Using KonfAI Apps¶
KonfAI Apps are packaged workflows that expose a stable user interface on top of KonfAI’s low-level prediction, evaluation, uncertainty, and fine-tuning logic.
Use apps when the low-level YAML workflow is already stable and you want a smaller, more repeatable interface for end users.
The konfai-apps CLI¶
The app CLI currently exposes these subcommands:
inferevaluncertaintypipelinefine-tune
The main command pattern is:
.. code-block:: bash
konfai-apps
App identifiers¶
Apps can be local or remote repository identifiers. The repository examples and tests show Hugging Face style identifiers such as:
VBoussot/ImpactSynth:MRVBoussot/ImpactSynth:CBCTVBoussot/TotalSegmentator-KonfAI:total
Common app workflows¶
These commands all use the same app package, but they expose different levels of workflow orchestration.
Inference:
.. code-block:: bash
konfai-apps infer VBoussot/ImpactSynth:CBCT
-i input.mha -o ./Output –gpu 0
Evaluation:
.. code-block:: bash
konfai-apps eval VBoussot/ImpactSynth:CBCT
-i prediction.mha –gt ct.mha –mask mask.mha –gpu 0
Pipeline:
.. code-block:: bash
konfai-apps pipeline VBoussot/ImpactSynth:CBCT
-i input.mha –gt ct.mha –mask mask.mha -o ./Output -uncertainty
Grouped inputs¶
The CLI accepts grouped inputs by repeating --inputs / -i. This matches the
grouping behavior documented in konfai_apps.cli.add_common_konfai_apps().
Use this when an app expects:
multiple input groups
multiple files per group
paired inputs such as image + mask
Fine-tuning¶
Fine-tuning is available through:
.. code-block:: bash
konfai-apps fine-tune
Under the hood, the app installs training assets, links the dataset, then calls the low-level training flow in resume mode.
Local vs remote¶
If you add --host, the same command switches from local execution to
client/server mode automatically. The CLI still looks the same; only the
execution backend changes.
See also¶
- doc:
remote-server
- doc:
../concepts/apps
- doc:
../reference/cli