Aitrepreneur
patreon
OBJECT LORA Training Presets Kohya
🕑 Added 2023-10-30 16:05:52 +0000 UTCHere you will find 5 optimized training presets for OBJECT LORA inside Kohya SS. Whether you're training locally or using Runpod.
What's inside?
3 Local Training Presets: Designed for those who love to train on their own systems.
- Normal Object Training: For your everyday needs.
- Fast Object Training: For those in a hurry
- LowVRAM: Optimized for local systems with limited VRAM.
2 Runpod Presets: Perfect for those using Runpod's GPU renting platform. These presets automatically locate the location of the stable diffusion XL model for you.
- Normal Object Training: The classic training preset.
- Fast Object Training: Train your models as fast as possible
Why These Presets?
For LORA training, every little optimization can make a huge difference. These presets are the culmination of countless hours of experimentation and refinement. By using them, you're ensuring that your LORA OBJECT training is both efficient and effective, regardless of where or what you're training.
And as always, supporting me on Patreon allows me to keep creating helpful resources like this for the AI art community. Thank you for your support - now go train some awesome LORAs!
👉Download Object Presets👈
password: OBJECTPRESETS
Comments
Kagemaboy
Can we get updated presets for Lora Training in Kohya? Specifically for Illustrious models
Aitrepreneur
You probably input the wrong path for the training images, the path to the images should point to the parent C:/training/images not into the folder itself D:/training/images/20_xxx
Shaun Bass
Hey! I am following your Object training tutorial, and i believe that I followed every step exactly as you explained however when i click "Start Training" i get this text in the terminal ... " INFO [Dataset 0] config_util.py:550 INFO loading image sizes. train_util.py:794 0it [00:00, ?it/s] INFO make buckets train_util.py:800 WARNING min_bucket_reso and max_bucket_reso are ignored if bucket_no_upscale is train_util.py:817 set, because bucket reso is defined by image size automatically / bucket_no_upscaleが指定された場合は、bucketの解像度は画像サイズから自動計 算されるため、min_bucket_resoとmax_bucket_resoは無視されます INFO number of images (including repeats) / train_util.py:846 各bucketの画像枚数（繰り返し回数を含む） C:\Users\shaun\AppData\Local\Programs\Python\Python310\lib\site-packages\numpy\core\fromnumeric.py:3504: RuntimeWarning: Mean of empty slice. return _methods._mean(a, axis=axis, dtype=dtype, C:\Users\shaun\AppData\Local\Programs\Python\Python310\lib\site-packages\numpy\core\_methods.py:129: RuntimeWarning: invalid value encountered in scalar divide ret = ret.dtype.type(ret / rcount) INFO mean ar error (without repeats): nan train_util.py:856 ERROR No data found. Please verify arguments (train_data_dir must be the train_network.py:213 parent of folders with images) / " does this make any sense to you, are you able to explain where i am going wrong? Thank you so much for your help.
Aitrepreneur
watch my lora training videos, you use them when selecting the training preset files inside kohya
Carlos.exe
Just joined and diving into everything prettty fast.. where do we put these files? I'm just trying to set up everything then dive into it lol.
Aitrepreneur
you can't use more Vram and make the training faster except indeed increasing batch size but if you don't have a lot of images (like at least 30+) then increasing that number will make the final model worse, just use my presets, they are optimized for the best results for max and low vram
Menthu Rae
Hey mate - thank you for being so kind with your responses in other threads. I have an RTX A6000 with 48GB VRAM and using your preset I am only seeing 19.6GB VRAM in use. What parameter(s) should I change to get the most out of my VRAM but retain good training results? I did some searching and people said while batch size should increase VRAM use, it could result in worse results depending on the dataset. Cheers mate.
Aitrepreneur
for the 1.5 the resolution is 512x512, the presets might work, you need to try them yourself
Demitri Grigori
Do these presets work for version 1.5 models by adjusting the resolution to 768x768, or would something else need to be adjusted?
