mirror of
https://github.com/LostRuins/koboldcpp.git
synced 2026-10-23 14:50:00 +00:00
* model: add native lazy-loading and MoE cache inputs * model: propagate lazy mode and target-only MoE cache * launcher: expose lazy tensors and MoE cache in GUI, CLI and configs * tests: cover memory option defaults, validation and config round trips * docs: explain lazy tensor loading and experimental expert cache * docs: trim trailing blank line * Delete tests/test_model_memory_options.py * docs: clarify temporary validation checks were removed * logging: keep scheduler reallocation traces in verbose mode * model: append new memory fields to native input struct * launcher: append memory fields in matching ctypes order * docs: remove guide already covered upstream |
||
|---|---|---|
| .. | ||
| bin | ||
| lib | ||
| packaging | ||
| expose.cpp | ||
| expose.h | ||
| gpttype_adapter.cpp | ||
| hash.cpp | ||
| json_to_gbnf.py | ||
| kcpp_backend.cpp | ||
| kcpp_backend.h | ||
| model_adapter.cpp | ||
| model_adapter.h | ||
| README.md | ||
| simplecpuinfo.cpp | ||
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/.