unsloth/studio
Roland Tannous 1c3b3b36bd Windows: pick PyTorch wheel from installed CUDA Toolkit, not driver max
nvidia-smi reports the driver's max-supported CUDA, which can outrun the
userland — picking cu130 off a "CUDA 13.0" driver with a 12.9 toolkit ships
a torch that silently falls back to CPU. Resolve the wheel from the actual
toolkit version instead, winget-installing the highest <= driver max if no
toolkit is present.
2026-05-18 11:53:14 +04:00
..
backend studio: extend offline DNS auto-detect to inference parent + training (#5512) 2026-05-18 00:31:33 -07:00
frontend studio/frontend: wire logout, singleflight refresh, shared 422 helper, current-password input (#5490) 2026-05-18 00:03:03 -07:00
src-tauri studio: expose launcher capability bits on unauth /api/health (#5486) 2026-05-17 21:29:24 -07:00
__init__.py Final cleanup 2026-03-12 18:28:04 +00:00
install_llama_prebuilt.py tests/studio: lock in Windows GPU detection fix (#5106) with a synthetic CI test (#5376) 2026-05-18 00:06:01 -07:00
install_python_stack.py studio: skip flash-attn install on Blackwell GPUs (sm_100+) (#5420) 2026-05-14 18:13:50 +04:00
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 Windows: pick PyTorch wheel from installed CUDA Toolkit, not driver max 2026-05-18 11:53:14 +04:00
setup.sh Route CPU-only Linux x86_64 to ggml-org/llama.cpp prebuilts (#5302) 2026-05-05 23:22:22 -07:00
Unsloth_Studio_Colab.ipynb studio: add --local to setup.sh + overlay unsloth-zoo from git main (#5252) 2026-05-02 08:51:56 +04:00