RealPLKSR 4x super-resolution for MLXEngine — the web-photo tier, trained on the provenance-clean Commons-SR corpus
-
Updated
Sep 29, 2026 - Swift
#
mlxengine
Here are 5 public repositories matching this topic...
FlashVSR v1.1 one-step streaming diffusion video super-resolution (x4) for Swift/MLX and MLXEngine — Metal block-sparse attention. Recommended for live action, not anime/graphics.
-
Updated
Oct 5, 2026 - Swift
NERVE x2/x4 super-resolution for MLXEngine (Swift/MLX): Phips/NERVE, Apache-2.0, CC0-trained — the provenance-clean fast imageUpscale tier
-
Updated
Sep 30, 2026 - Swift
VOSR2 one-step generative super-resolution (1.4B LightningDiT + DINOv2-L, Qwen-Image VAE) for Swift/MLX and MLXEngine
-
Updated
Oct 1, 2026 - Swift
HEART (Phips/HEART, Apache-2.0, CC0-trained HAT-iLN window-attention super-resolution) ported to Swift/MLX as an MLXEngine imageUpscale package
-
Updated
Sep 30, 2026 - Swift
Add this topic to your repo
To associate your repository with the mlxengine topic, visit your repo's landing page and select "manage topics."