like this demo
https://twitter.com/pika_labs/status/1678892871670464513

Posted in solution in #33

作为专业向使用，建议能够接入一些像Blender这样的生产软件中，只靠Gradio Interface这样的不是很适合于生产

如图，这是我在Blender接入了ComfyUI，3D软件可以input更精确的数据（而不是预处理器的不稳定计算），效果会更好，同时也更符合当前动画生产的需求

I'm trying to run the base model without any dreambooth lora but the results are random or meaningless stuff

.yaml config

NewModel:
  path: ""
  base: ""

  motion_module:
    - "models/Motion_Module/mm_sd_v14.ckpt"
    - "models/Motion_Module/mm_sd_v15.ckpt"
    
  steps:          25
  guidance_scale: 7.5

  prompt:
    - "a girl"

  n_prompt:
    - ""

if something is wrong can you tell me how to configure it well for sd 1.5

Thank you for building AnimateDiff!

@richardanderson001 created a review titled:

interesting experimental SD based animation

on repo-reviews.github.io to share their experience using AnimateDiff.

link to review

If you would like to help your super-users share their experiences using your repo, add a badge to your README.md.

We hope that sharing these experiences helps your users increase their productivity.

--
Please be kind,
I’m a human!

Hi! Thanks for a very interesting paper, I wonder if you've tried generating shorter/longer clips? I see that there is temporal_position_encoding_max_len=24 which limits the length to be 24 frames, but what about shorter clips?

Also I'm struggling to understand what is the shape of attention in Temporal Transformer? Here you resize (b f) d c -> (b d) f c where the batch (b) is probably 1 and the frames (f) is probably 16, (d) corresponds to reshaped features right? So each "super-pixel" is processed separately and the shape of attentions maps should be (B * D) x F x F which isn't really big. Why does then the inference take 60 Gb?

Hello, I really want to try AnimateDiff,
sadly I only have an 8Gb RTX 3070,

Do you think you will ever release a notebook?
Thanks

lyriel_v16:
base: "models/DreamBooth_LoRA/lyriel_v16.safetensors"
path: "models/DreamBooth_LoRA/DANWGFWL.safetensors"
motion_module:
- "models/Motion_Module/mm_sd_v14.ckpt"
- "models/Motion_Module/mm_sd_v15.ckpt"

seed: [10788741199826055526, 6520604954829636163, 6519455744612555650]
steps: 35
guidance_scale: 7

prompt:
- "1 girl solo, perfect_hand, (8k, RAW photo, best quality, masterpiece:1.2), (realistic, photo-realistic:1.4), (extremely detailed CG unity 8k wallpaper),(full body), (neon lights), machop, mechanical arms, hanfu, lora:DANWGFWL:0.6,Chinese clothes, dress, pretty face,(dark shot:1.1), epic realistic, RAW, analog, sharp focus, volumetric fog"
- "(masterpiece, best quality), 1girl, nude, closed eyes, upper body, splashing, abstract, psychedelic,"
- "(masterpiece, best quality), 1boy, muscular, beard, cyberpunk, (blurry, bokeh, fisheye lens), night, looking at viewer, contrast, contrapposto, neon oversized jacket,"

n_prompt:
- "(worst quality, low quality:2), monochrome, zombie,overexposure, watermark,text,bad anatomy,bad hand,extra hands,extra fingers,too many fingers,fused fingers,bad arm,distorted arm,extra arms,fused arms,extra legs,missing leg,disembodied leg,extra nipples, detached arm, liquid hand,inverted hand,disembodied limb, small breasts, loli, oversized head,extra body,completely nude, extra navel,easynegative,(hair between eyes),sketch, duplicate, ugly, huge eyes, text, logo, worst face, (bad and mutated hands:1.3), (blurry:2.0), horror, geometry, bad_prompt, (bad hands), (missing fingers), multiple limbs, bad anatomy, (interlocked fingers:1.2), Ugly Fingers, (extra digit and hands and fingers and legs and arms:1.4), ((2girl)), (deformed fingers:1.2), (long fingers:1.2),(bad-artist-anime), bad-artist, bad hand, extra legs ,(ng_deepnegative_v1_75t)"
- "(worst quality, low quality:2), monochrome, zombie,overexposure, watermark,text,bad anatomy,bad hand,extra hands,extra fingers,too many fingers,fused fingers,bad arm,distorted arm,extra arms,fused arms,extra legs,missing leg,disembodied leg,extra nipples, detached arm, liquid hand,inverted hand,disembodied limb, small breasts, loli, oversized head,extra body,completely nude, extra navel,easynegative,(hair between eyes),sketch, duplicate, ugly, huge eyes, text, logo, worst face, (bad and mutated hands:1.3), (blurry:2.0), horror, geometry, bad_prompt, (bad hands), (missing fingers), multiple limbs, bad anatomy, (interlocked fingers:1.2), Ugly Fingers, (extra digit and hands and fingers and legs and arms:1.4), ((2girl)), (deformed fingers:1.2), (long fingers:1.2),(bad-artist-anime), bad-artist, bad hand, extra legs ,(ng_deepnegative_v1_75t)"
- "(worst quality, low quality:2), monochrome, zombie,overexposure, watermark,text,bad anatomy,bad hand,extra hands,extra fingers,too many fingers,fused fingers,bad arm,distorted arm,extra arms,fused arms,extra legs,missing leg,disembodied leg,extra nipples, detached arm, liquid hand,inverted hand,disembodied limb, small breasts, loli, oversized head,extra body,completely nude, extra navel,easynegative,(hair between eyes),sketch, duplicate, ugly, huge eyes, text, logo, worst face, (bad and mutated hands:1.3), (blurry:2.0), horror, geometry, bad_prompt, (bad hands), (missing fingers), multiple limbs, bad anatomy, (interlocked fingers:1.2), Ugly Fingers, (extra digit and hands and fingers and legs and arms:1.4), ((2girl)), (deformed fingers:1.2), (long fingers:1.2),(bad-artist-anime), bad-artist, bad hand, extra legs ,(ng_deepnegative_v1_75t)"

is there a way to use controlnet ?

(animatediff) ubuntu@104-171-203-67:~/AnimateDiff$ python -m scripts.animate --config configs/prompts/2-Lyriel.yaml
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
Traceback (most recent call last):
File "/home/ubuntu/anaconda3/envs/animatediff/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/ubuntu/anaconda3/envs/animatediff/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/ubuntu/AnimateDiff/scripts/animate.py", line 159, in
main(args)
File "/home/ubuntu/AnimateDiff/scripts/animate.py", line 51, in main
text_encoder = CLIPTextModel.from_pretrained(args.pretrained_model_path, subfolder="text_encoder")
File "/home/ubuntu/anaconda3/envs/animatediff/lib/python3.10/site-packages/transformers/modeling_utils.py", line 2230, in from_pretrained
state_dict = load_state_dict(resolved_archive_file)
File "/home/ubuntu/anaconda3/envs/animatediff/lib/python3.10/site-packages/transformers/modeling_utils.py", line 386, in load_state_dict
with safe_open(checkpoint_file, framework="pt") as f:
OSError: No such device (os error 19)

大佬们好！ 这个稳定性已经很好了，工程库里面我没发现是如何提升帧与帧之间的稳定性的， prompt参数都是生成单帧的，请问如何做到如此大的相似性？

之前我有关注过下面这个，他们的思路给前景的关键词，用于锁定固定的内容。
https://github.com/ashen-sensored/sd_webui_masactrl
https://arxiv.org/abs/2304.08465

加油！

It seems that the google drive link that currently distributes the motion module is locked due to excessive access.
I suggest using huggingface instead.

Hi, is it possible to start with an existing image instead of generating from scratch?

Thanks for your great work! Could you provide scripts or configs to produce videos based on SD v1.5?

May we hope that you will offer this great functionality as an extension in for stable diffusion webui?

Hi,
Is it possible to generate a single character from the Pose for about 5 seconds?

I have a video of Pose ( openpose + hands + face) and i was wondering if it is possible to generate an output video withe the length of 5 seconds that has a consistent character/Avatar which plays Dance, .... from the controled (pose) input?

i want to generate human like animation (No matter what, but just a consistent Character/Avatar)
Sample Video

Thanks
Best regards

(base) C:\AnimateDiff>conda env create -f environment.yaml
Collecting package metadata (repodata.json): done
Solving environment: failed

ResolvePackageNotFound:

• xformers

pip install xformers didnt help

I can´t run it, when i try to run got those errors!, any know how to solve them?

(animatediff) F:\AnimateDiff>python -m scripts.animate --config configs/prompts/5-RealisticVision.yaml
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
Traceback (most recent call last):
File "C:\Users\Ryzen_Reaper\miniconda3\envs\animatediff\lib\runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "C:\Users\Ryzen_Reaper\miniconda3\envs\animatediff\lib\runpy.py", line 86, in _run_code
exec(code, run_globals)
File "F:\AnimateDiff\scripts\animate.py", line 15, in
from animatediff.models.unet import UNet3DConditionModel
ModuleNotFoundError: No module named 'animatediff'

(animatediff) F:\AnimateDiff>

huggingface_hub.utils._validators.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'models/StableDiffusion/stable-diffusion-v1-5'. Use repo_type argument if needed.

(animatediff) G:\AI\AnimateDiff>

win10 conda3 pytorch2.01 cuda118.
模型都下好，放在G:\AI\AnimateDiff\models\StableDiffusion,文件名为stable-diffusion-v1-5 .safetensors. 还是报这个错误。

I am pretty sure this is not currently supported by the API but I'm wondering if this would be possible with the current model architecture? The use case is for psychedelic animation (loopback, deforum, etc.), for example an animation which is always zooming or using some programmed camera movement with img2img - it would be truly spectacular if we could use AnimateDiff to inject lifelike movement! We currently have a lesser version of this with a model called FlowR, which predicts a flow map given the last 4 images. The result is very abstract, but it provides a subtle grounding.

I understand that the AnimateDiff model is trained on buffers of 16 frames, but in theory since it has been trained on random samples without start/end it should be possible to simply keep the buffer rolling and have infinitely long animations.

Not very sure about this
NewModel:
path: "[path to your DreamBooth/LoRA model .safetensors file]"
base: "[path to LoRA base model .safetensors file, leave it empty string if not needed]"

In Ghibli example case
GhibliBackground:
base: "models/DreamBooth_LoRA/CounterfeitV30_25.safetensors"
path: "models/DreamBooth_LoRA/lora_Ghibli_n3.safetensors"
it seems like the base means the base model, and the path means the lora, but in other example like
ToonYou:
base: ""
path: "models/DreamBooth_LoRA/toonyou_beta3.safetensors"

so I try to change it to like this, something goes strange
base: "models/DreamBooth_LoRA/toonyou_beta3.safetensors"
path: ""
I got this

It's not the style of Toonyou, I don't know.

And I did another try, I set the prompt like:
base: "models/DreamBooth_LoRA/toonyou_beta3.safetensors"
path: "models/DreamBooth_LoRA/smv1-10.safetensors"
the smv1-10.safetensors is a lora download from civitai, then I got an error.
omegaconf.errors.ConfigAttributeError: Missing key lora_alpha
full_key: ToonYou.lora_alpha
object_type=dict

so what's the different with base and path, and how to set it correct?

My problem is that conda can't install xformers on windows,could you provide pip requirements.txt dependent installation method.

I tried to move -xformers to pip, and finally created an environment error
cuda version 11.6, conda 22.9.0, WIN10

Can you save as video, I've noticed gif aminations are lower quality and have banding.

after banging my head with other issues, I'm now stuck here:
(animatediff) PS I:\AnimateDiff\AnimateDiff> python -m scripts.animate --config configs/prompts/1-ToonYou.yaml
loaded temporal unet's pretrained weights from models/StableDiffusion/stable-diffusion-v1-5\unet ...

missing keys: 560;

unexpected keys: 0;

Temporal Module Parameters: 417.1376 M

Traceback (most recent call last):
File "C:\Users\Lucas\miniconda3\envs\animatediff\lib\runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "C:\Users\Lucas\miniconda3\envs\animatediff\lib\runpy.py", line 86, in _run_code
exec(code, run_globals)
File "I:\AnimateDiff\AnimateDiff\scripts\animate.py", line 159, in
main(args)
File "I:\AnimateDiff\AnimateDiff\scripts\animate.py", line 56, in main
else: assert False
AssertionError

Got any ideas where I should start looking to figure out why the resulting GIF is just jittery diffusion noise and nothing else? Much appreciated!

Could this issue be related to the sampler being used? Is there a possibility to choose a different sampler?

cc @talesofai, that's related to this fork - https://github.com/talesofai/AnimateDiff
P.S. I'm facing this with original repo too

(animatediff) PS F:\animediff\AnimateDiff> python -m scripts.animate --config configs/prompts/1-ToonYou.yaml
C:\Users\dbodbo\miniconda3\envs\animatediff\lib\site-packages\torchvision\io\image.py:13: UserWarning: Failed to load image Python extension: Could not find module 'C:\Users\dbodbo\miniconda3\envs\animatediff\Lib\site-packages\torchvision\image.pyd' (or one of its dependencies). Try using the full path with constructor syntax.
warn(f"Failed to load image Python extension: {e}")
loaded temporal unet's pretrained weights from models/StableDiffusion\unet ...

missing keys: 560;

unexpected keys: 0;

Temporal Module Parameters: 417.1376 M

Traceback (most recent call last):
File "C:\Users\dbodbo\miniconda3\envs\animatediff\lib\runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "C:\Users\dbodbo\miniconda3\envs\animatediff\lib\runpy.py", line 86, in _run_code
exec(code, run_globals)
File "F:\animediff\AnimateDiff\scripts\animate.py", line 159, in
main(args)
File "F:\animediff\AnimateDiff\scripts\animate.py", line 56, in main
else: assert False
AssertionError

Hello and I have found an error that cannot continue. I make google colab and I have already prepared it but it shows an error for me.

[Errno 2] No such file or directory: '/content/animatediff'
/content
/usr/bin/python3: No module named scripts.animate

I've tried everything but it just doesn't work! question is, can you fix this?
I have already done with the jupyter code.

!python -m scripts.animate --config /content/AnimateDiff/configs/prompts/9-yiffymix31.yaml --pretrained_model_path /content/AnimateDiff/models/StableDiffusion --L 16 --W 512 --H 512

Is this AnimateDiff the same as PikaLabs where you input an text prompt+image to generate a video?

yaml generated results

same prompt,same seed,same cfg
gradio generated results
https://github.com/guoyww/AnimateDiff/assets/131337141/836f2bc5-b84f-44e2-9a3c-613dd28f083a

I am the author of the SAM extension and I wrote one extension for AnimateDiff.

https://github.com/continue-revolution/sd-webui-animatediff

Have fun with it!

(base) H:\Stablediff\aniumateddiff\AnimateDiff\AnimateDiff>conda env create -f environment.yaml
Collecting package metadata (repodata.json): done
Solving environment: failed

ResolvePackageNotFound:

• xformers

I tried using anaconda navigator as well for a premade venv, then i got:

ModuleNotFoundError: No module named 'diffusers.modeling_utils'

Nor can i download the other configs as you used google drive, download quota was reached, so can't download those.

In the examples in configs/prompts/, the prompts contain references to textual inversion embeddings like badhandv4, easynegative, ng_deepnegative_v1_75t:

  n_prompt:
    - "easynegative,bad_construction,bad_structure,bad_wail,bad_windows,blurry,cloned_window,cropped,deformed,disfigured,error,extra_windows,extra_chimney,extra_door,extra_structure,extra_frame,fewer_digits,fused_structure,gross_proportions,jpeg_artifacts,long_roof,low_quality,structure_limbs,missing_windows,missing_doors,missing_roofs,mutated_structure,mutation,normal_quality,out_of_frame,owres,poorly_drawn_structure,poorly_drawn_house,signature,text,too_many_windows,ugly,username,uta,watermark,worst_quality"

From looking at the source code I'm not sure where the support for textual inversion was implemented - is this functionality implemented in AnimateDiff?

I saw that it generates a sample.gif with two gif how do I generate only one I would like to keep only the 0-1.gif to leave the generation faster

Wow, this is really cool.
Is there a way to integrate it with automatic's webui? I mean to use external VAEs, advanced prompting with brackets and weights, controlnet etc?

UPDATE, oh I see it's a duplicate question.

Hello and if it also look good from gradio, you can also add on clip skip? Because if you set it to clip skip.. then you don't have to set it to 1, instead you can also set it to clip skip for 2, 3 or more.. then they won't have so much trouble about clip skip, without clip skip it would be dumb. you know. Thanks! :)

I wrote a community colab notebook for AnimateDiff, I can't say it's a demo because I add other fork to it.
https://colab.research.google.com/github/Linaqruf/sd-notebook-collection/blob/main/AnimateDiff.ipynb

Traceback (most recent call last):
File "/home/dmitry/miniconda3/envs/animatediff/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/dmitry/miniconda3/envs/animatediff/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/mnt/f/project/stable_animation/AnimateDiff/scripts/animate.py", line 159, in
main(args)
File "/mnt/f/project/stable_animation/AnimateDiff/scripts/animate.py", line 51, in main
text_encoder = CLIPTextModel.from_pretrained(args.pretrained_model_path, subfolder="text_encoder")
File "/home/dmitry/miniconda3/envs/animatediff/lib/python3.10/site-packages/transformers/modeling_utils.py", line 2230, in from_pretrained
state_dict = load_state_dict(resolved_archive_file)
File "/home/dmitry/miniconda3/envs/animatediff/lib/python3.10/site-packages/transformers/modeling_utils.py", line 386, in load_state_dict
with safe_open(checkpoint_file, framework="pt") as f:
safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooLarge

Did everything according to the instructions, what could be the problem? Perhaps I missed something?

When passing CLI arg --L to change video length. The program will crash if length is set more than 24.

for example, when passing --L 26:

RuntimeError: The size of tensor a (26) must match the size of tensor b (24) at non-singleton dimension 1

Like title. Is there an option for adjusting the resolution/size of animated images? If not, is this a feature that you may implement in the future? Thanks!

At the first version, which cost 60G GPU memory, I can run but I don't have such size memory, so I failed. but everything looks fine.
But now I updated the code, follow the new install guide, then I got this error.

RuntimeError: Detected that PyTorch and torchvision were compiled with different CUDA versions. PyTorch has CUDA Version=11.7 and torchvision has CUDA Version=11.8. Please reinstall the torchvision that matches your PyTorch install.

conda list

Name Version Build Channel

_libgcc_mutex 0.1 conda_forge conda-forge
_openmp_mutex 4.5 2_kmp_llvm conda-forge
accelerate 0.21.0 pypi_0 pypi
antlr4-python3-runtime 4.9.3 pypi_0 pypi
arrow 1.2.3 pypi_0 pypi
attrs 23.1.0 pypi_0 pypi
blas 1.0 mkl
brotli-python 1.0.9 py310hd8f1fbe_9 conda-forge
bzip2 1.0.8 h7f98852_4 conda-forge
ca-certificates 2023.5.7 hbcca054_0 conda-forge
certifi 2023.5.7 pyhd8ed1ab_0 conda-forge
charset-normalizer 3.2.0 pyhd8ed1ab_0 conda-forge
cudatoolkit 11.3.1 ha36c431_9 nvidia
diffusers 0.11.1 pypi_0 pypi
einops 0.6.1 pypi_0 pypi
ffmpeg 4.3 hf484d3e_0 pytorch
fqdn 1.5.1 pypi_0 pypi
freetype 2.12.1 hca18f0e_1 conda-forge
fsspec 2023.6.0 pypi_0 pypi
gdown 4.7.1 pypi_0 pypi
gmp 6.2.1 h58526e2_0 conda-forge
gnutls 3.6.13 h85f3911_1 conda-forge
huggingface-hub 0.16.4 pypi_0 pypi
icu 72.1 hcb278e6_0 conda-forge
idna 3.4 pyhd8ed1ab_0 conda-forge
imageio 2.27.0 pypi_0 pypi
importlib-metadata 6.8.0 pypi_0 pypi
isoduration 20.11.0 pypi_0 pypi
jpeg 9e h0b41bf4_3 conda-forge
jsonpointer 2.4 pypi_0 pypi
jsonschema 4.18.3 pypi_0 pypi
jsonschema-specifications 2023.6.1 pypi_0 pypi
lame 3.100 h166bdaf_1003 conda-forge
lcms2 2.15 hfd0df8a_0 conda-forge
ld_impl_linux-64 2.40 h41732ed_0 conda-forge
lerc 4.0.0 h27087fc_0 conda-forge
libdeflate 1.17 h0b41bf4_0 conda-forge
libffi 3.4.2 h7f98852_5 conda-forge
libgcc-ng 13.1.0 he5830b7_0 conda-forge
libhwloc 2.9.1 nocuda_h7313eea_6 conda-forge
libiconv 1.17 h166bdaf_0 conda-forge
libnsl 2.0.0 h7f98852_0 conda-forge
libpng 1.6.39 h753d276_0 conda-forge
libsqlite 3.42.0 h2797004_0 conda-forge
libstdcxx-ng 13.1.0 hfd8a6a1_0 conda-forge
libtiff 4.5.0 h6adf6a1_2 conda-forge
libuuid 2.38.1 h0b41bf4_0 conda-forge
libwebp-base 1.3.1 hd590300_0 conda-forge
libxcb 1.13 h7f98852_1004 conda-forge
libxml2 2.11.4 h0d562d8_0 conda-forge
libzlib 1.2.13 hd590300_5 conda-forge
llvm-openmp 16.0.6 h4dfa4b3_0 conda-forge
mkl 2021.4.0 h8d4b97c_729 conda-forge
mkl-service 2.4.0 py310ha2c4b55_0 conda-forge
mkl_fft 1.3.1 py310h2b4bcf5_1 conda-forge
mkl_random 1.2.2 py310h00e6091_0
ncurses 6.4 hcb278e6_0 conda-forge
nettle 3.6 he412f7d_0 conda-forge
numpy 1.24.3 py310hd5efca6_0
numpy-base 1.24.3 py310h8e6c178_0
omegaconf 2.3.0 pypi_0 pypi
openh264 2.1.1 h780b84a_0 conda-forge
openjpeg 2.5.0 hfec8fc6_2 conda-forge
openssl 3.1.1 hd590300_1 conda-forge
pillow 9.4.0 py310h023d228_1 conda-forge
pip 23.1.2 pyhd8ed1ab_0 conda-forge
pthread-stubs 0.4 h36c2ea0_1001 conda-forge
pysocks 1.7.1 pyha2e5f31_6 conda-forge
python 3.10.12 hd12c33a_0_cpython conda-forge
python_abi 3.10 3_cp310 conda-forge
pytorch 1.12.1 py3.10_cuda11.3_cudnn8.3.2_0 pytorch
pytorch-mutex 1.0 cuda pytorch
pyyaml 6.0 pypi_0 pypi
readline 8.2 h8228510_1 conda-forge
referencing 0.29.1 pypi_0 pypi
requests 2.31.0 pyhd8ed1ab_0 conda-forge
rpds-py 0.8.10 pypi_0 pypi
safetensors 0.3.1 pypi_0 pypi
setuptools 68.0.0 pyhd8ed1ab_0 conda-forge
six 1.16.0 pyh6c4a22f_0 conda-forge
tbb 2021.9.0 hf52228f_0 conda-forge
tk 8.6.12 h27826a3_0 conda-forge
torchaudio 0.12.1 py310_cu113 pytorch
torchvision 0.13.1 py310_cu113 pytorch
tqdm 4.65.0 pypi_0 pypi
transformers 4.25.1 pypi_0 pypi
typing_extensions 4.7.1 pyha770c72_0 conda-forge
tzdata 2023c h71feb2d_0 conda-forge
uri-template 1.3.0 pypi_0 pypi
urllib3 2.0.3 pyhd8ed1ab_1 conda-forge
webcolors 1.13 pypi_0 pypi
wheel 0.40.0 pyhd8ed1ab_0 conda-forge
xformers 0.0.20 py310_cu11.6.2_pyt1.12.1 xformers
xorg-libxau 1.0.11 hd590300_0 conda-forge
xorg-libxdmcp 1.1.3 h7f98852_0 conda-forge
xz 5.2.6 h166bdaf_0 conda-forge
zipp 3.16.2 pypi_0 pypi
zlib 1.2.13 hd590300_5 conda-forge
zstd 1.5.2 hfc55251_7 conda-forge

Hi everyone,
Running Windows 10 x64, followed the instructions for python, I am getting:

(animatediff) C:\Apps\AnimateDiff\scripts>python -m animate --config configs\prompts\10-InitImageYoimiya.yml
Traceback (most recent call last):
  File "<frozen runpy>", line 198, in _run_module_as_main
  File "<frozen runpy>", line 88, in _run_code
  File "C:\Apps\AnimateDiff\scripts\animate.py", line 15, in <module>
    from animatediff.models.unet import UNet3DConditionModel
ModuleNotFoundError: No module named 'animatediff'

Any help will be greatly appreciated! Thank you

Are you planning to implement image-to-animation pipeline?

I am refering to https://huggingface.co/docs/transformers/add_new_model

I wanted to see if there was a way to get the frames for upscaling.

I love the results generated with AnimateDiff so far, but most of the animations are fast and a lot of content changes in the animation which makes it a little incoherent. Is there any way to reduce the speed of animation or to tweak other settings of animation?

Is it intended for images to have barely any movement with the mm_sd_v15 ckpt?

See: