Jennifer

Model training - an open secret?!Well, some of you would say: "It's easy and self-explanatory." ...Yes, that's true. It's easy and almost self-explanatory to create a "model" ... for example known as LoRA.The open secret is: how do you train a model called a "checkpoint"? I've reviewed hundreds of pages and spent a lot of time trying to find a solution to this problem.The good thing is: I found a way to do it. Currently, I'm doing training for my own checkpoint. If the setup works properly, I'll share the configuration with you because I don't like the way other people have handled it.I don't like the idea of ​​paying for such information and I don't like answers that are wrong or incomplete, leaving you unable to create one yourself anyway.For the critical people among you: I'm not talking about "merging" a LoRA into checkpoint, but about the actual training.So if you're still looking for a solution, you can just check my profile from time to time.For anyone who wants to try it out for themselves before I finish my configuration, this article helped me get started:https://civitai.com/articles/7615/flux-dreambooth-finetune-settings-kohyassbut as the user himself mentioned, his configuration is not working properly yet. Nevertheless, I checked his configuration and tweaked a few points in it to make it a working configuration.So if anyone else has helpful information on this, feel free to share your information as well.Nevertheless, I will publish my configuration within the first checkpoint I can create properly. Since my PC is not the best, it will take some time to train a checkpoint... and then I have to evaluate its functionality. The first two iterations trained on Kohya_ss as expected, but ComfyUI could no longer identify the file as a checkpoint… So some more testing is needed.Stay tuned ;)

ComfyUI, Workflows and AI ToolsWhat is ComfyUI?ComfyUI is an open source user interface specifically designed for creating, customizing and managing workflows in AI-assisted media generation. The platform provides a modular environment in which you can link different steps of image or video creation together, similar to a visual programming tool. The intuitive drag-and-drop function makes it easy to visualize and customize complex processes.Application areas of ComfyUI workflowsImage generation and style transfersComfyUI enables users to create images of the highest quality, be it through stable diffusion, GANs or other AI models. Workflows can be configured to automate specific styles, colors or compositions. This is especially useful for artists, designers or content creators who want to achieve consistent results in their work.Video generation and editingComfyUI workflows also support the generation of video content. Using frame-by-frame style transfers or AI models for motion prediction, seamless animations or visual effects can be created. These tools open up new possibilities in film production, advertising or even video game creation.Personalized contentCompanies use ComfyUI to create personalized visual content, be it for marketing campaigns or interactive media. The ability to individually train AI models and optimize workflows ensures unique and engaging results.Benefits of ComfyUI-based toolsFlexibility: Workflows can be adapted to specific needs, from simple image editing to complex video projects.Modularity: Each element of a workflow is individually customizable, which allows iterations to be implemented quickly.Open source: As an open source project, ComfyUI allows the integration of other AI models and provides a platform for innovation and community contributions.Cost efficiency: Access to ComfyUI is free, which also enables smaller companies and individuals to generate high-quality media content.Challenges and development perspectivesDespite the many advantages, there are also challenges: Getting started with configuring workflows can be very complex for newbies. Furthermore, working with powerful AI models requires a certain amount of hardware power, which is still a hurdle for many users. This is where AI tools come into play. Here at Tensort.art, we can not only create these workflows, but also "convert" them into "user-friendly AI tools". The tool works with the created workflow and the user himself only has to enter the "necessary" inputs for the respective workflow to be able to use this workflow. Thanks to these AI tools, a user neither needs to be able to create nor understand a workflow.Conclusion: A tool for the creativity of the futureAI tools based on ComfyUI workflows are changing the way images and videos are created. They combine technical precision with creative freedom and offer both professional users and hobbyists new ways to implement their ideas. As the technology becomes more refined, ComfyUI will play a central role in creative media production.

Hi everyone.As I wrote in my last article, I am not an expert in using ComfyUI. However, in the last few days I have discovered that I can use it as I want, but others have not figured out how (yet). So this article is - really very short and sweet - about the possibility of using ComfyUI to create images using Tensor.Art's Flux checkpoint.In this article, I'm assuming you already have ComfyUI on your computer and we can start fine-tuning right away.Next, if you haven't already, please install ComfyUI Manager. This is a very useful add-on that allows you to easily add missing modules to ComfyUI for example. You can use it to install the missing modules I used in my example if you want. I'll come back to this later.Now that ComfyUI is already installed, two things are needed.Flux Checkpointput it in your ComfyUI/models/checkpoints/ folderFlux VAE Fileput it in your ComfyUI/models/vae/ folderOnce we've done that, we can create the workflow we want. I've tried to illustrate it as best as I can, you can move it around later, but for replication purposes, it was much easier to read.Notes on the workflow:If you cannot completely recreate this workflow, i.e. saving the file, then you are missing the Save Image Extended extension for ComfyUI. This can now be easily added via the manager.Please do not be put off by the warning, the module works wonderfully (with the setup shown here), although I would have liked to have saved more information when saving.You can of course add as many LoRA etc. as you like to the workflow, one extension was enough for me to start with and it is also sufficient for the example.Next we create the images. Adapt the workflow to your needs:Prompt (Positive Prompt only)LoRA(s) or not (just "Bypass" them if you don't need them)Height & Weight of the image you want to createGuidance Scale (FluxGuidance)Sampler (KSamplerSelect)Scheduler & Steps (BasicScheduler)If you used my setup in the Save Image module, you will find the image named COUNTER_SEED in your output directory. The counter is having 5 digits (if you copied my setup) and is used to know order them the same way they got created. That's why I wanted to use this extension, because I couldn't create such file names with the standard Save Image module.The seed is needed to create the same image again within Tensor.Art. I've added it to the filename, so i just need to copy the name and remove the counter part and i don't need to use other tools or workflows to get the needed information out of the meta-data of the file.Ok... The result. I created a file that I want. How do I post it here? That's not possible. But the good point: create it again without the other images you generated first until you get the image you like the most.Make sure you use the same setup in Tensort.Art as in your workflow. In my case, I would like to have this now:Then you need to add the prompt used and generate your image again... BUT...as I mentioned before, I'm not a professional in using this AI tool... so there is a difference between the generated images. The difference is soooo small that it shouldn't be a problem for you, but I'll show you an example between Tensor.Art and creating the same file locally:If you compare these images (the first one is the local version) you might see a difference in the ornament on her chest. These differences are very very small but if it is THAT detail that makes the image perfect for you it may give a different result on Tensor.Art.So if anyone reading this has a solution to this problem feel free to help me with it.Thanks for reading

Hello everyone.I have been working more intensively with various AI tools in the last few days and weeks. In this article I would like to briefly share my opinion on the "workflows" that you can create with ComfyUI.First of all, my computer is not the "more expensive, faster, better" type. It is a Ryzen 5 with a GForce 3060 Ti. So it is not bad, but by far not the best for training LoRAs, checkpoints or other AI things. It simply takes longer than with a Ryzen 9 and a GForce 4090 ;)But back to ComfyUI and the workflows.Since I have only been working with it for a few days, before that I used A1111 (Stable Diffusion), I am of course far from someone who can give you tips if you have problems. But one thing is certain: ComfyUI is definitely extremely faster than A1111 when creating images.With my current setup, I need over 2 minutes per XL image and almost 5 minutes for FLUX-based images with A1111. Anyone who can do a bit of math knows that this is really incredibly slow...ComfyUI, on the other hand, even with my setup, needs less than 20 seconds for an XL image and almost 60 seconds for a FLUX-based image. Of course, that depends on the workflow.The problem with ComfyUI, in my opinion, is that it is not at all beginner-friendly. There is a "standard" workflow, but that is not enough. After all, we want to integrate or test various checkpoints, LoRAs or other things.So you start and look at the different options... and then... then you don't know what to do next. So without looking at various documentation or examples, you will have an extremely difficult time understanding this tool.If we take the "fresh" installation of ComfyUI, after a long browse you will find that the things you actually want are "not" there. This includes things like using placeholders or a "better" way to save the files you create.This brings us to the possible extensions. Like in so many other communities, there are a huge number here. Unfortunately, this also makes things very confusing. Again, you have to look closely at what you want, need or expect, but even then it doesn't mean that the extension does what you want.The worst thing about ComfyUI in my opinion is the confusing menu and it gets worse with every extension. If you just look at the "Workflow" tool here in Tensort.Art, you immediately understand what I mean.Still. ComfyUI is a very good and powerful tool. Most importantly, it is much faster than the other tools I have tried so far. I also really like the flexibility of the tool. However, it could have a "better" menu to make it more user-friendly.If you haven't done it before: It's worth to check it out.