3.0 version: I've made some improvements. I've changed the way I use wildcards because sometimes the same seed kept appearing. I've also changed the caption method to prevent problems. Joycaption stopped working for me and I couldn't get it to recognize my model anymore, so I switched to Florence2, which doesn't have any issues.
2.0 version: I noticed that wildcards were being duplicated in the prompt text, so I fixed that. I've also changed node names to include both English and Spanish.
Hi! With this workflow, you just need to choose your character's Lora, change your character's description in the positive prompts, and click "generate." It will use wildcards to generate an image with Illustrious, then read that image and create a video from it. You'll practically have millions of different possibilities for creating a new video, With just one click!!
I've added a node between txt2img and img2video that clears the VRAM so the models load smoothly. I also added another one at the end so you can process multiple batches without any issues!
I recommend my wildcards package so you don't have to make any changes to the wildcards if you're a beginner, and I also recommend you check out my article to learn how to install things if you're a beginner: https://civarchive.com/articles/22254/eto-illustrious-txt2img-epic-comfyui-workflow
Description
Hi, I've made some improvements. I've changed the way I use wildcards because sometimes the same seed kept appearing. I've also changed the caption method to prevent problems. Joycaption stopped working for me and I couldn't get it to recognize my model anymore, so I switched to Florence2, which doesn't have any issues.
Hola, he hecho mejoras. He cambiado el método de usar las wildcards porque a veces se quedaba siempre la misma semilla. También he cambiado el método de caption para que no dé problemas. El Joycaption me dejó de funcionar y no fui capaz de que me reconociese el modelo nunca más, así que pasé al Florence2 que no da problemas.