SAFMN Super-Resolution (GGUF)

4x image super-resolution via SAFMN (ICCV 2023). 228K params, 948 KB.

Architecture: Conv3x3 โ†’ 8 AttBlocks (SAFM + CCM) โ†’ Conv3x3 + PixelShuffle(4).

Parity: cos=1.000000 vs Python reference.

Source: sunny2109/SAFMN (Apache-2.0). Trained on DF2K dataset.

File Size Description
safmn-x4-f32.gguf 948 KB F32, 228K params, 4x upscale
Downloads last month
63
GGUF
Model size
240k params
Architecture
safmn
Hardware compatibility
Log In to add your hardware

32-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support