Downloads · 30 days
6
40% of all-time downloads
vaev1-g2v-02-jit-b-clean-epoch200
karrykkk/vaev1-g2v-02-jit-b-clean-epoch200
vaev1-g2v-02-jit-b-clean-epoch200 is a machine learning model from karrykkk. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for pytorch.
Final checkpoint from experiment g2v-02-vae-jit-b-clean in the 2026-09-20 group20920vae run.
All-time downloads
15
Public
Repo size
2.1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt2.1 GB · 100%
At a glance
- Library
- pytorch
- Access
- Public
- Created
- Sep 21, 2026
- Updated
- Sep 21, 2026
- SHA
- 7dfa4c01
Model card
From the Hugging Face model README
VAEv1 g2v-02 JiT-B clean — epoch 200
Final checkpoint from experiment g2v-02-vae-jit-b-clean in the 2026-09-20
group2_0920_vae run.
- Model:
sihc.rae.models.RAEJiTB(JiT-B) - Prediction target: clean (
clean_eps: 0.05) - Encoder:
stabilityai/sd-vae-ft-ema - VAE setup: posterior sample, scaling factor
0.18215 - Latent shape:
4 x 32 x 32; transformer patch size:2 - Training precision: bf16
- Training epochs: 200
- Final step: 250200
- Global batch size: 1024
- EMA decay: 0.9999
- Sampling protocol: EMA, Euler-50, CFG 1.0, seed 12345
- Weights & Biases run
Files
checkpoints/ep-last.pt: full PyTorch training checkpoint saved withepoch == 200config.yaml: merged training and model configuration
For evaluation or inference, load the EMA state:
checkpoint = torch.load("checkpoints/ep-last.pt", map_location="cpu")
model.load_state_dict(checkpoint["ema"])
The external stabilityai/sd-vae-ft-ema weights are not included.
- Library
- pytorch
- Created
- Sep 21, 2026
- Updated
- Sep 21, 2026