anny-render-corpus-generated

Images generated by OmniGen2 from the constructed renders in chibifire/anny-render-corpus.

Code: weftspun/anny-render-corpus, on the 6-datasource side of the hexagon.

Why this is a separate repository

These are generated synthetic, not constructed. They were sampled from a model rather than rendered deterministically from a rig, so their labels are inferred and not true by construction. Our working agreement requires generated data to be stored and manifested separately and never merged into an undifferentiated pool. A separate repository enforces that where a subdirectory would only describe it: clone the corpus and these cannot come along.

They are not training data on their own, and evaluation does not use them.

Provenance

Every directory carries the JSON its run wrote: the resolved OmniGen2 commit df5dca8a981d74e6c3af214c145f5c735fe72367, the full prompt for each view, the seed, the step count, cfg_range and the guidance scales. bf16 throughout, because quantised weights do not produce corpus data here.

directory	contents
bench/	3 generated views
ladder/	14 generated views
lora/	8 generated views

What they show

ladder/ is the base model asked for eight camera azimuths: recovered azimuth tracks the request with a slope of 0.04. lora/ is the same eight prompts after training, at 0.10, with azimuth 90 moving from 97.6 degrees wrong to 13.3.

Two views in lora/ have no detectable person in them. They are kept rather than dropped: a set that quietly excludes its failures reports a better result than it earned.