Jay @memmaptensor

Independent researcher, 1st year computer engineering student bento.me/wootwoot Bangkok Joined December 2018

Tweets

26
Followers

519
Following

79
Likes

1

Jay @memmaptensor

a year ago

dehumidifiers are so good when it's cold and humid

0 0 1 662 0

I've set out some conditions for any future solvers to add: • No implicit solvers: those require root-finding, meaning a Jacobian has to be computed by running a backward pass through the model during LBFGS optimization. • No non-RK methods: this would cut off linear multistep…

0 0 2 674 1

Jay @memmaptensor

a year ago

Last major update! • Added solver settings for adaptive_scipy • Adaptive solvers now show the number of steps taken • Accurate 𝜎 timestep info is now displayed Check out the most comprehensive fixed and adaptive higher-order samplers on ComfyUI! github.com/wootwootwootwo…

0 3 21 2K 7

Download Image

Jay @memmaptensor

a year ago

Refactored and fixed some bugs with the progress bar! Also wrapped the solvers from scipy.integrate If you count the a-methods as 2 (since they work with both the adaptive_pid and fixed_scheduled controllers), then this node has (excluding forward euler) 31 new samplers! I…

0 0 6 550 2

Download Image

Jay @memmaptensor

a year ago

why do i prefer undersampled results 😭😭😭

1 0 1 545 0

Jay @memmaptensor

a year ago

diffusion models are definitely still not dead. sure, optimal transport conditional flow matching is provably better, but so much of the community was already built on discrete time diffusion. and with kolors out (an SDXL model trained with DDPM formulation and eps-pred…

1 0 16 1K 7

Jay @memmaptensor

a year ago

i have an idea for a slightly modified class of SDXL models that would mostly be compatible with existing finetunes and loras it's been proven to work well on SD1.5 with good results definitely next on my bucket list will post updates and releases soon, hopefully, if it works

2 0 16 876 0

Jay @memmaptensor

a year ago

i'd like to continue working on anime animation tech. version 1 is designed to be distilled for realtime inference. version 2 won't be concerned with realtime inference and would probably be based on a flow-matching mmdit with more fine-grained control and even better quality.

3 0 27 987 0

Jay @memmaptensor

a year ago

training is done in 2 days!!! as for inference compute requirements: it's basically the same burden as animatediff will probably work on realtime inference next month after figuring out life. realtime txt/vid2vid on distilled models is already empirically shown to be possible.

1 0 9 929 0

Jay @memmaptensor

a year ago

yo @EsotericCofe @anifusion_ai wanna join forces best manga tool + best character animation model = ???

12 32 263 26K 128

Download Video

Jay @memmaptensor

a year ago

I'm working on an architecture that can animate anime characters pretty consistently. It's about 40% trained. I still don't know what to do with the model once it's done.