koboldcpp/kcpp_src
Wagner Bruna aa034e1004
sd: sync with master-975-95b3090 (#2542)
* sd: sync with master-968-8cd731e

* feat: add TAEQI2.1 support for Qwen Image 2.1
* feat: add MoSR upscaler support
* feat: add Qwen Image 2.1 Turbo sigma preset
* fix: reclaim before accepting wrapped Vulkan iGPU memory estimates
* feat: add per-module automatic flash attention selection
* feat: add automatic direct convolution selection
* chore: move parameter and tokenizer diagnostics to debug
* fix: prevent core dump in CPU backend if using FP8 (e4m3) tensors
* fix: benchmark Vulkan direct convolution in auto mode
* feat: add FLUX.3 Image Klein and Klein Base support

* sd: update Makefile, fix flash attention workaround

* sd: add support for Flux 3

* sd: sync with master-973-f0aac52

* fix: open model files with UTF-8 paths on Windows
* fix: handle UTF-8 file paths on Windows
* chore: sync ggml and remove obsolete conv1d workarounds
* feat: add moss-image support

* sd: add support for MOSS-Image

* sd: sync with master-975-95b3090

* feat: support Diffusers-format MiniMax-H3 LoRAs
* feat: support MOSS split QKV weights and batch host loading

* sd: set conv_direct through the parameter list string
2026-10-20 17:35:33 +08:00
..
bin refactor: cleanup project root 3 2026-09-29 00:08:43 +08:00
lib refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
packaging relocate environment files for conda 2026-09-29 00:13:46 +08:00
expose.cpp refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
expose.h sd: sync with master-975-95b3090 (#2542) 2026-10-20 17:35:33 +08:00
gpttype_adapter.cpp improve kv allocation when using parallelrequests, allocates more kv but in a diminishing way 2026-10-20 00:20:25 +08:00
hash.cpp refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
json_to_gbnf.py refactor: cleanup project root 3 2026-09-29 00:08:43 +08:00
kcpp_backend.cpp refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
kcpp_backend.h refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
model_adapter.cpp refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
model_adapter.h refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00
README.md relocate environment files for conda 2026-09-29 00:13:46 +08:00
simplecpuinfo.cpp refactor: cleanup project root 1 2026-09-28 23:53:06 +08:00

KoboldCpp native sources

This directory contains KoboldCpp-owned implementation and build-support code. Upstream-synced llama.cpp and ggml sources remain in their original directories so upstream updates can continue to be merged without path-only conflicts.

Prebuilt helper programs and link inputs used by KoboldCpp builds live under bin/ and lib/, respectively. Release version metadata and its generation scripts live under packaging/; Conda environment definitions are kept in packaging/environments/.