Software Engineering Courses (SECourses)•10mo ago

RTX a6000 is not fast just big vram 😄

RTX a6000 is not fast just big vram

FFurkan Gözükara SECourses RTX a6000 is not fast just big vram 😄

cyleap•3/1/25, 10:04 AM

yeah, you reckon 4090 is faster?

Ccyleap yeah, you reckon 4090 is faster?

Furkan Gözükara SECoursesOP•3/1/25, 10:04 AM

yes

Furkan Gözükara SECoursesOP•3/1/25, 10:04 AM

but for this model i didnt compare

Furkan Gözükara SECoursesOP•3/1/25, 10:04 AM

5090 faster though i tested

cyleap•3/1/25, 10:05 AM

okk. ill give it a shot.

Furkan Gözükara SECoursesOP•3/1/25, 10:05 AM

make sure to get high ram machine

Furkan Gözükara SECoursesOP•3/1/25, 10:05 AM

if on runpod

FFurkan Gözükara SECourses yes

cyleap•3/1/25, 10:05 AM

another question. does fp8 makes it faster?

FFurkan Gözükara SECourses make sure to get high ram machine

cyleap•3/1/25, 10:05 AM

okk

Ccyleap another question. does fp8 makes it faster?

Furkan Gözükara SECoursesOP•3/1/25, 10:05 AM

nope only reduces RAM and VRAM

Furkan Gözükara SECoursesOP•3/1/25, 10:05 AM

because operations are still made in bf16

Furkan Gözükara SECoursesOP•3/1/25, 10:05 AM

only model weights loaded as fp8

cyleap•3/1/25, 10:05 AM

i see.

cyleap•3/1/25, 10:06 AM

thanks !

Aikage•3/1/25, 2:55 PM

I made a mistake and used the 720 image to video

1 hour later i'm only 65% done

Aikage•3/1/25, 3:06 PM

Does having a large base image resolution or the output resolution being different than the incoming image resolution make a difference in terms of speed for Wan i2v?

Zono50•3/1/25, 4:43 PM

message.txt2.42KB

Zono50•3/1/25, 6:35 PM

nevermind models havent' downloaded lol

Aikage•3/1/25, 7:48 PM

where would i clone qwen to if I want to use prompt enhance? For some reason even if I remove the line models won't download (probably due to VPN or something IDK)

Aikage•3/1/25, 7:48 PM

but I can manually clone repos

Aikage•3/1/25, 7:48 PM

just go in the model dir?

AAikage Does having a large base image resolution or the output resolution being differe...

Furkan Gözükara SECoursesOP•3/1/25, 8:03 PM

our app auto crops

AAikage where would i clone qwen to if I want to use prompt enhance? For some reason ev...

Furkan Gözükara SECoursesOP•3/1/25, 8:03 PM

i improved qwen thing

Furkan Gözükara SECoursesOP•3/1/25, 8:03 PM

now it uses way lesser vram but you need to reinstall

Furkan Gözükara SECoursesOP•3/1/25, 8:03 PM

havent announced yet

Furkan Gözükara SECoursesOP•3/1/25, 8:04 PM

we use this qwen exactly let me show

Furkan Gözükara SECoursesOP•3/1/25, 8:04 PM

Qwen/Qwen2.5-14B-Instruct

Furkan Gözükara SECoursesOP•3/1/25, 8:06 PM

this repo : https://huggingface.co/Qwen/Qwen2.5-14B-Instruct

Qwen/Qwen2.5-14B-Instruct · Hugging Face

Furkan Gözükara SECoursesOP•3/1/25, 8:07 PM

it is downloaded into hugging face cache folder

Furkan Gözükara SECoursesOP•3/1/25, 8:07 PM

i will add it to the model downloader as option

Aikage•3/1/25, 8:26 PM

ty very much

AAikage ty very much

Furkan Gözükara SECoursesOP•3/1/25, 8:40 PM

sorry for delay

Furkan Gözükara SECoursesOP•3/1/25, 8:40 PM

i am about to update

Furkan Gözükara SECoursesOP•3/1/25, 8:40 PM

adding FPS increase

Zono50•3/2/25, 1:16 PM

@Furkan Gözükara SECourses I downloaded the wan2.1 zip file, ran the massed compute, and then manually ran python Download_RIFE.py afterwards, and i'm still not seeing the updated fields in the gradio interface

Zono50•3/2/25, 2:06 PM

i also followed the massed compute instructions and still not seeing the updated gradio fields