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 | 1,899,391,810 | I_kwDOHa8MBc5xNmtC | 5,071 | https://github.com/huggingface/diffusers/issues/5071 | https://api.github.com/repos/huggingface/diffusers/issues/5071 | Add ControlNet for Würstchen and examples\train_controlnet_wuerstchen.py | [Introducing Würstchen: Fast Diffusion for Image Generation](https://huggingface.co/blog/wuerstchen).
**Describe the solution you'd like**
Please implement ControlNet pipeline and training script for Würstchen.
Würstchen requires only 12x12 latent space instead of 64x64 compared to Stable Diffusion. If this also m... | closed | completed | false | 10 | [
"stale"
] | [] | 2023-09-16T10:44:07Z | 2024-06-05T12:05:30Z | 2023-11-08T15:07:52Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | geroldmeisinger | 112,266,044 | U_kgDOBrELPA | User | false |
huggingface/diffusers | 1,899,416,584 | I_kwDOHa8MBc5xNswI | 5,072 | https://github.com/huggingface/diffusers/issues/5072 | https://api.github.com/repos/huggingface/diffusers/issues/5072 | pred_text is prompt pass the unet?pred_uncond is is negative_prompt pass the unet? | pred = pred_uncond + guidance_scale * (pred_text - pred_uncond)
why use this calculation? | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-16T12:12:14Z | 2023-10-30T15:06:17Z | 2023-10-30T15:06:17Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | henbucuoshanghai | 37,647,094 | MDQ6VXNlcjM3NjQ3MDk0 | User | false |
huggingface/diffusers | 1,899,597,595 | I_kwDOHa8MBc5xOY8b | 5,074 | https://github.com/huggingface/diffusers/issues/5074 | https://api.github.com/repos/huggingface/diffusers/issues/5074 | Support for PyTorch 2.1 torch.export To Load / Save Compiled Stable Diffusion Models | Diffusers supports PyTorch 2.0's torch.compile which yields significant performance gains. However, the torch2.0 implementation did not include support for saving and loading the compiled result, which unfortunately limits its use in some environments. An example would be a production environment where new GPUs need to... | closed | completed | false | 8 | [
"stale"
] | [] | 2023-09-16T23:34:21Z | 2024-04-12T22:13:15Z | 2023-11-20T15:06:19Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | djj0s3 | 1,120,027 | MDQ6VXNlcjExMjAwMjc= | User | false |
huggingface/diffusers | 1,899,722,558 | I_kwDOHa8MBc5xO3c- | 5,075 | https://github.com/huggingface/diffusers/issues/5075 | https://api.github.com/repos/huggingface/diffusers/issues/5075 | Using Prompt Embeddings with SDXL Refiner Triggers RuntimeError: mat1 and mat2 shapes cannot be multiplied (154x2048 and 1280x768) | ### Describe the bug
I'm following SDXL code provided in the documentation here: [Base + Refiner Model](https://huggingface.co/docs/diffusers/using-diffusers/sdxl#base-refiner-model), except that I'm combining it with Compel to get the prompt embeddings. The refiner inference triggers the error:
> RuntimeError: m... | closed | completed | false | 9 | [
"bug"
] | [] | 2023-09-17T09:41:38Z | 2024-01-12T05:06:39Z | 2023-09-24T08:20:12Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | alexisrolland | 13,064,696 | MDQ6VXNlcjEzMDY0Njk2 | User | false |
huggingface/diffusers | 1,900,029,940 | I_kwDOHa8MBc5xQCf0 | 5,079 | https://github.com/huggingface/diffusers/issues/5079 | https://api.github.com/repos/huggingface/diffusers/issues/5079 | Support for Loading Lyrocis by load_lora_weights | **Is your feature request related to a problem? Please describe.**
I am trying to load Lyrocis from https://civitai.com/models/76404/lycoris-couture-an-edg-collection by load_lora_weights. It shows the error.
**Describe the solution you'd like**
We can successfully load LyROCIS
![image](https://github.com/... | closed | completed | false | 5 | [] | [] | 2023-09-18T01:43:31Z | 2023-09-20T09:30:20Z | 2023-09-20T09:30:20Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | George0726 | 38,740,075 | MDQ6VXNlcjM4NzQwMDc1 | User | false |
huggingface/diffusers | 1,900,414,163 | I_kwDOHa8MBc5xRgTT | 5,081 | https://github.com/huggingface/diffusers/issues/5081 | https://api.github.com/repos/huggingface/diffusers/issues/5081 | LPW XL : mat1 and mat2 shapes cannot be multiplied | ### Describe the bug
When trying to use SDXLLongPromptWeightingPipeline, and more than 2 samples, it gives the below error
> mat1 and mat2 shapes cannot be multiplied
### Reproduction
```
from diffusers import DiffusionPipeline
import torch
pipe = DiffusionPipeline.from_pretrained(
"stabilityai/st... | closed | completed | false | 5 | [
"bug",
"stale"
] | [] | 2023-09-18T08:20:03Z | 2023-10-30T15:06:14Z | 2023-10-30T15:06:14Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | adhikjoshi | 11,740,719 | MDQ6VXNlcjExNzQwNzE5 | User | false |
huggingface/diffusers | 1,900,464,160 | I_kwDOHa8MBc5xRsgg | 5,082 | https://github.com/huggingface/diffusers/issues/5082 | https://api.github.com/repos/huggingface/diffusers/issues/5082 | RuntimeError: mat1 and mat2 shapes cannot be multiplied (154x1024 and 768x320) | ### Describe the bug
I build a model named sd-dreambooth-library/snapp-g-data via this notebook:
https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_training.ipynb#scrollTo=30lu8LWXmg5j
i want to give it to the ControlNet model via this code:
https://colab.research.g... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2023-09-18T08:50:17Z | 2023-10-30T15:06:12Z | 2023-10-30T15:06:12Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | hosnasn1987 | 145,333,136 | U_kgDOCKmbkA | User | false |
huggingface/diffusers | 1,900,937,237 | I_kwDOHa8MBc5xTgAV | 5,084 | https://github.com/huggingface/diffusers/issues/5084 | https://api.github.com/repos/huggingface/diffusers/issues/5084 | FlaxStableDiffusionXLPipeline Import Error | I've tried to run a Stable Diffusion XL with a script with its pipeline in a TPU instance.
diffusers version:0.21.1
Python version:Python 3.8.10
I do my works in a Linux Ubuntu, there is no conda or other environment.
I take an import error in import line `from diffusers import FlaxStableDiffusionXLPipeline`... | closed | completed | false | 1 | [] | [] | 2023-09-18T13:24:28Z | 2023-09-19T20:45:07Z | 2023-09-19T20:45:07Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | gokerguner | 20,363,072 | MDQ6VXNlcjIwMzYzMDcy | User | false |
huggingface/diffusers | 1,901,507,579 | I_kwDOHa8MBc5xVrP7 | 5,091 | https://github.com/huggingface/diffusers/issues/5091 | https://api.github.com/repos/huggingface/diffusers/issues/5091 | console input, 2x GPU and enable_model_cpu_offload() | ### Describe the bug
I am using a script with an infinite loop and console input.
pipe.enable_model_cpu_offload(gpu_id=1)
The second time the proc() function is executed, the model is uploaded to GPU 0, although only GPU 1 is set to be used.
![Screenshot 2023-09-18 211026](https://github.com/huggingface/diffuse... | closed | completed | false | 6 | [
"bug",
"stale"
] | [] | 2023-09-18T18:24:50Z | 2023-11-08T15:07:46Z | 2023-11-08T15:07:46Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Yakonrus | 4,978,662 | MDQ6VXNlcjQ5Nzg2NjI= | User | false |
huggingface/diffusers | 1,901,564,003 | I_kwDOHa8MBc5xV5Bj | 5,092 | https://github.com/huggingface/diffusers/issues/5092 | https://api.github.com/repos/huggingface/diffusers/issues/5092 | list of outdated schedulers we need to remove/deprecate | I'm compiling a list of outdated schedulers we can remove/deprecate here. Below are just a few I have in mind. There are many schedulers we have that I'm not super familiar with. So please help me complete this list! 🤗🤗🤗
* [`KarrasVeScheduler`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/sche... | closed | completed | false | 2 | [
"stale"
] | [
"yiyixuxu"
] | 2023-09-18T19:00:00Z | 2023-11-08T15:07:44Z | 2023-11-08T15:07:44Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | yiyixuxu | 12,631,849 | MDQ6VXNlcjEyNjMxODQ5 | User | false |
huggingface/diffusers | 1,373,029,011 | I_kwDOHa8MBc5R1sKT | 510 | https://github.com/huggingface/diffusers/issues/510 | https://api.github.com/repos/huggingface/diffusers/issues/510 | SD ONNX remaining pipelines | It has caught my attention that you guys already made an onnx pipeline for SD but it seems to be txt2img.
To my knowledge no one has done it for img2img and inpaint yet.
Is that possible? If so, I'd like to request that feature.
Thanks for everything you have been doing so far! | closed | completed | false | 2 | [
"enhancement"
] | [] | 2022-09-14T13:53:28Z | 2022-11-07T14:47:46Z | 2022-11-07T14:47:46Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | GreenLandisaLie | 80,653,235 | MDQ6VXNlcjgwNjUzMjM1 | User | false |
huggingface/diffusers | 1,903,375,388 | I_kwDOHa8MBc5xczQc | 5,103 | https://github.com/huggingface/diffusers/issues/5103 | https://api.github.com/repos/huggingface/diffusers/issues/5103 | zero_snr DDIM causing black image: | I am also having this issue - to reproduce the black image:
```
pipe = StableDiffusionXLPipeline.from_pretrained(SDXL_PATH)
pipe.scheduler = DDIMScheduler.from_config(
pipe.scheduler.config, rescale_betas_zero_snr=True, timestep_spacing="trailing"
)
pipe.to("cuda")
prompt = "A lion in galaxi... | closed | completed | false | 11 | [
"stale"
] | [] | 2023-09-19T16:48:00Z | 2024-10-10T11:55:11Z | 2023-11-08T15:07:39Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | jloganolson | 14,350,126 | MDQ6VXNlcjE0MzUwMTI2 | User | false |
huggingface/diffusers | 1,903,489,661 | I_kwDOHa8MBc5xdPJ9 | 5,108 | https://github.com/huggingface/diffusers/issues/5108 | https://api.github.com/repos/huggingface/diffusers/issues/5108 | T2I Multiadapter XL exceptions | ### Describe the bug
I'm trying to replicate the t2i Multi-Adapter similar to the example here:
https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/adapter
but with the XL pipeline, but without much success. I always get one of two exceptions:
```
Traceback (most recent call last):
File "... | closed | completed | false | 6 | [
"bug"
] | [] | 2023-09-19T18:06:36Z | 2023-09-22T00:46:03Z | 2023-09-21T23:08:08Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | reimager | 20,918,094 | MDQ6VXNlcjIwOTE4MDk0 | User | false |
huggingface/diffusers | 1,904,012,859 | I_kwDOHa8MBc5xfO47 | 5,109 | https://github.com/huggingface/diffusers/issues/5109 | https://api.github.com/repos/huggingface/diffusers/issues/5109 | load_lora_weights for StableDiffusionXLControlNetImg2ImgPipeline raises exception | ### Describe the bug
calling load_lora_weights on a StableDiffusionXLControlNetImg2ImgPipeline instance raises an AttributeError.
### Reproduction
`wget https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_offset_example-lora_1.0.safetensors`
from diffusers import ControlNetModel... | closed | completed | false | 2 | [
"bug"
] | [] | 2023-09-20T03:03:12Z | 2023-09-25T10:08:07Z | 2023-09-25T10:08:07Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | melancholy | 231,413 | MDQ6VXNlcjIzMTQxMw== | User | false |
huggingface/diffusers | 1,904,059,944 | I_kwDOHa8MBc5xfaYo | 5,110 | https://github.com/huggingface/diffusers/issues/5110 | https://api.github.com/repos/huggingface/diffusers/issues/5110 | huggingface_hub.utils._validators.HFValidationError: Repo id must use alphanumeric chars | ### Describe the bug
I loaded the stable diffusion XL 1.0 model on my disk and followed instructions from
https://huggingface.co/docs/diffusers/v0.8.0/en/using-diffusers/loading#loading-pipelines-locally
The code is simply load the pretrained model.
```
import torch
from diffusers import DiffusionPipeline
fr... | closed | completed | false | 7 | [
"bug"
] | [] | 2023-09-20T03:49:55Z | 2023-09-25T18:22:42Z | 2023-09-25T18:22:42Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | kevint324 | 8,800,468 | MDQ6VXNlcjg4MDA0Njg= | User | false |
huggingface/diffusers | 1,904,157,188 | I_kwDOHa8MBc5xfyIE | 5,112 | https://github.com/huggingface/diffusers/issues/5112 | https://api.github.com/repos/huggingface/diffusers/issues/5112 | StableDiffusionXLControlNetImg2ImgPipeline Pipeline: torch.Size() takes an iterable of 'int' (item 4 is 'float') | ### Describe the bug
Error while using StableDiffusionXLControlNetImg2ImgPipeline with Refiner model and Depth ControlNet
```
812 def _get_add_time_ids(
813 self,
814 original_size,
(...)
822 dtype,
823 ):
824 if self.config.requires_aesthetics_score:
--> 825 add_time_... | closed | completed | false | 4 | [
"bug"
] | [] | 2023-09-20T05:16:47Z | 2023-09-20T16:31:11Z | 2023-09-20T16:31:11Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | shubhdotai | 6,880,985 | MDQ6VXNlcjY4ODA5ODU= | User | false |
huggingface/diffusers | 1,904,215,111 | I_kwDOHa8MBc5xgARH | 5,113 | https://github.com/huggingface/diffusers/issues/5113 | https://api.github.com/repos/huggingface/diffusers/issues/5113 | add clip skip for StableDiffusionLongPromptWeightingPipeline+LORA | **Is your feature request related to a problem? Please describe.**
StableDiffusionLongPromptWeightingPipeline.__call__() got an unexpected keyword argument 'clip_skip'
**Describe the solution you'd like**
add CLIP skip as described in #4834
no support for StableDiffusionLongPromptWeightingPipeline?
diffusers =... | closed | completed | false | 2 | [
"stale"
] | [] | 2023-09-20T06:01:11Z | 2023-11-08T15:07:37Z | 2023-11-08T15:07:37Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Charlse-s | 66,292,682 | MDQ6VXNlcjY2MjkyNjgy | User | false |
huggingface/diffusers | 1,904,266,286 | I_kwDOHa8MBc5xgMwu | 5,114 | https://github.com/huggingface/diffusers/issues/5114 | https://api.github.com/repos/huggingface/diffusers/issues/5114 | convert_original_stable_diffison script have issue | https://github.com/huggingface/diffusers/blob/8263cf00f832399bca215e29fa7572e0b0bde4da/scripts/convert_original_stable_diffusion_to_diffusers.py#L157C9-L157C40
config_files command line argment is not include in this script.
I think we need to add this code
parser.add_argument(
"--config_files",
... | closed | completed | false | 0 | [] | [] | 2023-09-20T06:38:46Z | 2023-09-20T14:17:48Z | 2023-09-20T14:17:48Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | nullarch | 8,098,724 | MDQ6VXNlcjgwOTg3MjQ= | User | false |
huggingface/diffusers | 1,904,466,786 | I_kwDOHa8MBc5xg9ti | 5,116 | https://github.com/huggingface/diffusers/issues/5116 | https://api.github.com/repos/huggingface/diffusers/issues/5116 | T2IAdapter | ### Describe the bug
in the class MultiAdapter(), the forward input xs should be chunked into n chunks (n=number of adapters)
### Reproduction
x_list = torch.chunk(xs, self.num_adapter, dim=1)
accume_state = None
for x, w, adapter in zip(x_list, adapter_weights, self.adapters):
in the line 88 in the adapter.py
... | closed | completed | false | 2 | [
"bug",
"stale"
] | [] | 2023-09-20T08:30:34Z | 2023-11-08T15:07:35Z | 2023-11-08T15:07:35Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | wanglaofei | 26,437,644 | MDQ6VXNlcjI2NDM3NjQ0 | User | false |
huggingface/diffusers | 1,904,665,616 | I_kwDOHa8MBc5xhuQQ | 5,118 | https://github.com/huggingface/diffusers/issues/5118 | https://api.github.com/repos/huggingface/diffusers/issues/5118 | how to use controlnet's reference_only fuction with diffusers?? | ### Model/Pipeline/Scheduler description
can anyone help me to understand how to use controlnet's reference_only fuction with diffusers
### Open source status
- [ ] The model implementation is available
- [ ] The model weights are available (Only relevant if addition is not a scheduler).
### Provide useful links fo... | closed | completed | false | 4 | [
"stale"
] | [] | 2023-09-20T10:17:53Z | 2023-11-08T15:07:34Z | 2023-11-08T15:07:34Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sudip550 | 74,753,373 | MDQ6VXNlcjc0NzUzMzcz | User | false |
huggingface/diffusers | 1,904,758,625 | I_kwDOHa8MBc5xiE9h | 5,120 | https://github.com/huggingface/diffusers/issues/5120 | https://api.github.com/repos/huggingface/diffusers/issues/5120 | Critical bug introduced in Train Dreambooth with prior preservation! | ### Describe the bug
`train_dreambooth.py` needs `num_samples` to not be 0 even though `with_prior_preservation` is False.
### Reproduction
I'm running the latest `train_dreambooth.py` as follows
```
import os
from smart_open import open
dreambooth_file = f"{MODEL_CACHE_FOLDER}/train_dreambooth... | closed | completed | false | 0 | [
"bug"
] | [] | 2023-09-20T11:11:10Z | 2023-09-20T11:26:53Z | 2023-09-20T11:26:53Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | landmann | 12,756,081 | MDQ6VXNlcjEyNzU2MDgx | User | false |
huggingface/diffusers | 1,905,153,919 | I_kwDOHa8MBc5xjld_ | 5,121 | https://github.com/huggingface/diffusers/issues/5121 | https://api.github.com/repos/huggingface/diffusers/issues/5121 | Help writing eval script | Hello,
I would like to write an evaluation script outside the auto111 ui that allows me to quick eval some trained sd models
I started looking in debug mode the webui code but it's not very commented and hard to follow
The features I need are the following:
- base model SDXL1.0 and refiner starting from 0.8
- DPM... | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-20T14:42:54Z | 2023-11-08T15:07:32Z | 2023-11-08T15:07:32Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | segalinc | 9,353,106 | MDQ6VXNlcjkzNTMxMDY= | User | false |
huggingface/diffusers | 1,905,812,252 | I_kwDOHa8MBc5xmGMc | 5,124 | https://github.com/huggingface/diffusers/issues/5124 | https://api.github.com/repos/huggingface/diffusers/issues/5124 | How to fine tune checkpoint .safetensor | ### Describe the bug
I tried to fine tuning a model from a checkpoint (i.e https://civitai.com/models/119202/talmendoxl-sdxl-uncensored-full-model)I converted the checkpoint to diffuser format using this library:
https://github.com/waifu-diffusion/sdxl-ckpt-converter/
The model converted works fine for inference a... | closed | completed | false | 3 | [
"bug",
"stale"
] | [
"sayakpaul"
] | 2023-09-20T22:45:38Z | 2023-11-22T15:06:19Z | 2023-11-22T15:06:19Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | EnricoBeltramo | 22,332,651 | MDQ6VXNlcjIyMzMyNjUx | User | false |
huggingface/diffusers | 1,906,004,973 | I_kwDOHa8MBc5xm1Pt | 5,125 | https://github.com/huggingface/diffusers/issues/5125 | https://api.github.com/repos/huggingface/diffusers/issues/5125 | CUDA error: invalid configuration argument when using XFormersAttnProcessor | ### Describe the bug
Hi, I'm trying to make a custom attention module by inheriting `diffusers.models.attention_processor.Attention`. The attention module works fine in the default case when the `AttnProcessor` is being used. However, when I tried to use xformers and the `Attention` class changed the processor to the ... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2023-09-21T03:07:31Z | 2023-11-08T15:07:28Z | 2023-11-08T15:07:28Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | wren93 | 87,381,590 | MDQ6VXNlcjg3MzgxNTkw | User | false |
huggingface/diffusers | 1,906,022,145 | I_kwDOHa8MBc5xm5cB | 5,126 | https://github.com/huggingface/diffusers/issues/5126 | https://api.github.com/repos/huggingface/diffusers/issues/5126 | TypeError: Conv2d.forward() takes 2 positional arguments but 3 were given | ### Describe the bug
Hi, I'm trying to do fine-tuning using [this repository](https://github.com/ExponentialML/Text-To-Video-Finetuning), but The error above comes up every time I execute `python3 train.py --config ./configs/v2/train_config.yaml` which is the command for starting training.
### Reproduction
After clo... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2023-09-21T03:31:55Z | 2023-11-09T15:05:26Z | 2023-11-09T15:05:26Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | kenkenissocool | 103,943,822 | U_kgDOBjIOjg | User | false |
huggingface/diffusers | 1,906,115,605 | I_kwDOHa8MBc5xnQQV | 5,128 | https://github.com/huggingface/diffusers/issues/5128 | https://api.github.com/repos/huggingface/diffusers/issues/5128 | error loading a local .safetensors using XLPipeline.from_single_file | ### Describe the bug
I have downloaded the sdxl-base-1.0 *.safetensors from https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/tree/main
I am trying to load the model following
```
sdxl_model_path = "/xxx/xxx/sd_xl_base_1.0.safetensors"
pipe = StableDiffusionXLPipeline.from_single_file(
... | closed | completed | false | 2 | [
"bug",
"stale"
] | [] | 2023-09-21T05:09:26Z | 2023-11-09T15:05:24Z | 2023-11-09T15:05:24Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | garychan22 | 108,175,311 | U_kgDOBnKfzw | User | false |
huggingface/diffusers | 1,906,292,102 | I_kwDOHa8MBc5xn7WG | 5,130 | https://github.com/huggingface/diffusers/issues/5130 | https://api.github.com/repos/huggingface/diffusers/issues/5130 | Question regarding DDPMScheduler | ### Describe the bug
The `current_beta_t` is computed from alphas_cumprod and alphas_cumprod_prev at the following line:
[https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_ddpm.py#L287](https://github.com/huggingface/diffusers/blob/main/src/diffusers/schedulers/scheduling_ddpm... | closed | completed | false | 7 | [
"bug",
"stale"
] | [] | 2023-09-21T07:23:09Z | 2023-11-13T15:06:09Z | 2023-11-13T15:06:09Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | xiankgx | 4,113,258 | MDQ6VXNlcjQxMTMyNTg= | User | false |
huggingface/diffusers | 1,907,483,686 | I_kwDOHa8MBc5xseQm | 5,133 | https://github.com/huggingface/diffusers/issues/5133 | https://api.github.com/repos/huggingface/diffusers/issues/5133 | `LoRAAttnProcessor2_0` getting deprecated and consequences for Dreambooth training | **What API design would you like to have changed or added to the library? Why?**
I'm getting the following warning while running the dreambooth sdxl lora example:
> diffusers/models/attention_processor.py:1570: FutureWarning: `LoRAAttnProcessor2_0` is deprecated and will be removed in version 0.26.0. Make sure us... | closed | completed | false | 6 | [] | [] | 2023-09-21T17:53:48Z | 2024-02-15T06:44:45Z | 2023-10-11T11:02:44Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | rvorias | 21,060,408 | MDQ6VXNlcjIxMDYwNDA4 | User | false |
huggingface/diffusers | 1,907,595,879 | I_kwDOHa8MBc5xs5pn | 5,135 | https://github.com/huggingface/diffusers/issues/5135 | https://api.github.com/repos/huggingface/diffusers/issues/5135 | LoRA quality degradation -> xformers? | ### Describe the bug
I have a forked diffusers repo that I try to keep up to date. My latest sync was from around ~3 weeks ago. This week I synced again and tried to do some quality tests with my Stable Diffusion XL LoRA inference engine. However, I noticed some really strange quality degradation.
This table corr... | closed | completed | false | 31 | [
"bug"
] | [] | 2023-09-21T19:04:49Z | 2023-09-27T16:16:34Z | 2023-09-27T16:16:34Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | rvorias | 21,060,408 | MDQ6VXNlcjIxMDYwNDA4 | User | false |
huggingface/diffusers | 1,907,681,092 | I_kwDOHa8MBc5xtOdE | 5,136 | https://github.com/huggingface/diffusers/issues/5136 | https://api.github.com/repos/huggingface/diffusers/issues/5136 | Possible discrepancy in instructpix2pix example script wrt dataset format | Hi,
I'm following the instruct_pix2pix train script [here](https://github.com/huggingface/diffusers/blob/main/examples/instruct_pix2pix/train_instruct_pix2pix.py) and facing some issues with getting a custom dataset to work with it. I checked the demo dataset link provided in the readme and it seems to be a parquet da... | closed | completed | false | 14 | [] | [] | 2023-09-21T20:06:39Z | 2024-12-31T02:00:06Z | 2023-09-28T12:18:03Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | gunshi | 8,274,753 | MDQ6VXNlcjgyNzQ3NTM= | User | false |
huggingface/diffusers | 1,907,724,404 | I_kwDOHa8MBc5xtZB0 | 5,137 | https://github.com/huggingface/diffusers/issues/5137 | https://api.github.com/repos/huggingface/diffusers/issues/5137 | Resnet_flax Downsample2D implementation is not backward compatible with the pre-trained Stable Diffusion v1.4 model | ### Describe the bug
TLDR: Resnet_flax Downsample2D implementation is not backward compatible with the pre-trained Stable Diffusion v1.4 model.
I downloaded the Stable diffusion v1.4 model (under the flax branch) and ran model inference with the current implementation. I ended up with very poor generation results.
... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2023-09-21T20:40:40Z | 2023-12-02T15:06:25Z | 2023-12-02T15:06:25Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | xcyan | 8,760,875 | MDQ6VXNlcjg3NjA4NzU= | User | false |
huggingface/diffusers | 1,907,819,435 | I_kwDOHa8MBc5xtwOr | 5,138 | https://github.com/huggingface/diffusers/issues/5138 | https://api.github.com/repos/huggingface/diffusers/issues/5138 | [Training] Graduate compute_snr function to `training_utils.py` | Originally posted by @sayakpaul in https://github.com/huggingface/diffusers/pull/5107#discussion_r1332162601.
| closed | completed | false | 0 | [] | [
"sayakpaul"
] | 2023-09-21T21:53:57Z | 2023-09-27T16:12:22Z | 2023-09-27T16:12:22Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sayakpaul | 22,957,388 | MDQ6VXNlcjIyOTU3Mzg4 | User | false |
huggingface/diffusers | 1,373,341,341 | I_kwDOHa8MBc5R24ad | 514 | https://github.com/huggingface/diffusers/issues/514 | https://api.github.com/repos/huggingface/diffusers/issues/514 | pipe.enable_attention_slicing()o throwing a error | ### Describe the bug
When I activate the pipe=pipe.enable_attention_slicing() for low memory
I get "TypeError: 'NoneType' object is not callable"
I'm using a model I trained with textual inversion does that impact that function ?
### Reproduction
_No response_
### Logs
_No response_
### System In... | closed | completed | false | 2 | [
"bug"
] | [] | 2022-09-14T17:25:43Z | 2022-09-17T12:26:46Z | 2022-09-17T12:26:46Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | UglyStupidHonest | 68,284,461 | MDQ6VXNlcjY4Mjg0NDYx | User | false |
huggingface/diffusers | 1,908,006,550 | I_kwDOHa8MBc5xud6W | 5,140 | https://github.com/huggingface/diffusers/issues/5140 | https://api.github.com/repos/huggingface/diffusers/issues/5140 | BLIP-Diffusion don't support multi controlnet | how to use multiple controlnet in blip-diffusion case | closed | completed | false | 7 | [
"stale"
] | [] | 2023-09-22T01:11:02Z | 2023-11-13T15:06:07Z | 2023-11-13T15:06:06Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | yanchaoguo | 11,827,802 | MDQ6VXNlcjExODI3ODAy | User | false |
huggingface/diffusers | 1,908,034,985 | I_kwDOHa8MBc5xuk2p | 5,141 | https://github.com/huggingface/diffusers/issues/5141 | https://api.github.com/repos/huggingface/diffusers/issues/5141 | Configure for checkpoints saving number | Checkpoint files saving number can be controlled by modifying [this line](https://github.com/huggingface/diffusers/blob/main/examples/instruct_pix2pix/train_instruct_pix2pix.py#L387) to
```
accelerator_project_config = ProjectConfiguration(
total_limit=args.checkpoints_total_limit,
automatic_checkpo... | closed | completed | false | 4 | [
"stale"
] | [] | 2023-09-22T01:31:21Z | 2023-11-13T15:06:05Z | 2023-11-13T15:06:05Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | deJQK | 25,779,973 | MDQ6VXNlcjI1Nzc5OTcz | User | false |
huggingface/diffusers | 1,908,159,520 | I_kwDOHa8MBc5xvDQg | 5,142 | https://github.com/huggingface/diffusers/issues/5142 | https://api.github.com/repos/huggingface/diffusers/issues/5142 | dimension mismatch errror | ### Describe the bug
images = self._pipe(
prompt_embeds=conditioning_prompt,
pooled_prompt_embeds=pooled_prompt,
negative_prompt_embeds=conditioning_neg_prompt,
negative_pooled_prompt_embeds=pooled_neg_prompt,
height=res[0... | closed | completed | false | 1 | [
"bug"
] | [] | 2023-09-22T04:25:32Z | 2023-09-24T06:01:28Z | 2023-09-24T06:01:28Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | linnanwang | 4,675,611 | MDQ6VXNlcjQ2NzU2MTE= | User | false |
huggingface/diffusers | 1,908,414,531 | I_kwDOHa8MBc5xwBhD | 5,143 | https://github.com/huggingface/diffusers/issues/5143 | https://api.github.com/repos/huggingface/diffusers/issues/5143 | same data input to the unet network，the result is tiny different，which cause the different of the output img。 | ### Describe the bug
same data input to the unet network，the result is tiny different，which cause the different of the output img。
### Reproduction
![image](https://github.com/huggingface/diffusers/assets/1204... | closed | completed | false | 2 | [
"bug",
"stale"
] | [] | 2023-09-22T08:13:55Z | 2023-11-13T15:06:03Z | 2023-11-13T15:06:03Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | tnickMoxuan | 120,445,480 | U_kgDOBy3aKA | User | false |
huggingface/diffusers | 1,908,981,326 | I_kwDOHa8MBc5xyL5O | 5,149 | https://github.com/huggingface/diffusers/issues/5149 | https://api.github.com/repos/huggingface/diffusers/issues/5149 | Provide safetensors verisons for StableDiffusionXLInstructPix2PixPipeline and StableDiffusionXLInpaintPipeline | ### Describe the bug
Both `StableDiffusionXLInstructPix2PixPipeline` and `StableDiffusionXLInpaintPipeline` pipelines do not implement `from_single_file` correctly
Failure is identical in both:
```log
│ /home/vlado/dev/sdnext/modules/sd_models.py:779 in load_diffuser ... | closed | completed | false | 11 | [
"enhancement"
] | [] | 2023-09-22T14:07:14Z | 2025-02-12T16:21:40Z | 2025-02-12T16:21:40Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | vladmandic | 57,876,960 | MDQ6VXNlcjU3ODc2OTYw | User | false |
huggingface/diffusers | 1,373,474,583 | I_kwDOHa8MBc5R3Y8X | 515 | https://github.com/huggingface/diffusers/issues/515 | https://api.github.com/repos/huggingface/diffusers/issues/515 | Cannot load custom pipeline modules defined in the main script | ### Describe the bug
`pipeline_util.register_modules` tries to retrieve `__module__` from pipeline modules and crashes for modules defined in the main class because the module `__main__` does not contain a `.`
### Reproduction
Create a pipeline with an own safetychecker class, e.g.:
```
class DummyChecker:
... | closed | completed | false | 10 | [] | [
"patrickvonplaten"
] | 2022-09-14T19:22:06Z | 2022-11-04T18:38:14Z | 2022-11-04T18:38:14Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | allo- | 1,482,341 | MDQ6VXNlcjE0ODIzNDE= | User | false |
huggingface/diffusers | 1,908,992,901 | I_kwDOHa8MBc5xyOuF | 5,150 | https://github.com/huggingface/diffusers/issues/5150 | https://api.github.com/repos/huggingface/diffusers/issues/5150 | StableDiffusionUpscalePipeline does not implement `from_single_file` method | ### Describe the bug
Exactly as title suggests - pipeline `StableDiffusionUpscalePipeline` does not implement `from_single_file` method so loading from safetensors is not possible.
### Reproduction
attempt to load a model from safetensors.
### Logs
```shell
N/A
```
### System Info
diffusers==0.21.2
### Who can... | closed | completed | false | 5 | [
"bug"
] | [
"DN6"
] | 2023-09-22T14:14:19Z | 2023-10-06T12:46:23Z | 2023-10-06T12:46:23Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | vladmandic | 57,876,960 | MDQ6VXNlcjU3ODc2OTYw | User | false |
huggingface/diffusers | 1,909,456,090 | I_kwDOHa8MBc5xz_za | 5,152 | https://github.com/huggingface/diffusers/issues/5152 | https://api.github.com/repos/huggingface/diffusers/issues/5152 | Würstchen bug in 0.21.3 alpha | ### Describe the bug
Generating an image with Würstchen 1024x1024 causes the error in the log.
### Reproduction
Generating an image with Würstchen 1024x1024 causes the error in the log.
### Logs
```shell
File "C:\Users\45239\AppData\Roaming\Python\Python310\site-packages\diffusers\configuration_utils.py", line 636... | closed | completed | false | 2 | [
"bug"
] | [] | 2023-09-22T20:01:52Z | 2023-09-28T11:08:41Z | 2023-09-28T11:08:41Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | tin2tin | 1,322,593 | MDQ6VXNlcjEzMjI1OTM= | User | false |
huggingface/diffusers | 1,909,540,453 | I_kwDOHa8MBc5x0UZl | 5,153 | https://github.com/huggingface/diffusers/issues/5153 | https://api.github.com/repos/huggingface/diffusers/issues/5153 | FreeU: Free Lunch in Diffusion U-Net | ### Model/Pipeline/Scheduler description
FreeU, a method that substantially improves diffusion model sample quality at no costs: no training, no additional parameter introduced, and no increase in memory or sampling time.
Promising results on image and video generation tasks demonstrate that our FreeU can be read... | closed | completed | false | 2 | [
"stale"
] | [] | 2023-09-22T21:30:07Z | 2023-11-09T15:05:20Z | 2023-11-09T15:05:20Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | tin2tin | 1,322,593 | MDQ6VXNlcjEzMjI1OTM= | User | false |
huggingface/diffusers | 1,909,931,558 | I_kwDOHa8MBc5x1z4m | 5,160 | https://github.com/huggingface/diffusers/issues/5160 | https://api.github.com/repos/huggingface/diffusers/issues/5160 | Dreambooth/LoRA Masking | **Is your feature request related to a problem? Please describe.**
Replicate implemented a [script](https://github.com/replicate/cog-sdxl/tree/main) to train SDXL that creates a mask over the face of a person, and applies that mask during the loss function of dreambooth training.
**Describe the solution you'd lik... | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-23T16:45:36Z | 2023-11-09T15:05:16Z | 2023-11-09T15:05:16Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | landmann | 12,756,081 | MDQ6VXNlcjEyNzU2MDgx | User | false |
huggingface/diffusers | 1,910,006,502 | I_kwDOHa8MBc5x2GLm | 5,162 | https://github.com/huggingface/diffusers/issues/5162 | https://api.github.com/repos/huggingface/diffusers/issues/5162 | Script to convert a folder checkpoint to .safetensors | **Is your feature request related to a problem? Please describe.**
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
I have trained a stable diffusion but the text2image.py script doesn't return a model_xxx.safetensors for the checkpoint but all separate .bin and unet, optim... | closed | completed | false | 6 | [
"stale"
] | [] | 2023-09-23T21:30:56Z | 2023-11-09T15:05:14Z | 2023-11-09T15:05:14Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | segalinc | 9,353,106 | MDQ6VXNlcjkzNTMxMDY= | User | false |
huggingface/diffusers | 1,910,014,934 | I_kwDOHa8MBc5x2IPW | 5,163 | https://github.com/huggingface/diffusers/issues/5163 | https://api.github.com/repos/huggingface/diffusers/issues/5163 | Inconsistent Results Between AUTOMATIC1111 UI and Diffusers Package Implementation of ControlNet | **Summary**
ControlNet does not seem to function identically between AUTOMATIC1111's UI and an implementation using the Diffusers package, despite using identical configuration and models. When using Diffusers, shapes overflow unexpectedly from the main object being inpainted, which is not the case when using AUTOMA... | closed | completed | false | 12 | [] | [
"yiyixuxu"
] | 2023-09-23T22:06:52Z | 2024-01-12T05:45:10Z | 2023-11-26T15:06:39Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sebi75 | 36,008,268 | MDQ6VXNlcjM2MDA4MjY4 | User | false |
huggingface/diffusers | 1,910,114,922 | I_kwDOHa8MBc5x2gpq | 5,165 | https://github.com/huggingface/diffusers/issues/5165 | https://api.github.com/repos/huggingface/diffusers/issues/5165 | SDXL Pretrain on 400K dataset | ### Describe the bug
Hello there,
I have collected 400K images and prompts and trying to pretrain a SDXL using the diffuser scripts. However, the current script is crazily slow in mapping stage. When I add num_procs in map to 10, there is another cuda error. Then I set spawn process, another error.
So, does th... | closed | completed | false | 1 | [
"bug"
] | [] | 2023-09-24T06:05:18Z | 2023-09-24T20:43:11Z | 2023-09-24T20:43:11Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | linnanwang | 4,675,611 | MDQ6VXNlcjQ2NzU2MTE= | User | false |
huggingface/diffusers | 1,910,146,758 | I_kwDOHa8MBc5x2obG | 5,166 | https://github.com/huggingface/diffusers/issues/5166 | https://api.github.com/repos/huggingface/diffusers/issues/5166 | Missing scaling factor of vqvae in unconditional ldm pipeline | ### Describe the bug
After encoding the images by the vae, the resulting latent is multiplied by its scaling factor. However, when latents are created after denoising they don't get multiplied by the inverse of the scaling factor in the unconditional pipeline which leads to garbage results if it is not added manually.... | closed | completed | false | 4 | [
"bug",
"stale"
] | [] | 2023-09-24T08:07:21Z | 2023-11-09T15:05:12Z | 2023-11-09T15:05:12Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | asparius | 72,755,761 | MDQ6VXNlcjcyNzU1NzYx | User | false |
huggingface/diffusers | 1,910,248,236 | I_kwDOHa8MBc5x3BMs | 5,167 | https://github.com/huggingface/diffusers/issues/5167 | https://api.github.com/repos/huggingface/diffusers/issues/5167 | Several small suggestions for examples\controlnet\train_controlnet.py | I know that quality-of-life improvements are low-priority for an "example" but I want to compile list of suggestions anyway in case someone wants to tackle it.
**Handle missing resources**
Some steps are pretty heavy (loading SD, loading CN) only to fail because of a missing resource down the line, which may not ... | closed | completed | false | 2 | [
"stale"
] | [] | 2023-09-24T14:13:21Z | 2023-11-08T15:07:26Z | 2023-11-08T15:07:26Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | geroldmeisinger | 112,266,044 | U_kgDOBrELPA | User | false |
huggingface/diffusers | 1,910,320,445 | I_kwDOHa8MBc5x3S09 | 5,168 | https://github.com/huggingface/diffusers/issues/5168 | https://api.github.com/repos/huggingface/diffusers/issues/5168 | ControlNet-XS support | ### Model/Pipeline/Scheduler description
Hello all,
Are you planning on supporting ControlNet-XS ? It's promising a better architecture with a much lower number of parameters for the same quality of results.
Thanks !
### Open source status
- [X] The model implementation is available
- [X] The model weights are... | closed | completed | false | 10 | [
"contributions-welcome"
] | [] | 2023-09-24T17:49:44Z | 2023-12-06T22:33:48Z | 2023-12-06T22:33:48Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | scarbain | 2,834,956 | MDQ6VXNlcjI4MzQ5NTY= | User | false |
huggingface/diffusers | 1,910,598,974 | I_kwDOHa8MBc5x4W0- | 5,170 | https://github.com/huggingface/diffusers/issues/5170 | https://api.github.com/repos/huggingface/diffusers/issues/5170 | Remove Lora after loading it | **Is your feature request related to a problem? Please describe.**
Hello there, not sure if this feature has been implemented or not, is there an api that removes lora after loading it? Thanks. | closed | completed | false | 4 | [
"stale"
] | [] | 2023-09-25T03:44:35Z | 2023-11-20T15:06:07Z | 2023-11-20T15:06:07Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | linnanwang | 4,675,611 | MDQ6VXNlcjQ2NzU2MTE= | User | false |
huggingface/diffusers | 1,910,624,333 | I_kwDOHa8MBc5x4dBN | 5,171 | https://github.com/huggingface/diffusers/issues/5171 | https://api.github.com/repos/huggingface/diffusers/issues/5171 | TypeError: StableDiffusionXLPipeline.__init__() got an unexpected keyword argument 'safety_checker | ### Describe the bug
```
Traceback (most recent call last):
File "/workspace/diffusers/scripts/convert_original_stable_diffusion_to_diffusers.py", line 160, in <module>
pipe = download_from_original_stable_diffusion_ckpt(
File "/usr/local/lib/python3.10/dist-packages/diffusers/pipelines/stable_diffusion/co... | closed | completed | false | 1 | [
"bug"
] | [] | 2023-09-25T04:16:21Z | 2023-09-25T17:29:43Z | 2023-09-25T17:29:43Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | andysingal | 20,493,493 | MDQ6VXNlcjIwNDkzNDkz | User | false |
huggingface/diffusers | 1,911,825,339 | I_kwDOHa8MBc5x9CO7 | 5,178 | https://github.com/huggingface/diffusers/issues/5178 | https://api.github.com/repos/huggingface/diffusers/issues/5178 | DPM SDE fails when resolution changes after the first run | ### Describe the bug
DPM SDE fails to generate when resolution changes after the first run.
Resetting the entire scheduler or resetting the BrownianTreeNoiseSampler after every generation works as a workaround.
But resetting after every generation kills the reproducibility.
As a 2nd method, changing this line... | closed | completed | false | 14 | [
"bug"
] | [] | 2023-09-25T15:55:32Z | 2023-10-02T17:01:20Z | 2023-10-02T17:01:20Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Disty0 | 47,277,141 | MDQ6VXNlcjQ3Mjc3MTQx | User | false |
huggingface/diffusers | 1,373,636,881 | I_kwDOHa8MBc5R4AkR | 519 | https://github.com/huggingface/diffusers/issues/519 | https://api.github.com/repos/huggingface/diffusers/issues/519 | logging.remove_handler() has a faulty assertion, doesn't allow registered handlers to be removed | ### Describe the bug
in `utils/logging.py` there the function remove_handler seems to have a faulty assertion in it.
```py
def add_handler(handler: logging.Handler) -> None:
"""adds a handler to the HuggingFace Diffusers' root logger."""
_configure_library_root_logger()
assert handler is not N... | closed | completed | false | 4 | [
"stale"
] | [] | 2022-09-14T22:03:54Z | 2022-10-31T15:03:10Z | 2022-10-31T15:03:10Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | SimplyLinn | 1,394,292 | MDQ6VXNlcjEzOTQyOTI= | User | false |
huggingface/diffusers | 1,914,142,336 | I_kwDOHa8MBc5yF36A | 5,191 | https://github.com/huggingface/diffusers/issues/5191 | https://api.github.com/repos/huggingface/diffusers/issues/5191 | CIVITAI Safetensors LoRA works worse on latest diffusers compared to 0.18.2 for SD1.5 | ### Describe the bug
I noticed that when I use some LoRA from CIVITAI, it
- breaks the image when using latest diffusers
- but does not necessarily work on 0.18.2 (but does not break the image)
- in 0.18.2, loading LoRA **but not using it in prompt** does not **automatically** gets applied to the image
- in lates... | closed | completed | false | 7 | [
"bug",
"stale"
] | [] | 2023-09-26T19:11:18Z | 2023-11-08T15:07:19Z | 2023-11-08T15:07:19Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sarmientoj24 | 8,830,319 | MDQ6VXNlcjg4MzAzMTk= | User | false |
huggingface/diffusers | 1,914,618,750 | I_kwDOHa8MBc5yHsN- | 5,193 | https://github.com/huggingface/diffusers/issues/5193 | https://api.github.com/repos/huggingface/diffusers/issues/5193 | How can I use a chinese clip model for CLIP Guided Stable Diffusion | As the title suggests, my code is as follows. I attempted to use a Chinese clip model to guide stable diffusion to generate images, but an error was reported. Several Chinese clip models were replaced, but they did not work. Do you have any requirements for clip model parameters. My code is written like this,
```p... | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-27T03:22:58Z | 2023-12-26T15:10:38Z | 2023-12-26T15:10:38Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | p-moon | 16,488,309 | MDQ6VXNlcjE2NDg4MzA5 | User | false |
huggingface/diffusers | 1,914,827,699 | I_kwDOHa8MBc5yIfOz | 5,195 | https://github.com/huggingface/diffusers/issues/5195 | https://api.github.com/repos/huggingface/diffusers/issues/5195 | Unloading specific LORA from loaded LORAs | I know that there is an `unload_lora_weights` function that exist. But is it possible to unload specific LORAs in the pipeline?
Too many LORAs loaded mess the image.
Also, it seems like Loading the LORA and not using it in the prompt already changes the image that it seemingly use the LoRA. | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-27T06:55:43Z | 2023-11-08T15:07:15Z | 2023-11-08T15:07:15Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sarmientoj24 | 8,830,319 | MDQ6VXNlcjg4MzAzMTk= | User | false |
huggingface/diffusers | 1,915,135,698 | I_kwDOHa8MBc5yJqbS | 5,197 | https://github.com/huggingface/diffusers/issues/5197 | https://api.github.com/repos/huggingface/diffusers/issues/5197 | Multi-GPU Training Runs Slower Than Single-GPU Training for SD1.5 text-to-image finetuning | ### Describe the bug
I am attempting to train SD1.5 ([model](runwayml/stable-diffusion-v1-5)) for text-to-image generation using the [train_text_to_image.py](https://github.com/huggingface/diffusers/blob/main/examples/text_to_image/train_text_to_image.py) script on a custom dataset. Given the large size of the datas... | closed | completed | false | 16 | [
"bug",
"stale"
] | [] | 2023-09-27T09:55:41Z | 2024-04-04T11:49:06Z | 2023-11-08T15:07:14Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | garg-aayush | 17,342,823 | MDQ6VXNlcjE3MzQyODIz | User | false |
huggingface/diffusers | 1,915,219,629 | I_kwDOHa8MBc5yJ-6t | 5,198 | https://github.com/huggingface/diffusers/issues/5198 | https://api.github.com/repos/huggingface/diffusers/issues/5198 | Add 1D diffusion support to UniPC scheduler | UniPC has been impressive in generating diffusion audio for both singing and speaking. However, the current UniPC implementation in the diffusers repo only supports 2D dimensions. Adding 1D support to the UniPC scheduler would be a great enhancement.
Example projects for reference:
- [DiffSinger](https://github.com... | closed | completed | false | 1 | [] | [] | 2023-09-27T10:33:15Z | 2023-10-02T17:17:47Z | 2023-10-02T17:17:47Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | leng-yue | 25,119,060 | MDQ6VXNlcjI1MTE5MDYw | User | false |
huggingface/diffusers | 1,373,767,858 | I_kwDOHa8MBc5R4giy | 520 | https://github.com/huggingface/diffusers/issues/520 | https://api.github.com/repos/huggingface/diffusers/issues/520 | Multiplying the Latents by Sigmas during LMS Scheduler | null | closed | completed | false | 2 | [
"bug"
] | [] | 2022-09-15T01:15:16Z | 2022-09-17T21:11:19Z | 2022-09-15T03:40:37Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Hritikbansal | 28,641,936 | MDQ6VXNlcjI4NjQxOTM2 | User | false |
huggingface/diffusers | 1,915,508,562 | I_kwDOHa8MBc5yLFdS | 5,200 | https://github.com/huggingface/diffusers/issues/5200 | https://api.github.com/repos/huggingface/diffusers/issues/5200 | Using pre-trained stable diffusion checkpoint to train temporal layers for text-to-video | Hey!
I've been trying to see if I can use the text-to-video-synthesis pipelines (or some of the modules they use) in diffusers to recreate common video diffusion methods from papers that all have the common feature of starting with a pertained stable diffusion model, and adding temporal attention layers that add resi... | closed | completed | false | 1 | [
"stale"
] | [] | 2023-09-27T12:45:13Z | 2023-11-06T16:36:29Z | 2023-11-06T16:36:29Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | gunshi | 8,274,753 | MDQ6VXNlcjgyNzQ3NTM= | User | false |
huggingface/diffusers | 1,915,992,599 | I_kwDOHa8MBc5yM7oX | 5,205 | https://github.com/huggingface/diffusers/issues/5205 | https://api.github.com/repos/huggingface/diffusers/issues/5205 | Repaint + ControlNet (+ LoRA/text inversion) with StableDiffusion | ### Model/Pipeline/Scheduler description
I have implemented a pipeline in order to be able to use at the same time ControlNet and Repaint (a method that doesn't require a specific inpainting model).
Specifically, I started from StableDiffusionRepaintPipeline and adapted it to ControlNet.
Would it be interesting ... | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-27T17:04:16Z | 2024-10-31T16:15:06Z | 2023-11-08T15:07:12Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | uminaty | 56,725,662 | MDQ6VXNlcjU2NzI1NjYy | User | false |
huggingface/diffusers | 1,916,207,490 | I_kwDOHa8MBc5yNwGC | 5,209 | https://github.com/huggingface/diffusers/issues/5209 | https://api.github.com/repos/huggingface/diffusers/issues/5209 | SDXL Controlnet&T2IAdapter Script generates black images for validation set | ### Describe the bug
Running the train script for SDXL t2i_adapter with the default setting and getting black results for the validation images.
### Reproduction
Here's the notebook I'm using:
https://colab.... | closed | completed | false | 3 | [
"bug"
] | [] | 2023-09-27T19:44:32Z | 2023-09-28T17:44:19Z | 2023-09-28T17:44:18Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | nbardy | 1,278,972 | MDQ6VXNlcjEyNzg5NzI= | User | false |
huggingface/diffusers | 1,916,454,244 | I_kwDOHa8MBc5yOsVk | 5,212 | https://github.com/huggingface/diffusers/issues/5212 | https://api.github.com/repos/huggingface/diffusers/issues/5212 | I get a message that I need to specify `pad_to_multiple_of` when loading an embed. | ```python
pipe.load_textual_inversion("embed/negative",weight_name="EasyNegativeV2.safetensors",token="EasyNegative")
```
When I ran the code above, I got the following message How can I resolve this issue?
```
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means ... | closed | completed | false | 8 | [
"stale"
] | [] | 2023-09-27T23:00:45Z | 2024-10-05T05:34:19Z | 2023-11-08T15:07:10Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | suzukimain | 131,413,573 | U_kgDOB9U2RQ | User | false |
huggingface/diffusers | 1,916,859,620 | I_kwDOHa8MBc5yQPTk | 5,215 | https://github.com/huggingface/diffusers/issues/5215 | https://api.github.com/repos/huggingface/diffusers/issues/5215 | StableDiffusionUpscalePipeline changes brightness significantly when source image is dark | ### Describe the bug
When using the StableDiffusionUpscalePipeline with the model "stabilityai/stable-diffusion-x4-upscaler" to upscale image of low brightness, the overall brightness in the upscaled result changes a lot.
input image:
![test](https://github.com/huggingface/diffusers/assets/136558089/0baf5d6f-e... | closed | completed | false | 4 | [
"bug"
] | [] | 2023-09-28T07:12:50Z | 2023-10-02T03:16:03Z | 2023-10-02T01:12:59Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | omegayal1234 | 136,558,089 | U_kgDOCCO2CQ | User | false |
huggingface/diffusers | 1,916,877,188 | I_kwDOHa8MBc5yQTmE | 5,216 | https://github.com/huggingface/diffusers/issues/5216 | https://api.github.com/repos/huggingface/diffusers/issues/5216 | 🐛[`core`/`GLIGEN`]: TypeError when iterating over 0-d tensor with In-painting mode when EulerAncestralDiscreteScheduler is used | ### Describe the bug
TypeError raised when trying to directly iterate over a 0-dimension tensor in the denoising stage of GLIGEN In-painting operation.
The error occurs when using diffusion noise schedulers that iterate over timesteps
(e.g. [EulerAnchestralDiscreteScheduler](https://huggingface.co/docs/diffusers/v... | closed | completed | false | 0 | [
"bug"
] | [] | 2023-09-28T07:24:39Z | 2023-12-12T18:45:18Z | 2023-10-09T07:54:03Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | chuzhdontcode | 50,682,966 | MDQ6VXNlcjUwNjgyOTY2 | User | false |
huggingface/diffusers | 1,917,123,330 | I_kwDOHa8MBc5yRPsC | 5,219 | https://github.com/huggingface/diffusers/issues/5219 | https://api.github.com/repos/huggingface/diffusers/issues/5219 | unable to load the sdxl safetensors model which was converted from SDXL diffusers | I converted "stablediffusionapi/juggernaut-xl-v5" this model to a single safetensors file using this (https://github.com/huggingface/diffusers/blob/main/scripts/convert_diffusers_to_original_sdxl.py). Conversion happened properly but unable to load it using StableDiffusionXLPipeline.from_single_file()
Below is the e... | closed | completed | false | 1 | [
"stale"
] | [] | 2023-09-28T09:50:37Z | 2023-11-08T15:07:06Z | 2023-11-08T15:07:06Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sumanthdonapati | 74,263,775 | MDQ6VXNlcjc0MjYzNzc1 | User | false |
huggingface/diffusers | 1,374,667,437 | I_kwDOHa8MBc5R78Kt | 522 | https://github.com/huggingface/diffusers/issues/522 | https://api.github.com/repos/huggingface/diffusers/issues/522 | [Community] Add dropout to Flax UNet | ### Describe the bug
Flax UNet currently doesn't support dropout
### Reproduction
_No response_
### Logs
_No response_
### System Info
Main | closed | completed | false | 3 | [
"bug",
"good first issue"
] | [] | 2022-09-15T14:57:13Z | 2023-07-07T09:38:17Z | 2023-07-07T09:38:17Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,917,232,665 | I_kwDOHa8MBc5yRqYZ | 5,221 | https://github.com/huggingface/diffusers/issues/5221 | https://api.github.com/repos/huggingface/diffusers/issues/5221 | BUG/Regression: VideoToVideoSDPipeline produces garbage | ### Describe the bug
Using the example from here, but changing the model to VideoToVideoSDPipeline for the latter part(as intended?), results in garbage. I guess this must be a regression, since it used to work for me?
https://huggingface.co/docs/diffusers/main/en/api/pipelines/text_to_video#diffusers.VideoToVideoS... | closed | completed | false | 1 | [
"bug"
] | [] | 2023-09-28T10:55:26Z | 2023-09-28T11:25:06Z | 2023-09-28T11:25:05Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | tin2tin | 1,322,593 | MDQ6VXNlcjEzMjI1OTM= | User | false |
huggingface/diffusers | 1,918,123,731 | I_kwDOHa8MBc5yVD7T | 5,224 | https://github.com/huggingface/diffusers/issues/5224 | https://api.github.com/repos/huggingface/diffusers/issues/5224 | Flax Stable Diffusion 2.1 error | ### Describe the bug
When trying to load Stable Diffusion 2.1 using Flax, I am getting the following error:
```bash
Traceback (most recent call last):
File "/home/jfacevedo/infer.py", line 120, in <module>
run(opt)
File "/home/jfacevedo/infer.py", line 30, in run
pipeline, params = FlaxStableDiffus... | closed | completed | false | 3 | [
"bug",
"stale"
] | [
"pcuenca",
"patrickvonplaten"
] | 2023-09-28T19:19:01Z | 2023-11-08T15:07:02Z | 2023-11-08T15:07:02Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | entrpn | 6,580,675 | MDQ6VXNlcjY1ODA2NzU= | User | false |
huggingface/diffusers | 1,918,245,335 | I_kwDOHa8MBc5yVhnX | 5,225 | https://github.com/huggingface/diffusers/issues/5225 | https://api.github.com/repos/huggingface/diffusers/issues/5225 | SDXL: Cannot run forward pass on 24G VRAM even when using Adafactor | ### Describe the bug
With khoya_ss, who have their own unet and flash attn mechanisms, it seems that it's perfectly possible to load the Adafactor optimiser and AdafactorSchedule, and run a forward pass in under 23G of VRAM, about 22.2G.
In Diffusers, this does not seem to be possible.
### Reproduction
The ... | closed | completed | false | 8 | [
"bug",
"stale"
] | [
"sayakpaul"
] | 2023-09-28T20:43:33Z | 2023-11-08T15:07:00Z | 2023-11-08T15:07:00Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | bghira | 59,658,056 | MDQ6VXNlcjU5NjU4MDU2 | User | false |
huggingface/diffusers | 1,918,629,383 | I_kwDOHa8MBc5yW_YH | 5,226 | https://github.com/huggingface/diffusers/issues/5226 | https://api.github.com/repos/huggingface/diffusers/issues/5226 | StableDiffusionXLAdapterPipeline does not work with load_lora_weights | ### Describe the bug
StableDiffusionXLAdapterPipeline does not work with load_lora_weights
### Reproduction
```
import torch
from diffusers import StableDiffusionXLAdapterPipeline, T2IAdapter
from PIL import Image
adapter = T2IAdapter.from_pretrained(
"TencentARC/t2i-adapter-openpose-sdxl-1.0", ... | closed | completed | false | 5 | [
"bug"
] | [
"yiyixuxu"
] | 2023-09-29T05:41:46Z | 2023-10-06T09:47:40Z | 2023-10-06T09:47:40Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ziniuwan | 88,946,368 | MDQ6VXNlcjg4OTQ2MzY4 | User | false |
huggingface/diffusers | 1,918,806,282 | I_kwDOHa8MBc5yXqkK | 5,227 | https://github.com/huggingface/diffusers/issues/5227 | https://api.github.com/repos/huggingface/diffusers/issues/5227 | load_lora_weight does not work with torch.compile | ### Describe the bug
load_lora_weight does not work with torch.compile. StableDiffusionPipeline & StableDiffusionXLPipeline both failed.
It's fine in version 0.18.2, and fails since 0.19.0.
### Reproduction
```
from diffusers import StableDiffusionPipeline
from PIL import Image
import torch
from safeten... | closed | completed | false | 5 | [
"bug"
] | [
"sayakpaul"
] | 2023-09-29T08:14:55Z | 2023-10-06T15:14:49Z | 2023-10-06T15:14:49Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ziniuwan | 88,946,368 | MDQ6VXNlcjg4OTQ2MzY4 | User | false |
huggingface/diffusers | 1,918,958,422 | I_kwDOHa8MBc5yYPtW | 5,228 | https://github.com/huggingface/diffusers/issues/5228 | https://api.github.com/repos/huggingface/diffusers/issues/5228 | [HacktoberFest][Missing Docstrings] UNet2DConditionModel | This feature request is included in this years Diffusers' Hacktoberfest 👨💻!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixes simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description fu... | closed | completed | false | 2 | [] | [] | 2023-09-29T09:49:30Z | 2023-10-09T15:12:36Z | 2023-10-09T15:12:36Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,918,966,759 | I_kwDOHa8MBc5yYRvn | 5,229 | https://github.com/huggingface/diffusers/issues/5229 | https://api.github.com/repos/huggingface/diffusers/issues/5229 | [HacktoberFest][Missing Docstrings] Autoencoder `encode` | This feature request is included in this years Diffusers' Hacktoberfest 👨💻!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixes simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description fu... | closed | completed | false | 1 | [] | [] | 2023-09-29T09:53:22Z | 2023-10-02T17:17:36Z | 2023-10-02T17:17:36Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,374,668,635 | I_kwDOHa8MBc5R78db | 523 | https://github.com/huggingface/diffusers/issues/523 | https://api.github.com/repos/huggingface/diffusers/issues/523 | Flax<>PT conversion script | ### Describe the bug
We need to match the weight naming and add a Flax<>PyTorch conversion script.
### Reproduction
_No response_
### Logs
_No response_
### System Info
main | closed | completed | false | 1 | [
"bug"
] | [
"pcuenca",
"patrickvonplaten"
] | 2022-09-15T14:58:02Z | 2022-10-26T08:29:29Z | 2022-10-26T08:29:29Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,918,967,784 | I_kwDOHa8MBc5yYR_o | 5,230 | https://github.com/huggingface/diffusers/issues/5230 | https://api.github.com/repos/huggingface/diffusers/issues/5230 | [HacktoberFest][Missing Docstrings] Autoencoder `decode` | This feature request is included in this years Diffusers' Hacktoberfest 👨💻!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixes simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description fu... | closed | completed | false | 1 | [] | [
"yiyixuxu",
"sayakpaul"
] | 2023-09-29T09:54:07Z | 2023-10-02T16:42:34Z | 2023-10-02T16:42:34Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,918,977,128 | I_kwDOHa8MBc5yYURo | 5,231 | https://github.com/huggingface/diffusers/issues/5231 | https://api.github.com/repos/huggingface/diffusers/issues/5231 | [HacktoberFest][Missing Docstrings] Adapters | This feature request is included in this years Diffusers' Hacktoberfest 👨💻!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixes simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description fu... | closed | completed | false | 3 | [] | [] | 2023-09-29T10:00:50Z | 2023-10-02T16:14:43Z | 2023-10-02T16:14:43Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,918,986,135 | I_kwDOHa8MBc5yYWeX | 5,232 | https://github.com/huggingface/diffusers/issues/5232 | https://api.github.com/repos/huggingface/diffusers/issues/5232 | Hacktoberfest! | 👨💻 **Hacktoberfest** is starting soon :heart_eyes: !
The purpose of Hacktoberfest is to give beginning, upcoming developers the opportunity to easily contribute to larger codebases and by doing so learning how to contribute to the open-source system.
Diffusers needs a 🧹 clean-up for the winter (doc strings, ... | closed | completed | false | 7 | [
"stale"
] | [] | 2023-09-29T10:07:20Z | 2023-12-26T15:10:36Z | 2023-12-26T15:10:36Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | patrickvonplaten | 23,423,619 | MDQ6VXNlcjIzNDIzNjE5 | User | false |
huggingface/diffusers | 1,918,987,907 | I_kwDOHa8MBc5yYW6D | 5,233 | https://github.com/huggingface/diffusers/issues/5233 | https://api.github.com/repos/huggingface/diffusers/issues/5233 | Wrong step value in parameters sent to callback function | ### Describe the bug
At least StableDiffusionXLPipeline and StableDiffusionXLImg2ImgPipeline sends step to callback function, starting at 1 and incrementing by 2 instead of incrementing by 1.
### Reproduction
```python
import torch
from diffusers import AutoencoderKL, StableDiffusionXLPipeline, StableDiffu... | closed | completed | false | 6 | [
"bug"
] | [] | 2023-09-29T10:08:32Z | 2023-10-02T17:52:55Z | 2023-10-02T17:52:55Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | TimothyAlexisVass | 55,708,319 | MDQ6VXNlcjU1NzA4MzE5 | User | false |
huggingface/diffusers | 1,374,734,583 | I_kwDOHa8MBc5R8Mj3 | 524 | https://github.com/huggingface/diffusers/issues/524 | https://api.github.com/repos/huggingface/diffusers/issues/524 | on the textual inversion training colab, a way to output .bin file before the training is fully completed | since the training usually takes few hours, and for different datasets, objects or styles, you can adapt with different training parameters, is there a way to output the .bin -let's say- every 1000 steps or so?
that would be very useful to check if the training is going in a good direction and maybe you can economiz... | closed | completed | false | 8 | [] | [
"patil-suraj"
] | 2022-09-15T15:43:26Z | 2022-10-26T13:33:45Z | 2022-10-26T13:33:45Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ciaobnsr | 113,621,298 | U_kgDOBsW5Mg | User | false |
huggingface/diffusers | 1,919,646,509 | I_kwDOHa8MBc5ya3st | 5,241 | https://github.com/huggingface/diffusers/issues/5241 | https://api.github.com/repos/huggingface/diffusers/issues/5241 | [HacktoberFest][Missing Docstrings] Würstchen | This feature request is included in this year's Diffusers Hacktoberfest 🎃!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixing simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description furt... | closed | completed | false | 1 | [] | [] | 2023-09-29T17:14:10Z | 2023-10-06T21:51:14Z | 2023-10-06T21:51:14Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | stevhliu | 59,462,357 | MDQ6VXNlcjU5NDYyMzU3 | User | false |
huggingface/diffusers | 1,920,085,109 | I_kwDOHa8MBc5ycix1 | 5,245 | https://github.com/huggingface/diffusers/issues/5245 | https://api.github.com/repos/huggingface/diffusers/issues/5245 | The program freezes at the import stage in pipenv | ### Describe the bug
I used pipenv to set up a virtual environment and tried running the simple example provided in [Quickstart](https://github.com/huggingface/diffusers#quickstart). However, the program just freezes without any response.
### Reproduction
```sh
$ pipenv --python 3.10
$ pipenv install diffu... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2023-09-30T03:32:25Z | 2023-11-08T15:06:57Z | 2023-11-08T15:06:57Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | parksb | 6,410,412 | MDQ6VXNlcjY0MTA0MTI= | User | false |
huggingface/diffusers | 1,920,103,818 | I_kwDOHa8MBc5ycnWK | 5,246 | https://github.com/huggingface/diffusers/issues/5246 | https://api.github.com/repos/huggingface/diffusers/issues/5246 | "module diffusers has no attribute CLIPImageProjection" happens again | ### Describe the bug
When trying to run the given demo example,
"module diffusers has no attribute CLIPImageProjection" error occurs again.
### Reproduction
the example in https://huggingface.co/docs/diffusers/main/en/api/pipelines/stable_diffusion/gligen#diffusers.StableDiffusionGLIGENTextImagePipeline
### Logs
... | closed | completed | false | 3 | [
"bug",
"stale"
] | [] | 2023-09-30T04:02:31Z | 2023-11-08T15:06:56Z | 2023-11-08T15:06:55Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | lyleguo | 21,014,363 | MDQ6VXNlcjIxMDE0MzYz | User | false |
huggingface/diffusers | 1,920,109,650 | I_kwDOHa8MBc5ycoxS | 5,247 | https://github.com/huggingface/diffusers/issues/5247 | https://api.github.com/repos/huggingface/diffusers/issues/5247 | [WinError 126] | ### Describe the bug
hello i cant solve this problem
OSError: [WinError 126] Le module spécifié est introuvable. Error loading "C:\ai\stable-diffusion-webui\venv\lib\site-packages\torch\lib\c10.dll" or one of its dependencies.
Appuyez sur une touche pour continuer...
### Reproduction
OSError: [WinError 126] Le mo... | closed | completed | false | 2 | [
"bug",
"stale"
] | [] | 2023-09-30T04:12:04Z | 2023-11-08T15:06:54Z | 2023-11-08T15:06:54Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | luffylaf | 146,497,048 | U_kgDOCLteGA | User | false |
huggingface/diffusers | 1,920,253,866 | I_kwDOHa8MBc5ydL-q | 5,249 | https://github.com/huggingface/diffusers/issues/5249 | https://api.github.com/repos/huggingface/diffusers/issues/5249 | Missing docs in multiple diffusers/models files | Multiple classes such as [this](https://github.com/huggingface/diffusers/blob/78a78515d64736469742e5081337dbcf60482750/src/diffusers/models/unet_2d_blocks.py#L1114) in the models directory are missing documentation or an explanation for what the different parameters mean, and sometimes type hints. Has it been done this... | closed | completed | false | 5 | [] | [] | 2023-09-30T13:13:37Z | 2023-10-09T14:29:30Z | 2023-10-09T14:29:30Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | a-r-r-o-w | 72,266,394 | MDQ6VXNlcjcyMjY2Mzk0 | User | false |
huggingface/diffusers | 1,375,100,934 | I_kwDOHa8MBc5R9mAG | 525 | https://github.com/huggingface/diffusers/issues/525 | https://api.github.com/repos/huggingface/diffusers/issues/525 | Issue with diffusers training notebook when trying to load images from local disk | ### Describe the bug
### Reproduction
https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/training_example.ipynb
^notebook i used
### Logs... | closed | completed | false | 5 | [
"bug"
] | [
"anton-l"
] | 2022-09-15T21:19:19Z | 2022-09-17T21:13:30Z | 2022-09-16T07:48:38Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | GeorvityLabs | 58,555,432 | MDQ6VXNlcjU4NTU1NDMy | User | false |
huggingface/diffusers | 1,920,299,465 | I_kwDOHa8MBc5ydXHJ | 5,251 | https://github.com/huggingface/diffusers/issues/5251 | https://api.github.com/repos/huggingface/diffusers/issues/5251 | Website should have times when the pages were last updated | I'm looking at the huggingface documentation [here](https://huggingface.co/docs/diffusers/using-diffusers/weighted_prompts#blending), but it seems this library is quite outdated and confusing. Sometimes it uses tuples, sometimes it uses lists, it's not clear. It also uses `blend`, but this [compel code](https://github.... | closed | completed | false | 3 | [
"stale"
] | [] | 2023-09-30T15:30:11Z | 2023-11-20T15:06:03Z | 2023-11-20T15:06:03Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | landmann | 12,756,081 | MDQ6VXNlcjEyNzU2MDgx | User | false |
huggingface/diffusers | 1,920,585,224 | I_kwDOHa8MBc5yec4I | 5,252 | https://github.com/huggingface/diffusers/issues/5252 | https://api.github.com/repos/huggingface/diffusers/issues/5252 | [HacktoberFest][Documentation] Diffusers documentation translation to Thai | Hello!
Bringing the documentation for diffusers to the Thai language would be pretty awesome for the Thai-speaking community, and I would like to contribute to translating as much as I can. (Adding the issue as a non-code contribution for this year's hacktoberfest 🎃 would also be a great way of encouraging more peo... | closed | completed | false | 6 | [
"stale"
] | [] | 2023-10-01T06:07:52Z | 2023-11-20T15:06:01Z | 2023-11-20T15:06:01Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | memmaptensor | 61,973,036 | MDQ6VXNlcjYxOTczMDM2 | User | false |
huggingface/diffusers | 1,920,860,481 | I_kwDOHa8MBc5yfgFB | 5,254 | https://github.com/huggingface/diffusers/issues/5254 | https://api.github.com/repos/huggingface/diffusers/issues/5254 | Support for Multi-Controlnet in Controlnet Inpaint Pipeline | ### Describe the bug
```bash
0%| | 0/50 [00:00<?, ?it/s]
Traceback (most recent call last):
File "c:\Users\kadir\Desktop\kadirnar\wor... | closed | completed | false | 9 | [
"bug"
] | [
"yiyixuxu"
] | 2023-10-01T16:48:00Z | 2023-11-10T16:53:21Z | 2023-11-10T16:53:21Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | kadirnar | 36,204,372 | MDQ6VXNlcjM2MjA0Mzcy | User | false |
huggingface/diffusers | 1,921,071,823 | I_kwDOHa8MBc5ygTrP | 5,256 | https://github.com/huggingface/diffusers/issues/5256 | https://api.github.com/repos/huggingface/diffusers/issues/5256 | Add Insta-Flow Training support | **Is your feature request related to a problem? Please describe.**
For hacktoberfest, I wanted to add in the Rectified Flow training objective which is what the instaflow paper, which can generate images in one forward pass, used to finetune stable diffusion.
**Describe the solution you'd like**
I'll make a small ... | closed | completed | false | 10 | [
"community-examples",
"stale"
] | [] | 2023-10-02T00:52:39Z | 2024-04-22T06:42:40Z | 2023-11-20T15:05:59Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | isamu-isozaki | 23,430,101 | MDQ6VXNlcjIzNDMwMTAx | User | false |
huggingface/diffusers | 1,921,770,175 | I_kwDOHa8MBc5yi-K_ | 5,257 | https://github.com/huggingface/diffusers/issues/5257 | https://api.github.com/repos/huggingface/diffusers/issues/5257 | [HacktoberFest] remove outdated schedulers | This feature request is included in this year's Diffusers Hacktoberfest 🎃!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixing simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description furt... | closed | completed | false | 5 | [
"stale"
] | [
"a-r-r-o-w"
] | 2023-10-02T11:53:58Z | 2023-11-20T15:05:57Z | 2023-11-20T15:05:57Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | yiyixuxu | 12,631,849 | MDQ6VXNlcjEyNjMxODQ5 | User | false |
huggingface/diffusers | 1,921,928,410 | I_kwDOHa8MBc5yjkza | 5,258 | https://github.com/huggingface/diffusers/issues/5258 | https://api.github.com/repos/huggingface/diffusers/issues/5258 | StableDiffusionXLControlNetInpaintPipeline with refiner model | ### Describe the bug
I can use StableDiffusionXLControlNetInpaintPipeline with the base SDXL 1.0 model, but not the refiner model.
### Reproduction
```python
from diffusers import ControlNetModel, StableDiffusionXLControlNetInpaintPipeline
import torch
from PIL import Image
import requests
from io import ... | closed | completed | false | 4 | [
"bug",
"stale"
] | [] | 2023-10-02T13:30:47Z | 2023-11-20T15:05:55Z | 2023-11-20T15:05:55Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | rb-synth | 135,021,519 | U_kgDOCAxDzw | User | false |
huggingface/diffusers | 1,921,993,319 | I_kwDOHa8MBc5yj0pn | 5,259 | https://github.com/huggingface/diffusers/issues/5259 | https://api.github.com/repos/huggingface/diffusers/issues/5259 | OSError: Error no file named model_index.json found in directory | ### Describe the bug
I download the weight from civitai, and the format is safetensor, the folder structure is like this:
A/cuteyukimixAdorable.safetensors
B/Koreandoll.safetensors
### Reproduction
from diffusers import StableDiffusionPipeline
import torch
# print(torch.__version__)
base_model_id = ... | closed | completed | false | 13 | [
"bug",
"stale"
] | [] | 2023-10-02T14:06:34Z | 2023-11-20T15:05:53Z | 2023-11-20T15:05:53Z | NONE | null | 20260407T133413Z | 2026-04-07T13:34:13Z | Louis24 | 18,719,360 | MDQ6VXNlcjE4NzE5MzYw | User | false |
huggingface/diffusers | 1,375,135,929 | I_kwDOHa8MBc5R9ui5 | 526 | https://github.com/huggingface/diffusers/issues/526 | https://api.github.com/repos/huggingface/diffusers/issues/526 | include input parameters in PipelineOutput | It's useful to be able to associate a pipeline result with the parameters that generated it. It provides a record for the experimenter themselves, other people looking to learn from them, those interested in reproducible results, and community-wide surveys.
Current pipelines have a number of values they set by defau... | closed | completed | false | 3 | [
"stale"
] | [] | 2022-09-15T22:03:19Z | 2022-11-01T15:03:35Z | 2022-11-01T15:03:35Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | keturn | 83,819 | MDQ6VXNlcjgzODE5 | User | false |
huggingface/diffusers | 1,922,133,831 | I_kwDOHa8MBc5ykW9H | 5,260 | https://github.com/huggingface/diffusers/issues/5260 | https://api.github.com/repos/huggingface/diffusers/issues/5260 | [HacktoberFest] make fast test for `StableDiffusionXLPipeline` smaller when applicable | This feature request is included in this year's Diffusers Hacktoberfest 🎃!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixing simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description furt... | closed | completed | false | 0 | [
"enhancement"
] | [] | 2023-10-02T15:21:50Z | 2023-10-09T09:19:58Z | 2023-10-09T09:19:58Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sayakpaul | 22,957,388 | MDQ6VXNlcjIyOTU3Mzg4 | User | false |
huggingface/diffusers | 1,922,165,940 | I_kwDOHa8MBc5ykey0 | 5,262 | https://github.com/huggingface/diffusers/issues/5262 | https://api.github.com/repos/huggingface/diffusers/issues/5262 | [HacktoberFest] make fast test for `StableDiffusionPipeline` smaller when applicable | This feature request is included in this year's Diffusers Hacktoberfest 🎃!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixing simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description furt... | closed | completed | false | 0 | [] | [] | 2023-10-02T15:39:41Z | 2023-10-17T09:16:44Z | 2023-10-17T09:16:44Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sayakpaul | 22,957,388 | MDQ6VXNlcjIyOTU3Mzg4 | User | false |
huggingface/diffusers | 1,922,173,856 | I_kwDOHa8MBc5ykgug | 5,263 | https://github.com/huggingface/diffusers/issues/5263 | https://api.github.com/repos/huggingface/diffusers/issues/5263 | [HacktoberFest] make fast test for `StableDiffusionControlNetPipeline` smaller when applicable | This feature request is included in this year's Diffusers Hacktoberfest 🎃!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixing simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description furt... | closed | completed | false | 3 | [] | [] | 2023-10-02T15:43:53Z | 2023-10-09T09:26:28Z | 2023-10-09T09:26:28Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sayakpaul | 22,957,388 | MDQ6VXNlcjIyOTU3Mzg4 | User | false |
huggingface/diffusers | 1,922,178,343 | I_kwDOHa8MBc5ykh0n | 5,264 | https://github.com/huggingface/diffusers/issues/5264 | https://api.github.com/repos/huggingface/diffusers/issues/5264 | [HacktoberFest] make fast test for `StableDiffusionControlNetImg2ImgPipeline` smaller when applicable | This feature request is included in this year's Diffusers Hacktoberfest 🎃!
Hacktoberfest's purpose is to give upcoming open-source developers the opportunity to learn about open-source and programming by fixing simple issues (like this one).
**How to solve the issue**:
- 1. Every issue has a description furt... | closed | completed | false | 2 | [] | [] | 2023-10-02T15:46:38Z | 2023-10-26T10:32:36Z | 2023-10-26T10:32:36Z | MEMBER | null | 20260407T133413Z | 2026-04-07T13:34:13Z | sayakpaul | 22,957,388 | MDQ6VXNlcjIyOTU3Mzg4 | User | false |
huggingface/diffusers | 1,922,317,156 | I_kwDOHa8MBc5ylDtk | 5,265 | https://github.com/huggingface/diffusers/issues/5265 | https://api.github.com/repos/huggingface/diffusers/issues/5265 | [HacktoberFest] Update Docs/README.md | **Is your feature request related to a problem? Please describe.**
For a REAME.md file, I think we can make use of the documentation styling in the programming language that is being used in the project, i.e. Python here.
**Describe the solution you'd like**
I think we can make the documentation better if we u... | closed | completed | false | 3 | [] | [] | 2023-10-02T17:17:50Z | 2023-10-06T21:50:20Z | 2023-10-06T21:50:20Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | ShubhamJagtap2000 | 63,872,951 | MDQ6VXNlcjYzODcyOTUx | User | false |
huggingface/diffusers | 1,922,552,458 | I_kwDOHa8MBc5yl9KK | 5,268 | https://github.com/huggingface/diffusers/issues/5268 | https://api.github.com/repos/huggingface/diffusers/issues/5268 | Possibility to add filters to the generation process (modifying latents at each step) | **Is your feature request related to a problem? Please describe.**
I've added filters to the inference process with a workaround solution.
It happens in the latent space.
It would be great if there was some similar solutio as `callback` called something like `modifying_callback` which would modify the latents at eac... | closed | completed | false | 11 | [
"stale"
] | [
"yiyixuxu"
] | 2023-10-02T19:41:22Z | 2023-11-09T15:23:33Z | 2023-11-09T15:12:03Z | CONTRIBUTOR | null | 20260407T133413Z | 2026-04-07T13:34:13Z | TimothyAlexisVass | 55,708,319 | MDQ6VXNlcjU1NzA4MzE5 | User | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.