambooth script throw an exception
### Reproduction
_No response_
### Logs
```shell
RuntimeError: "upsample_nearest2d_out_frame" not implemented for 'BFloat16'
```
### System Info
- `diffusers` version: 0.4.0.dev0
- Platform: Linux-5.15.0-47-generic-x86_64-with-glibc2... | closed | completed | false | 2 | [
"bug"
] | [
"patil-suraj"
] | 2022-10-07T13:37:37Z | 2022-10-11T10:02:13Z | 2022-10-11T10:02:13Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | MalumaDev | 12,870,488 | MDQ6VXNlcjEyODcwNDg4 | User | false |
huggingface/diffusers | 2,252,005,782 | I_kwDOHa8MBc6GOuGW | 7,716 | https://github.com/huggingface/diffusers/issues/7716 | https://api.github.com/repos/huggingface/diffusers/issues/7716 | OSError: Error no file named config.json found in directory ./RunDiffusion_Juggernaut-XL-v9. | ### Describe the bug
when I load the sdxl model --- [RunDiffusion_Juggernaut-XL-v9](https://huggingface.co/RunDiffusion/Juggernaut-XL-v9),The following error was reported:
```
import torch
from diffusers import DiffusionPipeline, AutoencoderKL, UNet2DConditionModel
from diffusers import StableDiffusionXLPipelin... | closed | completed | false | 9 | [
"bug"
] | [] | 2024-04-19T03:31:24Z | 2024-11-20T00:11:50Z | 2024-11-20T00:11:47Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ScottishFold007 | 36,957,508 | MDQ6VXNlcjM2OTU3NTA4 | User | false |
huggingface/diffusers | 2,252,559,709 | I_kwDOHa8MBc6GQ1Vd | 7,722 | https://github.com/huggingface/diffusers/issues/7722 | https://api.github.com/repos/huggingface/diffusers/issues/7722 | Error using pipeline_animatediff_controlnet and IPAdapter | ### Describe the bug
when trying to use IPAdapter with pipeline_animatediff_controlnet i get an error.
### Reproduction
import torch
from diffusers.utils import load_image
from PIL import Image
from transformers import CLIPVisionModelWithProjection
from diffusers import ControlNetModel, AutoencoderKL, MotionAdap... | closed | completed | false | 4 | [
"bug"
] | [] | 2024-04-19T09:47:53Z | 2024-04-22T08:48:38Z | 2024-04-22T08:48:37Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | lulu1315 | 13,677,315 | MDQ6VXNlcjEzNjc3MzE1 | User | false |
huggingface/diffusers | 2,253,016,808 | I_kwDOHa8MBc6GSk7o | 7,724 | https://github.com/huggingface/diffusers/issues/7724 | https://api.github.com/repos/huggingface/diffusers/issues/7724 | RuntimeError: Error(s) in loading state_dict for AutoencoderKL: Missing Keys! How to solve? | ### Describe the bug
I am trying to get a Lora to run locally on my computer by using this code: https://github.com/hollowstrawberry/kohya-colab and changing it to a local format. When I get to the loading of the models, it gives an error, It seems that the AutoEncoder model has changed but I do not know how to adjust... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-04-19T13:27:17Z | 2024-04-22T08:45:24Z | 2024-04-22T08:45:24Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | veraburg | 44,125,234 | MDQ6VXNlcjQ0MTI1MjM0 | User | false |
huggingface/diffusers | 2,253,693,466 | I_kwDOHa8MBc6GVKIa | 7,726 | https://github.com/huggingface/diffusers/issues/7726 | https://api.github.com/repos/huggingface/diffusers/issues/7726 | Интеграция ControlNet, IPAdapter в pipe StableDiffusionControlNetImg2ImgPipeline | There are such introductory data: take the basic module (https://civitai.com/models/4384 ?modelVersionId=128713), this is the control network (https://huggingface.co/lllyasviel/sd-controlnet-depth ), ip adapter “ip-adapter-plus-face_sd15" (https://huggingface.co/h94/IP-Adapter/tree/main/models ). This time we decided t... | closed | completed | false | 7 | [] | [] | 2024-04-19T18:51:46Z | 2024-11-30T02:51:12Z | 2024-11-30T02:51:12Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Eduard6315 | 145,983,309 | U_kgDOCLOHTQ | User | false |
huggingface/diffusers | 2,254,348,850 | I_kwDOHa8MBc6GXqIy | 7,728 | https://github.com/huggingface/diffusers/issues/7728 | https://api.github.com/repos/huggingface/diffusers/issues/7728 | Accelerate error when training with train_dreambooth_lora_sdxl_advanced.py | ### Describe the bug
Encountered this error with zero information, when using 'train_dreambooth_lora_sdxl_advanced.py',
Traceback (most recent call last):
File "/usr/local/bin/accelerate", line 8, in
sys.exit(main())
File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/accelerate_cli.py", line 46... | closed | completed | false | 15 | [
"bug"
] | [] | 2024-04-20T03:22:20Z | 2024-11-25T13:12:07Z | 2024-11-25T13:12:07Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | nayan-dhabarde | 24,761,808 | MDQ6VXNlcjI0NzYxODA4 | User | false |
huggingface/diffusers | 1,401,311,821 | I_kwDOHa8MBc5ThlJN | 773 | https://github.com/huggingface/diffusers/issues/773 | https://api.github.com/repos/huggingface/diffusers/issues/773 | Support for load hypernet finetune .pt file | does the txt2img pipe support load a finetune .pt file to improve the output performance of checkpoint model ? | closed | completed | false | 4 | [] | [] | 2022-10-07T14:51:23Z | 2022-10-12T05:42:22Z | 2022-10-12T05:42:22Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | CrazyBoyM | 35,400,185 | MDQ6VXNlcjM1NDAwMTg1 | User | false |
huggingface/diffusers | 2,254,618,234 | I_kwDOHa8MBc6GYr56 | 7,730 | https://github.com/huggingface/diffusers/issues/7730 | https://api.github.com/repos/huggingface/diffusers/issues/7730 | Use of requests call without `timeout` parameter | ### Describe the bug
At few places in the codebase, where `requests.get()` is used, the `timeout` parameter is not specified.
https://github.com/huggingface/diffusers/blob/d1e3f489e909d351ea6abced05b0cdaf3a41be7a/src/diffusers/utils/loading_utils.py#L28
The `timeout` parameter is used to set the maximum time to ... | closed | completed | false | 9 | [
"bug",
"good first issue",
"contributions-welcome"
] | [] | 2024-04-20T16:17:02Z | 2025-04-04T06:23:16Z | 2025-04-04T06:23:16Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Sai-Suraj-27 | 87,087,741 | MDQ6VXNlcjg3MDg3NzQx | User | false |
huggingface/diffusers | 2,254,969,290 | I_kwDOHa8MBc6GaBnK | 7,731 | https://github.com/huggingface/diffusers/issues/7731 | https://api.github.com/repos/huggingface/diffusers/issues/7731 | StreamingT2V: Consistent, Dynamic, and Extendable Long Video Generation from Text | ### Model/Pipeline/Scheduler description
Text-to-video diffusion models enable the generation of high-quality videos given text prompts, making it easy to create diverse and individual content. However, existing approaches mostly focus on short video generation (typically 16 or 24 frames), requiring hard cuts when n... | open | null | false | 4 | [
"community-examples",
"Good second issue",
"contributions-welcome"
] | [] | 2024-04-21T08:39:56Z | 2024-11-20T00:15:24Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | clarencechen | 8,482,341 | MDQ6VXNlcjg0ODIzNDE= | User | false |
huggingface/diffusers | 2,255,021,923 | I_kwDOHa8MBc6GaOdj | 7,732 | https://github.com/huggingface/diffusers/issues/7732 | https://api.github.com/repos/huggingface/diffusers/issues/7732 | StableDiffusionXL RuntimeError with ControlNet and Refiner | Hi, I'm trying to use a controlNet in combination with a Refiner but I'm not able to get it working. I'm having this error: `pipeline_stable_diffusion_xl_img2img.py line 496: Number of dimensions of repeat dims can not be smaller than number of dimensions of tensor.`
The canny_image I'm using is a png image already ... | closed | completed | false | 2 | [] | [] | 2024-04-21T11:10:19Z | 2024-04-21T13:03:11Z | 2024-04-21T13:03:11Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Taone974 | 14,542,882 | MDQ6VXNlcjE0NTQyODgy | User | false |
huggingface/diffusers | 2,255,102,991 | I_kwDOHa8MBc6GaiQP | 7,734 | https://github.com/huggingface/diffusers/issues/7734 | https://api.github.com/repos/huggingface/diffusers/issues/7734 | documentation update in v0.24 deleted DeepFloyd info | ### Describe the bug
#5512 refactored the Dreambooth documentation to clarify and reduce overload, but in the process, deleted information on loading DeepFloyd-IF LoRAs, eg. the predicted variance parameter.
I wasn't able to find where this information moved to, so it feels like it's actually just gone now. Can o... | closed | completed | false | 3 | [
"bug"
] | [] | 2024-04-21T14:19:35Z | 2024-05-29T16:27:38Z | 2024-05-29T16:27:38Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | bghira | 59,658,056 | MDQ6VXNlcjU5NjU4MDU2 | User | false |
huggingface/diffusers | 2,255,466,478 | I_kwDOHa8MBc6Gb6_u | 7,735 | https://github.com/huggingface/diffusers/issues/7735 | https://api.github.com/repos/huggingface/diffusers/issues/7735 | Tryng to train on custome dataset but not working “–caption_column` value ‘text’ not found” | ### Describe the bug
I have encountered the problem as described as the link [here](https://discuss.huggingface.co/t/tryng-to-train-on-custome-dataset-but-not-working-caption-column-value-text-not-found/45258).
The dataset I used is [Levi980623/ControlNet_ade20k](https://huggingface.co/datasets/Levi980623/ControlNe... | closed | completed | false | 4 | [
"bug"
] | [] | 2024-04-22T02:48:26Z | 2024-04-24T07:53:47Z | 2024-04-24T07:53:47Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | abovzv | 72,337,405 | MDQ6VXNlcjcyMzM3NDA1 | User | false |
huggingface/diffusers | 2,255,997,297 | I_kwDOHa8MBc6Gd8lx | 7,736 | https://github.com/huggingface/diffusers/issues/7736 | https://api.github.com/repos/huggingface/diffusers/issues/7736 | Add Diffusers supported callbacks | **Is your feature request related to a problem? Please describe.**
We currently support passing in callbacks to pipelines in order to manipulate/customise aspects of the diffusion process. It would be nice to have a set of officially supported callbacks in Diffusers that would allow us to support functionality such as... | closed | completed | false | 4 | [
"help wanted",
"contributions-welcome",
"advanced"
] | [
"asomoza"
] | 2024-04-22T09:05:12Z | 2024-05-13T03:10:31Z | 2024-05-13T03:10:31Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | DN6 | 7,529,846 | MDQ6VXNlcjc1Mjk4NDY= | User | false |
huggingface/diffusers | 2,256,618,978 | I_kwDOHa8MBc6GgUXi | 7,740 | https://github.com/huggingface/diffusers/issues/7740 | https://api.github.com/repos/huggingface/diffusers/issues/7740 | How to get config of single_file | Hi,
Is there any way to get the equivalent of model_index.json from a single_file? | closed | completed | false | 3 | [] | [] | 2024-04-22T14:00:21Z | 2024-04-22T23:26:50Z | 2024-04-22T23:26:50Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | suzukimain | 131,413,573 | U_kgDOB9U2RQ | User | false |
huggingface/diffusers | 1,401,546,648 | I_kwDOHa8MBc5TieeY | 775 | https://github.com/huggingface/diffusers/issues/775 | https://api.github.com/repos/huggingface/diffusers/issues/775 | Still problem with fp16 and without autocast | ### Describe the bug
Hello!
There was often discussed about this issue and as far as I understand it is meanwhile fixed. But I stll get a
`RuntimeError: expected scalar type Half but found Float`
when I try to run the model with fp16 but without autocast.
My code:
```
pipe0 = StableDiffusionPipeline.f... | closed | completed | false | 5 | [
"bug"
] | [] | 2022-10-07T18:40:44Z | 2022-10-08T04:04:13Z | 2022-10-08T04:04:13Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Marcophono2 | 22,599,855 | MDQ6VXNlcjIyNTk5ODU1 | User | false |
huggingface/diffusers | 2,258,899,036 | I_kwDOHa8MBc6GpBBc | 7,751 | https://github.com/huggingface/diffusers/issues/7751 | https://api.github.com/repos/huggingface/diffusers/issues/7751 | errors while trying to "convert_animatediff_motion_module_to_diffusers" | ### Describe the bug
i'm trying to use different motion adapter for animatediff . these ones especially :
https://huggingface.co/ByteDance/AnimateDiff-Lightning/blob/main/animatediff_lightning_8step_comfyui.safetensors
https://civitai.com/models/331718/bubbling-rings-or-animatediff-motion-lora
when i try to conv... | closed | completed | false | 4 | [
"bug"
] | [] | 2024-04-23T13:32:20Z | 2024-11-19T09:12:48Z | 2024-11-19T09:12:41Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | lulu1315 | 13,677,315 | MDQ6VXNlcjEzNjc3MzE1 | User | false |
huggingface/diffusers | 2,258,928,375 | I_kwDOHa8MBc6GpIL3 | 7,752 | https://github.com/huggingface/diffusers/issues/7752 | https://api.github.com/repos/huggingface/diffusers/issues/7752 | What should be the value to be used for the "num_class-mages" while training the stable diffusion model? | I read from the hugging face docs that the parameter "num_class_images" refers to the "Minimal" class images for prior preservation loss. If with_prior_preservation = True and there are not enough images already present in class_data_dir, additional images will be sampled with class_prompt.
```
export MODEL_NAME="... | closed | completed | true | 2 | [
"stale"
] | [] | 2024-04-23T13:45:24Z | 2024-11-30T02:52:54Z | 2024-11-30T02:52:54Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sivaramakrishnan-rajaraman | 45,852,019 | MDQ6VXNlcjQ1ODUyMDE5 | User | false |
huggingface/diffusers | 2,259,544,257 | I_kwDOHa8MBc6GrejB | 7,756 | https://github.com/huggingface/diffusers/issues/7756 | https://api.github.com/repos/huggingface/diffusers/issues/7756 | Analysis of Classifier-Free Guidance Weight Schedulers | ### Model/Pipeline/Scheduler description
The paper's authors performs an analysis and proposes 1 line change in order to make Classifier-Free Guidance looks better
I personally run some test to confirm
**SD1.... | open | null | false | 12 | [
"stale",
"consider-for-modular-diffusers"
] | [] | 2024-04-23T18:43:34Z | 2024-12-25T15:06:36Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | rootonchair | 23,548,268 | MDQ6VXNlcjIzNTQ4MjY4 | User | false |
huggingface/diffusers | 2,259,610,726 | I_kwDOHa8MBc6Gruxm | 7,757 | https://github.com/huggingface/diffusers/issues/7757 | https://api.github.com/repos/huggingface/diffusers/issues/7757 | train_controlnet_sdxl.py Multi-GPU OutOfMemory | ### Describe the bug
Hi there. I've reliably used the train_controlnet_sdxl.py on single gpu on GCP (A100 - 40 GB). I have had to switch to AWS and am presently using a p3.8xlarge which has 4 V100 gpus w/ 64 GB GPU memory total.
Whenever I run my workflow on AWS I get a Cuda out of memory error just loading the ... | closed | completed | false | 13 | [
"bug",
"stale"
] | [] | 2024-04-23T19:23:16Z | 2024-11-30T02:47:50Z | 2024-11-30T02:47:49Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | asutermo | 561,333 | MDQ6VXNlcjU2MTMzMw== | User | false |
huggingface/diffusers | 2,260,032,716 | I_kwDOHa8MBc6GtVzM | 7,759 | https://github.com/huggingface/diffusers/issues/7759 | https://api.github.com/repos/huggingface/diffusers/issues/7759 | PixArtAlphaPipeline does not handle negative_prompt list correctly | ### Describe the bug
Passing negative_prompt as list to PixArtAlphaPipeline will eventually cause a "list object has no attribute lower" error.
The encode_prompt function documentation shows that it can be a list as well:
https://github.com/huggingface/diffusers/blob/7404f1e9dcdac286caf6ca3cfa43b294f61583bf/src/... | open | null | false | 8 | [
"bug",
"stale"
] | [] | 2024-04-24T00:51:01Z | 2024-12-24T15:06:30Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Trojaner | 1,809,172 | MDQ6VXNlcjE4MDkxNzI= | User | false |
huggingface/diffusers | 1,401,565,378 | I_kwDOHa8MBc5TijDC | 776 | https://github.com/huggingface/diffusers/issues/776 | https://api.github.com/repos/huggingface/diffusers/issues/776 | Stable-diffusion Image-to-image pipeline is broken for batch_size > 1 | ### Describe the bug
When multiple images are passed to the `init_image` argument as a tensor, `init_latents` will have the same dimension 0 as `init_images`. But this line is further duplicating the `init_latents` by the batch size, which results in a mismatched tensor size.
https://github.com/huggingface/diffus... | closed | completed | false | 1 | [
"bug"
] | [] | 2022-10-07T19:03:14Z | 2022-10-12T11:00:37Z | 2022-10-12T11:00:37Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | jorgemcgomes | 3,987,574 | MDQ6VXNlcjM5ODc1NzQ= | User | false |
huggingface/diffusers | 2,260,152,058 | I_kwDOHa8MBc6Gty76 | 7,760 | https://github.com/huggingface/diffusers/issues/7760 | https://api.github.com/repos/huggingface/diffusers/issues/7760 | Implement: Align Your Steps: Optimizing Sampling Schedules in Diffusion Models | Align Your Steps: Optimizing Sampling Schedules in Diffusion Models is a general and principled approach to optimizing the sampling schedules of DMs for high-quality outputs.
This work is presented by Nvidia labs in this paper:
https://arxiv.org/abs/2404.14507
and the project page is here
https://research.nvidia.... | closed | completed | false | 14 | [] | [
"yiyixuxu"
] | 2024-04-24T02:27:58Z | 2024-11-20T00:03:58Z | 2024-11-20T00:03:57Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | joe-aivatarz | 112,866,199 | U_kgDOBrozlw | User | false |
huggingface/diffusers | 2,260,459,293 | I_kwDOHa8MBc6Gu98d | 7,762 | https://github.com/huggingface/diffusers/issues/7762 | https://api.github.com/repos/huggingface/diffusers/issues/7762 | Bad result than A1111 sd webui | ### Describe the bug
Task: generate a new hairstyle for a photo with inpaint+controlnet+mask
Bug: webui can generate satisfied images while diffusers pipeline can not
webui result:
diffuers result: ... | closed | completed | false | 5 | [
"bug",
"inpainting"
] | [] | 2024-04-24T06:29:04Z | 2024-05-08T18:02:32Z | 2024-05-08T17:04:33Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | weiweiwang | 822,994 | MDQ6VXNlcjgyMjk5NA== | User | false |
huggingface/diffusers | 2,260,613,637 | I_kwDOHa8MBc6GvjoF | 7,765 | https://github.com/huggingface/diffusers/issues/7765 | https://api.github.com/repos/huggingface/diffusers/issues/7765 | FlaxStableDiffusionXLPipeline outputs 'NAN' in fp16 inference. | ### Describe the bug
The data of the image generated by FlaxStableDiffusionXLPipeline is all 'nan' with fp16 datatype, as displayed by the following pseudocode:
```
pipeline, params = FlaxStableDiffusionXLPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", dtype=jnp.float16)
images = pipeline(pro... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-04-24T07:45:26Z | 2024-04-25T05:38:03Z | 2024-04-25T05:38:03Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | yisonzhu | 107,918,054 | U_kgDOBm6y5g | User | false |
huggingface/diffusers | 2,260,640,693 | I_kwDOHa8MBc6GvqO1 | 7,766 | https://github.com/huggingface/diffusers/issues/7766 | https://api.github.com/repos/huggingface/diffusers/issues/7766 | IP-Adapter FaceID PLus How to use questions | https://github.com/huggingface/diffusers/blob/9ef43f38d43217f690e222a4ce0239c6a24af981/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py#L492
## error msg:
pipe.unet.encoder_hid_proj.image_projection_layers[0].clip_embeds = clip_embeds.to(dtype=torch.float16)
AttributeError: 'list' obje... | closed | completed | false | 10 | [] | [] | 2024-04-24T07:56:38Z | 2024-11-20T00:02:30Z | 2024-11-20T00:02:30Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Honey-666 | 76,274,942 | MDQ6VXNlcjc2Mjc0OTQy | User | false |
huggingface/diffusers | 2,261,912,115 | I_kwDOHa8MBc6G0goz | 7,770 | https://github.com/huggingface/diffusers/issues/7770 | https://api.github.com/repos/huggingface/diffusers/issues/7770 | Loading model and LoRA in Python causes tensor shape mismatch | ### Describe the bug
Hi.
Here, I am trying to do inpaint using **realisticStockPhoto** with **SDXL_FILM_PHOTOGRAPHY_STYLE** LoRA.
```
device = "cuda"
model_path = "realisticStockPhoto_v20.safetensors"
pipe = StableDiffusionInpaintPipeline.from_single_file(model_path,torch_dtype=torch.float16,safetensors=Tru... | closed | completed | false | 5 | [
"bug",
"stale"
] | [] | 2024-04-24T18:22:13Z | 2024-11-30T02:45:21Z | 2024-11-30T02:45:21Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | levoz92 | 162,665,819 | U_kgDOCbIVWw | User | false |
huggingface/diffusers | 2,262,607,654 | I_kwDOHa8MBc6G3Kcm | 7,772 | https://github.com/huggingface/diffusers/issues/7772 | https://api.github.com/repos/huggingface/diffusers/issues/7772 | Noised Image | Hi,
I'm using text_to_image_lora, and I found the variavle "X_T" is removed which prevent us feeding the noised image to diffusion. Do you think it is better to add this part? Or I miss something. It is different with previous version. | closed | completed | false | 3 | [] | [] | 2024-04-25T04:07:48Z | 2024-11-20T00:01:03Z | 2024-11-20T00:01:03Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | massyzs | 45,664,796 | MDQ6VXNlcjQ1NjY0Nzk2 | User | false |
huggingface/diffusers | 2,262,738,388 | I_kwDOHa8MBc6G3qXU | 7,773 | https://github.com/huggingface/diffusers/issues/7773 | https://api.github.com/repos/huggingface/diffusers/issues/7773 | Stable Video Diffusion generation bug when using batch size > 1 | ### Describe the bug
When generating multiple videos (say batch size = 2), SVD will give 1 correct generation and 1 incorrect one.
### Reproduction
video = pipeline(..., num_videos_per_prompt=2, ...)
### Logs
_No response_
### System Info
Any (latest)
### Who can help?
_No response_ | closed | completed | false | 5 | [
"bug"
] | [] | 2024-04-25T05:59:43Z | 2024-11-19T23:59:33Z | 2024-11-19T23:59:30Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | liyy201912 | 59,664,299 | MDQ6VXNlcjU5NjY0Mjk5 | User | false |
huggingface/diffusers | 2,263,028,647 | I_kwDOHa8MBc6G4xOn | 7,775 | https://github.com/huggingface/diffusers/issues/7775 | https://api.github.com/repos/huggingface/diffusers/issues/7775 | How to input gradio settings in Python | Hi.
I use **realisticStockPhoto_v20** on Fooocus with **sdxl_film_photography_style** lora and I really like the results.
Fooocus and other gradio implementations come with settings inputs that I want to utilize in Python as well. In particular, if this is my code:
```
device = "cuda"
model_path = "weights/reali... | closed | completed | false | 5 | [] | [] | 2024-04-25T08:43:20Z | 2024-11-20T00:07:26Z | 2024-11-20T00:07:26Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | levoz92 | 162,665,819 | U_kgDOCbIVWw | User | false |
huggingface/diffusers | 2,263,331,660 | I_kwDOHa8MBc6G57NM | 7,777 | https://github.com/huggingface/diffusers/issues/7777 | https://api.github.com/repos/huggingface/diffusers/issues/7777 | confusion when train dreambooth via script train_dreambooth_lora_sdxl_advanced | I used the train_dreambooth_lora_sdxl_advanced script downloaded from diffusers' official exmaples to train my own images, however, both the validation images generated and the images generated by trained model are of bad quality. They are always very indistinct. I used the script offered by diffsuers examples and didn... | closed | completed | false | 19 | [] | [] | 2024-04-25T11:19:48Z | 2024-05-13T02:52:03Z | 2024-05-13T02:52:03Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | joey0922 | 12,229,405 | MDQ6VXNlcjEyMjI5NDA1 | User | false |
huggingface/diffusers | 1,401,811,815 | I_kwDOHa8MBc5TjfNn | 778 | https://github.com/huggingface/diffusers/issues/778 | https://api.github.com/repos/huggingface/diffusers/issues/778 | Add `v`-prediction | Like in the Imagen Video paper. According to Katherine Crowson it's more stable and faster: https://twitter.com/rivershavewings/status/1578193039423852544?s=21&t=oIToMoHbtU1U8VH3q4Mzng | closed | completed | false | 17 | [
"good first issue"
] | [] | 2022-10-08T03:59:53Z | 2023-03-18T18:41:16Z | 2023-03-18T18:41:16Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | pcuenca | 1,177,582 | MDQ6VXNlcjExNzc1ODI= | User | false |
huggingface/diffusers | 2,264,209,030 | I_kwDOHa8MBc6G9RaG | 7,781 | https://github.com/huggingface/diffusers/issues/7781 | https://api.github.com/repos/huggingface/diffusers/issues/7781 | Get error 'Can only serialize PIL.Image.Image, got <class 'dict'>' | ### Describe the bug
I am trying to use a model hosted on huggingface and I have created an inference endpoint for it. I have attached the code I used to access the endpoint and generate images. However, after the image is generated I get a base 64 string which when decoded translates to: 'Can only serialize PIL.Ima... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-04-25T18:10:01Z | 2024-06-25T13:44:44Z | 2024-04-30T12:10:56Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | NaveenGTK | 115,793,005 | U_kgDOBubcbQ | User | false |
huggingface/diffusers | 2,264,750,889 | I_kwDOHa8MBc6G_Vsp | 7,782 | https://github.com/huggingface/diffusers/issues/7782 | https://api.github.com/repos/huggingface/diffusers/issues/7782 | Converting safetensors file with "_pickle.UnpicklingError: invalid load key, ' '." | ### Describe the bug
I'm trying to convert safetensors files on civitai to diffusers using the convert_original_stable_diffusion_to_diffusers.py script, and I'm getting <_pickle.UnpicklingError: invalid load key,> ' ' error, my pytorch version is 2.22 and cuda is 12.1. i have tried on both windows and liunx, but it's ... | closed | completed | false | 1 | [
"bug"
] | [] | 2024-04-26T01:08:04Z | 2024-04-28T01:35:01Z | 2024-04-28T01:35:01Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | hoyAng96 | 50,062,948 | MDQ6VXNlcjUwMDYyOTQ4 | User | false |
huggingface/diffusers | 2,265,648,549 | I_kwDOHa8MBc6HCw2l | 7,784 | https://github.com/huggingface/diffusers/issues/7784 | https://api.github.com/repos/huggingface/diffusers/issues/7784 | StableDiffusionUpscaleLDM3DPipeline object has no attribute 'upcast_vae' | ### Describe the bug
[miniconda3/lib/python3.12/site-packages/diffusers/configuration_utils.py:143](https://vscode-remote+ssh-002dremote-002b10-002e155-002e234-002e172.vscode-resource.vscode-cdn.net/data2/miniconda3/lib/python3.12/site-packages/diffusers/configuration_utils.py:143), in ConfigMixin.__getattr__(self, na... | open | null | false | 4 | [
"bug",
"stale"
] | [] | 2024-04-26T12:08:01Z | 2024-12-14T15:06:44Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | uwGZQ | 148,398,394 | U_kgDOCNhhOg | User | false |
huggingface/diffusers | 2,265,739,845 | I_kwDOHa8MBc6HDHJF | 7,785 | https://github.com/huggingface/diffusers/issues/7785 | https://api.github.com/repos/huggingface/diffusers/issues/7785 | Very Slow first inference with diffusers 0.27.X | ### Describe the bug
Hello diffusers team !
I face an annoying issue since I upgraded the diffusers version to 0.27.X
The first call (and only the first) of pipeline(...) takes now a lot of time **before** to start inference (like a minute)
Moreover the call of compel(prompts) takes 30 seconds versus instant in 0... | closed | completed | false | 7 | [
"bug"
] | [] | 2024-04-26T12:59:21Z | 2024-10-15T21:30:31Z | 2024-10-15T21:30:31Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | nesscube | 23,378,955 | MDQ6VXNlcjIzMzc4OTU1 | User | false |
huggingface/diffusers | 2,266,137,603 | I_kwDOHa8MBc6HEoQD | 7,787 | https://github.com/huggingface/diffusers/issues/7787 | https://api.github.com/repos/huggingface/diffusers/issues/7787 | Combined loss term for VQ-VAE (`diffusers.VQModel`) | For training the VQ-VAE component of a latent diffusion model a la `CompVis/ldm-celebahq-256` (which uses `diffusers.VQModel`), is there a combined loss term for each of the losses as described by the authors: reconstruction loss, vq loss, and commitment loss?
I see the vq loss term is collected in `VectorQuantizer`... | open | null | false | 3 | [
"stale"
] | [] | 2024-04-26T16:38:06Z | 2024-09-14T15:14:16Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | asy51 | 90,255,040 | MDQ6VXNlcjkwMjU1MDQw | User | false |
huggingface/diffusers | 2,267,033,165 | I_kwDOHa8MBc6HIC5N | 7,789 | https://github.com/huggingface/diffusers/issues/7789 | https://api.github.com/repos/huggingface/diffusers/issues/7789 | [mps] training / inferencing deepfloyd must be done in float32 | ### Describe the bug
_**Just to keep track of this issue, because I'm not sure if I've done something wrong or if it's due to the current issues in Pytorch's MPS support.**_
I'm having bad results running inference on Apple MPS with DeepFloyd stage I / II in Diffusers:
# Stage I (400M)
## CPU (float32)
![i... | open | null | false | 4 | [
"bug",
"stale"
] | [] | 2024-04-27T14:13:08Z | 2024-09-14T15:14:13Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | bghira | 59,658,056 | MDQ6VXNlcjU5NjU4MDU2 | User | false |
huggingface/diffusers | 1,401,924,209 | I_kwDOHa8MBc5Tj6px | 779 | https://github.com/huggingface/diffusers/issues/779 | https://api.github.com/repos/huggingface/diffusers/issues/779 | Stable Diffusion: providing a list of positive prompts and a list of negative prompts does not work as expected | ### Describe the bug
See this forum post: https://discuss.huggingface.co/t/stable-diffusion-bs-1-uses-negative-as-prompt/24130.
In short:
```Python
_ = pipe(["frog"]*2, negative_prompt=["bird"]*2)
```
Reaches this condition https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/stable_... | closed | completed | false | 4 | [
"bug",
"stale"
] | [
"pcuenca",
"patrickvonplaten"
] | 2022-10-08T12:00:43Z | 2022-11-18T15:32:05Z | 2022-11-18T15:32:05Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | pcuenca | 1,177,582 | MDQ6VXNlcjExNzc1ODI= | User | false |
huggingface/diffusers | 2,267,161,401 | I_kwDOHa8MBc6HIiM5 | 7,792 | https://github.com/huggingface/diffusers/issues/7792 | https://api.github.com/repos/huggingface/diffusers/issues/7792 | Issue with reproducing ip_adapter_sdxl_controlnet_demo.ipynb example from doc | Hi all!
I met an issue when I tried to use diffusers/hugging face syntax to reproduce an [example](https://github.com/tencent-ailab/IP-Adapter/blob/main/ip_adapter_sdxl_controlnet_demo.ipynb).
I am obtaining different/worse result and wondering why...
![image](https://github.com/huggingface/diffusers/assets/90... | open | null | false | 3 | [
"stale"
] | [] | 2024-04-27T19:47:26Z | 2024-09-14T15:14:07Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | katarzynasornat | 9,012,956 | MDQ6VXNlcjkwMTI5NTY= | User | false |
huggingface/diffusers | 2,267,227,820 | I_kwDOHa8MBc6HIyas | 7,793 | https://github.com/huggingface/diffusers/issues/7793 | https://api.github.com/repos/huggingface/diffusers/issues/7793 | [deepfloyd] unloading lora weights results in error | ### Describe the bug
I'm trying to run a stage I LoRA and then unload those fused weights, but it results in an error.
### Reproduction
```py
from diffusers import DiffusionPipeline
import torch
# Your LoRA paths.
deepfloyd_lora_path = "ptx0/deepcinema"
deepfloyd_stage1_lora_weights_name = "pytorch_lora... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-04-27T23:01:06Z | 2024-04-28T12:19:07Z | 2024-04-28T12:19:07Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | bghira | 59,658,056 | MDQ6VXNlcjU5NjU4MDU2 | User | false |
huggingface/diffusers | 2,267,287,318 | I_kwDOHa8MBc6HJA8W | 7,794 | https://github.com/huggingface/diffusers/issues/7794 | https://api.github.com/repos/huggingface/diffusers/issues/7794 | Tensor Device mismatch error | Hello. I am getting the following error:
```
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument index in method wrapper_CUDA__index_select)
```
This is my code:
```
import inspect, torch, PIL, random, os
import numpy a... | closed | completed | false | 1 | [
"bug"
] | [] | 2024-04-28T01:32:10Z | 2024-05-08T15:37:41Z | 2024-04-28T01:56:21Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | levoz92 | 162,665,819 | U_kgDOCbIVWw | User | false |
huggingface/diffusers | 2,267,310,106 | I_kwDOHa8MBc6HJGga | 7,796 | https://github.com/huggingface/diffusers/issues/7796 | https://api.github.com/repos/huggingface/diffusers/issues/7796 | Controlnet inpainting change too much of masking area | ### Describe the bug
mask
i got mask with this
```
def make_inpaint_condition (image... | open | null | false | 6 | [
"bug",
"stale"
] | [] | 2024-04-28T02:59:45Z | 2024-09-14T15:14:04Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | crapthings | 1,147,704 | MDQ6VXNlcjExNDc3MDQ= | User | false |
huggingface/diffusers | 2,267,406,111 | I_kwDOHa8MBc6HJd8f | 7,798 | https://github.com/huggingface/diffusers/issues/7798 | https://api.github.com/repos/huggingface/diffusers/issues/7798 | Bug of PixArtAlphaPipeline | ### Describe the bug
When I use
with torch.autocast("cuda"):
images = pipe(...)
the output is noise
### Reproduction
import torch
from diffusers import PixArtAlphaPipel... | closed | completed | false | 3 | [
"bug"
] | [] | 2024-04-28T07:31:04Z | 2024-10-15T21:28:48Z | 2024-10-15T21:28:48Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | CuddleSabe | 61,224,076 | MDQ6VXNlcjYxMjI0MDc2 | User | false |
huggingface/diffusers | 2,267,410,409 | I_kwDOHa8MBc6HJe_p | 7,799 | https://github.com/huggingface/diffusers/issues/7799 | https://api.github.com/repos/huggingface/diffusers/issues/7799 | Mismatching size in matmul when using StableDiffusionInstructPix2Pix pipeline with IP-Adapter | ### Describe the bug
I've tried to combine InstructPix2Pix model and IP-Adapter (`pipe.load_ip_adapter("h94/IP-Adapter", subfolder="models", weight_name="ip-adapter_sd15.bin"`), but I always get `RuntimeError: mat1 and mat2 shapes cannot be multiplied (771x1280 and 1024x3072)`. When I use other models i.e. Stable Diff... | closed | completed | false | 3 | [
"bug"
] | [] | 2024-04-28T07:44:03Z | 2024-05-01T16:27:44Z | 2024-05-01T16:27:44Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | misshimichka | 92,110,572 | U_kgDOBX1-7A | User | false |
huggingface/diffusers | 2,267,426,028 | I_kwDOHa8MBc6HJizs | 7,800 | https://github.com/huggingface/diffusers/issues/7800 | https://api.github.com/repos/huggingface/diffusers/issues/7800 | A strange time cost in denoising loop | ### Describe the bug
I found a stange time cost when I inference StableVideoDiffusionPipeline with GPU. On 3090, the UNet only costs 1.6s, however, a denoising loop costs 3.3s.
It is strange. So I analyse the time cost for all parts in the denoising loop with Viztracer (a tool for tracing and visualizing execution ti... | closed | completed | false | 5 | [
"bug"
] | [] | 2024-04-28T08:26:37Z | 2024-05-07T10:13:34Z | 2024-05-07T10:13:34Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | FEIFEIEIAr | 77,225,830 | MDQ6VXNlcjc3MjI1ODMw | User | false |
huggingface/diffusers | 2,267,470,706 | I_kwDOHa8MBc6HJtty | 7,801 | https://github.com/huggingface/diffusers/issues/7801 | https://api.github.com/repos/huggingface/diffusers/issues/7801 | set mixed_precision="fp16",but the model is not fp16 | ### Describe the bug
I trained the control net of SDXL using the mixed prediction="fp16" parameter, but the trained model did not have the suffix. fp16. safetensors. But it's a 5G file. I see that the Controlnet on HuggFace will have a 5G file and a 2.5G file with fp16.
### Reproduction
accelerate launch train_cont... | open | null | false | 5 | [
"bug",
"stale"
] | [] | 2024-04-28T10:12:53Z | 2024-09-21T07:35:54Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | starinskycc | 6,244,330 | MDQ6VXNlcjYyNDQzMzA= | User | false |
huggingface/diffusers | 2,267,495,507 | I_kwDOHa8MBc6HJzxT | 7,802 | https://github.com/huggingface/diffusers/issues/7802 | https://api.github.com/repos/huggingface/diffusers/issues/7802 | Support OLSS Scheduler | OLSS is the most advanced sampler. Can you consider adding it to the scheduler library of diffusers
(https://github.com/alibaba/EasyNLP/tree/master/diffusion/olss_scheduler)
| open | null | false | 4 | [
"stale"
] | [] | 2024-04-28T11:21:23Z | 2024-12-14T15:06:37Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ZihaoW123 | 47,708,655 | MDQ6VXNlcjQ3NzA4NjU1 | User | false |
huggingface/diffusers | 2,267,552,290 | I_kwDOHa8MBc6HKBoi | 7,805 | https://github.com/huggingface/diffusers/issues/7805 | https://api.github.com/repos/huggingface/diffusers/issues/7805 | Something went wrong | I'm having an issue in UNet training cell. After trying all types of parameters facing the same something went wrong .
i'm using 6 images of dogs of 512*512 resolution. please help me in this. | closed | completed | false | 3 | [] | [] | 2024-04-28T13:01:51Z | 2024-10-20T23:18:06Z | 2024-10-20T23:18:06Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | SadiaAman443 | 168,347,778 | U_kgDOCgjIgg | User | false |
huggingface/diffusers | 2,267,790,015 | I_kwDOHa8MBc6HK7q_ | 7,808 | https://github.com/huggingface/diffusers/issues/7808 | https://api.github.com/repos/huggingface/diffusers/issues/7808 | Make the sampling loop of pipelines modular, with designated inputs that can be handled iteratively with functions and a default sampler step function | ### Model/Pipeline/Scheduler description
Related to #7761 .
This is an effective replacement for the existing sampling loop function and the many, many kwargs that were made to allow the user to control it or inject callbacks into it.
```py
class SamplingInput:
def __init__(self, img, text_embedding, une... | open | null | false | 3 | [
"stale"
] | [] | 2024-04-28T20:02:00Z | 2024-09-14T15:13:49Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | AmericanPresidentJimmyCarter | 110,263,573 | U_kgDOBpJ9FQ | User | false |
huggingface/diffusers | 2,267,848,001 | I_kwDOHa8MBc6HLJ1B | 7,809 | https://github.com/huggingface/diffusers/issues/7809 | https://api.github.com/repos/huggingface/diffusers/issues/7809 | Feature Request: new release | Hello, in nixpkgs we are about to start stabilizing our packages for a new stable release of nixos. We would love to ship diffusers and dependent packages with python 3.12 (context: https://github.com/huggingface/diffusers/issues/7441), but the necessary fixes have not been tagged in a release yet.
**There is no ne... | closed | completed | false | 3 | [] | [] | 2024-04-28T22:16:24Z | 2024-04-30T03:03:02Z | 2024-04-30T03:03:02Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | pbsds | 140,964 | MDQ6VXNlcjE0MDk2NA== | User | false |
huggingface/diffusers | 2,267,924,511 | I_kwDOHa8MBc6HLcgf | 7,810 | https://github.com/huggingface/diffusers/issues/7810 | https://api.github.com/repos/huggingface/diffusers/issues/7810 | Inquiry About Using Non-Square Images for ControlNet Training |
Subject: Inquiry About Using Non-Square Images for ControlNet Training
Dear [Team/Developer],
First of all, thank you for providing the training code for ControlNet. I have recently been utilizing this code to train the ControlNet for an sdxl model, and I am quite satisfied with the results. However, I believe ... | open | null | false | 2 | [
"stale"
] | [] | 2024-04-29T00:58:08Z | 2024-09-14T15:13:46Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | starinskycc | 6,244,330 | MDQ6VXNlcjYyNDQzMzA= | User | false |
huggingface/diffusers | 2,268,061,468 | I_kwDOHa8MBc6HL98c | 7,813 | https://github.com/huggingface/diffusers/issues/7813 | https://api.github.com/repos/huggingface/diffusers/issues/7813 | I feel confused about this TODO issue. how to pass timesteps as tensors? | https://github.com/huggingface/diffusers/blob/235d34cf567e78bf958344d3132bb018a8580295/src/diffusers/models/unets/unet_2d_condition.py#L918
| closed | completed | false | 3 | [
"stale"
] | [] | 2024-04-29T03:46:21Z | 2024-11-23T00:19:17Z | 2024-11-23T00:19:16Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ghost | 10,137 | MDQ6VXNlcjEwMTM3 | User | false |
huggingface/diffusers | 2,268,187,419 | I_kwDOHa8MBc6HMcsb | 7,814 | https://github.com/huggingface/diffusers/issues/7814 | https://api.github.com/repos/huggingface/diffusers/issues/7814 | Multi-controlnet formatting issue | ### Describe the bug
Hi.
There is an inconsistency between `from_pretrained` and `save_pretrained` within the Multicontrolnet class.
The from_pretrained function returns a directory structure like this: controlnet, controlnet_1, controlnet_2,
whereas save_pretrained is like this: controlnet, controlnet_1, contro... | closed | completed | false | 1 | [
"bug"
] | [] | 2024-04-29T05:57:17Z | 2024-05-13T19:32:07Z | 2024-05-13T19:32:07Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | rebel-kblee | 119,555,851 | U_kgDOByBHCw | User | false |
huggingface/diffusers | 1,402,005,132 | I_kwDOHa8MBc5TkOaM | 782 | https://github.com/huggingface/diffusers/issues/782 | https://api.github.com/repos/huggingface/diffusers/issues/782 | Model download broken after removal of use_auth_token | ### Describe the bug
After updating my git checkout to a version that removed `use_auth_token` parameters, diffusers cannot download `CompVis/stable-diffusion-v1-4` anymore with an error 401.
I am logged in with the same auth token as before. Downloading another model that is not protected with an auth token usin... | closed | completed | false | 10 | [
"bug"
] | [
"patil-suraj"
] | 2022-10-08T17:05:31Z | 2022-11-16T00:28:35Z | 2022-10-09T17:39:07Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | allo- | 1,482,341 | MDQ6VXNlcjE0ODIzNDE= | User | false |
huggingface/diffusers | 1,402,012,859 | I_kwDOHa8MBc5TkQS7 | 783 | https://github.com/huggingface/diffusers/issues/783 | https://api.github.com/repos/huggingface/diffusers/issues/783 | problem when training, The following values were not passed to `accelerate launch` and had defaults used instead: | The following values were not passed to `accelerate launch` and had defaults used instead:
`--num_cpu_threads_per_process` was set to `6` to improve out-of-box performance
To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`.
Traceback (most recent call last)... | closed | completed | false | 4 | [
"stale"
] | [
"patil-suraj"
] | 2022-10-08T17:35:50Z | 2022-11-17T16:05:00Z | 2022-11-17T16:05:00Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ivaxsirc | 48,472,866 | MDQ6VXNlcjQ4NDcyODY2 | User | false |
huggingface/diffusers | 2,274,682,098 | I_kwDOHa8MBc6HlOTy | 7,836 | https://github.com/huggingface/diffusers/issues/7836 | https://api.github.com/repos/huggingface/diffusers/issues/7836 | CLIP Training Example Bug - Overfitting | ### Describe the bug
After reading numerous posts on overfitting, I am convinced that the example run_clip.py has bugs. it doesn't converge. Earlier I posted it for discussion [here](https://github.com/huggingface/diffusers/discussions/7835)
### Reproduction
Execute the run_clip.py and watch for conv/divg.
### ... | closed | completed | false | 1 | [
"bug"
] | [] | 2024-05-02T06:04:47Z | 2024-05-11T14:14:44Z | 2024-05-11T14:14:44Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | humanely | 6,862,868 | MDQ6VXNlcjY4NjI4Njg= | User | false |
huggingface/diffusers | 1,402,066,456 | I_kwDOHa8MBc5TkdYY | 784 | https://github.com/huggingface/diffusers/issues/784 | https://api.github.com/repos/huggingface/diffusers/issues/784 | DeprecationWarning: BICUBIC is deprecated and will be removed in Pillow 10 (2023-07-01). Use Resampling.BICUBIC instead. | ### Describe the bug
Textual inversion uses a deprecated import for the scaling methods in Pillow
DeprecationWarning: BICUBIC is deprecated and will be removed in Pillow 10 (2023-07-01). Use Resampling.BICUBIC instead.
### Reproduction
Use the textual inversion script with a recent Pillow version
### Logs
`... | closed | completed | false | 4 | [
"bug",
"stale"
] | [] | 2022-10-08T21:21:56Z | 2022-11-16T15:03:50Z | 2022-11-16T15:03:50Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | allo- | 1,482,341 | MDQ6VXNlcjE0ODIzNDE= | User | false |
huggingface/diffusers | 2,275,525,776 | I_kwDOHa8MBc6HocSQ | 7,841 | https://github.com/huggingface/diffusers/issues/7841 | https://api.github.com/repos/huggingface/diffusers/issues/7841 | with XL ，When the image is enlarged for viewing，The drawn image will appear as red dots in most cases | ### Describe the bug
with XL ，When the image is enlarged for viewing，The drawn image will appear as red dots in most cases
### Reproduction
```
from diffusers import StableDiffusionXLPipel... | closed | completed | false | 8 | [
"bug"
] | [] | 2024-05-02T13:18:48Z | 2024-05-03T01:14:55Z | 2024-05-03T01:14:55Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | zhaoyun0071 | 35,762,050 | MDQ6VXNlcjM1NzYyMDUw | User | false |
huggingface/diffusers | 2,275,868,046 | I_kwDOHa8MBc6Hpv2O | 7,842 | https://github.com/huggingface/diffusers/issues/7842 | https://api.github.com/repos/huggingface/diffusers/issues/7842 | deepfloyd stage 2 crashes with tensor size mismatch when input image size is not divisible by 8 | ### Describe the bug
DeepFloyd's upstream code supports 8px-aligned inputs for stage II, which I believe the Diffusers implementation is based upon. However, it seems that for certain sizes, there is some unfortunate interaction between the hidden states and the residual hidden states.
I'm not sure if this is som... | open | null | false | 3 | [
"bug",
"stale"
] | [] | 2024-05-02T15:48:39Z | 2024-09-14T15:13:40Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | bghira | 59,658,056 | MDQ6VXNlcjU5NjU4MDU2 | User | false |
huggingface/diffusers | 2,276,094,137 | I_kwDOHa8MBc6HqnC5 | 7,848 | https://github.com/huggingface/diffusers/issues/7848 | https://api.github.com/repos/huggingface/diffusers/issues/7848 | examples/community/lpw_stable_diffusion_xl.py Not correctly decoded | ### Describe the bug
When I used `lpw_stable_diffusion_xl` on a text2img model (`playgroundai/playground-v2.5-1024px-aesthetic`), I found the image gray, seemed not decoded correctly.
To solve this I copied ... | closed | completed | false | 3 | [
"bug"
] | [] | 2024-05-02T17:25:10Z | 2024-11-20T00:06:31Z | 2024-11-20T00:06:30Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | HACLINE | 142,033,181 | U_kgDOCHdBHQ | User | false |
huggingface/diffusers | 1,402,137,417 | I_kwDOHa8MBc5TkutJ | 785 | https://github.com/huggingface/diffusers/issues/785 | https://api.github.com/repos/huggingface/diffusers/issues/785 | PNDMScheduler has a considerable cost when running StableDiffusionPipeline in diffuser-0.4.1 | ### Describe the bug
## Enviroment
GPU: A10, CUDA 11.6, cuDNN 8.4.0
Torch: 1.12.1
diffuser: 0.4.1
## Phenomenon
When I ran the StableDiffusionPipeline with fp16 precision, I found the time cost of PNDMScheduler increase a lot after I upgraded the diffuser to 0.4.1. It costs about **4.2 seconds** while unet cost... | closed | completed | false | 2 | [
"bug",
"stale"
] | [] | 2022-10-09T03:37:27Z | 2022-11-09T21:09:01Z | 2022-11-09T21:09:01Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | joey12300 | 10,826,371 | MDQ6VXNlcjEwODI2Mzcx | User | false |
huggingface/diffusers | 2,276,155,287 | I_kwDOHa8MBc6Hq1-X | 7,850 | https://github.com/huggingface/diffusers/issues/7850 | https://api.github.com/repos/huggingface/diffusers/issues/7850 | Severe difference with A1111 | ### Describe the bug
I am using a custom model. On A1111 it is far more colorful than on diffusers. I am aware that it's impossible to replicate images between the 2 with the same input, but my observation is across many examples
here's results for diffusers with 1 prompt accross guidance scale:
![image](https://g... | closed | completed | false | 6 | [
"bug"
] | [] | 2024-05-02T17:59:30Z | 2024-07-08T11:56:30Z | 2024-05-04T19:26:10Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | alexblattner | 15,870,094 | MDQ6VXNlcjE1ODcwMDk0 | User | false |
huggingface/diffusers | 2,276,720,071 | I_kwDOHa8MBc6Hs_3H | 7,851 | https://github.com/huggingface/diffusers/issues/7851 | https://api.github.com/repos/huggingface/diffusers/issues/7851 | Index 11 is out of bounds for dimension 0 with size 11 on concurrent pipeline calls | ### Describe the bug
Hello,
I've tried to create a web server which handles multiple requests from users. When I run multiple requests at the same time, the executions always aborts with the following error:
```
index 11 is out of bounds for dimension 0 with size 11
```
The number depends on the value I've ... | closed | completed | false | 1 | [
"bug"
] | [] | 2024-05-03T00:51:22Z | 2024-05-03T01:05:31Z | 2024-05-03T01:05:11Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | dnish | 7,636,313 | MDQ6VXNlcjc2MzYzMTM= | User | false |
huggingface/diffusers | 2,277,609,117 | I_kwDOHa8MBc6HwY6d | 7,854 | https://github.com/huggingface/diffusers/issues/7854 | https://api.github.com/repos/huggingface/diffusers/issues/7854 | RuntimeError: Input type (c10::Half) and bias type (float) should be the same | ### Describe the bug
```bash
File "/usr/local/lib/python3.10/dist-packages/diffusers/pipelines/ledits_pp/pipeline_leditspp_stable_diffusion_xl.py", line 1422, in encode_image
x0 = self.vae.encode(image).latent_dist.mode()
File "/usr/local/lib/python3.10/dist-packages/diffusers/utils/accelerate_utils.py", li... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-05-03T12:45:42Z | 2024-05-08T02:45:03Z | 2024-05-08T02:45:03Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | kadirnar | 36,204,372 | MDQ6VXNlcjM2MjA0Mzcy | User | false |
huggingface/diffusers | 2,278,934,698 | I_kwDOHa8MBc6H1ciq | 7,856 | https://github.com/huggingface/diffusers/issues/7856 | https://api.github.com/repos/huggingface/diffusers/issues/7856 | Multi-vector Token already in tokenizer vocabulary. Please choose a different token name | ### Describe the bug
I'm trying to load a embedding in a pipe then i unload it after that the same embedding i'm loading so it's giving me the error
![image](https://github.com/huggingface/diffusers/assets/87... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-05-04T10:58:46Z | 2024-05-05T09:14:49Z | 2024-05-05T09:14:48Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | 7iFinalBoSS | 87,640,946 | MDQ6VXNlcjg3NjQwOTQ2 | User | false |
huggingface/diffusers | 2,279,251,568 | I_kwDOHa8MBc6H2p5w | 7,859 | https://github.com/huggingface/diffusers/issues/7859 | https://api.github.com/repos/huggingface/diffusers/issues/7859 | FlashFace implementation in diffusers | you can implement flashface in img2img pipeline or inpainting
https://github.com/ali-vilab/FlashFace | open | null | false | 1 | [
"community-examples",
"Good second issue",
"contributions-welcome"
] | [] | 2024-05-05T00:20:50Z | 2024-11-20T00:08:44Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | loboere | 15,265,895 | MDQ6VXNlcjE1MjY1ODk1 | User | false |
huggingface/diffusers | 2,279,314,561 | I_kwDOHa8MBc6H25SB | 7,860 | https://github.com/huggingface/diffusers/issues/7860 | https://api.github.com/repos/huggingface/diffusers/issues/7860 | Support Residual Classifier-Free Guidance (RCFG) ? | **Is your feature request related to a problem? Please describe.**
In text2img/img2img pipelines, we use Classifier-free guidance (CFG) to enhance the effect of the condition, which requires 2n computations of model (U-Net): unconditioning term and the original conditioning term.
**Describe the solution you'd like.... | closed | completed | false | 0 | [] | [] | 2024-05-05T04:15:27Z | 2024-06-06T08:50:25Z | 2024-06-06T08:50:25Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | soryxie | 78,358,442 | MDQ6VXNlcjc4MzU4NDQy | User | false |
huggingface/diffusers | 2,279,659,834 | I_kwDOHa8MBc6H4Nk6 | 7,863 | https://github.com/huggingface/diffusers/issues/7863 | https://api.github.com/repos/huggingface/diffusers/issues/7863 | diffusers soft inpainting support | SDXL inpaint quality is really poor
but it can solve with like A1111 webui support soft inpainting feature
is there any support plan to soft inpaint?
| closed | completed | false | 5 | [] | [] | 2024-05-05T18:15:38Z | 2024-05-10T18:12:08Z | 2024-05-10T18:12:08Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | nullarch | 8,098,724 | MDQ6VXNlcjgwOTg3MjQ= | User | false |
huggingface/diffusers | 2,279,875,985 | I_kwDOHa8MBc6H5CWR | 7,864 | https://github.com/huggingface/diffusers/issues/7864 | https://api.github.com/repos/huggingface/diffusers/issues/7864 | MotionMaster: Training-free Camera Motion Transfer For Video Generation | ### Model/Pipeline/Scheduler description
Currently, most existing camera motion control methods for video generation with denoising diffusion models rely on training a temporal camera module, and necessitate substantial computation resources due to the large amount of parameters in video generation models.
The auth... | open | null | false | 1 | [
"community-examples",
"Good second issue",
"contributions-welcome"
] | [] | 2024-05-06T01:54:10Z | 2024-11-20T00:08:56Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | clarencechen | 8,482,341 | MDQ6VXNlcjg0ODIzNDE= | User | false |
huggingface/diffusers | 1,402,238,518 | I_kwDOHa8MBc5TlHY2 | 787 | https://github.com/huggingface/diffusers/issues/787 | https://api.github.com/repos/huggingface/diffusers/issues/787 | DefaultCPUAllocator : not enough memory | > RuntimeError: [enforce fail at C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\c10\core\impl\alloc_cpu.cpp:81] data. DefaultCPUAllocator: not enough memory: you tried to allocate 1073741824 bytes.
I have 16gb of RAM (12.9 usable) which i increased by making a pagefile but still receiving the above ... | closed | completed | false | 0 | [] | [] | 2022-10-09T11:03:01Z | 2022-10-09T11:03:39Z | 2022-10-09T11:03:39Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | brentleywilson | 91,980,756 | U_kgDOBXuD1A | User | false |
huggingface/diffusers | 2,281,368,586 | I_kwDOHa8MBc6H-uwK | 7,871 | https://github.com/huggingface/diffusers/issues/7871 | https://api.github.com/repos/huggingface/diffusers/issues/7871 | Error when using blockwise scales with sd_xl_offset_example-lora_1.0.safetensors | ### Describe the bug
When loading the [sd_xl_offset_example-lora_1.0.safetensors](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/sd_xl_offset_example-lora_1.0.safetensors) from the [official repository](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/) and setting the scales ... | closed | completed | false | 3 | [
"bug"
] | [] | 2024-05-06T17:25:45Z | 2024-05-07T11:54:58Z | 2024-05-07T11:54:58Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | asomoza | 5,442,875 | MDQ6VXNlcjU0NDI4NzU= | User | false |
huggingface/diffusers | 2,281,659,038 | I_kwDOHa8MBc6H_1qe | 7,872 | https://github.com/huggingface/diffusers/issues/7872 | https://api.github.com/repos/huggingface/diffusers/issues/7872 | diffusers0.19.3 | ### Describe the bug
diffusers0.19.3，in pipeline_controlnet.py，
in \__call__(),`cross_attention_kwargs: Optional[Dict[str, Any]] = None,`,
but pipe(cross_attention_kwargs={"scale": 0.8}) has problem:
File "/lib/python3.8/site-packages/diffusers/models/attention_processor.py", line 322, in forward
return self.p... | closed | completed | false | 3 | [
"bug"
] | [] | 2024-05-06T20:04:18Z | 2024-10-20T23:13:22Z | 2024-10-20T23:13:22Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | icedCoffe001 | 116,945,973 | U_kgDOBvh0NQ | User | false |
huggingface/diffusers | 2,282,714,516 | I_kwDOHa8MBc6ID3WU | 7,877 | https://github.com/huggingface/diffusers/issues/7877 | https://api.github.com/repos/huggingface/diffusers/issues/7877 | Support out_dim argument for Attention block | **Is your feature request related to a problem? Please describe.**
When i feed the `out_dim` argument in `__init__` in [Attention block](https://github.com/huggingface/diffusers/blob/b69fd990ad8026f21893499ab396d969b62bb8cc/src/diffusers/models/attention_processor.py#L114) it will raise the shape error, because the `... | open | null | false | 4 | [
"stale"
] | [] | 2024-05-07T09:14:36Z | 2024-12-14T15:06:29Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | tigerlittle1 | 44,134,547 | MDQ6VXNlcjQ0MTM0NTQ3 | User | false |
huggingface/diffusers | 2,283,040,346 | I_kwDOHa8MBc6IFG5a | 7,879 | https://github.com/huggingface/diffusers/issues/7879 | https://api.github.com/repos/huggingface/diffusers/issues/7879 | pokemon-blip-captions dataset | ### Describe the bug
[The pokemon-blip-captions dataset](https://huggingface.co/datasets/lambdalabs/pokemon-blip-captions) is no longer available. I thought that this was not about this library but documentation is written by using it, so it is a documentation issue.
### Who can help?
@stevhliu | closed | completed | false | 7 | [
"bug"
] | [] | 2024-05-07T11:38:57Z | 2024-05-08T06:50:41Z | 2024-05-07T16:36:40Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | tolgacangoz | 46,008,593 | MDQ6VXNlcjQ2MDA4NTkz | User | false |
huggingface/diffusers | 2,283,317,613 | I_kwDOHa8MBc6IGKlt | 7,881 | https://github.com/huggingface/diffusers/issues/7881 | https://api.github.com/repos/huggingface/diffusers/issues/7881 | lpw_stable_diffusion pipeline not working when "from_single_file" is used | ### Describe the bug
I'm using `StableDiffusionPipeline` with [lpw_stable_diffusion](https://github.com/huggingface/diffusers/blob/main/examples/community/lpw_stable_diffusion.py) to do txt2img generations from long prompts.
When loading the safetensors checkpoint file using `.from_single_file`, prompts longer th... | closed | completed | false | 8 | [
"bug"
] | [] | 2024-05-07T13:19:43Z | 2025-05-13T13:41:31Z | 2024-05-08T23:15:14Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | digitardz | 168,777,958 | U_kgDOCg9Y5g | User | false |
huggingface/diffusers | 2,283,837,124 | I_kwDOHa8MBc6IIJbE | 7,882 | https://github.com/huggingface/diffusers/issues/7882 | https://api.github.com/repos/huggingface/diffusers/issues/7882 | stable diffusion adapter pipeline for t2i adapter missing `from_single_file` | **Is your feature request related to a problem? Please describe.**
I want to use and test out the t2i adapter with my safetensor model. I used to use diffusers models, but none work since the update. Many of the other adapters use `from_single_file` to load safetensors correctly. I keep bringing this up, but the curre... | open | null | false | 18 | [
"help wanted",
"Good second issue",
"contributions-welcome",
"single_file"
] | [
"yiyixuxu"
] | 2024-05-07T17:02:09Z | 2024-11-20T00:08:24Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | JemiloII | 5,192,788 | MDQ6VXNlcjUxOTI3ODg= | User | false |
huggingface/diffusers | 2,284,725,205 | I_kwDOHa8MBc6ILiPV | 7,884 | https://github.com/huggingface/diffusers/issues/7884 | https://api.github.com/repos/huggingface/diffusers/issues/7884 | changing stable cascade sampler | Hi
According to this [link](https://civitai.com/articles/5111/stable-cascade-sampler-config) . I want to test other samplers in stable cascade. but I get the below error.
my in... | closed | completed | false | 1 | [] | [] | 2024-05-08T05:42:52Z | 2024-08-20T19:56:54Z | 2024-08-20T19:56:54Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | saeedkhanehgir | 65,589,645 | MDQ6VXNlcjY1NTg5NjQ1 | User | false |
huggingface/diffusers | 2,284,866,838 | I_kwDOHa8MBc6IME0W | 7,887 | https://github.com/huggingface/diffusers/issues/7887 | https://api.github.com/repos/huggingface/diffusers/issues/7887 | Strange Results in Latent to Diffusion | Hello. I am trying to use text to image in the XL model to output latents and then use those latents to perform diffusion again. I used the output_type="latent" provided by diffusers to receive the latent as a return value and then used it to generate an image, but strange results are appearing. The strange results are... | closed | completed | false | 4 | [] | [] | 2024-05-08T07:16:46Z | 2024-05-08T11:21:33Z | 2024-05-08T11:21:33Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | nullarch | 8,098,724 | MDQ6VXNlcjgwOTg3MjQ= | User | false |
huggingface/diffusers | 2,285,000,922 | I_kwDOHa8MBc6IMlja | 7,889 | https://github.com/huggingface/diffusers/issues/7889 | https://api.github.com/repos/huggingface/diffusers/issues/7889 | single_file_utils.py#load_single_file_model_checkpoint bugs | ### Describe the bug
Loading a checkpoint from file crashes as it tries to print a warning with unexpected keys because is accessing `__name__` on incorrect target (instance instead of class)
https://github.com/huggingface/diffusers/blob/d50baf0c632342b9576a24352244c4235ce8b875/src/diffusers/loaders/single_file_u... | open | null | false | 5 | [
"bug",
"stale"
] | [] | 2024-05-08T08:25:20Z | 2024-09-14T15:13:12Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | knoopx | 100,993 | MDQ6VXNlcjEwMDk5Mw== | User | false |
huggingface/diffusers | 2,285,542,409 | I_kwDOHa8MBc6IOpwJ | 7,891 | https://github.com/huggingface/diffusers/issues/7891 | https://api.github.com/repos/huggingface/diffusers/issues/7891 | [Exception]Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! | ### Describe the bug
Scenery: Developing a face editing server using fastapi+celery+diffusers, each request will generate a celery task.
Concurrence: 2 workers for fastapi, 1 worker for celery
Diffusers usage:
- 1 StableDiffusionControlNetPipeline + 3 StableDiffusionControlNetInpaintPipeline
- share two che... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-05-08T13:00:11Z | 2024-05-08T18:09:13Z | 2024-05-08T17:03:47Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | weiweiwang | 822,994 | MDQ6VXNlcjgyMjk5NA== | User | false |
huggingface/diffusers | 2,286,720,006 | I_kwDOHa8MBc6ITJQG | 7,895 | https://github.com/huggingface/diffusers/issues/7895 | https://api.github.com/repos/huggingface/diffusers/issues/7895 | Flax's use_memory_efficient_attention is broken | ### Describe the bug
I accidentally introduced a bug in this [PR](https://github.com/huggingface/diffusers/pull/5181) by making a condition on [this line](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_flax.py#L238), which is needed when `use_memory_efficient_attention=True`.
Addi... | closed | completed | false | 1 | [
"bug",
"stale"
] | [] | 2024-05-09T01:16:38Z | 2024-12-14T16:45:46Z | 2024-12-14T16:45:46Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | entrpn | 6,580,675 | MDQ6VXNlcjY1ODA2NzU= | User | false |
huggingface/diffusers | 2,286,924,393 | I_kwDOHa8MBc6IT7Jp | 7,897 | https://github.com/huggingface/diffusers/issues/7897 | https://api.github.com/repos/huggingface/diffusers/issues/7897 | text_to_image multi-gpu not working | We are training text_to_image on Google cloud platform, the jupyterlab instance has 2 GPUs (NVIDIA Tesla P100) with a total memory of 32GB (16GB each). I tried using accelerate for training the text_to_image model for multi_gpu support. But still getting out of memory error. Even with 32GB, I don't understand why its o... | open | null | false | 6 | [
"stale"
] | [] | 2024-05-09T05:31:09Z | 2024-09-14T15:13:05Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Sunflower54 | 169,332,708 | U_kgDOChfP5A | User | false |
huggingface/diffusers | 1,402,432,085 | I_kwDOHa8MBc5Tl2pV | 790 | https://github.com/huggingface/diffusers/issues/790 | https://api.github.com/repos/huggingface/diffusers/issues/790 | lack of memory clear causes some GPU tests to fail in test_models_unet.py | ### Describe the bug
running pytest if I suppress test_attention_block_default, two additional unet tests pass
```
test_layers_utils.py ................. ... | closed | completed | false | 10 | [
"bug"
] | [
"anton-l"
] | 2022-10-09T23:27:52Z | 2022-10-27T12:57:42Z | 2022-10-27T12:57:41Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Thomas-MMJ | 112,830,596 | U_kgDOBrmohA | User | false |
huggingface/diffusers | 2,287,809,119 | I_kwDOHa8MBc6IXTJf | 7,900 | https://github.com/huggingface/diffusers/issues/7900 | https://api.github.com/repos/huggingface/diffusers/issues/7900 | Importing error; modules not added to __init__.py | ### Describe the bug
Dear Hugging Face Team,
I wanted to ask if there is another way to import the `FromOriginalControlnetMixin`, `UNetMidBlockSpatioTemporal`, and `get_down_block functions`. They aren’t included in the `__init__.py files` in the current `0.27.2v`, so I can't access them through the main package an... | closed | completed | false | 2 | [
"bug"
] | [] | 2024-05-09T14:25:27Z | 2024-05-09T14:58:25Z | 2024-05-09T14:58:19Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | oooolga | 5,509,539 | MDQ6VXNlcjU1MDk1Mzk= | User | false |
huggingface/diffusers | 2,288,349,861 | I_kwDOHa8MBc6IZXKl | 7,902 | https://github.com/huggingface/diffusers/issues/7902 | https://api.github.com/repos/huggingface/diffusers/issues/7902 | new from_single_file implementation is always internet-first and using local files only on timeout | ### Describe the bug
new `from_single_file` implementation is always internet-first and using local config files only after timeout even if they are already present locally as they were previously downloaded.
code flow is `from_single_file` -> `_download_diffusers_model_config_from_hub` -> `snapshot_download`
... | closed | completed | false | 14 | [
"bug"
] | [] | 2024-05-09T19:51:37Z | 2024-05-22T12:05:41Z | 2024-05-22T12:05:41Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | vladmandic | 57,876,960 | MDQ6VXNlcjU3ODc2OTYw | User | false |
huggingface/diffusers | 2,288,767,667 | I_kwDOHa8MBc6Ia9Kz | 7,904 | https://github.com/huggingface/diffusers/issues/7904 | https://api.github.com/repos/huggingface/diffusers/issues/7904 | pipe.scheduler.sigmas change after forward call | ### Describe the bug
I noticed this issue when trying to manually modify the sigmas in my pipeline scheduler. Despite changing the values in this tensor, the sigmas tensor shrinks to the number of inference steps after one inference instance. Subsequent calls no longer change the values of the sigmas.
### Reproducti... | open | null | false | 6 | [
"bug",
"stale"
] | [] | 2024-05-10T02:01:06Z | 2024-09-14T15:12:59Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | neuron-party | 96,799,331 | U_kgDOBcUKYw | User | false |
huggingface/diffusers | 2,289,327,304 | I_kwDOHa8MBc6IdFzI | 7,908 | https://github.com/huggingface/diffusers/issues/7908 | https://api.github.com/repos/huggingface/diffusers/issues/7908 | Add aspect ratio bucketing to training scripts | **Is your feature request related to a problem? Please describe.**
When fine tuning SDXL, images are required to be a fixed size (1024x1024) which involves a lot of cropping that both takes time/resources, and often causes important parts of the image to get cropped out, which lowers model quality.
**Describe the ... | open | null | false | 11 | [
"stale"
] | [] | 2024-05-10T09:07:33Z | 2024-09-14T15:12:56Z | null | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | jferments | 158,022,198 | U_kgDOCWs6Ng | User | false |
huggingface/diffusers | 2,289,638,537 | I_kwDOHa8MBc6IeRyJ | 7,909 | https://github.com/huggingface/diffusers/issues/7909 | https://api.github.com/repos/huggingface/diffusers/issues/7909 | Support Lumina T2I 5B flow matching T2I DiT model | # Lumina-T2X: Transforming Text into Any Modality, Resolution, and Duration via Flow-based Large Diffusion Transformers
> Sora unveils the potential of scaling Diffusion Transformer (DiT) for generating photorealistic images and videos at arbitrary resolutions, aspect ratios, and durations, yet it still lacks suffic... | open | null | false | 7 | [
"stale"
] | [] | 2024-05-10T12:27:05Z | 2024-12-14T15:06:21Z | null | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | AmericanPresidentJimmyCarter | 110,263,573 | U_kgDOBpJ9FQ | User | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.