Results for “nvdinov2”

DINOv2 for Dense Prediction Tasks

Train, evaluate, export, distill, quantize, or run inference for a TAO DINO 2D object detector using transformer-based detection with denoising training and multi-scale features.

Build video analytics pipelines using NVIDIA DeepStream SDK 9.0 with Python pyservicemaker API, including GStreamer-based video processing, TensorRT inference integration, object detection/tracking, and Kafka/message broker integration.

Finetunes the NV-Generate-CTMR MR-brain diffusion UNet from user-supplied NIfTI training volumes using a wrapper that stages configs and delegates to upstream scripts.

Train, evaluate, distill, quantize, export, and run inference for RT-DETR object detection models using NVIDIA TAO.

Start or patch Dynamo router modes and run router endpoint smoke checks for round-robin, KV-aware, least-loaded, or device-aware routing.

Deploy, debug, and operate the RTVI-CV 2D detection/tracking microservice and call its REST API for stream management, health checks, and metrics.

Segments abdominal organs from CT NIfTI volumes using the NV-Segment-CT VISTA3D model, producing label maps and structured evidence JSON.

Routes NVIDIA Nemotron Speech (Riva) NIM tasks for ASR, TTS, and NMT, covering cloud-hosted inference, self-hosted Docker deployment, and custom model builds.

Provides deep expertise in Node.js internals: V8 engine, libuv event loop, native addon development, build systems, and performance debugging.

NLVR2: A Visual Reasoning Benchmark for Natural Language

Development with Nvdla: tools and best practices

Use when converting ViennaRNA-style sequence-plus-dot-bracket records on stdin into RNA connectivity-table output.

Implements the NOWAIT technique for efficient reasoning in R1-style LLMs. Use when optimizing inference of reasoning models (QwQ, DeepSeek-R1, Phi4-Reasoning, Qwen3, Kimi-VL, QvQ), reducing chain-of-thought token usage by 27-51% while preserving accuracy. Triggers on "optimize reasoning", "reduce thinking tokens", "efficient inference", "suppress reflection tokens", or when working with verbose CoT outputs.

Use when working from the local Evo 2 repository for DNA-sequence scoring, embeddings, generation, or phage-genome design experiments.

Expert en Obsidian avancé (vaults, plugins, dataview, templates, graph view, PKM workflows, sync, publish)

Validate n8n expression syntax and fix common errors when writing {{}} expressions, accessing $json/$node variables, and mapping data between nodes.

PyTorch, TensorFlow, neural networks, CNNs, transformers, and deep learning for production

CommunityToolkit.Mvvm (the MVVM Toolkit) core: source generators ([ObservableProperty], [RelayCommand], [NotifyPropertyChangedFor], [NotifyCanExecuteChangedFor], [NotifyDataErrorInfo]), base classes (ObservableObject / ObservableValidator / ObservableRecipient), commands (RelayCommand / AsyncRelayCommand), and validation. Companion skills: mvvm-toolkit-messenger for pub/sub, mvvm-toolkit-di for Microsoft.Extensions.DependencyInjection wiring. Works across WPF, WinUI 3, MAUI, Uno, and Avalonia.

Implements the NOWAIT technique for efficient reasoning in R1-style LLMs. Use when optimizing inference of reasoning models (QwQ, DeepSeek-R1, Phi4-Reasoning, Qwen3, Kimi-VL, QvQ), reducing chain-of-thought token usage by 27-51% while preserving accuracy. Triggers on "optimize reasoning", "reduce thinking tokens", "efficient inference", "suppress reflection tokens", or when working with verbose CoT outputs.

A versatile input component for text, password, and textarea inputs with comprehensive features including validation, icons, and various states

Reclaim DRAM on NVIDIA Jetson devices by disabling unused display, camera, and DMA subsystems across MB1 BCT, MB2 BCT, kernel reserved-memory, and SWIOTLB layers for headless or no-camera deployments.

Train, evaluate, export, quantize, and run inference on Mask2Former models for panoptic, instance, and semantic segmentation using NVIDIA TAO.

Profile a DeepStream pipeline with Nsight Systems and derive its configs from the measurement.

Generate cinematic videos with Seedance 2.0 using director-level prompts, camera grammar, and multi-mode generation across Chinese, Global, and VIP tiers.

A component for displaying a timeline of events with customizable items, icons, and layout options including horizontal and vertical orientations.

Analyze .NET malware by decompiling and debugging assemblies with dnSpy, deobfuscating with de4dot, and extracting C2 configurations and IOCs.

Guides users through Nonviolent Communication to turn emotionally charged thoughts into constructive messages, structuring responses with observation, feeling, need, and request.

BISON v2.0 — Conviction Holder (Hardened). Top 10 assets by volume. All signals are score contributors — no hard gates. Scanner enters via create_position internally (Wolverine pattern). RatchetStop exits. Thesis exit REMOVED. v2.0: every hard gate converted to score contributor, ensureExecutionAsTaker=false, conviction-scaled margin 25-37%.

Animate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (Arena #1, native audio, identity preservation) for general animations, Wan 2.7 with `audio_url` for custom-voiceover lip-sync, or Seedance 2.0 Pro for multi-modal animation from image + reference video + reference audio. Bundles each model's documented prompting patterns so the caller gets sharper output without burning iterations on the wrong model. Calls `runcomfy run <vendor>/<model>/image-to-video` (or endpoint variant) through the local RunComfy CLI. Triggers on "image to video", "image-to-video", "i2v", "animate image", "make this move", or any explicit ask to turn a still into video.

Scaffold, install, and compose shadcn-vue components built on Reka UI v2, including Vue variants, form patterns with vee-validate/Zod, and adaptation of existing components.

Provides a command-line workflow for binary analysis with radare2, covering reconnaissance, disassembly, function analysis, string and import inspection, patching, binary diffing, and r2pipe scripting across PE, ELF, Mach-O, DEX, and WASM formats.

SNLI-VE: Visual Entailment Dataset

Build the proof-of-concept automation that backs a freelance bid — turns a drafted gig's solution shape into a real, validated n8n workflow via the n8n Cloud MCP (SDK flow), publishes it, and writes the live workflow URL back onto the gig's Notion row. The Prove phase of the Inbound Gig Engine.

Develop quantitative trading strategies with the vn.py framework, covering CTA, spread, and options strategies with support for 20+ broker gateways.