Mamen Mendizábal is a Spanish journalist and TV presenter.
This is a 140-step TI trained on a dataset of 15 images with these settings.
Curious about my work process? I have summarized it here.
Do you have a specific idea for a TI in mind? Visit my website and let me know.
Building a good prompt with my TIs
You're obviously free to experiment, but bear in mind that my TIs are trained with a more or less fixed phrasing, that normally starts with:
"photo of EMBEDDING_NAME, a woman"
So I recommend always starting your prompt like that and then building the rest of the prompt from there. For instance, "photo of beautiful (mamenmdz:0.99), a woman with beautiful hair, as a movie star in a (movie premiere), premiere gala, (near a movie theatre), natural skin texture, (elegant {red|yellow|blue|green|black|turquoise} sweater off-shoulder with suspenders), 24mm, 4k textures, soft cinematic light, adobe lightroom, photolab, hdr, intricate, elegant, highly detailed, sharp focus, ((((cinematic look)))), soothing tones, insane details, intricate details, hyperdetailed, low contrast, soft cinematic light, exposure blend, hdr, faded, (paparazzi in background), (painted lips:1.1), ((looking straight at the camera:1.1)), (smiling)"
Description
140-step TI trained on a dataset of 15 images with these settings.
FAQ
Comments (6)
"Wow! Very Good. I'm currently very satisfied with the Monica Bellucci one from before, and this current work looks very much alike. However, I find calculations like Epochs, batch size, and gradient to be complex and hard to understand. So, instead of explanations or calculations, I would appreciate it if you could provide me with the numbers I should use. Typically, when I have 44 photos ready.
I have 44 data photos, and I'm working on embedding 18 of them. These are pictures of my own face, not someone else's. My batch size is 3, and the Gradient accumulation steps are 44. Max steps: 160. I'm not sure what the appropriate values should be after this. It's not 10, 10, is it? I would be grateful if you could let me know." T,.T
Hey! With a dataset of 18 pics, I'd probably go for batch size 3 and gradient 18. That's the ratio I always use for my TIs--my gradient always equals the number of pics in the dataset. What that means is that you run 3 full epochs per step. In my case, my TIs usually peak around epoch 400, which means that you should check steps from around 120 to 140 or so in order to find the best version of your embedding, if you follow the settings I've mentioned above. And yes, I save a copy of the embedding every 10 steps (and also generate a training pic every 10 steps). I normally do that until step 120 and then I change the values to 5 and 5 respectively until step 140.
Thank you for your kind response
beautiful!
What a hot lady! Thx!
I knew you'd like her! ^^
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.