This should get you started generating with Anima on Google Colab. This will be useful for those of us that don't have any sort of GPU - all you need is a computer that can run a web browser, a Google account, and internet access.
First, sign up for a Google account if you don't have one already. We will be storing ~7GB of files on the Drive service of that account (free-tier Google Drive provides 15GB of storage).
Second, I'm HIGHLY recommending using that Google account (or just the regular way) to sign up for an Ngrok account at https://ngrok.com. This is a service that allows you to 'tunnel' in to colab and display the interface that runs there that we'll use to generate images. Ngrok (at least as of this writing) is free to use if you sign up and tell them you're not a business. There are other similar services available that I've used in the past (cloudflare and localtunnel), but they are getting less and less reliable as time goes on.
Now, you'll download the 2 attachments - Anima.zip and Anima-workflow.json. Unzip Anima.zip and get the Anima.ipynb file out (that's the 'notebook' file that you use with Colab)
Go to https://colab.google.com, Choose the "Upload" option and give it the Anima.ipynb file. Once it opens, you'll see a screen with 3 sections, 'Install Dependencies', 'Download Resources', and 'Run ComfyUI'. The first time through, we have to do things a little differently so we can get ComfyUI set up (that's the interface I was talking about that you use to generate images). After that, you skip a couple steps each time you start Colab back up.
Expand the 'Install Dependencies' section and look to the right for the 4 checkboxes - check all of them. You do need to run this every time you start colab back up, but after this first run, you'll only need the first and last checkboxes checked, and you can uncheck the middle 2 to save some time starting up. Once those checkboxes are checked correctly, click the Run button (the little triangle inside the circle to the top left of all the code). It takes awhile, and will ask you to connect to your Google Drive. Once it's complete, make sure to uncheck the second and third checkboxes before moving on.
Now, expand the 'Download Resources' section and run it by clicking the Run button. We only need to run this once - this downloads the Anima v1.0 base model, as well as the necessary text encoder, VAE, and Turbo LoRAs from huggingface.com and saves them to your Google Drive. You don't HAVE to use the Turbo LoRAs, but I'm including them anyway because they're useful and popular. They allow you to generate images in 12-ish steps instead of the usual 30-50 - meaning a 1024x1024 image will take 16-17 seconds to generate, once the model is loaded.
Finally, expand the 'Run ComfyUI' section. Look to the right for the 2 checkboxes and the text box. If you decided to use Ngrok, you'll need an Authtoken from Ngrok to use the service. Log in at https://ngrok.com, then look for "your Authtoken" in the list of links to the left. Copy your Authtoken from that page and paste it in the "NGROK_TOKEN" box. If you don't want to use NGROK, I've had the best luck lately with localtunnel. Either way, you'll get a link to ComfyUI in the output once it starts up, you click on it and confirm, and that should start up ComfyUI. Once you've got the token pasted in or have chosen another option, click the Run button to start up ComfyUI. This takes a LONG TIME. Watch the output for a link to the tunneling service you chose and click on it, and ComfyUI should start loading. Loading takes awhile, especially the first time. I had to shut it down and restart a couple of times (the Run button turns into a 'Stop' button with a square instead of a triangle, you can click that to stop the section of code from running), and restart. Sometimes, if you get a '403' error, clicking in the address bar of the window and hitting 'Enter' to do a hard reload has worked for me. This is the hokiest, most annoying part of the whole setup, but, eventually, you'll get a running ComfyUI window that by default the first time opens a list of templates.
If the list of templates is open, close it, then open the Anima-workflow.json file that you downloaded from the attachments. The File/Edit/whatever menus should be along the left side of the ComfyUI window near the top, or they're hidden under the button with the ComfyUI 'C' logo on it. Once you have that workflow open, it should already be configured with all the models you downloaded. I think I even included a test prompt with the workflow that should generate a picture of a certain comic book heroine. I tried to color-code all the nodes that you would need to adjust in the workflow Green. Specifically, there's a node marked 'Positive Prompt' that you will use to tell the model what you want it to draw.
Once you've generated your images, you can right-click on them in the preview window to open or save them. This workflow works like any other ComfyUI workflow, so you can add nodes if you want (just remember, the Save Image nodes will save things to your Google Drive - you'll have to go into your Google Drive and download them from there). Also, I've included the civicomfy plugin that you can use to download other LoRAs and things from civitai.red or civitai.com as the mood may strike you. You can also save the workflow to your Google Drive so that ComfyUI will have it the next time you start it up - there's a Save option in the File menu, or just push Ctrl+S.
Important Safety Tip: You must shut down your Colab instance when you are done using it - just closing the window doesn't do that, and you only get a certain amount of time for free each day. You shut down by clicking the down arrow in the very upper-right of the Google Colab window and choosing "Disconnect and Delete Runtime". Make sure you do that when you are done.
Please see my other articles that are similar to this one for more in-depth instructions and screenshots that should be pretty close to what you're trying to accomplish here. the SDXL article is a good place to find screenshots if you're confused on what I'm talking about in a step - I typed these instructions up very quickly.
The workflow I'm including is very bare-bones, but it should work as a starting point. The first image right after starting ComfyUI up will take a really long time, due to the models getting pulled from Google Drive, but images you generate after that will go much more quickly, once the models are all stored in your colab instance's memory.