It should work. I've tested it on anything4.5, foxynsfw and midnight maple and all work. Nsfw works well too. I assume it'll work with orangemix nsfw because I recall midnight maple was a mix of that

4 repeats.

I heard that the repeats should be such that if I had to multiply it by the number of pictures then the value should be between 200 and 400

Thank you :)

I can definitely relate to the nostalgia

Sorry just to clarify, you want a guide on how I trained the LoRA right? If so I can share the process I used, just I'll do it later today most likely

Before I explain specifics I want to say I largely followed the guide by HollowStrawberry (https://civitai.com/models/22530) and used their Google colab for training (https://colab.research.google.com/github/hollowstrawberry/kohya-colab/blob/main/Lora_Trainer.ipynb). I did certain things differently so that's what I'll largely explain even though it may not be the best.

I personally find having an optimal dataset as one of the more important processes, so I honestly spend most of my time on it. I'll explain this step first.

1) Creating Data Set

1.1) Sourcing images

I usually get no less than 70 pictures and occasionally use about 200. I number depends on the quality, variety, character and more so there's no rule I don't think.

I prioritise pictures that clearly show the character but make sure that many of the pictures are varied. So I make sure to have them in different clothing, different facial expressions, different angles, different backgrounds.

My personal experience showed me that including nsfw pictures helps the ai know exactly that their clothes in the pictures can be removed/changed and not "stuck" permanently to the character. Having nude pictures is essentially the same as having the character in the "nude clothing style" but I think it helps because there's a lot of source images to compare the nudes with so it better understands the shape of the character. I could be wrong about this though, I'm no expert 🙃

1.2) Editing images

Training LoRAs is great because you don't have to focus so much on cropping images like you would for a textual inversion because LoRAs use something called bucketing. With bucketing you can use any of your pictures as is.

However because I don't know how bucketing works exactly and I know my images are cropped to a square during training. I manually crop pictures to a perfect square provided the character fits reasonably well in it.

Additionally, the longest part is manually going through each picture and cropping out thing I think the ai may get confused with. I remove people too close or touching the character in case it gets their arms mixed up for example. I remove watermarks, text and signatures using ai removal tools. (I use my phone's built in photo editor since I can remove unwanted objects. It's tedious though)

Rename your first picture as "project (1)" where "project" is whatever you want to call your lora. Then the next pictures will be "project (2)", "project (3)" etc. I use "file tools" an app for android to do it, but windows has the feature built in I think, just search it.

Convert png pictures to jpg because I think some colab don't work with it or something.

1.3) Captioning

Open stable diffusion and load a tagging extension, I use this ( https://github.com/toriato/stable-diffusion-webui-wd14-tagger.git ).

Before captioning all pictures, I load and interrogate one picture with the character with a threshold of 0.25 using wd-v1-4-swinv2-tagger-v2. I then find all tags that are going to be essential to the character in all scenarios. Eg: Gwen has orange hair and green eyes would always be used when generating Gwen so I add those to the list of tags associated with the character.

Then add the list of tags always associated with the character and put them in the box for "tags to be excluded" essentially, we don't want any of our captions to contain any of those tags. Then add a trigger prompt in a box to append tags to your captions, eg: Gwendolyn_Tennyson. That way, all those now missing tags in your captions will be entirely represented by that one trigger tag.

Now you can batch process all captions for your images and get your txt files.

Save your txt files and place them in the folder where you keep your images and they will correspond to the pictures they're named after. This folder is your dataset.

2) Actual Training

Going to say this quickly because this is fairly standard and not too different from HollowStrawberry. Its pretty much the default settings in HollowStrawberry, but here's what I change.

> number of pictures multiplied by repeats should be between 200 and 400. So I usually just use the number 300 as my goal to aim for. For example if I have 75 pictures, then I'd take 300 divided by 75 to give me the needed repeats. In that example it's roughly 4 repeats. HollowStrawberry explains this.

> max_train_epochs = 20, though in many instances an epoch of 10 was actually fine too.

> I usually change my unet_lr to 1e-4 because I have a lot of pictures.

> make sure to turn "flip_aug" off if your character has asymmetrical stuff you want to keep asymmetrical.

> I change network_dim and network_alpha on occasion but usually leave them as 16 and 8

I can't make any promises as of right now, but I'll look into it if I have a chance 👍

@asdzxczzzz no problem and thanks for understanding :)

Hi thanks for commenting :)

I probably won't make an alien LoRA anytime soon because I've got a bit of a backlog. Also my process can be slow, so I tend to pick projects that I personally will benefit from too.

Place your requests here though👌 If I somehow have some time and I find a way to do it reasonably quickly I don't mind creating one. Also someone else may see it and make one for you :)

Don't feel bad about requesting, it's nothing to feel bad about.

Also I'm honestly not sure when or if I'll be able to get to your requests but even if I don't hopefully someone sees this and can do some.

BTW I noticed you mentioned your potato pc, I'm in the same boat 😂... I actually make all my LoRAs from my phone on Google colab. Search up hollowstrawberry on civitai. They're a user who has a full explanation on how to use it

Hi zlingerfinger :)

I figured I should let you know that I'm open for commissions now on ko-fi. I don't like charging but tbh the lora thing is cutting into my IRL work time a bit and the income from commissions helps. I understand if you're not in a position to to pay for commissions though, just wanted to let you know 🙏

Also if you do decide to request a comission I can link you to my more sfw civitai account that also has lora commissions but at reduced prices, since you're requesting sfw

I can consider it. Could you give an example on which cartoon styles you're referring to?

Because Ben 10 almost has a more "Anime" feel than something like Adventure Time would.

Hi thanks for the comment :)

I was also thinking it would be nice to make one of Alien Force Gwen, just haven't had much time since making these LoRAs takes me a while. I've been quite busy with my irl job as of late, so not sure when I'll get down to LoRAs

@userviewer77 thanks so much for the compliments, I hope I'll find time to continue too :)

You mentioned you've been trying to learn so I'm going to just paste a message I sent a few people who asked how I make my LoRAs. You may find it helpful, I figured I'd send it now in case I get even more busy in the coming days. Good luck with it all

It's really long btw 😅

"Before I explain specifics I want to say I largely followed the guide by HollowStrawberry (https://civitai.com/models/22530) and used their Google colab for training (https://colab.research.google.com/github/hollowstrawberry/kohya-colab/blob/main/Lora_Trainer.ipynb). I did certain things differently so that's what I'll largely explain even though it may not be the best.

I personally find having an optimal dataset as one of the more important processes, so I honestly spend most of my time on it. I'll explain this step first.

1) Creating Data Set

1.1) Sourcing images

I usually get no less than 70 pictures and occasionally use about 200. I number depends on the quality, variety, character and more so there's no rule I don't think.

I prioritise pictures that clearly show the character but make sure that many of the pictures are varied. So I make sure to have them in different clothing, different facial expressions, different angles, different backgrounds.

My personal experience showed me that including nsfw pictures helps the ai know exactly that their clothes in the pictures can be removed/changed and not "stuck" permanently to the character. Having nude pictures is essentially the same as having the character in the "nude clothing style" but I think it helps because there's a lot of source images to compare the nudes with so it better understands the shape of the character. I could be wrong about this though, I'm no expert 🙃

1.2) Editing images

Training LoRAs is great because you don't have to focus so much on cropping images like you would for a textual inversion because LoRAs use something called bucketing. With bucketing you can use any of your pictures as is.

However because I don't know how bucketing works exactly and I know my images are cropped to a square during training. I manually crop pictures to a perfect square provided the character fits reasonably well in it.

Additionally, the longest part is manually going through each picture and cropping out thing I think the ai may get confused with. I remove people too close or touching the character in case it gets their arms mixed up for example. I remove watermarks, text and signatures using ai removal tools. (I use my phone's built in photo editor since I can remove unwanted objects. It's tedious though)

Rename your first picture as "project (1)" where "project" is whatever you want to call your lora. Then the next pictures will be "project (2)", "project (3)" etc. I use "file tools" an app for android to do it, but windows has the feature built in I think, just search it.

Convert png pictures to jpg because I think some colab don't work with it or something.

1.3) Captioning

Open stable diffusion and load a tagging extension, I use this ( https://github.com/toriato/stable-diffusion-webui-wd14-tagger.git ).

Before captioning all pictures, I load and interrogate one picture with the character with a threshold of 0.25 using wd-v1-4-swinv2-tagger-v2. I then find all tags that are going to be essential to the character in all scenarios. Eg: Gwen has orange hair and green eyes would always be used when generating Gwen so I add those to the list of tags associated with the character.

Then add the list of tags always associated with the character and put them in the box for "tags to be excluded" essentially, we don't want any of our captions to contain any of those tags. Then add a trigger prompt in a box to append tags to your captions, eg: Gwendolyn_Tennyson. That way, all those now missing tags in your captions will be entirely represented by that one trigger tag.

Now you can batch process all captions for your images and get your txt files.

Save your txt files and place them in the folder where you keep your images and they will correspond to the pictures they're named after. This folder is your dataset.

2) Actual Training

Going to say this quickly because this is fairly standard and not too different from HollowStrawberry. Its pretty much the default settings in HollowStrawberry, but here's what I change.

> number of pictures multiplied by repeats should be between 200 and 400. So I usually just use the number 300 as my goal to aim for. For example if I have 75 pictures, then I'd take 300 divided by 75 to give me the needed repeats. In that example it's roughly 4 repeats. HollowStrawberry explains this.

> max_train_epochs = 20, though in many instances an epoch of 10 was actually fine too.

> I usually change my unet_lr to 1e-4 because I have a lot of pictures.

> make sure to turn "flip_aug" off if your character has asymmetrical stuff you want to keep asymmetrical.

> I change network_dim and network_alpha on occasion but usually leave them as 16 and 8

"

This is very true in hindsight 😂

I've made a couple realistic LoRAs and they all turned out well for me and I followed a similar process as I did with my anime LoRAs. I don't have the LoRA training scripts but I'll paste a long guide I sent one other person a while back... Though be aware that recently I've realised that different ways of training may be best depending on what character/person you're training (I could be wrong)

Before I explain specifics I want to say I largely followed the guide by HollowStrawberry (https://civitai.com/models/22530) and used their Google colab for training (https://colab.research.google.com/github/hollowstrawberry/kohya-colab/blob/main/Lora_Trainer.ipynb). I did certain things differently so that's what I'll largely explain even though it may not be the best.

I personally find having an optimal dataset as one of the more important processes, so I honestly spend most of my time on it. I'll explain this step first.

1) Creating Data Set

1.1) Sourcing images

I usually get no less than 70 pictures and occasionally use about 200. I number depends on the quality, variety, character and more so there's no rule I don't think.

I prioritise pictures that clearly show the character but make sure that many of the pictures are varied. So I make sure to have them in different clothing, different facial expressions, different angles, different backgrounds.

My personal experience showed me that including nsfw pictures helps the ai know exactly that their clothes in the pictures can be removed/changed and not "stuck" permanently to the character. Having nude pictures is essentially the same as having the character in the "nude clothing style" but I think it helps because there's a lot of source images to compare the nudes with so it better understands the shape of the character. I could be wrong about this though, I'm no expert 🙃

1.2) Editing images

Training LoRAs is great because you don't have to focus so much on cropping images like you would for a textual inversion because LoRAs use something called bucketing. With bucketing you can use any of your pictures as is.

However because I don't know how bucketing works exactly and I know my images are cropped to a square during training. I manually crop pictures to a perfect square provided the character fits reasonably well in it.

Additionally, the longest part is manually going through each picture and cropping out thing I think the ai may get confused with. I remove people too close or touching the character in case it gets their arms mixed up for example. I remove watermarks, text and signatures using ai removal tools. (I use my phone's built in photo editor since I can remove unwanted objects. It's tedious though)

Rename your first picture as "project (1)" where "project" is whatever you want to call your lora. Then the next pictures will be "project (2)", "project (3)" etc. I use "file tools" an app for android to do it, but windows has the feature built in I think, just search it.

Convert png pictures to jpg because I think some colab don't work with it or something.

1.3) Captioning

Open stable diffusion and load a tagging extension, I use this ( https://github.com/toriato/stable-diffusion-webui-wd14-tagger.git ).

Before captioning all pictures, I load and interrogate one picture with the character with a threshold of 0.25 using wd-v1-4-swinv2-tagger-v2. I then find all tags that are going to be essential to the character in all scenarios. Eg: Gwen has orange hair and green eyes would always be used when generating Gwen so I add those to the list of tags associated with the character.

Then add the list of tags always associated with the character and put them in the box for "tags to be excluded" essentially, we don't want any of our captions to contain any of those tags. Then add a trigger prompt in a box to append tags to your captions, eg: Gwendolyn_Tennyson. That way, all those now missing tags in your captions will be entirely represented by that one trigger tag.

Now you can batch process all captions for your images and get your txt files.

Save your txt files and place them in the folder where you keep your images and they will correspond to the pictures they're named after. This folder is your dataset.

2) Actual Training

Going to say this quickly because this is fairly standard and not too different from HollowStrawberry. Its pretty much the default settings in HollowStrawberry, but here's what I change.

> number of pictures multiplied by repeats should be between 200 and 400. So I usually just use the number 300 as my goal to aim for. For example if I have 75 pictures, then I'd take 300 divided by 75 to give me the needed repeats. In that example it's roughly 4 repeats. HollowStrawberry explains this.

> max_train_epochs = 20, though in many instances an epoch of 10 was actually fine too.

> I usually change my unet_lr to 1e-4 because I have a lot of pictures.

> make sure to turn "flip_aug" off if your character has asymmetrical stuff you want to keep asymmetrical.

> I change network_dim and network_alpha on occasion but usually leave them as 16 and 8

Go nuts😂