1-Click INSTALL Ai-TOOLKIT WAN LORA TRAINING!
Added 2025-07-20 16:50:43 +0000 UTCHey everyone! I've created a 1-click installer for the “ULTIMATE WAN LORA TRAINING AI‑TOOLKIT WEBUI.” Hands‑down the BEST AI model training I have ever made! VIDEOS & IMAGES!!! Just fricking insane! This really is the best Ai model currently available and Ai-Toolkit makes training a Lora super ultra easy as well, basically the ultimate combination, perfect for complete beginners! With it you can train any LoRA you want for WAN: people, styles, and even full video LoRAs… it’s just nuts!
You can check out the video right here: https://youtu.be/lRg5sPBXTZE
The installer of course automates the entire install process!
LINK TO THE WORKFLOW + LORAS: https://www.patreon.com/posts/134551107
LOCAL INSTALL!!! RECOMMENDED TO HAVE AT LEAST 24GB OF VRAM FOR THE WAN TEXT TO IMAGE LORA TRAINING! if it's an issue, definitely use the Runpod installer instead as recommended in the video!
Local install:
Requisites:
– CUDA 12.8 (RTX-50) OR CUDA 12.6 (RTX-40/30/20/GTX)
– Git, Python 3.10 64-bit, Node 18/20 LTS
1. Download the AI-TOOLKIT_AUTO_INSTALL.bat and place it in a folder WITHOUT any spaces in the path!!!
2. Run the bat file and select either option 1 or 2 depending on your GPU
3. Wait for the install to finish
4. ???
5. Profit 😎
Use then one of the launcher once you have closed the ui and you want to restart it.
RUNPOD:
1. Create an account if you haven't already: Runpod
2. Click on Pod (on the left side) then click deploy
3. Choose a 5090 GPU (if you intend on training a text to image lora OR a text to video lora with text encoder unloaded and 512 resolutions) otherwise choose the 6000 pro blackwell GPU, and chose the pytorch 2.8.0 then edit the template and choose 100GB for both the container and volume disk, then under expose ports, edit:
8888 into 8888, 8675. Then go into environment variables below, click add and then type:
AI_TOOLKIT_AUTH on the left then type a password on the right
then scroll then and deploy.
4. Go to my pods, wait for everything to finish and then click "connect", then "Connect to HTTP SERVICE port 8888"
5. Then drag and drop the AI-TOOLKIT_AUTO_INSTALL-RUNPOD.sh file on the left side of the UI then click on the "Terminal" icon on the right side on the UI
6. Copy and paste these two lines then press enter:
chmod +x AI-TOOLKIT_AUTO_INSTALL-RUNPOD.sh
./AI-TOOLKIT_AUTO_INSTALL-RUNPOD.sh
7. Wait for everything to be installed
8. Go back to the port 8675 and click on it to launch the webui then enter your password you created previously
IF YOU WANT TO RESTART THE UI AFTER STOPPING AND RESTARTING THE POD
Use the same
chmod +x AI-TOOLKIT_AUTO_INSTALL-RUNPOD.sh
./AI-TOOLKIT_AUTO_INSTALL-RUNPOD.sh
command as you did for the install
Chatgpt prompt:
I will upload a few images, wait until I finished uploading them all.
These are photos of Margot Robbie, analyze those images and caption them correctly for a Lora training using "Margot Robbie" as the caption token and convert each single caption into a .txt file and name them all in the same order, Margot Robbie_0001.txt, Margot Robbie_0002.txt, etc... and then finally zip them up all into a single zip file I can download
portrait of margot robbie, clean studio headshot, 85 mm lens, soft beauty lighting, photorealistic
margot robbie, film‑noir portrait, dramatic hard key light, deep shadows, black‑and‑white photo
oil painting portrait of margot robbie in the style of John Singer Sargent, loose brushwork, warm palette
margot robbie cyberpunk portrait, neon rim‑lights, rain‑soaked city backdrop, vivid colors, photoreal
1940s Hollywood glamour portrait of margot robbie, sepia tone, classic butterfly lighting, vintage lens flare
pop‑art portrait of margot robbie, bold halftone dots, saturated primary colors, Andy Warhol style
ukiyo‑e woodblock portrait of margot robbie, flat colors, delicate linework, Edo‑period aesthetic
renaissance oil portrait of margot robbie, chiaroscuro lighting, rich velvet dress, baroque background
futuristic holographic portrait of margot robbie, volumetric light, translucent interface elements, sci‑fi realism
candid Polaroid portrait of margot robbie, pastel color shift, soft film grain, natural window light
As always, supporting me on Patreon allows me to keep creating helpful resources like this for the community. Thank you for your support - now go have some fun😉!
Comments
CUDA error: misaligned address CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
nurtu
2025-10-24 00:05:10 +0000 UTCI'm trying to use this on runpod with an RTX pro 6000. i get this error. "Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu!"
Matthew Smallwood
2025-10-09 11:43:35 +0000 UTC"[ERROR] curl not found ΓÇô install Windows 10 v1803+ or add it to PATH." This pops up immediately after choosing which RTX-series card (on the installer). I definitely have curl.exe on my computer, and it sure seems to be added to PATH. ETA> After over 2 years of obsessing over this stuff, today I learned that I have been doing the PATH thing wrong. Having corrected that (those, actually), it installed as expected.
totally_not_a_weirdo
2025-09-20 00:49:37 +0000 UTCwill i have better results if i train the model using videos? or does it not matter?
Kent Umeki
2025-08-26 18:59:30 +0000 UTCgetting this error on runpod CUDA out of memory. Tried to allocate 68.00 MiB. GPU 0 has a total capacity of 31.37 GiB of which 58.69 MiB is free. Including non-PyTorch memory, this process has 31.30 GiB memory in use. Of the allocated memory 30.57 GiB is allocated by PyTorch, and 144.35 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
D Dunn
2025-08-22 19:29:35 +0000 UTCI want to train NSFW but for the descriptions, GPT chat doesn't give them to me.
SAMIRA
2025-08-20 20:32:47 +0000 UTCthis error actually happens because you don't have enough space on the pod
Aitrepreneur
2025-08-14 23:05:12 +0000 UTCI did this last night and trained a Lora for Lake Bell and it turned out amazing! Thank you! However, today, I'm receiving an error when I try to train a new Lora. The error message is: "An error occurred while downloading using `hf_transfer`. Consider disabling HF_HUB_ENABLE_HF_TRANSFER for better error handling." I've tried multiple ways to disable this including setting this variable to '0' in the terminal as well as with a python script and can't seem to get it to download the files without erroring out. I even nuked my pod and created a brand new one and am getting the same error. Any idea for how to disable this so I can continue my training?
Patrick Brown
2025-08-14 21:35:35 +0000 UTCno but check that you are running the correct model, keep in mind that if you use other loras than lightx2v it might change the final result. You can also increase the weight of the lora as well
Aitrepreneur
2025-08-12 18:39:44 +0000 UTCSo, while the training seems to go great, the samples become good quite quickly, when I try to use the lora in the Wan t2i workflows, the result looks nothing like the samples - the effects are greatly diminished and often barely visible. For character loras it works fine, but for more complex concept loras - I keep running into this problem. Like a tried a battle-scarred-look lora and it worked great in the training samples, but in the workflows, the effects were much much weaker. I tried fiddling with samplers, cfg, etc, but still no luck. And the training itself seems to be working, as the samples do look good, but it just doesn't work in-workflow later. Do I need to train on a different model than just the base one or something?
DFined
2025-08-12 08:22:41 +0000 UTCas soon as it's updated
Aitrepreneur
2025-08-12 00:11:01 +0000 UTCAre we going to be able to have wan 2.2 config on Aitoolkit locally here soon?
mikki
2025-08-11 23:24:32 +0000 UTC
