feat: add support for multiple model files

[anhappdev]

• Closes https://github.com/mlcommons/mobile_app_open/issues/902
• The function signature will stay the same:

  mlperf_backend_ptr_t mlperf_backend_create(
  const char *model_path, 
  mlperf_backend_configuration_t *configs,
  const char *native_lib_path
  )

• Here is an example how to have multiple models:

  delegate_choice: {
  delegate_name: "NNAPI"
  accelerator_name: "npu"
  accelerator_desc: "NPU"
  model_file: {
    model_path: "https://file1.tflite"
    model_checksum: "abc"
  }
  model_file: {
    model_path: "https://file2.tflite"
    model_checksum: "def"
  }
  }

  • If there is only one model file, the model_path will point to the model file like before.
  • If there are multiple model files, the model_path will point to a single directory where all files are located in there.

[github-actions[bot]]

MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅

[sonarcloud[bot]]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud

[anhappdev]

@freedomtan @mohitmundhragithub @AhmedTElthakeb Please take a look at the first comment to know how to define multiple models in the backend settings.