数据与 AI技能

Run an Isaac for Healthcare workflow from robot recordings through dataset preparation, policy training, and checkpoint validation. Uses the maintained pipeline driver, checks the planned commands first, and verifies artifacts and logs at each stage.

中文简介正在准备中。当前显示英语简介。

• Running the full recording-to-validation pipeline
• Checking planned stages and artifact paths before execution
• Reviewing datasets, training logs, and checkpoints

Fine-tune GR00T or openpi robot policies using compatible LeRobot data. Check dataset requirements, preview the training configuration, run training, and verify that a saved checkpoint, a file containing the trained model, loads through the inference backend.

中文简介正在准备中。当前显示英语简介。

• Fine-tuning a robot policy on LeRobot data
• Checking dataset compatibility before training
• Training with specified steps, batch size, and GPU count

Prepare BOP datasets and run depth estimation, image segmentation, and FoundationPose to estimate object positions and orientations. Evaluate completed runs, compare accuracy and coverage, or change scoring settings without repeating inference.

中文简介正在准备中。当前显示英语简介。

• Adapt a BOP dataset and test a single scene
• Estimate object poses without ground-truth annotations
• Re-score a completed run while preserving its original report

Grade recorded workflow episodes against a written success criterion using a vision model, which interprets sampled camera images. Summarize pass and fail counts, or optionally extract successful episodes into a separate HDF5 dataset.

中文简介正在准备中。当前显示英语简介。

• Label recorded episodes against a written success criterion
• Extract successful episodes into a separate dataset
• Check sampled camera frames before sending images

Find suitable Earth2Studio sources for weather and climate variables at specified times, verify variable support, and generate a Python script that retrieves historical or forecast data as an xarray DataArray, an array with labeled dimensions.

中文简介正在准备中。当前显示英语简介。

• Retrieve ERA5 temperature data for a specific date
• Fetch forecasts for selected lead times
• Compare sources by coverage and resolution

Replay an episode stored in HDF5, a structured data file, using its original workflow Scene. Check the full recorded motion visually, inspect segment boundaries and camera views, and report agreement or the first mismatch.

中文简介正在准备中。当前显示英语简介。

• Visually verify a recorded motion trajectory
• Replay a specific episode from an HDF5 recording
• Inspect camera views and segment boundaries

Record human-operated robot demonstrations in workflow HDF5 files, a structured data format. The skill guides supported input-device selection, foreground recording, and checks of robot motion, camera observations, saved episodes, and task outcomes.

中文简介正在准备中。当前显示英语简介。

• Recording keyboard-controlled robot demonstrations
• Recording demonstrations with supported leader or VR devices
• Checking saved actions and camera observations

Expand successful workflow recordings stored in HDF5 by copying episodes and adding small variations to recorded actions. Limit changes to a requested node segment and check generated counts, data dimensions, success flags, and preserved segments.

中文简介正在准备中。当前显示英语简介。

• Creating additional variants from successful recordings
• Applying action variations only within a specified node
• Checking episode counts and preserved segment information

Build Earth2Studio wrappers for transformations at a single time point, from calculating wind speed to using pretrained or diffusion models. The workflow covers input and output coordinates, model loading, tests, and supporting documentation.

中文简介正在准备中。当前显示英语简介。

• Calculate wind speed from u10m and v10m
• Add a pretrained precipitation estimator
• Wrap a diffusion model for higher-resolution output

Find Earth2Studio models, data sources, and examples for weather or climate tasks. Filters options by region, forecast type, and GPU memory, then checks whether data sources supply the variables required by each model using current documentation and source files.

中文简介正在准备中。当前显示英语简介。

• Choosing weather models for a region and forecast horizon
• Comparing models against available GPU memory
• Checking model and data source variable compatibility

Convert recorded robot workflows from HDF5 into local LeRobot datasets for training or browser inspection. Preserves actions, state, camera recordings, and task labels, then checks episode counts, metadata, videos, and data dimensions.

中文简介正在准备中。当前显示英语简介。

• Convert successful robot episodes into a training dataset
• Prepare recorded workflows for browser inspection
• Check converted videos, metadata, and data dimensions

Open a completed LeRobot dataset in a browser to inspect camera videos, state and action timelines, and episode details. The skill manages a local server and reports observations together with the command to stop it.

中文简介正在准备中。当前显示英语简介。

• Inspect the latest converted LeRobot dataset
• Check that every camera video plays
• Review state and action timelines

Create a new data source wrapper, code that connects remote weather data to Earth2Studio. The skill covers source selection, variable mapping, implementation, tests, and validation with real data, with further steps for documentation and pull request submission.

中文简介正在准备中。当前显示英语简介。

• Connect a new remote data store to Earth2Studio
• Add support for gridded forecasts or station observations
• Map remote variable names to Earth2Studio names

Build Python scripts for a single weather forecast with Earth2Studio. Choose a model, compatible input data, and an output format based on the forecast period, region, requested variables, and available GPU memory.

中文简介正在准备中。当前显示英语简介。

• Build a script for a five-day weather forecast
• Save only temperature and wind variables
• Choose a weather model that fits available GPU memory

Prepare an environment for evaluating speech recognition on clinical terms. Checks the API key and Python dependencies, then tests NVIDIA's hosted services by generating speech from one synthetic sentence and transcribing it before the next stage.

中文简介正在准备中。当前显示英语简介。

• Preparing a clinical speech recognition evaluation
• Checking API access from a fresh environment
• Testing speech synthesis and transcription together

Integrate a supported object detection model from HuggingFace or NVIDIA NGC into NVIDIA DeepStream. Automates model acquisition and conversion, TensorRT engine building, detection output parsing, multi-stream performance benchmarks, and reports with charts in Markdown, HTML, and PDF.

中文简介正在准备中。当前显示英语简介。

• Run the default RT-DETR model in DeepStream
• Integrate a supported HuggingFace or NVIDIA NGC detection model
• Compare performance across multiple video streams

Predict the structure of a single protein chain from its amino acid sequence using OpenFold2 NIM. Run through NVIDIA's hosted API or local Docker, optionally add sequence alignments and structural templates, and save predictions with confidence information.

中文简介正在准备中。当前显示英语简介。

• Predicting a single protein chain's structure from its sequence
• Adding A3M sequence alignments or mmCIF structural templates
• Comparing predictions from multiple model parameter sets

Explore competing goals such as cost and service quality through repeated cuOpt solves. Organize efficient alternatives into a Pareto frontier, a set of tradeoff choices, and explain what each improvement costs in another objective.

中文简介正在准备中。当前显示英语简介。

• Compare cost and service quality
• Explore return and risk tradeoffs
• Compare delivery distance and vehicle count

Turn molecules listed in a SMILES CSV into numerical feature vectors using a KERMT model. The workflow checks the model and input, cleans molecular notation, and saves four sets of vectors with validity information.

中文简介正在准备中。当前显示英语简介。

• Extract molecular feature vectors from a SMILES CSV
• Reuse a local KERMT checkpoint
• Extract features with a released pretrained model

Calibrate live cameras by recording their RTSP video streams through VIOS and processing the clips with AutoMagicCalib. The workflow imports settings and alignment files, checks readiness, tracks calibration progress, and retrieves results.

中文简介正在准备中。当前显示英语简介。

• Calibrate cameras from live RTSP streams
• Import settings and alignment files for a camera scene
• Retrieve evaluation statistics using ground truth data

Monitor KERMT jobs running in the background by reading run records, Docker container status, and logs. Summarize training progress and validation metrics, stream logs, or report final test metrics when a finetune run completes successfully.

中文简介正在准备中。当前显示英语简介。

• Check progress on a background pretraining run
• Watch training logs and validation metrics live
• Review test metrics after finetuning completes

Predict protein structures by connecting MSA-Search with OpenFold3. It finds related sequences, aligns them to provide evolutionary context, and saves predicted structures with confidence scores. It also supports comparing predictions with and without that context.

中文简介正在准备中。当前显示英语简介。

• Predict protein structures using related sequences
• Save alignments, structures, and confidence scores
• Compare predictions with and without alignment context

Predict structures of proteins, DNA, RNA, and complexes with small molecules using OpenFold3. Guides hosted API requests or local Docker setup, saves structure files, and helps interpret confidence scores supplied with the predictions.

中文简介正在准备中。当前显示英语简介。

• Predicting protein structures through the hosted API
• Predicting complexes with DNA, RNA, or small molecules
• Setting up local inference with Docker

Fine-tune Parakeet TDT v2 on clinical vocabulary using NeMo, then compare key-term error rates before and after training. The results guide whether to keep the model, expand the dataset, or prepare it for deployment.

中文简介正在准备中。当前显示英语简介。

• Reducing transcription errors in medication names
• Checking whether training improved clinical term recognition
• Deciding whether to expand the training dataset