executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
runtime-opsRoute Det3D installation, import, CUDA extension, PyTorch ABI, optional dependency, distributed-runtime, and checkpoint filesystem troubleshooting.
Independent GitHub snapshot · Automated classification · Original documentation governs
Use this route whenever Det3D fails during installation, package import, CUDA
operator loading/building, distributed initialization, or filesystem setup.
Run scripts/check_runtime.py first; it is read-only and reports JSON.
nvcc, compiler, extension, and spconv availability.Do not call a CPU import a successful GPU capability check. Do not repair a user-owned environment by silently changing torch, setuptools, CUDA, or compiled packages. Training/evaluation and data conversion remain separate side-effectful workflows.