Skip to content

INSTALL.md

Latest commit

 

History

History
25 lines (19 loc) · 896 Bytes

INSTALL.md

File metadata and controls

25 lines (19 loc) · 896 Bytes
Raw

Installing python3-flagquantum

After apt install python3-flagquantum (Debian/Ubuntu) or dnf install python3-flagquantum (Fedora), install the ML runtime separately — it is intentionally not declared as a hard Depends/Requires because the distro versions are CPU-only (torch) or too old (triton) for GPU workloads.

FlagQuantum needs PyTorch (GPU build) at runtime:

pip install --index-url https://download.pytorch.org/whl/cu128 torch==2.9.0+cu128

A venv is recommended to isolate pip-installed packages from the system Python:

python3 -m venv ~/.venv/flagos
source ~/.venv/flagos/bin/activate
# then the pip install lines above

Note: the distro python3-torch package (CPU-only build) is intentionally not pulled in — FlagOS workloads need a GPU build, which PyTorch upstream distributes via PyPI (per-CUDA-version wheels), not as a .deb / .rpm.