order could be affecting as well
order could be affecting as well
Fnot sure though
Cit does, but you have to remember that you are most likely using too few assumptions anyway and thus overtraining is virtually impossible
CUhm I mean, thus NOT overtraining is virtually impossible, we are human
CBut no combination of the numbers
[5812, 49345, 2415, 2042, 10075, 30354, 269, 268, 672, 578, 28746, 890, 2266, 4190, 4950, 10075, 9833, 11, 30787, 220, 10747, 11, 4047, 6496, 12653, 4875, 12036, 11, 3683, 4429, 416, 978, 459, 958, 7842, 41476, 23197, 11277, 259, 27321, 31341, 47983, 8746, 271, 952, 33629, 1355, 354, 12381, 33728, 371, 5374, 64, 16251, 396, 269, 2417, 346, 686, 11993, 3487, 3881, 4053, 11, 40544, 4135, 1986, 2045, 62, 265, 62, 1177, 263, 11, 7117, 513, 35, 410, 2433, 22865, 807, 74, 21480, 67, 11, 43150, 8034, 7786, 33675, 25237, 220, 220, 198]
Will have a HIGHER weight than [5812]
[5812, 49345, 2415, 2042, 10075, 30354, 269, 268, 672, 578, 28746, 890, 2266, 4190, 4950, 10075, 9833, 11, 30787, 220, 10747, 11, 4047, 6496, 12653, 4875, 12036, 11, 3683, 4429, 416, 978, 459, 958, 7842, 41476, 23197, 11277, 259, 27321, 31341, 47983, 8746, 271, 952, 33629, 1355, 354, 12381, 33728, 371, 5374, 64, 16251, 396, 269, 2417, 346, 686, 11993, 3487, 3881, 4053, 11, 40544, 4135, 1986, 2045, 62, 265, 62, 1177, 263, 11, 7117, 513, 35, 410, 2433, 22865, 807, 74, 21480, 67, 11, 43150, 8034, 7786, 33675, 25237, 220, 220, 198]
Will have a HIGHER weight than [5812]
Fbut we also cant know what kind of affect the combination makes. also remember, after training, you have to use those prompts trained on usually to get your subject likeliness
CI want to associate looking at viewer
Cbut probably smarter to train and finetune with looking at viewer facing left
CWell of course because the images will lose weight >(
CIf I train with token
[5812, 49345, 2415, 2042, 10075, 30354, 269, 268, 672, 578, 28746, 890, 2266, 4190, 4950, 10075, 9833, 11, 30787, 220, 10747, 11, 4047, 6496, 12653, 4875, 12036, 11, 3683, 4429, 416, 978, 459, 958, 7842, 41476, 23197, 11277, 259, 27321, 31341, 47983, 8746, 271, 952, 33629, 1355, 354, 12381, 33728, 371, 5374, 64, 16251, 396, 269, 2417, 346, 686, 11993, 3487, 3881, 4053, 11, 40544, 4135, 1986, 2045, 62, 265, 62, 1177, 263, 11, 7117, 513, 35, 410, 2433, 22865, 807, 74, 21480, 67, 11, 43150, 8034, 7786, 33675, 25237, 220, 220, 198]
[5812, 49345, 2415, 2042, 10075, 30354, 269, 268, 672, 578, 28746, 890, 2266, 4190, 4950, 10075, 9833, 11, 30787, 220, 10747, 11, 4047, 6496, 12653, 4875, 12036, 11, 3683, 4429, 416, 978, 459, 958, 7842, 41476, 23197, 11277, 259, 27321, 31341, 47983, 8746, 271, 952, 33629, 1355, 354, 12381, 33728, 371, 5374, 64, 16251, 396, 269, 2417, 346, 686, 11993, 3487, 3881, 4053, 11, 40544, 4135, 1986, 2045, 62, 265, 62, 1177, 263, 11, 7117, 513, 35, 410, 2433, 22865, 807, 74, 21480, 67, 11, 43150, 8034, 7786, 33675, 25237, 220, 220, 198]
Cand then use token 5812, it will not be as affected
Cmakes perfect sense
NOk! i think its not risk then
Fyep
Funless there is some auto update command inside it. what you have inside that file
NI didnt add anything. Didnt add Git pull for example
Fkk you should be totally safe
C
C
Fman poor girl getting into many different shapes 
CI hope its ok?
Fye as long as not nswf
Fyou know this is school discord
COhwx woman black and white cyberpunk cenobite intricate long red hair beautiful cybernetic, realism landscape, highly detailed realistic digital painting, Art station by Alastair magnaldo adam martinakis alejandro burdisio Alison Bechdel Andreas Rocha cartoonist cyril rolando normal rockwell,beautiful face looking at viewer, Engine 3D vray unreal 8k uhd, Lighting studio sharp HDR ambient
prompt is not haram
prompt is not haram
CYes, of course is school discord
CI am trying to be scientific 
Fstill able to keep face shape
F100% stylzing
Fvery good success
CI am showing the images which I will keep for attempt at fine tuning
Fi see
Cthis one will not be accepted
Cbecause for instance, no red intricate hair
Fyes looks not related
Cit is because I have long red hair
Token count: 3
1538, 736, 2225
Token count: 3
1538, 736, 2225
CThe relative WEIGHT probably doesnt hurt at all
Cbut let say I use only 3 tokens
CLong red hair
CThen you can see we already have an ambiguous error because we will have a very tough understanding of what do you mean by long red hair
CIf i keep sending images say of a woman with blue hair for token 1538, 736, 2225
Cthen you will overtrain theoretically as you approach infinite time
NMay i ask you a question about graphics cards? i have just purchased a 3090 but i doubt if is more future proof a 4080 for training (because it seems less vram is needed each time). Should i keep the 3090? What do you think?
CVRAM is king is my understanding
C3090 has more vram
Ckeep it
Cif this is what you want to do?
FKeep more vram
Nsure, id like to keep training without issues in future
FThere is also one another thing
FUnet training
FI am still not sure how all works yet