Verified: a month ago
SafeTensor
The FLUX.1 [dev] Model is licensed by Black Forest Labs. Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs. Inc.
IN NO EVENT SHALL BLACK FOREST LABS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
Use "L1m1n4l" as the trigger w1rd.
One could also try "photograph of an empty l1m1n4l 90s interior, liminal space, photo from 1991, vintage thriller movie still" if one wanted to.
Search "liminal space" on Reddit to get an idea of the goal. Trained from images from the early 90s.
----
(づ｡◕‿‿◕｡)づ 🗯𝐓𝐡𝐚𝐧𝐤 𝐲𝐨𝐮 𝐟𝐨𝐫 𝐠𝐞𝐭𝐭𝐢𝐧𝐠 𝐒𝐥𝐢𝐦𝐞𝐝!
𝐖𝐞 𝐤𝐧𝐨𝐰 𝐲𝐨𝐮 𝐡𝐚𝐯𝐞 𝐚 𝐜𝐡𝐨𝐢𝐜𝐞 𝐢𝐧 𝐯𝐢𝐬𝐜𝐨𝐬𝐢𝐭𝐲 𝐚𝐧𝐝 𝐚𝐞𝐬𝐭𝐡𝐞𝐭𝐢𝐜 𝐰𝐡𝐞𝐧 𝐲𝐨𝐮 𝐜𝐡𝐨𝐨𝐬𝐞 𝐰𝐡𝐨 𝐭𝐨 𝐬𝐥𝐢𝐦𝐞 𝐰𝐢𝐭𝐡.
Helpful Nuggets:
The fire walks with you. Trained on images that pertain to the first sentence, you could say. You could also say, "those two mountain-tops looks exactly alike."
Both interior and exterior spaces can be conjured and explored without much prompt. While the images don't contain people, AI has learned that humans almost always place a subject in their photos.
For that reason, you will end up with some fairly rad 1990s fashion when humans are generated. Try "empty l1m1n4l space" to avoid the phenomenon.