Choosing the model in SD generators is a double edged sword. I understand the power of having the exact model for the thing you want, but having to find a model and choose the model among all the models that are out there is a massive ball-ache. Much prefer MJ and just being able to type the style you want and have it work (most) of the time :)
You've got the trade off in a nutshell. Stable Diffusion models give you lots of freedom, but there are so many! Midjourney gives you it's style, and it's great. If this is the style you like, you don't have to look very far.
Speaking as a person who is 0 in 2 for downloading Stable Diffusion on my laptop, Tensor is a great program. Still not easy, it is possible to get images. Thanks for this introduction into the different settings on Tensor. I appreciate that Tensor is free (for now) allowing for a lot of experimentation and mistakes. Thank you for this video and all your hard work.
once again thank you Jenn, this really helped, I put this on loop for about an hour whilst working through your guide and it accelerated my outputs ten fold, with a much clearing understanding of what I'm doing whilst still experimenting.Makes a big difference knowing what your tweaking 🤩 @@MakingPhoto
I'm absolutely in love with Tensor Art . The ability to upload my checkpoint file created in colab is just freaking GREAT. Especially for me who dont have a high end pc with a gpu.
@@MakingPhotoNow I want to make AI arts of an anime character which already exists. But that face model is not in the Tensor Art. How to get that particular face and hairstyle for that character to generate the images?
The AI won't reliably make any characters. There are usually changes. But if the AI knows a celebrity or character, you'll get many of their features. But if that particular character isn't known by the AI, try adding an image prompt to show the AI what you want.
Incredible video! You levelled up my knowledge of AI image generation about 20 times in the course of explaining the platform. I would have missed out on so much of Tensor by just poking around. I'm very excited about this platform now. Thank you!
I know this is my second comment, but I literally used this video as a step-by-step running of Tensor (yes, your video is that good). This was my second shot at generating an image. While not easy, your video made the process easier to understand. I also was able to go back and use your invitation code. I hope it helps you.
Thanks! I'm glad it helped. Stable Diffusion has a lot of power, but you have to know how to drive it (like a fast car). Thanks for letting me know about the Invitation Code. I'm glad they got that fixed.
Thanks, Jen! It's great to learn about new Ai generative platforms! It would be great if you could to a head to head comparison of the different free Ai generative platforms to see which ones create the best images.
@@MakingPhoto I know right? So many new playgrounds to play in! It's a lot of fun but it's hard to know which ones to focus on, as it's hard to know which ones produce the best images. A comparison video of the different free platforms would be awesome! :) Only if you would want to make another video like that though. I'm sure they're probably a bit difficult to make and can be pretty time consuming. It would definitely be helpful to the community though, I think - if you're up to it! :D
@@MakingPhoto hahah True that! AI stuff is changing at lightning speeds! Very cool new things and new text to image sites keep popping up! I'm really enjoying SeaArt right now though. Clipdrop is also pretty awesome!
I would try Tensor and Leonardo if it weren't for the _slightly_ daunting learning curve. This video is highly appreciated. A deeper dive into this and Leonardo would be very helpful. Going to watch the negative prompting video. Thank you so much for this. 🙏
@@MakingPhoto I've been using MJ now for about 6 months. The lack of control is what many wish to get rid of. Most of the SD videos I've watched are either not comprehensive enough, or so indepth for 30 minutes that even setting it up seems difficult. Thank you for replying, and I do hope you consider videos about SD in your presentation style. 👍
@@AG_before I had the same experience when I was first starting. Now, there are programs like Tensor Art that help create Stable Diffusion images. When I first started, it was all done locally through Github! My brain about exploded!
Thanks for this awesome video! I've learned a lot from it! Do you mind please explaining to me about the models that I see on Tensor Art? For example, I see a lot of different "selections"? (for a lack of a better term). Those selections all have users/creators that you can follow. And most of these selections or filters or models that you chose, the creator says you have to credit them, or cannot download their model, etc. My question is, how do you create what they've created from scratch? or how did these people create what they have the rights to at the first place? I want to know more about that. For example, some faces look very real but I am not sure if those are real people, because someone used Jenna Ortega's photo, then I know for sure that's a real person (Jenna Ortega, because she's famous), but some other faces look very realistic and I don't feel comfortable using them because they may be real people. So my question is how do I create a model from scratch that I know I can comfortably use knowing it's not real? Thank you very much for your time and sorry for such a long wordy question.
With Stable Diffusion-based AI, you can train your own model using similar images. The models are designed to make very specific things or people or settinsg. I did a very basic video to help folks understand Create Stunning Images with Stable Diffusion Models ua-cam.com/video/cV5QNFX_XLA/v-deo.html Tensor art doesn't make these models, but it hosts them. The creators let you use their models.
I want to generate so many images with different prompts of a person which i generated earlier, the face of each image should be the same, how can i do this?
This is tricky because the AI "generates" or changes things each time it makes an image. There are some tricks for character development. Try using image prompts. I have an older video (which I should update) about making characters in Midjourney. "Re-imagining Storytelling with Midjourney: Creating a Recurring Character" ua-cam.com/video/jdFo-QDhBl0/v-deo.html
I'm a huge fan of this channel. Just subscribed a few days ago :-) Thanks for creating this super informative and helpful video, Jen! I have an idea for an AI art project, but I'm not sure which AI art generator to use. I'd love to see you create a comparison/contrast video between Stable Diffusion and Midjourney. I liked your other video about using Midjourney and Dupdub to create consistent characters. It it also possible to create consistent characters on Stable Diffusion? I'm trying to decide which of the two I should start learning to use! Any advice would be appreciated. Thanks in advance! -Eric
Thanks! With Stable Diffusion, there's a tool called ControlNet that helps with creating consistent characters. Let me see if I can make that into a video, too.
Short answer - you can't. The AI image generators are always "generating" or "creating". I have some tricks for consistent characters in an upcoming video. But try adding an image prompt of you and highlight your features in the text prompt.
Thanks Jen! When I try to explore Tensor Art, a message pop-up "MakingThePhoto" Invitation Code is invalid, can you make it right or provide another code?
help please how do you convert images i have all ready for into a model set to use on tensor please as it want file formats of these .ckpt | .pt | .safetensors | .pth to upload onto site kind regards
I have a question, I made an ai persona using tensor art and the next day I couldn’t continue using the same. Is there any chance that I can get back the person I created before?
My tennsort ar errored, I couldn't enter the prompt I wanted, all I saw was a blank window, there is no prompt box, just a words generate. do you know how to fix it?
Good Job! Anyway i'm still struggling with creating an AI Avatar that similar to my own face...It always look big different although i uploaded my own facial image & try different setting
Yes. It will do this. The AI "generates" - or creates. Make sure the denoising slider is all the way up. In Stable Diffusion, you almost have to train a model with multiple images of your face. In Midjourney, you can use multiple image prompts, which helps.
Dear Friends, may I ask for advise. I have error message on my workspace and cannot continue to work on images. Anyone knows how I can contact TENSOR for help to resolve this problem. Thanks.
How can I train the Ai to generate images based on my face for example generating images of the same person like myself, I got photos of me which I wanna generate something using tensor with it, is there a way for me to do it?
I ❤ TensorArt a lot, but today I've found out that I can't login and access my account. Data won't load. And since there is no support, I have to create a new account and start all over again😢
Just as I go pro, "Models not Found" comes up. After a. couple of months without problems. Is there any help anywhere? Nothing, not nowhere. So buyer beware
I find it so sad that I almost always have to write "bare breasts, bare stomach, bare legs" in the negative prompt in order to not get an almost pornographic image.
I tried using systems with models and was disappointed Not because the images weren't good, but because they all felt the same The EXACT same tone, same sensibility, same aesthetic. BORING! This is why I like Gabby/Gab AI, it's more random in the output, so there's lots of variation
It seems like all the images are borderline pornographic images of half naked women! Why are there no images of cute animals or even ONE image of a .... male? 🤔
Tensor Have Alot of NSFW pic. To be honest i tried creating one and boi it did not disappoint. But tensor is great over all just turn on the nsfw filter
I am starting to think that Tensor art is racist. I put in prompt brown skin, Carmel skin, latin lady and they keep giving me a white girl lol. Do I need to use an add lora that is black already to get a person of color?
I'll check this out! It all depends on how the AI was trained. For a long while Midjourney and other AI did this, too. Try Asian. Some of the major AI companies are based in Asia.
I'm sorry you hate my voice . It's the only one I have. I'm not crying , I'm excited to teach you about tensor art. I'm sure I can find some physical feature about you that is sub-par. Let me call you out on that. Why don't you find some other channel to follow - if you can find content like mine, go find it somewhere else.
Oh damn it! Not only does this say zero helpful info about the ADetails subsection, it also ignores the entire list of hires upscalers! So every bloody beginner is doomed to waste all daily free credits in useless guessing game clicks bc most upscale models won't enhance the image we currently work on. I say this after wasting 2 days with useless upscalers and most of the time only some ultra sharp did any nice result.
Thumbs down for such a plain overview that anybody can guess by using tensor website for 1 day. What is missing is the ACTUAL deep info about what nuances the cryptic bs named sub-options do. How many damn videos do we have to watch for all those?!
If you're getting this information elsewhere, than go find it. I've watched a lot of videos on SD and knowing what each tool does is key to getting good results. If you know it all - good for you. But a lot of creators don't.
Choosing the model in SD generators is a double edged sword.
I understand the power of having the exact model for the thing you want, but having to find a model and choose the model among all the models that are out there is a massive ball-ache.
Much prefer MJ and just being able to type the style you want and have it work (most) of the time :)
You've got the trade off in a nutshell. Stable Diffusion models give you lots of freedom, but there are so many! Midjourney gives you it's style, and it's great. If this is the style you like, you don't have to look very far.
MJ is great, but theres so much you cant do. I prefer MJ for aesthetic and creative images but tensor for images I have in mind.
@@MakingPhotothe invite code is expired
probably best guide for someone who has just started with ai image generation, thank you. no one talks about these setting and levers and what the do.
When I was first learning, I had a hard time finding info. SD is only good if you know how to fly the controls!
@@MakingPhoto youre absolutely right, are there gonna be more videos like that in future?
Thank you so much. Now I have an understanding of what I am doing.
Speaking as a person who is 0 in 2 for downloading Stable Diffusion on my laptop, Tensor is a great program. Still not easy, it is possible to get images. Thanks for this introduction into the different settings on Tensor. I appreciate that Tensor is free (for now) allowing for a lot of experimentation and mistakes. Thank you for this video and all your hard work.
Great! That was always my struggle, too. TA seemed to have all the tools
Wonderful clear overview, thank you Jenn
You are so welcome!
once again thank you Jenn, this really helped, I put this on loop for about an hour whilst working through your guide and it accelerated my outputs ten fold, with a much clearing understanding of what I'm doing whilst still experimenting.Makes a big difference knowing what your tweaking 🤩 @@MakingPhoto
That's terrific! Stable Diffusion can be more personalized than other tools, BUT you need to know what all the knobs do.
I'm absolutely in love with Tensor Art . The ability to upload my checkpoint file created in colab is just freaking GREAT. Especially for me who dont have a high end pc with a gpu.
Glad this is working! Tensor art seemed to have all the tools you'd want.
@@MakingPhotoNow I want to make AI arts of an anime character which already exists. But that face model is not in the Tensor Art. How to get that particular face and hairstyle for that character to generate the images?
The AI won't reliably make any characters. There are usually changes. But if the AI knows a celebrity or character, you'll get many of their features. But if that particular character isn't known by the AI, try adding an image prompt to show the AI what you want.
Incredible video! You levelled up my knowledge of AI image generation about 20 times in the course of explaining the platform. I would have missed out on so much of Tensor by just poking around. I'm very excited about this platform now. Thank you!
Great to hear!
Outstanding informative video. This has been well put together and the most helpful that I have came across. Happily sharing it.
Glad it was helpful!
I learn more from your video than from exploring the tool on my own.
Thanks! I love that you do both - watch videos and explore!
I know this is my second comment, but I literally used this video as a step-by-step running of Tensor (yes, your video is that good). This was my second shot at generating an image. While not easy, your video made the process easier to understand. I also was able to go back and use your invitation code. I hope it helps you.
Thanks! I'm glad it helped. Stable Diffusion has a lot of power, but you have to know how to drive it (like a fast car). Thanks for letting me know about the Invitation Code. I'm glad they got that fixed.
Is it for commercial use?
Excellent... thank you Jen 👍
You bet!
Thanks a lot for this helpful video. Much appreciated.
No problem! Hope this helped.
Thank you for this!
My pleasure! Hope it was helpful
Excellent , very helpful , just what i was looking for , subscribed liked and hit bell, thank you 💓💞💕
Thanks! Hope this helps you!
Thanks, Jen! It's great to learn about new Ai generative platforms! It would be great if you could to a head to head comparison of the different free Ai generative platforms to see which ones create the best images.
Great idea! I haven't done one of these videos for a while! New AI image generators coming on to the market so fast!
@@MakingPhoto I know right? So many new playgrounds to play in! It's a lot of fun but it's hard to know which ones to focus on, as it's hard to know which ones produce the best images. A comparison video of the different free platforms would be awesome! :) Only if you would want to make another video like that though. I'm sure they're probably a bit difficult to make and can be pretty time consuming. It would definitely be helpful to the community though, I think - if you're up to it! :D
I did a video like this only a few months ago - and things have already changed so much!!
@@MakingPhoto hahah True that! AI stuff is changing at lightning speeds! Very cool new things and new text to image sites keep popping up! I'm really enjoying SeaArt right now though. Clipdrop is also pretty awesome!
I would try Tensor and Leonardo if it weren't for the _slightly_ daunting learning curve. This video is highly appreciated. A deeper dive into this and Leonardo would be very helpful. Going to watch the negative prompting video. Thank you so much for this. 🙏
You're right. Stable Diffusion has a lot of flexibility but there is a learning curve. Other AI like MJ are less daunting.
@@MakingPhoto I've been using MJ now for about 6 months. The lack of control is what many wish to get rid of. Most of the SD videos I've watched are either not comprehensive enough, or so indepth for 30 minutes that even setting it up seems difficult. Thank you for replying, and I do hope you consider videos about SD in your presentation style. 👍
@@AG_before I had the same experience when I was first starting. Now, there are programs like Tensor Art that help create Stable Diffusion images. When I first started, it was all done locally through Github! My brain about exploded!
What a great video, thanks.
Glad it was helpful!
Thank you so much
thanks so much! 😍💖
No problem 😊
Tensor Art is the best AI generation platform. Even it can provide you with convenience in your work!😁😁
It worked really well for me. It's quick and has all the tools.
Wonderful video, thank you! Are long wait times for image generation normal? It takes quite a few minutes for mine to generate
I've had all sorts of times. I think it probably depends on how many people are on the site at the same time.
Thanks! Great video!
no problem!
LOVE ITITT!!!
Hope this helps!
Can these things be done on a mobile phone❤
Thanks for this awesome video! I've learned a lot from it!
Do you mind please explaining to me about the models that I see on Tensor Art? For example, I see a lot of different "selections"? (for a lack of a better term). Those selections all have users/creators that you can follow. And most of these selections or filters or models that you chose, the creator says you have to credit them, or cannot download their model, etc. My question is, how do you create what they've created from scratch? or how did these people create what they have the rights to at the first place? I want to know more about that. For example, some faces look very real but I am not sure if those are real people, because someone used Jenna Ortega's photo, then I know for sure that's a real person (Jenna Ortega, because she's famous), but some other faces look very realistic and I don't feel comfortable using them because they may be real people. So my question is how do I create a model from scratch that I know I can comfortably use knowing it's not real?
Thank you very much for your time and sorry for such a long wordy question.
With Stable Diffusion-based AI, you can train your own model using similar images. The models are designed to make very specific things or people or settinsg. I did a very basic video to help folks understand Create Stunning Images with Stable Diffusion Models ua-cam.com/video/cV5QNFX_XLA/v-deo.html Tensor art doesn't make these models, but it hosts them. The creators let you use their models.
@@MakingPhoto Ah gotcha! Thank you very much for the explanation!
hi, theres no way to upload an image whitout the ai to edit first; to inpait ?
I want to generate so many images with different prompts of a person which i generated earlier, the face of each image should be the same, how can i do this?
This is tricky because the AI "generates" or changes things each time it makes an image. There are some tricks for character development. Try using image prompts. I have an older video (which I should update) about making characters in Midjourney. "Re-imagining Storytelling with Midjourney: Creating a Recurring Character" ua-cam.com/video/jdFo-QDhBl0/v-deo.html
Are you allowed to use images generated from models for commercial use?
Yes.
TOOOOOOOOOOOOO DIFFICULT 🙏
sorry! Stable Diffusion can be head scratching. There are easier AI image generators out there. I often use Midjourney.
Can you please make a video about hosting a model😢???
Good idea! I'll see what I can do.
I'm a huge fan of this channel. Just subscribed a few days ago :-) Thanks for creating this super informative and helpful video, Jen! I have an idea for an AI art project, but I'm not sure which AI art generator to use. I'd love to see you create a comparison/contrast video between Stable Diffusion and Midjourney. I liked your other video about using Midjourney and Dupdub to create consistent characters. It it also possible to create consistent characters on Stable Diffusion? I'm trying to decide which of the two I should start learning to use! Any advice would be appreciated. Thanks in advance! -Eric
Thanks! With Stable Diffusion, there's a tool called ControlNet that helps with creating consistent characters. Let me see if I can make that into a video, too.
Looking forward to it! Thanks again for sharing your experience and techniques!
@@MakingPhoto
Would you say that you prefer Midjourney to Stable Diffusion? Which do you think is easier to use for a beginner?
@@MakingPhoto
how to get exact face of my reference image to selected art?
Short answer - you can't. The AI image generators are always "generating" or "creating". I have some tricks for consistent characters in an upcoming video. But try adding an image prompt of you and highlight your features in the text prompt.
@@MakingPhoto thnx, i m waiting
Thanks Jen! When I try to explore Tensor Art, a message pop-up "MakingThePhoto" Invitation Code is invalid, can you make it right or provide another code?
Thanks for letting me know. I'll reach out to the TensorArt team.
Is tensorart still free? Loaded by pro purchase comes up?
Tensorart has changed A LOT since I did this video!! Did you try CREATE button and then sign in?
Can't find anywhere to upgrade to PRO version, $15.....
What VAE is commonly used to add freckles to faces and skin?
Hmm. When I ask for freckles. I get freckles. Maybe something else in the prompt is negating this?
@@MakingPhoto any Loras that you know are decent for adding freckles? Should they show after image generation or after being processed in Adetailer
Diffusion's AI is in love with Megan Fox. It keeps generating Megan Fox lookalikes.
What are you promoting 😉
help please how do you convert images i have all ready for into a model set to use on tensor please as it want file formats of these .ckpt | .pt | .safetensors | .pth to upload onto site kind regards
I have a question, I made an ai persona using tensor art and the next day I couldn’t continue using the same. Is there any chance that I can get back the person I created before?
Sorry! This is a technical question for TensorArt.
My tennsort ar errored, I couldn't enter the prompt I wanted, all I saw was a blank window, there is no prompt box, just a words generate. do you know how to fix it?
Mine was acting up, too. Hopefully it's a temporary problem.
Good Job! Anyway i'm still struggling with creating an AI Avatar that similar to my own face...It always look big different although i uploaded my own facial image & try different setting
Yes. It will do this. The AI "generates" - or creates. Make sure the denoising slider is all the way up. In Stable Diffusion, you almost have to train a model with multiple images of your face. In Midjourney, you can use multiple image prompts, which helps.
Is it possible to generate dog images in this AI?
I don't see why not. Give it a try - let me know!
How can i find the seed?
Click on the image. TensorArt is really good at laying out all the details including seed. There's a box with the seed.
Dear Friends, may I ask for advise. I have error message on my workspace and cannot continue to work on images.
Anyone knows how I can contact TENSOR for help to resolve this problem.
Thanks.
Tensor has been a bit glitchy lately. Send them a note on their Discord server or X
Thank you@@MakingPhoto
How can I train the Ai to generate images based on my face for example generating images of the same person like myself, I got photos of me which I wanna generate something using tensor with it, is there a way for me to do it?
Try this video "Create Stunning Images with Stable Diffusion Models" ua-cam.com/video/cV5QNFX_XLA/v-deo.html
@@MakingPhoto ok thanks for the reply but that didn't help, I meant like how can I train Tensor Art Ai, is there is a way?
Tensor Art hosts models. You have to make the model first.
How to get more credits?
New credits every day. Plus, earn credits for various things in your account. Not sure if they're taking payments yet.
I click on the download button but but its not downloading. Its just download an error file. How to fix it❓️
It's probably a glitch. But if this happens, click on the image itself to make it larger. Right click and choose Save Image As...
invitation code not valid
Thanks! I've sent them an email about this.
need more tutorial
I'll work on it!
thanks chicken x
🐣?
image to image not Generating. Showing support jpg /jpeg/png.etc
Hmmm Img2Img was working for me today. Try it again. Maybe there was a glitch.
@@MakingPhoto yeaa its a glitch
I ❤ TensorArt a lot, but today I've found out that I can't login and access my account. Data won't load. And since there is no support, I have to create a new account and start all over again😢
Weird. Maybe there's a glitch. I'm not having any problems today.
Just as I go pro, "Models not Found" comes up. After a. couple of months without problems. Is there any help anywhere? Nothing, not nowhere. So buyer beware
I'm seeing some changes on the site. Maybe they're changing some things around. I'm using it and haven't seen this problem. But I'll keep my eye open.
👋
I find it so sad that I almost always have to write "bare breasts, bare stomach, bare legs" in the negative prompt in order to not get an almost pornographic image.
I tried using systems with models and was disappointed
Not because the images weren't good, but because they all felt the same
The EXACT same tone, same sensibility, same aesthetic. BORING!
This is why I like Gabby/Gab AI, it's more random in the output, so there's lots of variation
This is becoming a thing. The models are good and consistent - but that consistency is getting boring. I'm finding this on Midjourney as well.
am really sad as there is noinpainting here
Good point. No inpainting or outpainting.
It seems like all the images are borderline pornographic images of half naked women! Why are there no images of cute animals or even ONE image of a .... male? 🤔
You only saw two photos it seems
Some creators want to make ladies. But that's not all that's out there!
Tensor Have Alot of NSFW pic. To be honest i tried creating one and boi it did not disappoint. But tensor is great over all just turn on the nsfw filter
@@itsalexbitch9930 Platforms like Tensor Art let you have the freedom to create pretty much anything you want. I like the creativity.
@@MakingPhoto thanks for the A.I tutorials and special this tensor one. Definitely helps alot.
I am starting to think that Tensor art is racist. I put in prompt brown skin, Carmel skin, latin lady and they keep giving me a white girl lol. Do I need to use an add lora that is black already to get a person of color?
I'll check this out! It all depends on how the AI was trained. For a long while Midjourney and other AI did this, too. Try Asian. Some of the major AI companies are based in Asia.
Voice hears like someone is crying
I'm sorry you hate my voice . It's the only one I have. I'm not crying , I'm excited to teach you about tensor art. I'm sure I can find some physical feature about you that is sub-par. Let me call you out on that. Why don't you find some other channel to follow - if you can find content like mine, go find it somewhere else.
Oh damn it! Not only does this say zero helpful info about the ADetails subsection, it also ignores the entire list of hires upscalers! So every bloody beginner is doomed to waste all daily free credits in useless guessing game clicks bc most upscale models won't enhance the image we currently work on. I say this after wasting 2 days with useless upscalers and most of the time only some ultra sharp did any nice result.
Do you need a different video. Go for it! Make the video! I can only do so much in an entry level video.
only works once then unable to get credits total crap scam,,only if you invite people daily! LAME!
I don't know what you're talking about. It's working fine for me.
Thumbs down for such a plain overview that anybody can guess by using tensor website for 1 day. What is missing is the ACTUAL deep info about what nuances the cryptic bs named sub-options do. How many damn videos do we have to watch for all those?!
If you're getting this information elsewhere, than go find it. I've watched a lot of videos on SD and knowing what each tool does is key to getting good results. If you know it all - good for you. But a lot of creators don't.
@MakingPhoto the invite code is expired
Thanks for letting me know.