AngelBottomless
AngelBottomless
hopper minecart with option turned on tries to push item every tick, so it means its taking 3 / 2 ticks per block - yes, 8 block / second so...
**note :** its for tweakeroo or **pure direction property packets**, litematica *should* be able to already place double chests.
@gel-crabs I will do more test within 18 hours, but I guess this should work (as they share same structure) @FurkanGozukara [The XL Code ](https://github.com/horseee/DeepCache/commit/7c92363413cd93aeac3f22ea30b48eb7a56aff06)has released 5 hours ago, but...
@gel-crabs I guess we might have to adjust the indexes of in-out blocks, XL Unet is more deeper, so using shallow parts earlier would lead to cache 'noisy' semantic informations....
Temporary update : I think the implementation should be modified to follow the original paper again. Original paper says that we should sample the values for nearby steps, not by...
768x768 test Hypertile only  7.86it/s Index 0, 0, 0   Cache Rate 27.23%, 8.03it/s Index 8, 8, 8  Cache Rate 27.23%, 8.61it/s Index 0, 0, 5 ...
@FurkanGozukara yes, its quality is degraded in xl type models - it requires more experiments or.. maybe re-implementation. It did not happen to 1.5-types though.
@gel-crabs I fixed some explanations - for in types, it applies to after-index, thus -1-> all caching For out types, it applies to before-index, thus 9-> all Timestep is kinda...
I should also explain why quality gets degraded even when we use less cache than all caching- its about input-output mismatching. To summarize, there are corresponding pairs to each caches,...
New implementation - should be tested though https://github.com/aria1th/sd-webui-deepcache-standalone ## SD 1.5 512x704 test, with 40% disable for initial steps ``` Steps: 23, Sampler: DPM++ SDE Karras, CFG scale: 8, Seed:...