I just rendered my first fp16 video on wan 2.1 and quality is great (far better than fp8) but last 1
I just rendered my first fp16 video on wan 2.1 and quality is great (far better than fp8) but last 1,5h to generate it. I need the Teacache update
Fi think generated
Fyes i keep asking this to authors
Fbut nothing yet
Fi made ram optimization myself but i may try cache myself too 
T
Tiam stumped
Tflux doesn't work with forge and swarmUI gives me this
Fuse our unified downloader to download
Fand install
Fwhat is your gpu?
T4070
T572.60 driver
Fit should work perfect
Fuse our unified downloader
FPatreon
Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon
XI’m finally generating a grid for the first time ever ahahha after many downloads of swarm and training
Ti already used it man
Tcan i not use flux dev 1 ?
F@Tria ok then do this. make a fresh install and download and record screen how you are doing so i can comment. or upgrade gold tier and i can connect your pc and fix.
XGenerating a grid just previews which .safetensors file of my model fits my liking best, right? No new files or checkpoints are created?
ORunning WAN on 3090 - 25 steps - 368x208
Yesterday I was getting 17-25 s/iteration. So about 3-5 mins to create.
Windows update happened last night, I am now getting 100-185s /iteration
Anyone seeing the same?
Yesterday I was getting 17-25 s/iteration. So about 3-5 mins to create.
Windows update happened last night, I am now getting 100-185s /iteration
Anyone seeing the same?
Fcheck how much ram and vram you are using
Fbefore starting
Fbut doesnt make sense
Nnot sure the update to the WAN app is working here to address the long LoRA loading time... I've got a 3090 and 64GB RAM, still seeing a really long wait at "Adding LoRA to wan_video_dit (..." step. how long should this take? is there anything i can check\change to improve?
Neven though i have 24GB VRAM i tried the 16GB profile but that did not seem to help
VHuge THANK YOU, got swarm working for the first time on my 5080. All my past attempts failed and I was only able to get comfy ui working.
NI also tried changing torch type to torch.float8_e4m3fn and then it makes it past the 'Adding LoRA' step, but then after saving the auto pre processed image I get the error: RuntimeError: getCudnnDataTypeFromScalarType() not supported for Float8_e4m3fn
Bhttps://www.patreon.com/posts/click-to-open-post-used-in-tutorial-110879657
@Dr. Furkan Gözükara i am implementing your traininig flow for person training can i get your training folder of the images your data set ? that also a text file as i can see in your vedio
@Dr. Furkan Gözükara i am implementing your traininig flow for person training can i get your training folder of the images your data set ? that also a text file as i can see in your vedio
Patreon
Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon
B@Dr. Furkan Gözükara i guess that is a file which have the config or the captioning of the images or somethin would be please to know if you can give me your 15 image dataset folder with the file
BI just dropped a new model, the last piece of the puzzle for making Very Special, inspired by Saturday Morning Cartoons from the 80s and 90s. Enjoy! https://everlyheights.tv/tools/stable-diffusion-models/lora/everly-heights-saturday-morning-flux/
Dhey @dxqbYD how do you do negative prompting with flux? I thought you can't do it
DI also saw your other pictures where you said you can get better outputs with different prompting.. how are you doing that?
Fare you making complete fresh install
Fbecause i tested fp8 works both on rtx 3090 ti and 5090
Fdid you also follow requirements video;?
Four example dataset is here
FFFif i send you a py file
Fcan you replace it?
Ffor debuggng purposes of speed
Fok i will pm you so we can see what is taking time
Ni verified i have cuda 12.8 and have the latest working kohya and swarm. fp16 image2vid works, just takes a long time at the LoRA step.
B@Dr. Furkan Gözükara can you tell me something one this , like how he is able to do it with negative prompting
Fye for debugging reason
Fuse de distilled model
XERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/convert-caffe2-to-onnx
ERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/convert-onnx-to-caffe2
ERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/torchfrtrace
ERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/torchrun
ERROR: Could not install packages due to an OSError: [Errno 122] Disk quota exceeded
ERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/convert-onnx-to-caffe2
ERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/torchfrtrace
ERROR: Failed to restore /workspace/supir files/SUPIR/venv/bin/torchrun
ERROR: Could not install packages due to an OSError: [Errno 122] Disk quota exceeded
