transform-spec-validator

To validate a custom transform spec, you must ensure the Python file defines a string LIBRARY, a callable __call__, and a TRANSFORMS list of dictionaries with name and transform entries. This verifies compatibility before running image or video pipeline benchmarks.

The TRANSFORMS structure in a benchmark specification must be a list of dictionaries. Each dictionary within this list is required to contain both a name entry and a transform entry to properly define custom image or video transformations.

Yes, validating transform specifications requires a Python environment. The validator explicitly checks that the __call__ attribute is defined and callable, which necessitates an active Python runtime to evaluate custom transform definitions for benchmarking pipelines.

Your custom transform spec fails compatibility checks if LIBRARY is missing or not a string, if __call__ is not callable, or if TRANSFORMS lacks the required name and transform dictionary entries. The validator ensures correct structure before running benchmarks.

Yes, you can validate image and video pipeline transforms before running benchmarks. The validator checks your specification file to ensure developers verify required attributes like callable __call__ definitions and correct TRANSFORMS structures prior to execution.