Prompting 101 Context Is Key! | Playground AI Tutorial
Вставка
- Опубліковано 6 вер 2024
- In this Playground AI tutorial, we'll dive into Prompting 101 and why context is key in generating accurate and relevant AI results. Whether you're new to Playground AI or a seasoned user, this tutorial is a must-watch!
Check out our Prompt Guide, we'll be adding much more very soon! playground.com...
Find us on
Playground AI - ai image generator: playground.com
X (Formerly Twitter) https: // playground_ai
What are some of the "prompting challenges" you come across? How can we help?
I'm having a lot of trouble creating certain flags. Every now and then I want to add a German flag, the AI knows the colors but keeps mixing them wrong and in strange shapes.
I have already tried a few things. I tell the AI where which colors should be or that they should be horizontal, but the result doesn't get any better.
Well the biggest one is promp adherence,which still a problems ,please improve it and playground will better!!
👋
Yup, I discovered the important of context nearly 2 years ago in MJ ------ but it applies to all text to image AI models
I notice that even when I am able to get he full body, most of the time the overall quality of the image goes down. More distorted faces and limbs.
I have lots of trouble on creating maps
gracias, tks
Welcome!
Excellence as always!!!!
Thank you! Cheers!
I put this in the discord, but I thought I would also put it here:
It would be great if there was a feature to create folders to be able to keep track of different designs that might be like kind. Like if you created a series of dog pictures and cat pictures and some Pirates or something.
So I could put all my dog pictures in one folder, all my cat pictures in another folder, etc. I had created a series of horoscope pictures, couldn't quite get a few to turn out like I wanted, so I moved onto some different idea, came back and created more horoscope pics a month later, now I have some in the middle, some at the start of the pictures I created. I want to group them together for easier access to upscale them then download them for the horoscope project. Folders would make that easier.
I would happily pay more to be able to group together my pictures and have them categorized. I would settle for being able to tag the pictures and then search my pictures based on the tags. Something-anything-to be able to group them together better!
Something that would also be great, to be able to resize a picture after it is created and keep the resolution or DPI for what I need it for (which is Print on Demand).
Lastly, a few of us in Discord talked about this, Bows and Arrows...getting the the string of the bow to work is problematic...I was doing a horoscope series and got to Sagittarius and created really amazing pictures...except the bow string...
It also had a hard time with the stinger for a Scorpion and the legs for a crab...
Any recommendations or fixes for bow, stinger, or crab legs?
Noted and thanks for the feedback.
Not only bows and arrows, but also axes. Maybe all types of weapon except swords.
the key is the ability to lable them what ever we need not preset things like weapons ect@@zavulon422
I just found out that there is a some instruction for prompting. I follow this chanel but somehow I missed this till now.
Please add a link from the profile page to this guide! Maybe highlighted the link.
The link is in the description :) And it's new! I'll link it here anyway. playground.com/prompt-guide But yes we will put it into the product likely in the side menu.
I’ve been a MidJourney member for a year but now want to try this awesomeness, great video sir
Welcome and enjoy!
It would be great if Playground AI introduced a feature where two images could be merged in canvas mode. Many users create montages by incorporating objects from other creations. It's surprising that the site offers so many options but has overlooked basic features like working with layers.
We appreciate the feedback and yes we've had many requests for this and will be looking into it. 👍
love the rythm
Me too! 🙌😊
Cartoonists feels like
😁😊
This was helpful. I do caricature wood carvings. My problem is creating a pattern. I can't get a good prompt for playground to get a front view and side view done in pencil drawing.
Yeah side views can be tricky and often require multiple generations. What can sometimes work depending on the subject is using a wide width and prompting for "character turnaround, multiple poses, side, front, back". Might require some post production though.
@@playgroundai thanks for the reply. I'll play around with it.