Skip to main content

Conversion Outputs and Diagnostics

Overview​

This repository is a converter plugin. Its job is to accept a supported input model, produce a conversion-stage artifact, and hand that artifact to a later MLIA backend. That means the important outputs here are conversion results and diagnostics, not final target metrics.

What this converter contributes​

In a successful workflow, this converter contributes:

  • A TOSA-oriented intermediate representation derived from a .tflite input.
  • Either a .tosamlir text artifact or .tosa.mlirbc bytecode artifact, depending on the requested output format.
  • Conversion-stage logs or diagnostics during the MLIA run.
  • Artifacts that a downstream backend can consume for further analysis.

What success looks like​

A successful end-to-end run usually looks like this:

  1. The .tflite model is accepted.
  2. The converter produces an intermediate artifact.
  3. A downstream backend consumes that artifact.
  4. The final user-facing metrics come from the downstream backend.

What this repo does not own​

This repo does not define the final performance or target-level compatibility metrics for a run. If you are looking for values such as cycle counts, memory figures, or target-specific compatibility summaries, those belong to the backend that runs after conversion.

Useful signals to look for​

The most useful signals in this repo are usually:

  • Whether the .tflite model was accepted and converted.
  • Whether the produced artifact could be consumed by the next backend.
  • Whether legalization or conversion failed before downstream analysis started.
  • Whether the failure happened during conversion or after conversion completed.

How to interpret failures​

If conversion fails, investigate the model or conversion path first. If conversion succeeds and later analysis fails, move the investigation to the downstream target or estimator backend.