This is animate Diff with LCM distilled into the weights. Clean, fast renders. Works well at low steps, and in my opinion is the truest representation of what the base aesthetic model (SD1.5) has to offer
Description
FAQ
Comments (36)
Awesome model! Do you have workflow?
I will be doing a live stream on the official Civit Twitch this coming thursday
You can generate videos with this model and a UI called DiffEX. No workflow needed, you just need some settings that you can probably find yourself.
@M14w DiffEXl?
@machinedelusions I couldn't find any link. What time will be today?
@dezooomer It is "DiffEx", i don't know why there was an additional character. Follow this tutorial, if not, you are going to have hard time to find the correct settings:
https://civitai.com/articles/4393/generate-videos-with-paradigm-lcm-using-diffex
@machinedelusions It's a pity to miss this live , can you share the recorded video or link of the live broadcast?
Should this be used as an AnimatedDiff model (saved in that folder) or a typical Checkpoint model & attached regular AnimateDiff models to workflow?
Yes this should be used in the animate Diff loader, it is a motion module, not an aesthetic model (1.5)
Is this safe for me download if it's saying pickletensor next to the download button?
totally safe to use, myself and others have been using it for months with no compromises
Yeah, downloads here are generally safe. As far as I know, Civitai scans the files. Plus, even places like Hugging Face warn you about pickle files by showing what they import. Just a heads-up, pickle files can contain malicious code, so it's always good to be careful. Safer alternatives exist, like safetensors. For extra security, I sometimes use the VirusTotal website to scan files or URLs. It checks them with a bunch of different scanners to catch any nasty stuff. hope this helps.
@Matp Oh, thank you. I was unaware of the VirusTotal website! Thank you, it does help!
so we will not need LCM lora with it?
if you pair this model with a LCM model you will not no
I recommend my Photon-LCM, its very good with this model
@machinedelusions thank you!
@machinedelusions so a LCM model is mandatory? i guess also the LCM sampler right?
@loristecno960 not 100% needed, but it works very well with LCM loras or models
I loaded this a a motion model but generates an error and doesn't work. Any idea why? I don't have issues with any other motion models.
2024-03-20 01:29:45,811 - AnimateDiff - INFO - Loading motion module animatediffLCMMotion_v10.ckpt from F:\ai\sd\stable-diffusion-webui\extensions\sd-webui-animatediff\model\animatediffLCMMotion_v10.ckpt 2024-03-20 01:29:46,492 - AnimateDiff - INFO - Guessed animatediffLCMMotion_v10.ckpt architecture: MotionModuleType.AnimateDiffV2 *** Error running before_process: F:\ai\sd\stable-diffusion-webui\extensions\sd-webui-animatediff\scripts\animatediff.py Traceback (most recent call last): File "F:\ai\sd\stable-diffusion-webui\modules\scripts.py", line 776, in before_process script.before_process(p, *script_args) File "F:\ai\sd\stable-diffusion-webui\extensions\sd-webui-animatediff\scripts\animatediff.py", line 64, in before_process motion_module.inject(p.sd_model, params.model) File "F:\ai\sd\stable-diffusion-webui\extensions\sd-webui-animatediff\scripts\animatediff_mm.py", line 68, in inject self.load(model_name) File "F:\ai\sd\stable-diffusion-webui\extensions\sd-webui-animatediff\scripts\animatediff_mm.py", line 52, in load self.mm.load_state_dict(mm_state_dict) File "F:\ai\sd\stable-diffusion-webui\venv\lib\site-packages\torch\nn\modules\module.py", line 2152, in load_state_dict raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for MotionWrapper: Missing key(s) in state_dict: "down_blocks.0.motion_modules.0.temporal_transformer.transformer_blocks.0.attention_blocks.0.pos_encoder.pe", "down_blocks.0.motion_modules.0.temporal_transformer.transformer_blocks.0.attention_blocks.1.pos_encoder.pe", "down_blocks.0.motion_modules.1.temporal_transformer.transformer_blocks.0.attention_blocks.0.pos_encoder.pe", "down_blocks.0.motion_modules.1.temporal_transformer.transformer_blocks.0.attention_blocks.1.pos_encoder.pe", "down_blocks.1.motion_modules.0.temporal_transformer.transformer_blocks.0.attention_blocks.0.pos_encoder.pe", "down_blocks.1.motion_modules.0.temporal_transformer.transformer_blocks.0.attention_blocks.1.pos_encoder.pe", "down_blocks.1.m ...
I'm getting that same error, on windows and linux.
A1111 forge, sd-forge-animatediff, python 3.10.6, xformer 0.0.23.post1 on both
not sure what this could be from, iv only tested the motion module in comfyUI, its something with auto1111 it seems
The error still persist for forge.
try and the same error in comfyui, this dont work!
What motion model is this based on? AnimateDiff V1, 2, 3? TemporalDiff?
Are you working with Fu yun Wang ?
Hello, are you training based on AnimateDiff v3 Model?
Is that model is the same one as on hugging face :
https://huggingface.co/wangfuyun/AnimateLCM
same Q
Yes
Hello, can you please upload a workflow?
I placed it in the AnimateDiff model folder, and run AnimateDiff with this motion module, however, when I run, it prompts a warning say no modules detected (A1111 WebUI)... So basically, only ComfyUI can support?
Go Fill Go!
please share workflow
could you please share workflow or made a tutorial.
I want to get exactly your sample outputs. Share your workflow your settings your everything, enlighten us please! (im using comfyui btw)
What's the matter is you share this without sharing the workflow ???
Details
Files
animatediffLCMMotion_v10.ckpt
Mirrors
animatediffLCMMotion_v10.ckpt
AnimateLCM_sd15_t2v.ckpt
AnimateLCM_sd15_t2v.ckpt
AnimateLCM_sd15_t2v.ckpt
AnimateLCM_sd15_t2v.ckpt
AnimateLCM_sd15_t2v.ckpt
rcna-tv2-model.ckpt
AnimateLCM_sd15_t2v.ckpt
LCM_sd15_t2v_beta.ckpt
Diff-LCM.ckpt
AnimateLCM_sd15_t2v.ckpt
animatediffLCMMotion_v10.ckpt
AnimateLCM_sd15_t2v.ckpt
AnimateLCM_sd15_t2v.ckpt
AnimateLCM_sd15_t2v.ckpt
animatediffLCMMotion_v10.ckpt
animatediffLCMMotion_v10.ckpt
animatediffLCMMotion_v10.ckpt
sd15_t2v_beta_motion.ckpt
AnimateLCM_sd15_t2v.ckpt
animatediffLCMMotion_v10.ckpt
animatediffLCMMotion_v10.ckpt
sd15_t2v_beta.ckpt
animatediffLCMMotion_v10.ckpt
AnimateLCM_sd15_t2v.ckpt
animatediffLCMMotion_v10.ckpt
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.