Spandrel is installed but the model will not load: check architecture support

Installation is only the first checkpoint

Spandrel supports a defined set of model architectures, not every PyTorch model. Its upstream documentation distinguishes the base package from additional architectures and their licensing conditions.1

python -m pip show spandrel spandrel-extra-arches

Our suggested report combines that package record with the calling node revision and the model's original release, filename and revision. A supported extension such as .pth is not an architecture guarantee.

A missing import calls for dependency checks. An unrecognized model calls for a support check. Unexpected weight keys call for file-identity and model-structure investigation. Renaming a file is not the proposed solution.

Install additional architectures only when the task requires them and review the relevant model and architecture permissions separately. Test a small, known-source model supported by the selected version. Record import, architecture recognition and output as different outcomes; do not replace a model-specific failure with an indiscriminate environment reinstall.

Related checks

• Kornia import errors: check the base package, Torch and optional extras
• Missing safetensors package or unreadable weights? Diagnose them separately

Research draft. No local/GPU reproduction or independent editorial sign-off. Upstream statements and our recommendations are distinguished.