SD 1.5 IFT — foto asli, US, gender
UNet hasil train.py --sft di 280 foto asli Pexels (1400 baris train / 420 val,
dataset fassabilf/ift-train-us-real), 30 epoch, --seed 0.
VAE, text encoder, dan tokenizer identik dengan SD 1.5 base — tidak ikut diunggah.
Rakit pipeline-nya dengan memasang UNet di sini ke stable-diffusion-v1-5/stable-diffusion-v1-5
(lihat scripts/ift/ckpt_to_pipeline.py di repo kode).
Checkpoint yang diunggah: ep1, ep2, ep3, ep4, ep5, ep6, ep7, ep8, ep9, ep10, ep11, ep12, ep13, ep14, ep15, ep16, ep17, ep18, ep19, ep20, ep21, ep22, ep23, ep24, ep25, ep26, ep27, ep28, ep29, ep30. Setelan lengkap di hparams.yml; angka evaluasi,
error bar, dan lantai MAE di analysis/.
Model ini artefak penelitian tentang bias distribusi gender, bukan model produksi.
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for fassabilf/sd15-ift-us-real
Base model
stable-diffusion-v1-5/stable-diffusion-v1-5