Software Engineering Courses (SECourses)•10mo ago

Is it possible to train with videos on wan2.1 with 16gb VRAM? Are you going to make a tutorial on it

Is it possible to train with videos on wan2.1 with 16gb VRAM? Are you going to make a tutorial on it?

Yyi Is it possible to train with videos on wan2.1 with 16gb VRAM? Are you going to m...

Furkan Gözükara SECourses•3/24/25, 1:29 PM

yes

Furkan Gözükara SECourses•3/24/25, 1:29 PM

hopefully it is coming. i am finishing the app features

ChillPanda•3/24/25, 5:37 PM

Could someone help me with SDXL lora training?

hello123•3/24/25, 10:40 PM

https://youtu.be/sPgUtZKnWMY?si=6srIjffUSvv24VCY

Inoshi•3/25/25, 8:00 AM

https://github.com/Wan-Video/Wan2.1/issues/264#issuecomment-2747490626

GitHub

Start & end frames integration · Issue #264 · Wan-Video/Wan2.1

Hi do you think that can be implemented? https://github.com/raindrop313/ComfyUI-WanVideoStartEndFrames Thanks

IInoshi https://github.com/Wan-Video/Wan2.1/issues/264#issuecomment-2747490626

Furkan Gözükara SECourses•3/25/25, 10:34 AM

ye i shared this on reddit

uisato•3/25/25, 3:52 PM

https://www.instagram.com/p/DGlVKAJxjOo

uisato_

Found it.

Technique consisting in a new synthetically trained AI model [FLUX.D LORA], and some ComfyUI wizardry, with the objective of accurately reproducing a 'found footage/liminal' aesthetic.

Both music and visuals by myself.

You can access these [new FLUX.dev LORA] + ComfyUI workflow + 2350 images and prompts in metadata + 26 img-to-vid e...

Likes

1265

Instagram

Inoshi•3/25/25, 11:24 PM

https://x.com/bdsqlsz/status/1904477594545721815?s=33

青龍聖者 (@bdsqlsz) on X

Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models

a novel framework for direct ultra-high-resolution image synthesis using text-to-image diffusion models

Open Source Aesthetic-4K Benchmark datasets
#CVPR2025

•

3/25/25, 10:16 AM

IInoshi https://x.com/bdsqlsz/status/1904477594545721815?s=33

Furkan Gözükara SECourses•3/26/25, 12:15 AM

it has transformer let me test

Furkan Gözükara SECourses•3/26/25, 12:15 AM

will add to our downloader

Furkan Gözükara SECourses•3/26/25, 12:17 AM

https://github.com/zhang0jhon/diffusion-4k/issues/3

GitHub

Can you please publish a unified file so that we can directly use i...

The model size matching to FLUX DEV but i am not sure if you trained text encoders and VAE as well https://huggingface.co/zhang0jhon/flux_wavelet/tree/main/transformer

Hiramos_•3/26/25, 4:23 PM

Guys, what are your Wan2.1 img-to-vid 480 14B generation times like after the latest updates? On my 4060 Ti 16GB, I'm getting 22/50 [26:20<33:55, 72.71s/it]. Before, each generation used to take around 1 hour and 30 minutes.

cilium•3/26/25, 7:04 PM

1 hour and 30 minutes for 5 seconds ?

HHiramos_Guys, what are your Wan2.1 img-to-vid 480 14B generation times like after the la...

cilium•3/26/25, 7:05 PM

ad n ow its taking 1 minute or what ?

cilium•3/26/25, 8:02 PM

Is OneTrainer better than Flux Gym?

HHiramos_Guys, what are your Wan2.1 img-to-vid 480 14B generation times like after the la...

Furkan Gözükara SECourses•3/26/25, 10:34 PM

nice so it is way faster now

Furkan Gözükara SECourses•3/26/25, 10:34 PM

hopefully will be even faster with GGUF once arrived

Ccilium Is OneTrainer better than Flux Gym?

Furkan Gözükara SECourses•3/26/25, 10:34 PM

probably

Ccilium 1 hour and 30 minutes for 5 seconds ?

Hiramos_•3/26/25, 11:32 PM

Yes,

Ccilium ad n ow its taking 1 minute or what ? 😄

Hiramos_•3/26/25, 11:33 PM

hahaha now it's taking around 1h

Ccilium Is OneTrainer better than Flux Gym?

Weatherby•3/27/25, 12:27 PM

Flux gym and kohya allow training from fine tuned models, ai-toolkit and maybe onetrainer require raw dev from huggingface...i believe. I haven't looked in awhile. I'm still using ai toolkit on cloud, but I also run q8 so don't really use a finetune.

cilium•3/27/25, 12:53 PM

What is the easiest way to train a LoRA locally / and not locally?

Ccilium What is the easiest way to train a LoRA locally / and not locally? 😄

Furkan Gözükara SECourses•3/27/25, 12:53 PM

kohya

Furkan Gözükara SECourses•3/27/25, 12:53 PM

with our config

Furkan Gözükara SECourses•3/27/25, 12:53 PM

what gpu you have?

cilium•3/27/25, 1:07 PM

NVIDIA 4070

Ccilium NVIDIA 4070

Furkan Gözükara SECourses•3/27/25, 1:17 PM

it is a good one. if you have 64 gb ram you can train easily

cilium•3/27/25, 1:17 PM

i have 32, does the more RAM u have make it so u can train faster too? So its a mix of Video card and RAM?

Ccilium i have 32, does the more RAM u have make it so u can train faster too? So its a ...

Furkan Gözükara SECourses•3/27/25, 1:18 PM

without 64 gb ram sadly it fails

Furkan Gözükara SECourses•3/27/25, 1:19 PM

because we do block swapping

Furkan Gözükara SECourses•3/27/25, 1:19 PM

so when training it moves non trained blocks to ram

Furkan Gözükara SECourses•3/27/25, 1:19 PM

and get from ram going to get trained blocks

Furkan Gözükara SECourses•3/27/25, 1:19 PM

this way the model fits into gpu with minimal speed loss

HHiramos_Guys, what are your Wan2.1 img-to-vid 480 14B generation times like after the la...

J2_Nedo•3/27/25, 1:28 PM

what resolution are you using for the videos?

FFurkan Gözükara SECourses without 64 gb ram sadly it fails

Weatherby•3/27/25, 4:10 PM

Does cuddims at like, 8400 help this at all, or is it just more ram regardless of speed is better?

Weatherby•3/27/25, 4:11 PM

(Arrow lake doesn't like more than 2 dimms without memory issues. Went from 128 to 48.. but 6000 running at 5400 to 8400)

cilium•3/27/25, 6:37 PM

Are you guys playing with the new Chat GPT image generation

cilium•3/27/25, 6:37 PM

it seems kind of silly good

WWeatherby Does cuddims at like, 8400 help this at all, or is it just more ram regardless o...

Furkan Gözükara SECourses•3/27/25, 6:39 PM

well it makes little difference

Ccilium Are you guys playing with the new Chat GPT image generation

Furkan Gözükara SECourses•3/27/25, 6:39 PM

they will probably cripple it soon :d

cilium•3/27/25, 6:41 PM

Yeah they already announced 2 hours ago but the subtle changes it can make just from text, seem quite good

Ccilium Yeah they already announced 2 hours ago but the subtle changes it can make just ...

Furkan Gözükara SECourses•3/27/25, 6:49 PM

yes its text feature looking really good

Furkan Gözükara SECourses•3/27/25, 6:49 PM

i am expecting local similar model :d

cilium•3/27/25, 7:04 PM

Local damn

cilium•3/27/25, 7:04 PM

that would be a complete gamechanger

cilium•3/27/25, 7:04 PM

What is a good way to just simply get a good LoRA trained quickly by paying? like whatever it costs to get it fast and izi

Inoshi•3/27/25, 7:09 PM

4o model is crazy you can pretty much do any pose/packground/ character/ style and then move it to a good inpaint/depth/lora combination to get what you want.. but i cant wait until Open Source community catches up to them. maybe until the end of the year?

Ccilium What is a good way to just simply get a good LoRA trained quickly by paying? lik...

Furkan Gözükara SECourses•3/27/25, 7:16 PM

get a good gpu and train with our configs

IInoshi 4o model is crazy you can pretty much do any pose/packground/ character/ style a...

Furkan Gözükara SECourses•3/27/25, 7:16 PM

100% i expect

IInoshi 4o model is crazy you can pretty much do any pose/packground/ character/ style a...

cilium•3/27/25, 8:39 PM

it seems like even LoRAs might not be needed soon

Is it possible to train with videos on wan2.1 with 16gb VRAM? Are you going to make a tutorial on it

Similar Threads

Similar Threads

Similar Threads