NVIDIA Corporation

Generate small-molecule candidates from a starting structure, optimize supported molecular properties, and save results for validation. Local deployment also supports embeddings, numerical representations of molecules, and workflows that convert between molecular structures and learned internal representations.

中文简介正在准备中。当前显示英语简介。

• Generate molecular candidates around a seed structure
• Optimize candidates for QED or plogP
• Obtain molecular embeddings with a local deployment

Modify NVIDIA cuOpt itself, covering C++/CUDA solver code, Python bindings, server endpoints, and CI. Follow project conventions for environment setup, builds, tests, debugging, and contributions, including the required sign-off for submitted changes.

中文简介正在准备中。当前显示英语简介。

• Fixing solver internals and investigating numerical issues
• Changing Python APIs or server endpoints
• Setting up a development environment and running tests

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

Build English clinical speech-recognition benchmarks by selecting medical terms, checking pronunciations, and generating synthetic audio. Outputs include audio files and a NeMo manifest, a structured file listing recordings and metadata, ready for a separate scoring stage.

中文简介正在准备中。当前显示英语简介。

• Building a speech-recognition benchmark for a clinical specialty
• Checking synthesized pronunciations of drug and procedure names
• Generating evaluation audio across voices, noise, and contexts

Customize NVIDIA Nemotron Voice Agent’s Generic Pipecat example for appointment scheduling, five-field patient intake, or a custom healthcare conversation. Build prompts and tool handlers without a separate backend, then validate conversation behavior, services, and voice interaction.

中文简介正在准备中。当前显示英语简介。

• Build an appointment-scheduling voice agent
• Collect five patient-intake fields through conversation
• Design a custom healthcare conversation with tool calls

Build and troubleshoot a GPU-powered service that checks factory videos for missing or out-of-order procedure steps. Guides video segmentation, action classification with a vision-language model, deployment, evaluation, and latency measurement for files and live cameras.

中文简介正在准备中。当前显示英语简介。

• Building a service to check assembly-line procedure compliance
• Testing video files, RTSP streams, and Basler cameras
• Investigating action classification and step-order errors

Test a running AMC service with the bundled four-camera warehouse dataset. Run camera calibration from upload to completion, then check distance and image projection errors against ground-truth data to verify the installation.

中文简介正在准备中。当前显示英语简介。

• Verify a newly installed AMC service
• Check the sample's calibration errors against ground truth
• Walk through the sample workflow in Swagger UI

Extract selected metadata from a single DICOM medical imaging file into JSON, including scan type and study details. Flag personal health information in a limited set of standard tags for technical inspection.

中文简介正在准备中。当前显示英语简介。

• Inspecting metadata in a single DICOM file
• Exporting study and image details as JSON
• Checking selected standard tags for personal health information

Generate drug-like molecule candidates with GenMol NIM through NVIDIA's hosted API or local Docker deployment. Expand molecular scaffolds, modify existing candidates, rank results by drug-likeness or lipophilicity, and save the ranked molecules to a file.

中文简介正在准备中。当前显示英语简介。

• Generate new molecule candidates from scratch
• Add structures around a molecular scaffold
• Modify fragments of a lead molecule

Generate DNA sequences with NVIDIA Evo 2, check their length and alphabet, and save sequence files alongside probabilities and timing metrics. A local deployment also supports extracting numerical outputs from selected internal model layers.

中文简介正在准备中。当前显示英语简介。

• Generate DNA from a starting sequence
• Save generated sequences and metrics
• Extract internal layer outputs from a local model

Pretrain a new KERMT chemistry model from a molecular corpus, starting with random weights. The skill validates the data, builds fresh vocabularies, estimates runtime, and launches containerized training with logs and records for reproducing the run.

中文简介正在准备中。当前显示英语简介。

• Train a new model on a specialized chemistry corpus
• Run pretraining without an existing checkpoint
• Choose a training objective and batch size

Evaluate speech recognition on synthetic clinical audio listed in a NeMo manifest. Produce four error metrics and a five-section report highlighting missed medical terms, then use pronunciation and noise breakdowns to guide dataset improvements or fine-tuning decisions.

中文简介正在准备中。当前显示英语简介。

• Scoring medical term recognition on synthetic audio
• Generating an error report by term, category, and noise level
• Investigating accuracy gaps between pronunciation sources

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

Extend an existing grover_base model with a new decoder that reconstructs molecular structure strings, then continue pretraining on your corpus. The workflow validates inputs, rebuilds vocabulary components, and records the conversion and training outputs.

中文简介正在准备中。当前显示英语简介。

• Extend an existing grover_base model with cMIM
• Continue pretraining on a molecular corpus
• Trace checkpoint conversion and training records

Check one DICOM medical imaging series before conversion or AI inference by inspecting headers only. Produce a JSON report with file inventory, image orientation, personal information flags, findings, and a pass, warn, or fail verdict.

中文简介正在准备中。当前显示英语简介。

• Checking a DICOM series before conversion
• Reviewing image orientation before AI inference
• Recording preflight findings in JSON

Measure and tune a DeepStream video pipeline using Nsight Systems. The skill benchmarks inference batch sizes, derives configuration settings from measurements and GPU limits, and reports FPS, processing bottlenecks, and available timings for individual processing components.

中文简介正在准备中。当前显示英语简介。

• Benchmark video processing FPS
• Tune settings for multiple video streams
• Identify processing bottlenecks

Build DeepStream commands for image and video inference, object detection, tracking, and streaming. The skill asks about missing requirements, assembles a GStreamer processing pipeline, and checks its syntax, components, properties, and connections where supported.

中文简介正在准备中。当前显示英语简介。

• Build a pipeline for inference on local images or video
• Detect and track objects across multiple RTSP streams
• Save processed output to an MP4 file

Continue training a KERMT molecular model on a custom SMILES corpus, where text strings describe molecular structures. The skill validates data and checkpoints, prepares training inputs, and launches containerized training with saved logs and checkpoints.

中文简介正在准备中。当前显示英语简介。

• Continue molecular model pretraining on a custom SMILES corpus
• Start additional training with a new learning-rate schedule
• Resume an interrupted training run from a saved checkpoint

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

Turn written planning problems into mathematical optimization models. Separate fixed inputs, constraints, decisions, and objectives, then choose linear, mixed-integer, or quadratic formulations for tasks such as production planning, inventory management, and material allocation.

中文简介正在准备中。当前显示英语简介。

• Formulate production plans with capacity and cost constraints
• Model inventory limits across purchasing and sales periods
• Define multiple goals in priority order

Design protein backbones, the structural framework of proteins, with RFDiffusion through NVIDIA NIM. Generate new structures, build around existing structural motifs, or design binders, then save PDB files for subsequent amino acid sequence design.

中文简介正在准备中。当前显示英语简介。

• Generating a new protein backbone
• Building a backbone around an existing structural motif
• Designing a binder around specified target residues

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

Fine-tune a pretrained KERMT model on a CSV of molecular structures and labeled targets. The workflow checks inputs, prepares features, launches training in a container, and saves trained models, logs, and test results.

中文简介正在准备中。当前显示英语简介。

• Train a molecular model on labeled CSV data
• Select specific prediction targets for training
• Reuse prepared data to compare training settings

Predict how small molecules may bind to proteins using DiffDock through NVIDIA NIM. Prepare molecular inputs, run hosted API or local Docker requests, and save ranked binding poses with confidence scores for inspection.

中文简介正在准备中。当前显示英语简介。

• Predicting small-molecule binding poses for a protein
• Submitting SMILES or SDF ligands
• Saving ranked poses with confidence scores