Data & AI skills

Prepare a public CodonFM v1 environment and download supported Encodon checkpoints, the files containing trained model weights. Covers container or host setup, GPU access checks, and verification of downloaded weights and configuration files.

• Prepare a CodonFM development container
• Set up CodonFM directly on the host
• Download public Encodon model weights

Validate coding-sequence CSV files and extract embeddings, numerical features representing each sequence, with public CodonFM/Encodon models. Prepare extraction commands, choose suitable checkpoints, and verify outputs for downstream modeling of translation efficiency, expression, or mRNA stability.

• Validate coding-sequence CSV files before extraction
• Extract numerical sequence features with Encodon
• Choose a checkpoint for downstream property modeling

Guides fine-tuning public Encodon models on labeled protein-coding sequences or variants. Supports numerical prediction and classification, with data checks, LoRA adapters that train selected parameters, head-only training, or full training, followed by checkpoint and metric checks.

• Fine-tuning Encodon for sequence-level numerical prediction
• Training a classifier on labeled coding variants
• Preparing training data and a launch command

Prepare or run variant scoring with public CodonFM Encodon models. Validate coding sequences and codon positions in CSV inputs, assemble the scoring command, and check output arrays that compare reference and alternate codon probabilities in context.

• Validating a variant CSV before scoring
• Preparing an Encodon scoring command
• Running scoring and checking output arrays

Fine-tune NV-Segment-CT VISTA3D on CT images and region labels in NIfTI format. Choose standard training or fixed-class softmax training, save model checkpoints and evaluation records, and optionally track training with MLflow.

• Fine-tuning with paired CT images and region labels
• Checking the training workflow with a smoke run
• Training mutually exclusive classes with softmax

Provides shared guidance and templates used by NVFLARE conversion skills. Supports consistent dependency setup, site-specific data paths, validation, and custom model aggregation, combining results from participating sites. It is an internal companion rather than a standalone converter.

• Consulting dependency setup rules during conversion
• Checking data paths for participating sites
• Following shared validation and reporting guidance

Convert existing PyTorch Lightning training code into an NVFLARE job for federated learning, where sites train collaboratively while keeping data local. It integrates model exchange through the Lightning Client API and supports local validation or job export.

• Converting Lightning training code to federated learning
• Preparing collaborative training with data kept at each site
• Checking a converted job in a local simulation

Turn records from a stopped NVFLARE Auto-FL optimization campaign into a reproducible Markdown report and JSON summary. Summarize improvements, retained candidates, failed ideas, and literature outcomes, with a refreshed progress plot when plotting dependencies are available.

• Create a final report after an optimization campaign stops
• Summarize baseline scores and retained improvements
• Review successful and unsuccessful literature-based ideas

Fine-tune the MAISI VAE, a model that compresses and reconstructs CT or MRI volumes, using NIfTI datasets. Prepare training settings, check inputs before training, and save model checkpoints and loss histories for review.

• Check a NIfTI dataset before GPU training
• Fine-tune a VAE with CT or MRI volumes
• Compare reconstruction losses across training runs

Compute statistics across distributed NVFLARE sites without sharing raw data. Generate and validate jobs for numeric tables or image datasets, then report site-level and global summaries, missing data rates, and applied privacy settings.

• Comparing numeric data distributions across sites
• Computing means, variances, and quantiles across sites
• Summarizing image counts and pixel intensities

Generate synthetic 3D brain MRI volumes with NV-Generate-CTMR MR-Brain v1, choosing scan types such as T1 or FLAIR. The wrapper records generation settings and output checks for experimental review, with restrictions on clinical and production use.

• Generating synthetic T1 or T2 brain MRI volumes
• Trying FLAIR or SWI generation settings
• Specifying a random seed for generation

Generate synthetic CT volumes with paired masks that mark anatomical regions or tumors. Configure body regions and output dimensions, then inspect image slices, label mappings, and recorded generation details for synthetic research workflows.

• Generate synthetic chest CT volumes with lung masks
• Explore tumor size settings in synthetic CT generation
• Preview configuration and resource estimates before generation

Generate synthetic body MRI volumes, or three-dimensional images, with NV-Generate-CTMR rflow-mr. The skill stages configuration settings, runs the existing inference workflow, and records output geometry and data checks alongside the generated images.

• Generate synthetic body MRI volumes for experimentation
• Try supported MRI modalities such as T1 and T2
• Compare outputs with different settings and random seeds

Find PhysicsNeMo models, data loading pipelines, and examples for machine learning tasks involving physical systems. The skill checks current repository files, presents relevant alternatives, and suggests a reading order without writing training code.

• Finding model families for physical forecasting
• Comparing models for grids and unstructured meshes
• Finding data loading pipelines for a dataset

Run NV-Reason-CT on a supplied 3D chest or abdominal CT scan and a text prompt for research or engineering. Returns generated text and JSON records of input geometry, model identity, dependencies, and limitations.

• Testing inference on a chest CT scan
• Running abdominal CT inference with a focused question
• Checking local model assets and runtime readiness

Evaluate object detection predictions against labeled ground truth with TAO Data Services. Produce class-level precision, recall, accuracy, and AP, a detection quality measure, with CSV results and precision–recall plots for reviewing model performance.

• Score object detection predictions against ground truth
• Review precision and recall by class
• Generate metrics CSV files and precision–recall plots

Deploy and manage NVIDIA RAG Blueprint, a system that retrieves information to support AI answers. Check the running environment, configure models and search features, troubleshoot unhealthy services, and verify affected workflows after changes.

• Deploy NVIDIA RAG Blueprint
• Configure search and document ingestion
• Change models or enable image features

Run NV-Segment-CTMR to divide CT or MRI NIfTI volumes into labeled regions. Save the label map and records of region volumes, processing time, and geometry checks for technical verification, without clinical interpretation.

• Automatically segmenting CT body volumes
• Segmenting body MRI or preprocessed brain MRI
• Requesting regions by specific label IDs

Coordinate defect image generation, augmentation, and labeling for circuit board, metal surface, and glass inspection on OSMO. Select workflows that create initial board datasets from CAD scenes, process real images, or fine-tune NVIDIA Cosmos AnomalyGen models.

• Creating initial circuit board inspection datasets from CAD scenes
• Generating circuit board texture and component pose defects
• Running inference and labeling on real inspection images

Convert existing Hugging Face Trainer or TRL SFTTrainer code into an NVFLARE job for federated learning, where multiple sites train a shared model. Preserve training settings, validate locally, and export the resulting job.

• Converting Trainer-based training into a federated job
• Adapting PEFT or LoRA fine-tuning for federated learning
• Preserving evaluation metrics during conversion

Optimize an existing NVFLARE federated learning job, which trains across participating sites, through isolated candidate changes. Compare results under consistent training budgets and metric definitions, while preserving the best source and recording changes, scores, and artifacts.

• Compare changes to an existing federated learning job
• Evaluate two approaches under a fixed training budget
• Resume optimization using recorded candidates and results

Run video augmentation, which creates variations of source videos, and automatic labeling on OSMO. Select a workflow, check credentials and resources, monitor execution, download outputs, and compare augmented videos with their originals.

• Automatically label source videos
• Create video variations and label the outputs
• Process original and augmented videos in one workflow

Replace the object detector in VSS Alerts Blueprint verification mode with an ONNX model. This guide covers model placement, interpreting detection outputs, DeepStream configuration, TensorRT engine builds, redeployment, and common loading or mount failures.

• Replace the verification detector with a YOLO11 ONNX model
• Investigate DeepStream custom parser loading failures
• Fix ONNX mounts and TensorRT engine rebuild settings

Segment specified organs in CT NIfTI scans using NV-Segment-CT VISTA3D. Produces a label map identifying organ regions and JSON evidence covering labels, voxel counts, physical volumes, geometry checks, and model identity for engineering validation.

• Segmenting the liver, spleen, and both kidneys in CT scans
• Recording voxel counts and physical volumes for each organ
• Checking output geometry and unexpected labels