If you're missing the dropdown menu for VAE's - in the web Stable Diffusion go to Settings/User Interface - click User Interface - you'll see sd_model_checkpoint , add sd_vae next to it, apply settings, reload UI. tada
I recently switched to ED V3. All of my settings are pretty much the same. I used DDIM sampler as it seemed to render images the fastest on my old 980Ti. One benefit of that is that in the live preview, DDIM used to render the image with a lot of details already shown in the live preview, instead of the normal demosaicing process (don't know the exact description) where you only see color noise. It was like cheating the start. This helped my a lot since my GPU is slow, I could see right from the start if the image was to my liking, if not I could have cancelled it and let it render the next. Now this is all gone and the images are rendered from 0% like with the rest of the samplers. I need to wait for it till it reaches 50%-60% and I lose a lot of time. Can someone please enlighten my what I'm doing wrong in V3, or it isn't me and things changed with the new version. Thank you all!
Thank you very much for the video. Some great content here. However, please be advised- 1:12 the audio was really bad (not all sounds were clear). Having the music in the background (and so loud) made it really really hard to listen to this section. It is hard enough to hear words when there audio is that bad; having music like that just makes it torture, especially being that kind of music. Just had to say it. Big editing error.
I am using a MAC. When I type the command to download easy diffusion in the terminal (./start.sh), this error pops up: xcode-select: note: no developer tools were found at '/Applications/Xcode.app', requesting install. Choose an option in the dialog to download the command line developer tools.
This may not be the right place to ask this, but I have both Forge webUI and A1111 installed. I really don't want to - nor do I have the drive space to - have two different folders with modes and LoRAs. Is there a way to link them together so they share the same folder? That way I don't need to effectively double the amount of space they take up?
Great Question. Look how how to symlink folders. For example, I have ComfyUI and Auto1111. I use the same models and LoRAs in both. I keep all of them in my Comfy directory but symlink that folder to Auto1111 which then just reads that directory. This means I only need single copy of all my models and Loras :)
@0:57 Erm! Correction - A1111 also runs locally on Linux too. In fact it installs and runs better for me on Linux than on Windows.
If you're missing the dropdown menu for VAE's - in the web Stable Diffusion go to Settings/User Interface - click User Interface - you'll see sd_model_checkpoint , add sd_vae next to it, apply settings, reload UI. tada
Thanks for the video. Civitai is an awesome resource. I found it difficult to read text in this video though.
Sorry about that! Realized the bit rate was too low after the fact. We will re-record and re-upload!
@@civitai Yes, please!
I would love to see you update this with Forge instead of Automatic 1111.
The text in the footage are all pixelated...but still a good tip, thanks.
I recently switched to ED V3. All of my settings are pretty much the same. I used DDIM sampler as it seemed to render images the fastest on my old 980Ti. One benefit of that is that in the live preview, DDIM used to render the image with a lot of details already shown in the live preview, instead of the normal demosaicing process (don't know the exact description) where you only see color noise. It was like cheating the start. This helped my a lot since my GPU is slow, I could see right from the start if the image was to my liking, if not I could have cancelled it and let it render the next. Now this is all gone and the images are rendered from 0% like with the rest of the samplers. I need to wait for it till it reaches 50%-60% and I lose a lot of time.
Can someone please enlighten my what I'm doing wrong in V3, or it isn't me and things changed with the new version. Thank you all!
Thank you very much for the video. Some great content here.
However, please be advised- 1:12 the audio was really bad (not all sounds were clear). Having the music in the background (and so loud) made it really really hard to listen to this section. It is hard enough to hear words when there audio is that bad; having music like that just makes it torture, especially being that kind of music. Just had to say it. Big editing error.
We were def. still ironing out some bugs when recording this one! Thank you for the feedback!
Guys pls anybody can help me ? I receive an error said (Runtime error couldnt clone stable diffusion
I cant believe its that easy to get A1111 up and running nowadays, thats awesome for new users
I installed A 111111 but I didn't understand which version of SD I have, sd 1.5 or sd xl?
depends on the model u download on it
Hi ! I'm a begginer in all of this and my first issue is I am using AMD GPU so is there any stable version of Automatic 1111 for it ?
I have an AMD as well. Any help for us?
I am using a MAC. When I type the command to download easy diffusion in the terminal (./start.sh), this error pops up: xcode-select: note: no developer tools were found at '/Applications/Xcode.app', requesting install. Choose an option in the dialog to download the command line developer tools.
Does python and Git are not included to install when download stable diffusion?
In the particular version we highlighted, it no longer requires you to go through those steps and packages it all up for you
This may not be the right place to ask this, but I have both Forge webUI and A1111 installed. I really don't want to - nor do I have the drive space to - have two different folders with modes and LoRAs. Is there a way to link them together so they share the same folder? That way I don't need to effectively double the amount of space they take up?
Great Question. Look how how to symlink folders. For example, I have ComfyUI and Auto1111. I use the same models and LoRAs in both. I keep all of them in my Comfy directory but symlink that folder to Auto1111 which then just reads that directory. This means I only need single copy of all my models and Loras :)
@@civitai Thank you so much! You're a life (and space) saver.
@@civitai how to do it ? please reply I have both comfy and auto 11 in my pc
And why my GPU (4080 RTX) is 90+% while me doing nothing? Is it a miner?
My antivirus popped up with a warning on the Easy Diffusion installer. I am not sure to install it yet.
@@AshesForTrees-1972 I started to use ComfyUI and Stable Diffusion models there. Its cool
why don't you YTers put the link to the next video in the series in the description? Now I gotta go hunting for it.
Sorry about that! We haven't uploaded the next one yet. They will all be in the same playlist though
If you can't find the link to the next video on a channel with like 10 videos, this software might be a bit over your head.
@@timeformegamanhey dumbie it didn't exist at the time I posted it. So go jerk mode somewhere else.
does this support SDXl ?
Yes!