Furkan Gözükara patreon

The Very Best Kohya GUI Workflow & Config For SD 1.5 Based Models DreamBooth / Full Fine Tuning

🕑 Added 2024-08-11 22:58:00 +0000 UTC

Comments

Furkan Gözükara

you are welcome. after training you can use it on your pc just same as using regular FLUX with SwarmUI - it should work fine but slow :D you can also use massed compute really fast - 31 cents per hour

Marc

Excellent. Thank you!

Furkan Gözükara

hello for 8 gb cards you dont have much options. i dont think it can do fine tuning but only a lora perhaps. but i dont have a good lora option either. try tier_2_lowVRAM_10_2GB.json and see if works. what i would suggest you is train a flux model on cloud and use it. it learns body 100% with sufficient images https://youtu.be/-uhL2nW7Ddw https://medium.com/@furkangozukara/trained-myself-with-256-images-on-flux-results-mind-blowing-80e45aad4162

Marc

Hi, I would like to know which “tier” I should choose for the Onetrainer configuration. I have a pc 16gb Ram, Nvidia 3060 RTX Ti 8gb, A1111 SD 1.5 checkpoints, I work to obtain an embedding for consistent character face. My second question is: Is there a way to train embedding or lora on a whole body, in order to obtain a consistent character at the level of the whole body? Thank you in advance and also for your kindness and excellent work, which is a great help.

Furkan Gözükara

yes massed compute. 31 cents per hour. it has installed kohya as well : https://youtu.be/0t5l6CP9eBg

saint-mares gérard

In your opinion, is there a good way to train online today ? I used colab but all kohya scripts are outdated

Furkan Gözükara

what help you need? i can give you private lecture

min min

amazing work!can you touch us how to make the reg datasets?what's the thought!

Profile Photo

okay thank you

Furkan Gözükara

please watch this tutorial : https://youtu.be/16-b1AjvyBE you did set epochs 8 and save every n epoch 4

Profile Photo

https://drive.google.com/file/d/1PPks1aMG3US6bZWmp1HurKPd6enqfp_c/view?usp=drive_link This is a link to my configuration, could you check it too ?

Profile Photo

Yes I'm using, but my question about where do I input number of epoch, is it epoch parameter or something else ?

Furkan Gözükara

that means it is using shared vram. you must be doing something inaccurate. are you using regularization images?

Profile Photo

What number of epochs did you input? 150 epochs ? Because I chose 150 epochs with 15 photos and I have 72 hours of execution which is too much

Furkan Gözükara

Save as half precision. Like bf16 or fp16

Dmitrii

hyperrealism is 2GB size, but my trained output model is 4GB. Why is that? how to make output 2GB as well?

Furkan Gözükara

actually it downloads it is auto set. i uploaded it here as diffusers : https://huggingface.co/MonsterMMORPG/sd15_best_realism

Ella Bang

Says that the configs include the models hyper realism v3 to be downloaded from huggingface but it seems to be missing.

Furkan Gözükara

you are welcome

saint-mares gérard

nice thanks

Furkan Gözükara

hello just fixed the folder naming issue. also you are loading as dreambooth tab right? xformers not selected i verified

saint-mares gérard

maybe i'm wrong but when i'm loading tier1 config it uses xformers and model output name is completed by tier2

Furkan Gözükara

you guys are right. because of that I prepared this article (1 week ago). please check it out and let me know if anything is still not clear : https://www.patreon.com/posts/full-workflow-sd-98620163

Furkan Gözükara

you guys are right. because of that I prepared this article (1 week ago). please check it out and let me know if anything is still not clear : https://www.patreon.com/posts/full-workflow-sd-98620163

Furkan Gözükara

you guys are right. because of that I prepared this article (1 week ago). please check it out and let me know if anything is still not clear : https://www.patreon.com/posts/full-workflow-sd-98620163

Julius Seizure

yeah it honestly is confusing with the amount of videos there are

pravin khetan

there are so many video tutorials on your yourtube channel. Which one should I follow to use your .json file SD 1.5 lora. And how to get best output form automatic1111?

Furkan Gözükara

i would compare this. use captions vs non caption only ohwx man or woman. and see which one performs better

Tom Preece

Do captions make a difference with this setup? I already have a well captioned dataset which gives me good results with LoRA training but I want to see if Dreambooth Fine tuning works better. Should I include my captions or leave them out?

Furkan Gözükara

sadly not yet. therefore i suggest you to use Kaggle notebook. it is free 30 hours training every week. sd 1.5 works best settings there but of course runpod would be easiest and for SDXL runpod is much better. but for SD 1.5 kaggle works same quality as runpod https://www.patreon.com/posts/kohya-sdxl-lora-88397937

Patrick Major

Sorry, I just re-read the title of this wiki. It is for Dreambooth training and not LoRA training. I'm switching over to loading the config in dreambooth tab and trying again, sorry for the trouble. Even in the lowmem dreambooth, I can't get below 10 GB, which means a 17 hour train. I guess runpod will be the option. Did you ever find that you could get the Dreambooth training in your low mem config as low as 8 GB (or slightly lower?)

Furkan Gözükara

this is Kohya config and I just loaded and all set to 7e-7 accurately . this is not LoRA this is DreamBooth that is your mistake.

Patrick Major

The "tier_2_lowVRAM_10_2GB.json" is the config I grabbed from this page in relation to the SD 1.5 Koyha best settings training. I am training against SD1.5 as I don't have enough memory to train SDXL locally yet. If you do a quick check on the "tier_2_lowVRAM_10_2GB.json" file you'll notice the LR is "7e-07" and the Unet and TE are 0. I'm not sure if its intentional, but just raising the question as I suspect the values might be set too low. It's not the SDXL config, but the SD1.5 one. Specifically this setting from the json above is what I'm inquiring about: "learning_rate": 7e-07 Am I correct in that I load your above config into the LoRA tab of Koha_ss, or this intended for the fine tuning tab?

Furkan Gözükara

i dont have 7e-7 for SDXL config. you are confusing SDXL and SD 1.5 configs

Patrick Major

I hate to bug you again. I've tried all sorts of ways to get your "tier_2_speed_11_8GB.json" file to work, with the settings recommended. I've already got some good success with other methods, so took a decent dataset where the reality of likeness is around 90% and I used it with the settings. I am training on an Irish woman, the folder path repeats and token being "1_owsh woman" and I'm doing 150 epochs. There are 21 decent quality images. It should way over train with that number. Instead what is happening is during the training, in the sampling (with a prompt of "photo of a owsh woman") I sample every 50 steps and see the system showing me a sea wall, a harbour, then a car and eventually settles onto an older asian woman which has no likeness. I can confirm the folder paths are good, as I can change the LR from 7e-07 to 1e-05 and start to see a decent likeness. I'm using the custom model you recommend in the settings and more importantly the learning rate which you have set to 7e-07. I've tried a variety of image sets that I had prepared, no captions, just the folder tags. To fix it, I just changed the LR to 1e-05 for example. I am raising this more because I think its possible the base example has the LR value set wrong and was going to ask you to look at it as I suspect if you do a training at the exact base settings from the json (with of course modifications for the repeats/epoch combination you choose), you'll find it way undertrains. Unless 7e-7 is for very large datasets? Or do I have to set epochs towards 1000? Something is wrong with the config. It occurs to me now to ask. Maybe I'm using the configs in the wrong location. I'm loading them in kohya_ss in the LoRA tab, are these configs for fine tuning or other? I noticed some learning rate differences between the: tier_2_speed_11_8GB.json and tier_2_lowVRAM_10_2GB.json file, namely in the lowVRAM there are additional training settings that are missing from the tier_2_speed_11_8GB.json such as : "learning_rate_te": 7e-07, "learning_rate_te1": 3e-06, "learning_rate_te2": 0.0, I know the setup us using Adafactor as the optimizer, but the LR Scheduler is setup as "constant" instead of allowing the Adafactor to optimize the LR. Is this intentional? I really liked the results you have in your 6 pages of showing what you can do and am doing my earnest to make your config work. Thanks for any further advice. I'll try a number of settings and report back what I find.

Furkan Gözükara

thank you so much. if you dont use reg images yes make repeating 1 and use 150 epoch better

Furkan Gözükara

accurate

Furkan Gözükara

true it is 150_ohwx man. there is no _ between ohwx and man. so we are training with ohwx rare token and man class token

Patrick Major

if you were doing the prompt you are mentioning, in Kohya_ss it would be 150_ohwx man Notice the space after ohwx. The had confused me in the beginning. The images folder in kohya points to something like this: \YourAITrainingDirectory\Subject\Images But the actual images of your topic live one folder deeper: \YourAITrainingDirectory\Subject\Images\150_ohwx man

Khoa Vo

What do you use as the prompt for the dataset? Is it something like 150_ohwx_man?

Patrick Major

Thank you again for the clarification. So in your first example to my above question with 15 images and 150 epoch, that would be a good setup for a run without reg images. Are you setting the epoch high in that scenario to make it easier to setup more checkpoint saves? Even with some of my settings off, your base setup is starting to work for me. It's been an obsession of mine for months with some hit and miss results. Your work here is wonderful.

Furkan Gözükara

well epoch doesnt mean much since kohya utilized repeating logic. so as you increase repeat you are actually increasing epochs in reality. more repeating better since it uses more reg images. so 8 epoch 40 repeat vs 1 epoch 320 repeat, 320 repeat better

Patrick Major

Thank you for that advice, I'll go give it a try. I had noticed the epoch was set to the value 1 in the config file and had thought maybe it was one of those setups where the highest quality was running at 1 epoch and setting the steps high by way of repeats (and then just saving every 500 steps). I'll go give that a try.

Furkan Gözükara

actually i trained my model with only 15 images and 150 epoch for this config. so when reg images included it was total 15 * 150 * 2 = 4500 steps. so with 24 images 75 repeating and 1 epoch might not be sufficient. try 150 repeat and it may even become overtrained

Patrick Major

I have a question in regards to the "tier_2_speed_11_8GB.json" file. The default LR is 7e-7. I have 24 decent images, no captions and using an img folder name of "75_oshw woman". I also have your reg set for the woman 768x768 being used. When I train, the sample images never ever converge into a likeness of what I'm training on (like normally it would). I think the training rate is way too low, or in your config are you doing something closer to 15,000 repeats? Normally the LR is something between 1e4 and 1e5 so I just want to make sure. I'm running still in kohya as I've only got a 3070 8GB and none of the OneTrainer settings gets me below 10GB.

Furkan Gözükara

Ah you say Kaggle. SD 1.5 is almost same as in Kaggle but not SDXL. SDXL requires BF16 for better performance. or FP32

Furkan Gözükara

yes 100%. whatever we have can be used on RunPod

Art

apart from the EMA stuff, this is replicable on Kaggle Kohya notebook to train loras?

Furkan Gözükara

I tested it and I didnt get better results. I think it depends on other parameters too. So if you have a pre-shared config you can compare

Steve Bruno

Any thoughts on Prodigy over Adafactor? Been reading a lot more stuff from Civitai creators none of the YT "influencers" are pushing it.

Furkan Gözükara

probably you didnt set training images folder paths accurately

AIrtistry Contact

something wrong with the first preset tier_1_quality_slow_22GB.json, took me 2 minutes to finish with an rtx 4090 24 gb from runpod and also when I loaded the config, in the training parameters name it's was the name of the third .json

Furkan Gözükara

I used what is hosted on CivitAI which is FP16. I think it uses 840000 VAE as embedded but not sure

Arcon Septim

when you used hyper realism, what checkpoint you used for training? fp16 or fp32? also what vae it uses?

Furkan Gözükara

Well i prefer to use all same resolution but of course you can try bucketing too. i would compare though bucket vs no bucket

Wktra

no buckets??? 🤯

Furkan Gözükara

hello here you can see i have shown : https://youtu.be/EEV8RPohsbw and if you want learn more about kohya here https://youtu.be/sBFGitIvD2A

Nikoles Frances

Noob question how do I run the scripts?

Furkan Gözükara

please watch this tutorial gui settings part. where I set training dataset folders : https://youtu.be/16-b1AjvyBE - 13:23 Which parameters you need to set on Kohya GUI before starting training

Dogmaster

Thanks! a further question, im really new at kohya, where do I set the repeats? For reference ive got 191 images

Furkan Gözükara

I suggest to crop all images into 768x768 for SD 1.5 including reg images. if you are using our reg images we already have that resolution

Dogmaster

Question, are the reg images/traning images aspect ratio important/considered on this config? Should I crop my input images to be square or leave them in any resolution, thanks!