* sd: sync to master-509-4cdfff5
* sd: Anima support
* sd: sync to master-514-5792c66
* sd: additional workaround for Anima .safetensors model
* sd: sync to master-517-ba35dd7
* sd: sync to master-520-d950627
* sd: sync to master-504-636d3cb
* sd: fix and simplify limit calculation
- restore the "arbitrarily high" 8192 limit, since it's used to turn
off the img_hard_limit (and if each side was always limited by 2048,
we wouldn't need hard_megapixel_res_limit)
- avoid changing the config cfg_square_limit during a generation
- apply the hard_megapixel_res_limit only in the configuration-changed
path, since the default path uses constants
- clean up comments
The calculation itself remains the same:
- default area limit: 832² for SD1.5/SD2, 1024² otherwise
- configured limit always between 64 and 2048
* sd: sync to master-427-78e15bd
* add kl_optimal to the available schedulers list
* more robust workaround to avoid stb linkage issues
* sd: sync to master-431-23fce0b
* add TAEHV support and disable TAE if the model isn't found
* sd: clean up changes against stable-diffusion.cpp 90ef5f8
Clean up the diff, and include a few missing changes, mainly from
the upscaler and model weight type statistics.
* added line clear again
* remove excess spaces
---------
Co-authored-by: LostRuins Concedo <39025047+LostRuins@users.noreply.github.com>
Using KCPP_BAKE_SD_VOCAB to turn off the change to not embed the
vocabulary files makes testing new upstream merges harder, because
we then need to set that macro on the sd.cpp original build.
So, revert the tests, making the define turn the change on. Also,
since model.cpp is always built by Koboldcpp as part of the
sdtype_adapter.cpp, it's enough to set the macro on that file.
* Update stable-diffusion.cpp to 5900ef6605c6 (new API)
* Clean up pending LoRA code and simplify LoRA changes to upstream
* Move VAE tiling disabling for TAESD to sdtype_adapter.cpp
* Move auxiliary ctx functions to sdtype_adapter.cpp
* Use ref_images parameter for Kontext images
* Drop clip skip workaround (fixed upstream)
* Workaround for flash attention with img2img
leejet/stable-diffusion.cpp#756
* Workaround for Chroma with flash attention, debug prints
* Disable forcing CLIP weights to F32 for reduced memory usage
