domains / families / science / nv-segment-ct-finetune

nv-segment-ct-finetune

separate

Runs standard or fixed-channel softmax finetuning of NV-Segment-CT VISTA3D on CT NIfTI image/label datasets, with optional MONAI-native MLflow tracking and checkpoint evidence. Uses softmax for predefined, mutually exclusive classes; keeps the standard workflow when point prompts or runtime-variable classes are needed. Not for clinical validation.

When to use

• Used for smoke or dataset finetuning of NV-Segment-CT VISTA3D on CT NIfTI labels, including the upstream fixed-channel softmax workflow and optional MLflow tracking. Not for clinical validation.
• Wraps the upstream MONAI bundle entrypoint; do not replace it with handwritten training or inference code.
• Manifest inputs are `dataset_dir`, `datalist`, `target_anatomy`, `label_mapping`, `smoke`, `sanity`, `auto_seg`, `softmax`, `skip_formal_eval`, `mlflow_tracking_uri`, `mlflow_experiment_name`, and `mlflow_run_name`.
• Manifest outputs are `finetuned_ckpt` and schema-checked `result_json`.

When not to use

• Thin wrapper. Training, validation, transforms, and checkpointing are delegated to the upstream bundle in `bundle/`.
• Tensor comparison uses restricted `weights_only=True` checkpoint loading. Unsupported serialized objects produce a comparison error; they are not retried with unrestricted pickle loading. This comparison is not a security audit of the upst…
• Reproduction record only: the successful five-epoch Task06 run used Python
• The auto-derived plan is heuristic; caller-provided `--patch-size`, `--cache-rate`, `--epochs`, and `--learning-rate` win.
• `--softmax` is not compatible with `--sanity`: the Task06 reference scores and original-spacing pre/post evaluation belong to the standard VISTA3D continual-learning workflow. Softmax runs record the training validation trajectory but need…
• The Task06 sanity recipe intentionally forces single-GPU execution to match the DFW reference. Multi-GPU mode for other datasets requires host `torchrun` support.
• The paired verifier is CPU-only and audits the evidence pack; it does not re-run GPU segmentation.
• MLflow support is optional and uses MONAI's built-in tracking handlers. Tracking errors are part of the upstream MONAI run and can therefore fail the finetune command.

Skill page

Source