Thanks a lot for the quick rundown. Appreciated! Looks like my A1111 is finally going in the bin, WITHOUT me having to get to grips with ComfyUI hard core.
You're welcome! It still helps to know ComfyUI to take full advantage of SwarmUI. I go over a simple way to using custom workflows here. ua-cam.com/video/MMQcfvGJYZs/v-deo.htmlsi=w7vDHhm-5SLNYX8X I'll definitely go more in depth on this soon!
Sure that's an option too however not everyone likes to manage all their platforms with one installer. I like Pinokio since it can install other AI platforms other than text to image.
@@MonzonMedia Haha, I'm new to this so the number of LLM options in Pinokio intimidates me. I still have it, because it might be a better fit once I know better what I'm doing. For now I'm using Stability Matrix for Text2Img and LM Studio for chat/code generation. I haven't messed with any other categories of AI tools. But anyways, I definitely recognize that an installer isn't for everyone, or one particular installed even if they like the idea. The reason I brought it up is because I'm on Linux, and installers like this make it extremely easy to set up on an otherwise complicated system. This is very good to know for anyone else who's new to this and might not know how to install AI platforms manually, or doesn't want to manage updates and linking model folders manually. Stability Matrix doesn't even need to be installed on Linux. Just download the .AppImage file and run it. The file isn't an installer package, it's the actual app itself. Very convenient, harkening back to ye olden days of computing when you could download an exe and put it wherever you wanted. The only problem I've faced is when Stability Matrix itself gets an update, it just downloads the newest .AppImage and puts it in a subfolder. Then I have to manually move the downloaded .AppImage over the older version to replace it. Then it works fine. It would be a better user experience if it restarted itself and did the swap with a temporary script, or something. But, as it is, it's a little tedious, but not difficult, even for a newb like me. For the record, Pinokio and LM Studio are also available as .AppImages on Linux. So the same convenience of use applies to them too. I haven't used either long enough to know how they handle updates though, so I don't know if they have the same minor hoop jumping.
i already have sdxl as i was using a1111. can we use symlink so both a1111 and swarm can use same model instead of having multiple copies of same models?
Absolutely, personally I don't use Symlink but I've been meaning too since I've heard many users on this channel use it. I just point to the model directory within swarm to my models folder that's on another drive.
Have you found any way or workflow to use llm to image gen ? In ollama with open webui I can see a1111 and comfiui. Im not sure how to integrate both. It will be so much faster if i can tell LLM my requirement and it gives prompt directly to swarm or comfiui
LMAO! SAI said they are improving SD3 Medium so let's see. As I said in the video, you can use SwarmUI for SD1.5, SDXL, Pixart and soon I hear they may support Pony XL. I'm also looking forward to see what the OMI group does. Been following their progress.
Hi friend, this model is running slowly on my notebook configuration: Intel(R) Core(TM) i5-5200U CPU @ 2.20GHz 2.20 GHz Installed RAM 16.0 GB System Type 64-bit operating system, x64-based processor Windows 10 Home Single Language Edition Version 22H2
It's still in Beta but right now other than ComfyUI and SDNext, this is the only platform that supports SD3 and Pixart. There is more to come so let's see how it develops. Not sure if or when Fooocus will support SD3 and Pixart, I heard they are looking to support Pony models soon.
I'm editing the video on using SD3 Medium for Swarm UI. It's much simpler than ComfyUI. Also working on a video to use the Pixart models. Stay tuned!
Thanks a lot for the quick rundown. Appreciated!
Looks like my A1111 is finally going in the bin, WITHOUT me having to get to grips with ComfyUI hard core.
You're welcome! It still helps to know ComfyUI to take full advantage of SwarmUI. I go over a simple way to using custom workflows here. ua-cam.com/video/MMQcfvGJYZs/v-deo.htmlsi=w7vDHhm-5SLNYX8X I'll definitely go more in depth on this soon!
Quick Tip!! I had to right click bat file and run as Adminstrator to get the Swarm Ui installer to run.
It looks like Stability Matrix supports SwarmUI as well. Much easier to install, update, and share models that way.
Sure that's an option too however not everyone likes to manage all their platforms with one installer. I like Pinokio since it can install other AI platforms other than text to image.
@@MonzonMedia Haha, I'm new to this so the number of LLM options in Pinokio intimidates me. I still have it, because it might be a better fit once I know better what I'm doing. For now I'm using Stability Matrix for Text2Img and LM Studio for chat/code generation. I haven't messed with any other categories of AI tools.
But anyways, I definitely recognize that an installer isn't for everyone, or one particular installed even if they like the idea. The reason I brought it up is because I'm on Linux, and installers like this make it extremely easy to set up on an otherwise complicated system. This is very good to know for anyone else who's new to this and might not know how to install AI platforms manually, or doesn't want to manage updates and linking model folders manually.
Stability Matrix doesn't even need to be installed on Linux. Just download the .AppImage file and run it. The file isn't an installer package, it's the actual app itself. Very convenient, harkening back to ye olden days of computing when you could download an exe and put it wherever you wanted. The only problem I've faced is when Stability Matrix itself gets an update, it just downloads the newest .AppImage and puts it in a subfolder. Then I have to manually move the downloaded .AppImage over the older version to replace it. Then it works fine. It would be a better user experience if it restarted itself and did the swap with a temporary script, or something. But, as it is, it's a little tedious, but not difficult, even for a newb like me.
For the record, Pinokio and LM Studio are also available as .AppImages on Linux. So the same convenience of use applies to them too. I haven't used either long enough to know how they handle updates though, so I don't know if they have the same minor hoop jumping.
i already have sdxl as i was using a1111. can we use symlink so both a1111 and swarm can use same model instead of having multiple copies of same models?
Absolutely, personally I don't use Symlink but I've been meaning too since I've heard many users on this channel use it. I just point to the model directory within swarm to my models folder that's on another drive.
Have you found any way or workflow to use llm to image gen ?
In ollama with open webui I can see a1111 and comfiui. Im not sure how to integrate both.
It will be so much faster if i can tell LLM my requirement and it gives prompt directly to swarm or comfiui
Hello, yes, if you could show how the video part works. That would be great. Thank you for the videos. William
Seconded!
Will do!
Can we download models manually and add them later on in swarm UI as i already have flux model in my ssd
is there a text box for me to upload 100prompts at a time is swarm-ui?
As far as I know not in Swarm directly however I suppose if you are familiar with comfyui you can create a workflow to do that.
No one likes Stability AI anymore lol.
Even Comfy team is focusing on a new model with the Invoke guys. Hope they get really successful.
LMAO! SAI said they are improving SD3 Medium so let's see. As I said in the video, you can use SwarmUI for SD1.5, SDXL, Pixart and soon I hear they may support Pony XL. I'm also looking forward to see what the OMI group does. Been following their progress.
Hi friend, this model is running slowly on my notebook configuration: Intel(R) Core(TM) i5-5200U CPU @ 2.20GHz 2.20 GHz
Installed RAM 16.0 GB
System Type 64-bit operating system, x64-based processor
Windows 10 Home Single Language Edition
Version 22H2
Likely your onboard graphics card is too weak. Dedicated gpus are needed to run decently.
They write to us Flux works 10 times faster
nice idea ... but not as beautiful as you can get in a real comfyui workflow. Even Fooocus might do better than swarmUI
It's still in Beta but right now other than ComfyUI and SDNext, this is the only platform that supports SD3 and Pixart. There is more to come so let's see how it develops. Not sure if or when Fooocus will support SD3 and Pixart, I heard they are looking to support Pony models soon.
@@MonzonMedia I love pony models :D, I'm waiting for the SD3 pony as well :). Ok, I'll look into your video of SD3 with swarmui too