Home Artists Posts Import Register

Content

I've recently added all my raw workflows on my Google drive where you can see all my past workflows to study the "Prompts and settings"  I've used to make those videos ie the prompts, loras, models, controlnet settings, versions ... etc

A Folder with the respective name has been provided here : https://drive.google.com/drive/folders/1HoZxKUX7WAg7ObqP00R4oIv48sXCEryQ?usp=drive_link

It will be regularly updated once the proper version of that workflow is released.
Render Videos Link: https://www.youtube.com/@Ai_Davos/shorts

How to Load Workflow


1) Download the photo you want to see, drag and drop the photo into the comfy workspace to load up.

2) If you don't want to download the workflow, you can open the image in a new tab and directly drag and drop from the webpage.


How to Find Loras/Models


Go to https://civitai.com and search for loras from the Lora Stack list, (if you cannot find the loras from the stack that means it has been renamed or removed from site, then find something similar)

Good Luck Creating !!

My Discord Server : https://discord.gg/z9rgJyfPWJ

- Jerry Davos

Files

Comments

delly moore

Hi,Jerry.In this workflow, how do I add a new background image? I've used IPAdapter with Prompt control, but the results are still not satisfactory.

Jerry Davos

Yes, IP works best for single image, but when using for animations, its not up to the mark, but when using low ip weights(0.2- 0.5) and End percent (0.3 - 0.6) , results gets a little better... I am also testing, how to use IP properly in this workflow.

Alexandr Bondarchuk

Hello Jerry! Can you please help with this message in console? When I tried to refine images, this showed in console Using pytorch attention in VAE Working with z of shape (1, 4, 32, 32) = 4096 dimensions. Using pytorch attention in VAE missing {'cond_stage_model.clip_l.logit_scale', 'cond_stage_model.clip_l.text_projection'} left over keys: dict_keys(['model_ema.decay', 'model_ema.num_updates', 'cond_stage_model.clip_l.transformer.text_model.embeddings.position_ids']) I tried to find info, but only found opened issue in Github in Comfy repo. Thank you!

Jerry Davos

I have encountered something like this once, it may be that : You are using SDXL or other version of SD somewhere in this workflow, Check for 1) SDXL in Model loader 2) SDXL Vae 3) SDXL Lora 4) SDXL Embeddings Also this can happen if you are using some personal trained model... Try different SD model.... This workflow is only compatible with SD 1.5 models, other versions give dimension errors