#🍥|anime
1 messages · Page 208 of 1
if i took the time I could probably find just as many 1.5 models that don't generate young girls with that prompt
I'll just go back to generating cute maid girl doing her job
Just wanna bonk you 
@lyric zealot you are using this one here? https://civitai.com/models/288584/autismmix-sdxl
Mix of pony with some stuff. It's an attempt at making pony more predictable and less dependent on schizo negatives without removing its comprehens...
Not the pony ones
it's not a 1.5 vs XL thing, it's just "this model does, that model doesn't" thing
But the confetti ones
What's the difference
A different merge
ahh ok the same page im guessing that reads cofetti next to pony
ok thanks
I’m off computer so can’t grab the metadata for you
But if you have stealthpng, you should be able to get it from that pic
but the cover images are pretty gory .. i couldn't have guessed that model can produce that subtle cuteness
autismmix_confetti version has another model that's named confetti merged with the autismmix_pony version. autismmix_pony just pony with some loras merged into it.
pony and autism are popular because of the sfw content they can produce, but they do well at most things
cause i dont wanna end up with gory model and not be able to produce cute images
kinda referring to the facial look
score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, rating_safe,
1girl, white hair, short hair, head tilt, ringed eyes, red eyes, off shoulder, jacket, nape, long neck,
source_anime, (bee \(deadflow\):0.7), ligne claire,
Negative prompt: (worst quality, low quality:1.2), 3d, censored, abdominal muscle, artistic glow
Steps: 24, Sampler: Euler a, CFG scale: 7, Seed: 246167163, Size: 768x1024, Model hash: ac006fdd7e, Model: autismmixSDXL_autismmixConfetti, VAE hash: 66f78ad136, VAE: sdxl_vae-fp16fix-blessed.safetensors, Denoising strength: 0.4, Clip skip: 2, Hires upscale: 1.5, Hires steps: 10, Hires upscaler: 4x_fatal_Anime_500000_G, Version: f0.0.17v1.8.0rc-latest-269-gef35383b```
This was from previous previous
So just change the middle section to whatever you want
Okay nap time, before steamboat
ok thanks
You kept artistic glow in it

where do you guys get artist name?
But whatever
Bring me her

You need to ask from the lord and savior @wispy canopy
I don’t have the lists with me right now
it was posted here a couple days ago
ah, this one

thanks eface and smol sus
i guess i deleted pony model earlier too soon
Autism model is the model that manages to move me from sd1.5 to sdxl
they are, you just don't need to prompt the lewd parts 😄
these are all autismmix_confetti images with a few different styles.
Me when I see certain @wispy canopy generations
very interesting to hear that
Silent ping just so I can get back to this message
Loud ping
same

and here i am, hating autism model xD
nah, it's good model, just a bit too single styled
what do you mean single styled?
Yup, if you want more styles then maybe go pony instead
it really wants to do a single style only, artist names work a whole lot less good on that one
this one is also from autismmix_confetti
so my big question is ... how is pony / autism so popular?
hentai
my own personal blend with some silly stuff as prompting
it does everything
Like you tried for the tiny bit you used, it really knows those tags well
except being smart -> dumb as bricks model lol
it knows a ton of styles and characters, is good at multi-character interactions, and also it's very good at lewd images.
as far as the style bias though, I'm not sure it's a global thing, but I found using
score_9, score_6_up, score_5_up, score_4_up, source_anime,
works better than the full one
score_8_up specifically has a huge bias towards a specific style
For you

yeah, for me, i didnt like the style bias, so i removed it
(best masterpiece absurdres 4k hd 8k nbc mtv uber highest quality:6.4)
i took a random image from civitai, is it possible for autism mix to create a pic like this? https://civitai.com/images/7417218
probably
this was the prompt *Channel_42* *Anthropomorphic* *Skunk* *Kawaii*
but you'd have to do some trying
yeah that's a bogus prompt
lol ok so that's more leaned towards how the model was trained?
absolutely
i get it
but honestly all the faces i see except this one from smol sus the faces dont look too good
because i'm doing like the one thing pony models REALLY hate
and not using adetailer, or anything
just thowing out 1024x1024
Draw a city made of colored bubbles
with the right artist tags, maybe. it's not the model I would pick for more realistic styles though.
damn fascinating ... so this pony breed takes model specimen to an entirely different level
i have to see some more work from smol sus 🙂
hard agree here
i mean you guys are good with technical stuff but she's got some sweet touch to the images she creates
@nova remnant these are all autism, same base prompt/seed/sampler but with different artists swapped in the prompt
and that's a very small fraction of what it knows
small fraction of what it knows you say ....
yeah, that list eface posted is only the confirmed artist names, i chose a few from it
and these are all combinations of 2 or 3 artists
there is no way i'd have guessed that came from autismmix .. the examples you showed @vital raptor you do see the big difference tho?
this is using the artist tag from smol sus's image
It knows literally hundreds of artist tags as well as countless copyright tags. It can do a ton of different styles which also usually affect things like face structure and eye shapes.
again, everything else the same as the above images

can also change color tones, line 'cleanliness', shading, backgrounds, etc
just takes some running through the artists to find ones you like
color themes is pretty cool
I used the list as a wildcard and ran the same seed/prompt a few hundred times and picked out a few i liked
like "black/white theme"
then i did more random mixing of the ones i liked until i had a few mixes of styles i like and saved thsoe
oh, you sweet summerchild, i just ran the list like 100k times XD
yeah, i wasn't that bothered 😄
im kinda blown away with its potential ..fair to say i underestimated the model
tbh, i kind of wrote it off at first too
I assumed it was a meme when I first heard of it simply because of the name.
same, and then you discover -> hey, you can make pretty damn good images with it
but after playing with it more it's pretty much the only model i use unless i want realism
well... it IS pretty good at lewd ponies
if i could replace neurons that make it able to do that, and have it be smarter, i'd do it in a heartbeat
yeah, the name and the original nsfw pony images on the civit page put me off it
@nova remnant same seed, only changed artist prompt, using dynamic prompts (combinatorial generation)
score_9 score_8_up score_7_up, {[cutesexyrobutts:liang xing:0,0.9]|[sobchan:scott malin:0,0.9]},detailed official art, character single, cute, girl, leather armor, white shirt, red scarf, rain, petals in the air, wind, hurricane, moody, bust
eh, i fucked something up again
anyway, you get the point
im grabbing the AutismMix Lightning 4 step version
i'd go with the regular version, and then just use the lightnign lora's
but that's me
no big trade offs tho
i'm doing a lot of artist traveling, it doesn't work really well at low steps
ahh ok.. hmm
that's because it's interpolative on each step, and having the prompt be different on step 1 vs step 4, lighting doesn't like that, from my experience, especially if said tokens completely want to change image outcome
is sobchan involved in that one?
no, it's abmayo (its a NAI image, upscaled through kxl-delta though)
Im tired...
i hate it how dynamic prmopts doesn't respect "fixed seed" once you enable "unlink see from prompt" oh well, 4x different artist mixes
wd tagger exists and just recently got new versions. https://huggingface.co/spaces/SmilingWolf/wd-tagger
but we need to do some manual captiong at first right?
no, just drop the images in and it spits out the tags.
?
like this
You can do that if you want. Whether or not it helps at all will depend on the quality of your manually written captions.
you specifically don't want to crop images to the same aspect ratio.
that's going to make the model/lora worse at generating anything other than 1:1 images
https://civitai.com/user/Kisaku_KK77
this guy has pretty good character loras, and he let us download his dataset too, I peaked and saw how he done his loras, so Im gonna try it that way
Soushiki
no
you just made a homeless elf
lol
im suprised, you can make innocent girls too
this looks neat, how did you do it?
which part?
well the pose and since you asked, also the background
"Image Remove Background (Alpha): Remove the background from a image by threshold and tolerance."
uhh.. i dont use comfyui :/
any chance you are playing with pony models?
The pose is just WD image translater mixed with LLava 7b LCM
Using the new Anything XL https://civitai.com/models/9409/or-anything-xl?modelVersionId=384264
下载模型之前,请仔细查看模型介绍 Please review the model introduction carefully before downloading the model モデルをダウンロードする前に、モデル紹介をよく見てください ————————————————————————...
i grabbed the AutismMix cofetti
it has pretty good image output quality, doesn't it?
ahh, ok, im kinda tinkering with pony models
not sure, lol i just woke up
gonna get some coffee then start messing around
not usa tho.
yeah, i was thinking that just after i wrote it
im kinda on the east side .. far from us
that timezone is kinda uh... way bigger than just USA
usa is just most common where people online come from in english channels
i have a wacky internal timezone tbh
all my families are scattered all over the world and i interact online mostly
not bounded by any particular clock lol
@nova remnant still have to thank you for that single image you posted, i think as inspiration for these, i think they worked lovely
which one is that?
that's on my private model mix
ohh
heh
basically just a small pony mix
i posted several on your page
on bpn 1.3?
yeah....
geesh my hands
meant yeah...
last i posted with bpn model was few days ago
still, mixing the pony model did teach me some other tricks when mixing
knowing that the noise amounts need tuning too
you planning on the next version of BPN?
if you can fine tune it for prompt coherence and perhaps work on lightning model that could be fun
bpn is pretty versatile to play with as it is
but since you last worked on it, technology has changed a bit.. lot of authors are releasing turbo and lightning models
basically would be an application of the lora
dreamshaper's author lykon wouldn't release a regular version of xl models... he started with alpha when XL came out, then he released LCM, Turbo, now lightning
i'd much rather fix the damage i have in the IN06-IN08 blocks i know there is 😦
perhaps doing a rehash, and making it a lighting model too while im at it
if you can incorporate the new trends and work on improvements that'd be nice
( <- guy who absolutely doesn't keep up with any trend )
but yeah, lightning is nice ^^
:p
The work you posted is in violation of the Guidelines (linked below). Therefore, we have suspended your account.
balls
it is only a jujutsu kaisen geto drinking an orange juice why does pixiv hate me
pixiv allow chained up woman but not a bare chested man drinking orange juice
how did I miss this
Because its new?
is it?
anyway can it generate fine 16:9 images?
as of 3 hours ago
ouhh I didnt see I thought all of them are xl
probably based on animagine
u prob didnt tagged your img as adult or as an ai img
nice ones, im learning to walk again btw
baby steps ...
im kinda asking gemini for artists name ideas based on specific styles
name some popular and trendy anime artists famous for stunning and vivid visual styles
pony prompt concepts are very booru tag oriented
i wonder if i can posts explicit images here for educational purpose
nah, don't do that 🙂
keep them sfw
ifyou're on a pony model, add rating_safe and most images will be ok to post
how am i doing with prompt style?
1girl, park, blue t-shirt, petite, white shorts, barefoot, picnic basket, sunny day.
source_anime, Kyojurou Kira```
dunno if kyojurou kira works tho
but that seems about correct
well i got a dude instead of a girl with that prompt
but by the look of the image tones .. texture looks pretty ok
so i dunno how that generated a boy
probably a name that's more associated with males
most probably
i mean, have seen that with my testing a lot
im not fond of wildcards
but any suggestion how i can hone in on your long list of artists to get the best of it?
oh -> you have petite in your prompt -> that kinda pushed into nsfw territory because of it's association
this one has petite too
i was trying to imply a regular chest size
for girls
but i have rating_safe
all of them have mildly suggestive poses
lol but they are just sitting o
lol (sorry, got a message blocked because i was trying to help @nova remnant on a prompting so he would have regular chest size)
anyway, you might want to put large chest into negatives, that words better than putting smaller breasts in possive
i used a b word, but that word cannot be spoken here lol
right.
balls?
bouncy b can get you really funny results XD
oh heck ya, i have to try that out sometime
still, why does it want to do suggestive poses only
inflation too
meh
can send you inspirations on dm if u want xD
@nova remnant add rating_suggestive to your negative prompt, that really makes the images a whole lot less suggestive/sexual oriented
my priorities are like this lol.. i wanna get the hang of cute faces, then dainty figure and postures, and then the surroundings
this is really nice
my first tip, always would be -> REMOVE 1GIRL
but i just hate that fucking idea even
that one is more appealing than what you were calling suggestive btw lol
i've made quite a few images to test, but they were all in suggestive poses, which just ruins the whole vibe of the image 😦
how do you portray girl w/o 1girl?
ahh right
and if i really want 1 girl, it's girl, solo
but why remove 1girl specifically?
interesting
never thought of that as 2 tokens. but i was thinking if you were gonna say 1girl is more generic to lewed anime
and this
same as masterpiece on leaked model base
i got rid of masterpiece
masterpiece just increased bust size, not much else
yes, it's "more pleasing" but for all the wrong reasons
masterpiece ... bust size+++ amazing .. i thought that was about art
1girl gets you lewd poses a lot more
top left is beautiful ngl
why -> because 1girl is booru originated, and what kind of stuff do you find there?
yeah i guess, but i dont have much complain
as long as they are clothed and sfw they can all look adorable
the rest of the prompt will overwrite the lewd latent for sure, but when you work with weak latents to begin with (for example lyrics prompting), you don't want 1 token set to ruin everything
gotcha, good point
top left: [ki \(kk-sk-ray\):kaisen chuui:0,0.9] score_9 ,detailed official art, character single, cinematic, BREAK 1girl, park, blue t-shirt, white shorts, barefoot, picnic basket, storm, rain, wet
i like her pose on that one
i'm using prompt fusion for artist interpolation
i just don't like poses like that, i suppose
omigosh
but dont you think she is adorably sweet
the whole physique has such nice and sweet symmetry
oh shit, sorry, that's the wrong artists 😦
which artist is that tho?
i need to save that name
[ki \(kk-sk-ray\):kaisen chuui:0,0.9]
so that's ki \(kk-sk-ray\) which interpolates to kaisen chuui during generation
so at step 1 it's 100% the first artist
and at 90% of the generation, it's 100% the 2nd artist
and that goes in the positive box just like that format?
ahh yes what i was wondering that i'd need some extension
kek, she's the powerhouse 😄
i'm more inclined to the cartoony style of anime, i really love that
this is cartoony too, its pretty 2dish, but so sweet
so many artist variations in that wildcard file
might have been a wrong day for a picnic xD
that image above that i keep mentioning, it has so many talking points to appreciate beauty
Hello everyone. Excuse me we cant generate image ? It’s finished ?
there is stunning grace from head to feet
source_anime, ki \(kk-sk-ray\),
that list you posted some hours ago dont have ki \(kk-sk-ray\),
Since the 02/06/24 all server is down , someone have an idea when they will be back ?
this style is kaisen chuui
Or it’s totally finished ?
bots might not be active
check their status to be sure
i didnt have to describe hair style or pose with ki \(kk-sk-ray\) the image just came out beautiful
weird, i did reference the wildcard list but indeed, doesn't seem to have it
well ki \(kk-sk-ray\) goes on top of that txt file i saved
smol sus was using (bee \(deadflow\):0.7), ligne claire,
im curious about ligne claire
and this is what i got with ligne claire
@wispy canopy ligne claire isnt listed either
you could remove bee from deadflow
ligne claire is not on my list, that i know for sure
bee (deadflow) this comes as built in booru tag
that's how i tested all artist names
just generated an image for ALL artists and see which ones had actual activation values
nice
weeding out false positives is hard tho
but i also see ligne claire inside booru tag
you didnt choose to include that in txt file i guess
ah, probably one of the artists that earlier in danbooru.csv
my workflow wasn't as optimized at that point
ah. and now i see why -> it's an accidental
@nova remnant this is why i test names very extensively
the activation you see it purely concidental, but it doesn't have any real style attached to it
ligne claire is a kind of artstyle if im not wrong
something like outline
well, you can see what it did
for comparison, here's the middle with an actual activation value
"ligne clair" <=> "clear line"

got a few left to do tho 😦 don't feel like doing them, but i've been postponing for like 2 weeks now
nearly 18k images to compare
for probably around ~150 positives
just looking at this image 17k+ times lol
and then pikcing out those who have actual activation values
there's at least one of them on this image, can you spot it?
how did you choose to find ki (kk-sk-ray), and not like the style?
right in the middle
pretyt sure that's a false positive
oh wow
and it is -> the style didn't actually change
so there was a word that just changed the layout
but didn't so antying to the style, so you have to watch for those too
if you can look cross-eyed (like those 3d images), it makes searching a lot easier
a girl sitting on a couch in the living room, watching tv.
cushion be like "nuh uh, not gonna support this"
although, i think of th ethings i marked, probably all low activation value -> not even sure if to include, as they are probably near-incidental or completely incidental activations
oh well, lets give it a go XD
this is what's a strong activation value -> i'm pretty sure this name actually will do something
A joker
chatgpt wasnt a big help
the \( and \) is specific to prompting in a1111
it's to escape them being used attention weighting
it gave me bunch of relevant names but their sytle didnt feel right and some were not recognized as booru tags
heres your ligne claire art example 
nice image, but i need to return to source and breathe, i gotta spend some time on regular models
@supple raptor I trained that lora, but she looks like a child, do you know anything to do?
she doesnt look like a child
sadge
remove 1girl, add 1 woman???
i share your sadge view too
using woman makes the character look better in some ways
better as in age appropriate
deym
.
anything xl?
yup
was using the same prompt as what they were using
to test
is there a big difference between animagine and anything xl?
sadge
left animagine, right anything xl
not sure parameters are exact same I just remixed with civitai
what model is that please?
AnythingXL_xl
thanks
lol im struggling so much to get anything even remotely male
that glow tho
this is better
Well, remember. You're training a character LoRA, not a style LoRA. The goal is to have it generalize things like the outfit, eye color, hair color and style, etc to any style that can be prompted. It should have minimal effect on the style. If she looks too young to you, I would first try prompting 1girl, mature female, character tags with the current version.
I would love to get better at prompting
I do know that there are some Google spreadsheets out there that have some pre-made prompts for people to get an idea of how
Experience is the best teacher. Just start typing danbooru tags into the prompt, lock your seed and then change things in the prompt to see what effect they have.
I don't remember not making one
I only gen cute stuff

1boy
Also if it's hard, probably a bad model overtrained on girls
Or 1guy
1man seems to work too sometimes
Danbooru isn’t the only place tokens come from. It depends on how the model was trained
1guy, solo it likes to add a second character a lot if you don't use solo, with autism at least
Autism has more generic prompting than pony
So what you're saying works better on merges
And other models probably
like i said, depends on the model
2x neko 🙂
AI is the FUTUIRE!
lick
I will post another image in like 2 hours when it finishes rendering at slow af speeds

no
on a potato
RTX 3070
even with XL it shouldn't take over a minute on a 3070 unless you're doing some massive batch tile upscaling or something
Theres a good chance since i dont know what im doing on my own custom workflow that i threw together and hacking and slashing other workflows
The bot is gone?
no,im here
there's never been a bot here
Smh my head comfy
Hot

looks like you're training/merging a lora and something is broken
Where's it
We need more dragon girl butts


made a locon
too busy making more cat girls on the couch with as much exposed tummy as possible without being explicit...
lets marry
is it trained on a pony model? it's common to get errors if the model or lora/locon is pony and the other is not
nope, its just a 1.5sd model, yesmix
You can leave the explicit ones to me
I'll take care of them so that they don't bother your gens

😄
Too ugly 
I think i need to find a good discord that allows nsfw anime gens
so much goes to waste 😄

the waifu diffusion discord 
create a new server, add cat
a dm works for 1 person
I already own and run multiple servers, no desire to make another
Captain has a mafia of servers
We should be careful
join mine 🙂
😄
Maku kiu

based on account creation, @lyric zealot is more OG than me, by a few months at least
i meant his account was pretty old
banned


pretty much since the dawn of discord though
the max probably a few months before me
apparently went publicly live may 2015
i had another account before this one, but it was only maybe 1-2 months before this one so it's not much difference
And TS I couldn't use well
teamspeak was sick tbh
You had to talk to your Minecraft girlfriend
i remember u from TS u used to spam moans in our public lobbies

I only did that in battlefield I swear
looks like my personal server has been around since 2017
made it for gaming with friends
stooopid
I just make another
lol

Kek
i've had the same email address for like...17 years or something
i mean, i've had a lot working in IT, but my personal emails are the same
no bonk
yeah, oldest email in my gmail account is 17 years old 😄

Cursed hair
W
cursed hair just makes it a little messy, kind of disappointed
trained another lora, meh.. not good
Messy hair but with a lot of weight

Try nightmare, I had better success calling things nightmare before
Lmao
Made her have that girl from one punch man hairstyle
The flying green girl
yeah
thanks for this btw, never knew they were called dolphin shorts
never even thought of trying to look it up for some reason, they've always been one of my favorites
It's quite the gamble on sd1.5 though whether you'll get only the shorts or shorts + actual dolphin lol
sometimes both is ok
Reminds me of my favorite character
to beat up in a round
👍
its trying to create smthing, but I need sm sfw images
perfect 🙂
The first potential appearance for a TTRPG character I'm working on
ttrpg?
Aryetis has the right idea though wrong game
oh
She’s for a GM-less system called Starforged
the eyes look very pretty
that face looks mature and eyes look pretty, not something i get with most of the anime models
this image here too, expression is mature and facial look very sweet
its so close to figuring out hands 😦 yet so far away
Out of the suit
cool stuff Tora!
thanks creating with bing
oh interesting
thought you were using SD and some complex styles
nope
i subed but haven't had the access... yet
unless their email went to my spam folder
just doing bing create while I wait for a game to finish updating
so bored going to play Lord of rings online
😄
sounds fun
im listening to music on youtube and just taking things kinda chill
and lazy
while randomly generating some image
and yeah listenting to a vid on youtube
cool.. i have a long music track playing, its a remix of fashion models w/ music
have fun
im too lazy to put effort into casting skills atm .. so not planning on gaming lol
just sitting here listening to house music and watching pretty fashion models
if SD3 turns out to be what it's hyped up about then a lot of ppl would be delighted to ditch the models that came before it
it still a thing with every sd models to come across f*dup hands
drains the creative energy
right now everyone is trying to make the best of what they got, which is different from having the best tool in their box
as those in minnesota say, yeah no bud
not gonna play that game ew
sigh don't know what I want to play.
play guildwars2
best mmo out there
only game i can ever play cause im very fussy about qualities in a game
the mechanism, the visuals, the fluidity - all that important stuff
I mean I could start up star rail
and just do the main missions and ignore the side stuff that distracts me
or ESO
but have you looked into gw2?
but yeah I do have GW2
ahhhh
lol so you know then
that game has me spoiled
i cant settle with any other game after playing gw2 for years
shrugs, sorta in a blah gaming mood
it didn't. only companies with some significant foothold in ML have gotten invites.
yes seems so, lol i checked the folder
but how i feel about the whole SD stuff, its good to have witnessed its evolution with different txt2img products, and i can't wait to evolve past sd, xl
i take gaming bit differnetly, i dont just keep playing every new games, i do check their trailers tho, but i try to settle with a game that has healthy community and the game mechanics i can relax to and build upon
gw2 has won me
I basically only play path of exile
The rest that I play is when I'm done with path of exile cx
Roguelike games are very fun for a short while
is that a top down view game?
oh diablo style games can suck me in
Yes like Diablo or lost ark
ahh ok
But it's actually good x3
honestly i dont find top down games very immersize
thought of trying Tale of Immortal again, a cultivation game but my gawd so much to keep track of, so complex
it's not so much about immersion as keeping my brain busy and choices
that kinda game i would wanna enjoy their game world has big significance in that respect
GW2 and 2dlc's should be done in about 30 😄
sure, but i like to explore the game world thats a major attraction for me
wellyeah me too
thats why I like games like No man's sky. but wish they had a better built in story and pregression, maybe some auto random questions that popup instead of gaving to go search for them
lol tora its a mystery to me why ppl play so many games
i mean when you havea game that covers all .....
Amazing cultivation simulator is another game that is so complex I gave up lmao, they have that in common at least
cat, gw2 has pvp, pve, story quests, open world group events, immersive world, different metas of different levels of dificulty, gambling, so many stuff all rolled into one
Reading a Cultivation LitRPG book right now
crafting, gathering, dungeons, etc etc
What keeps me playing it is the amount of creativity it allows you that other games don't give
yes
gw2 is very good with creative contents too
I played gw2 with few friends in the past
Just the fact that a game is class locked turns me off lmao
ever looked at "Project Gorgon" it's an old skool style MMORPG
I really like being able to do anything in a game
well there are 5 races each with 9 different classes, what do you mean locked?
they even have a demo
The fact that you have to choose one
And are locked into that
humans, norns, asura, charr, sylvari all of them can be any class
so each of those race be elementalist or whatever class you enjoy
ponders reinstalling
be warned very old skool graphics and gameplay
and is being updated very often
gw2 runs on directx12 but the visual is spectacular
nods. I know
Have you seen throne and liberty
That game looks so good I want to play it just for that xd
gw1 is great too but their combat system is not as fluid as gw2
also key difference with gw1 is its intance based but gw2 is open world focused
Torqx
You tried world of warcraft?
yes
What did you think
i have to admit tho, the sub system discouraged me from playing
but a really old skool , in a maddening way, WURM
What does that mean
but cat, wow had this enchanting charm of game world
Oh subscription
the subscription fee to be able to keep playing
yeah, it felt like i have to force myself to play to get my money's worth, thats discomforting
WoW was one of the worst for me xd
wow had a enchanting game world tho when i was playing it
It tried to kill me of boredom, I didn't get past level 30
i mean mmo were kinda fresh too back then
I think the game is just too old for me
I don't have the nostalgia factor for it or the love for mmorpgs
yeah but im heavily absorbed with gw2
My early days of the internet I spent on cs 1.6

what's cs?
fy_iceworld
Counter strike
same and glass maps
de_dust
loved the vip
the mode that was forgotten
Community based modding stuff
surf_skyworld
i dont like pvp either, tho gw2 has it, it feels weird to go after fellow gamers
and to kill them :/
Super hero was you kill enemies to level up and get more powerful, you could have powers like Goku that charges up some Kamehameha's lmao
Made the gameplay extremely fun
All that inside counterstrike
it was the the same thing, lol
sounds like every pve gaming when lvling up
Yeah but in a first person shooter it's much more fun
Earth Special Forces mod was Epic too
I love first person games
yes i can imagine that
i would love it too if i wasnt going for real ppl
Dragon Ball Z: Earth's Special Forces! Playing as Cell, Goku and Vegeta. Trying to pull off the Spirit Bomb!
Check out the mod: http://www.esforces.com/
https://twitter.com/jdantastic
https://plus.google.com/+JdantasticGaming/posts
http://www.twitch.tv/jdantastic
http://instagram.com/jdantastic
http://www.jdantastic.com/
https://www.facebook.c...
I played cs for a few years, then went to league of legends, soon path of exile after
poe
its very likely you probably killed a lot of sweet girls behind those characters
err, what? lol
😛
and they curse you in the mic
Sticker Maker =
I'm lost 😄
lol that's how in gw2 pvp CAN be toxic
it keeps putting the monitor image on the back of the monitor lol
lol
lol i mean girls gamers in real life behing those game characters
I knew a very cute girl in league of legends who would cosplay as one piece characters, we loved to play together but when we played against each other I'd never hear as much cursing as any other of my friends

It was fun, friends vs friends private matches lmao
I'm so lost
yeah cause you two were close
just talking about games stuff
Your prompt has the word Game which is why its on the monitor
word association
everyone was talking about counter-strike 😄
then someone said I was killing something and I got lost 😛
Multiple conversations, Think of discord like a Think tank, a group of people with the same idea's throwing their thoughts
yeah prolly 🙂
i was coming at you when describing pvp where ppl kill each other
lol
text fail, but close enough 😄
you get what im saying, i dont like pvp 🙂
He's only Pve
Maybe I can generate an image of Minesweeper.
Time to shower 
i was working on music theory and tired to make bing generate C ionian mode, instead it genearted gibberish, wish those Ai could be improved to generate more meaningful stuff and how that can change learning
since you were using BING, google gemini has no cap to the number of images you can generate
draw a game world with cute creatures and lush forest. the sky is purple with bright blue moon.
left one is from bing and right one from gemini
i picked 1/4
SDXL
sdxl wins
the asura looks like Dobby- house elf
an eagle mid flight. a scenic and stunning view below
lol i like to mess with asuras
cause they are kinda tiny and smart

Too bad Bard can't generate people at the moment

ponders a name
yeah cause google screwed up big time with their woke guy in the deparment
dankit , used
moccha with moa



