Feat: z-image pixel space (model still training atm) (#12709)

* draft zeta (z-image pixel space)

* revert gitignore

* model loaded and able to run however vector direction still wrong tho

* flip the vector direction to original again this time

* Move wrongly positioned Z image pixel space class

* inherit Radiance LatentFormat class

* Fix parameters in classes for Zeta x0 dino

* remove arbitrary nn.init instances

* Remove unused import of lru_cache

---------

Co-authored-by: silveroxides <ishimarukaito@gmail.com>

comfy/latent_formats.py

@@ -776,3 +776,10 @@ class ChromaRadiance(LatentFormat):
 
     def process_out(self, latent):
         return latent
+
+
+class ZImagePixelSpace(ChromaRadiance):
+    """Pixel-space latent format for ZImage DCT variant.
+    No VAE encoding/decoding — the model operates directly on RGB pixels.
+    """
+    pass