ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
unsloth/studio
History
Daniel Han d7a78599e6 studio: optimize llama.cpp build with static linking, ccache, ninja 
Benchmarked BUILD_SHARED_LIBS ON vs OFF:
- Build time identical (~55s both on 192-core B200)
- Static: 75MB self-contained binary, no LD_LIBRARY_PATH needed
- Shared: 7.4MB + .so deps, needs lib path management
- Both work correctly

Optimizations applied to LlamaCppBuilder:
- BUILD_SHARED_LIBS=OFF: static binary, simpler deployment
- ccache: 27x faster rebuilds (55s -> 2s with warm cache)
- Ninja generator: parallel builds
- CMAKE_CUDA_ARCHITECTURES: build only for detected GPU arch
- CMAKE_CUDA_FLAGS=--threads=0: multi-threaded CUDA compilation
- GGML_NATIVE=ON: native CPU optimizations
- LLAMA_BUILD_TESTS=OFF, LLAMA_BUILD_EXAMPLES=OFF: skip unused targets
- LLAMA_BUILD_SERVER=ON: build only what we need
2026-03-16 14:15:14 +00:00
..
backend studio: optimize llama.cpp build with static linking, ccache, ninja 2026-03-16 14:15:14 +00:00
frontend studio: training progress visibility + deferred llama.cpp compilation 2026-03-16 14:02:20 +00:00
__init__.py Final cleanup 2026-03-12 18:28:04 +00:00
install_python_stack.py [pre-commit.ci] auto fixes from pre-commit.com hooks 2026-03-14 00:54:09 -07:00
LICENSE.AGPL-3.0 Add AGPL-3.0 license to studio folder 2026-03-09 19:36:25 +00:00
setup.bat Final cleanup 2026-03-12 18:28:04 +00:00
setup.ps1 PR: Windows Setup Improvements (#4299) 2026-03-14 23:59:49 +04:00
setup.sh Fix setup.sh crash on Mac with empty gitignore array 2026-03-15 22:33:04 -07:00
Unsloth_Studio_Colab.ipynb Fix/colab plugin editable install (#4281) 2026-03-16 01:34:37 +04:00