master-9578fdc

chore: remove rocm5.5 build temporarily

master-b5f4932

refactor: add some sd vesion helper functions

master-9b1d90b

fix: improve clip text_projection support (#397)

master-8f94efa

feat: add support for loading F8_E5M2 weights (#460)

master-8c7719f

fix: typo in clip-g encoder arg (#472)

master-6ea8122

feat: add flux 1 lite 8B (freepik) support (#474)

* Flux Lite (Freepik) support

* format code

---------

Co-authored-by: leejet <[email protected]>

master-65fa646

feat: add sd3.5 medium and skip layer guidance support (#451)

* mmdit-x

* add support for sd3.5 medium

* add skip layer guidance support (mmdit only)

* ignore slg if slg_scale is zero (optimization)

* init out_skip once

* slg support for flux (expermiental)

* warn if version doesn't support slg

* refactor slg cli args

* set default slg_scale to 0 (oops)

* format code

---------

Co-authored-by: leejet <[email protected]>

master-2b1bc06

feat: add PhotoMaker Version 2 support (#358)

* first attempt at updating to photomaker v2

* continue adding photomaker v2 modules

* finishing the last few pieces for photomaker v2; id_embeds need to be done by a manual step and pass as an input file

* added a name converter for Photomaker V2; build ok

* more debugging underway

* failing at cuda mat_mul

* updated chunk_half to be more efficient; redo feedforward

* fixed a bug: carefully using ggml_view_4d to get chunks of a tensor; strides need to be recalculated or set properly; still failing at soft_max cuda op

* redo weight calculation and weight*v

* fixed a bug now Photomaker V2 kinds of working

* add python script for face detection (Photomaker V2 needs)

* updated readme for photomaker

* fixed a bug causing PMV1 crashing; both V1 and V2 work

* fixed clean_input_ids for PMV2

* fixed a double counting bug in tokenize_with_trigger_token

* updated photomaker readme

* removed some commented code

* improved reconstructing class word free prompt

* changed reading id_embed to raw binary using existing load tensor function; this is more efficient than using model load and also makes it easier to work with sd server

* minor clean up

---------

Co-authored-by: bssrdf <[email protected]>

master-1c168d9

fix: repair flash attention support (#386)

* repair flash attention in _ext
this does not fix the currently broken fa behind the define, which is only used by VAE

Co-authored-by: FSSRepo <[email protected]>

* make flash attention in the diffusion model a runtime flag
no support for sd3 or video

* remove old flash attention option and switch vae over to attn_ext

* update docs

* format code

---------

Co-authored-by: FSSRepo <[email protected]>
Co-authored-by: leejet <[email protected]>

master-ac54e00

feat: add sd3.5 support (#445)