We currently develop a Glow backend for CEVA target DSP that supports AOT (offline) compilation mode.
During the bundle creation stage, our backend generates metadata which is required by the target DSP during runtime.
We need to access this data from the main application on the ARM host before it calls the bundle entry point to start the actual Glow inference process.
Is there an existing mechanism in Glow that allows the backend to store proprietary raw data inside the bundle or the weights file, and then access it during inference initialization?
If not, is it possible then for the backend to write the data to a “private” file and then open it for read during runtime?