Maskbit
Collection
11 items
•
Updated
This model is the MaskBit tokenizer with a vocabulary size of 16bits. It uses a downsampling factor of 16 and is trained on ImageNet for images of resolution 256.
You can find more details on the project page and in the paper.
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Dataset used to train markweber/maskbit_tokenizer_16bit
Space using markweber/maskbit_tokenizer_16bit 1
Collection including markweber/maskbit_tokenizer_16bit
Evaluation results
- rFID on ILSVRC/imagenet-1kself-reported1.290
- InceptionScore on ILSVRC/imagenet-1kself-reported193.600
- LPIPS on ILSVRC/imagenet-1kself-reported0.278
- PSNR on ILSVRC/imagenet-1kself-reported21.800
- SSIM on ILSVRC/imagenet-1kself-reported0.580
- CodebookUsage on ILSVRC/imagenet-1kself-reported1.000