ComfyUI AnimateDiff Prompt Travel: LoRAs and Embeddings
This video gets you fully set up with LoRAs, embeddings, as well as notes a couple tricks to make working in comfyUI easier: dragging png files to load work flows, git pull for updating your install with the most recent changes, and installing Tiny Terra Nodes for making your embeddings workflow drastically easier.
Tiny Terra Nodes Repo for making embeddings easier:
github.com/tinyterra/ComfyUI_...
The workflow file can be downloaded from here:
drive.google.com/file/d/1y26u...
The model (checkpoint) downloaded for this tutorial series are here:
civitai.com/models/25694/epic...
The VAE used can be downloaded from:
huggingface.co/AIARTCHAN/aich...
The motion_modules and motion_loras can be found on the original AnimateDiff repo where you will be offered different sources to download them from:
github.com/guoyww/AnimateDiff
Or here's a quick link to civitai:
civitai.com/models/108836/ani...
civitai.com/models/153022
Socials:
x.com/c0nsumption_
/ consumeem
Пікірлер: 32
Tutorials are lean and clean, I like it!
you are killing it with this tutorials !
I'm learning so much form you! Dead Ass! I love your simple, practical delivery. Its easy to follow.
@c0nsumption
8 ай бұрын
Thanks. Means a lot, it’s really hard keeping it to the point and concise. Usually have to refilm a bunch 😅
Love your tutorials, super clear and easy to follow. As a ComfyUI newcomer they have been super helpful, really appreciate you sharing your workflows with us!
yeah this works much better for animatediff like you stuff glad i subbed
@c0nsumption
9 ай бұрын
Thanks! More is for sure in the way 🙏🏽
thanks your tutorials are awesome
Thanks!
Very useful. Thank you.
Thanks a lot! Still waiting for the best practices on using Animatediff either control nets
@c0nsumption
9 ай бұрын
I have a lot coming. Just working right now sorry 🧍🏽♂️ Most definitely will have some ControlNet stuff up by Sunday!
Definitely learning a lot from this channel so far. I seem to be getting better, sharper images in Automatic 1111 than I do with comfy ui (using this workflow) but I am still a noob on comfy. I noticed some models don't look as good in Comfy as in Auto for me.
@c0nsumption
9 ай бұрын
Could be the sampler you’re using. Also don’t have comfy save as GIF, it diminishes the quality a lot. I say: just keep fiddling with it. How much VRAM you have?
Love the output, but I cannot figure a way to change the model (actual model not checkpoint) usually there's like a load video or load image to work with,
Thanks! Next vid2vid with AnimDiff Comfy UI?
@c0nsumption
9 ай бұрын
Next 3 videos are IPAdapter, 3 Init Image ControlNet, and Interpolation/Refiner. After that Vid2Vid. I promise. I’m working right now but will film tonight so that I can post in the morning 🙏🏽
@automioai
9 ай бұрын
YOu are doing it @@c0nsumption , amazing well explained.
Is there a way to autocomplete lora names in the prompt field just like you did with embeddings?
@c0nsumption
9 ай бұрын
2:38 LoRAs are not added into text fields. Please watch video fully before asking such questions 🙏🏽
thank you so much ! I installed everything in ComfyUI but there's no box to put the HEAD PROMPT, do I have to install another node (tail prompt aswell)
@c0nsumption
9 ай бұрын
pre_text (head) and app_text (tail) sections in the BatchPromptSchedule node
@atlasv2562
9 ай бұрын
thank you so much !@@c0nsumption
@kenrock2
8 ай бұрын
@@c0nsumption thanks, i was about to ask the same question too
Is there any way to get the process using Mac?
@c0nsumption
8 ай бұрын
You would have to reference the documentation: github.com/Kosinkadink/ComfyUI-AnimateDiff-Evolved
stable diffusion animatediff extension got update and now support similar features would be cool if you could take a look and make a vid
@c0nsumption
9 ай бұрын
In the webUI? Id say stick with comfyUI or the CLI. Auto1111 isn’t really developer friendly anymore. I can but it’ll take a while as I’m primarily focused on CLI and ComfyUI tutorials for a while. I work full time so it’s a bit difficult to find time windows 😅
@wagmi614
9 ай бұрын
@@c0nsumption check video by AI in Wonderland, it's in Japanese but most likely you'll figure out what they're doing on the screen, it's quite detailed would love simplified version of that in English on your channel
Oh my that UI is the opposite of comfy. That is hands down the ugliest and sketchiest ui i have ever seen.
@c0nsumption
9 ай бұрын
????????