Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
nvidia avatar

Tao Train Dino

nvidia/skills

  • 1.4k installs
  • 2.8k repo stars
  • Updated August 4, 2026
  • nvidia/skills

tao-train-dino is an agent skill that trains, evaluates, distills, quantizes, exports, and deploys NVIDIA TAO DINO transformer object detectors for developers who need COCO-format 2D detection with denoising anchor boxes

About

tao-train-dino is an NVIDIA agent skill (version 0.1.0) for DINO (DETR with Improved DeNoising Anchor Boxes), a transformer-based 2D object detector inside NVIDIA TAO Toolkit. The skill enforces COCO or coco_raw datasets, mandatory validation data sources, and spec overrides for train, evaluate, export, distill, quantize, inference, and TensorRT gen_trt_engine deploy actions. Default training uses 10 epochs, batch_size 4, learning_rate 2e-4, and num_classes 91 for COCO, with ResNet-50 backbone options and AutoML HPO routing via tao-run-automl. Developers reach for tao-train-dino when adapting production object detectors on custom COCO datasets and exporting ONNX or TensorRT engines for inference APIs or edge devices. Reference docs cover multi-GPU DDP/FSDP, error patterns for CUDA OOM and num_classes mismatches, and deploy spec templates.

  • NVIDIA TAO DINO training workflow
  • Custom dataset and export guidance
  • Vision detection model lifecycle support
  • Pairs with NVIDIA skills repo conventions
  • Prepares models for inference deployment

Tao Train Dino by the numbers

  • 1,449 all-time installs (skills.sh)
  • +26 installs in the week ending Aug 5, 2026 (Skillselion tracking)
  • Ranked #186 of 2,064 Data Science & ML skills by installs in the Skillselion catalog
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/nvidia/skills --skill tao-train-dino

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs1.4k
repo stars★ 2.8k
Last updatedAugust 4, 2026
Repositorynvidia/skills ↗

How do you train DINO object detection in NVIDIA TAO?

Train and adapt NVIDIA TAO DINO object-detection models on custom datasets, then export artifacts for inference services or edge deployment pipelines.

Who is it for?

ML engineers training transformer-based 2D object detectors on COCO-format datasets who need TAO-guided spec overrides, AutoML, and TensorRT export paths.

Skip if: Teams needing open-vocabulary text-prompted detection without fixed class labels, which should use tao-train-grounding-dino instead.

When should I use this skill?

The user asks to train DINO, run DETR object detection in TAO, configure COCO val_data_sources, or export a 2D detector with distillation or quantization.

What you get

Fine-tuned DINO checkpoints, mAP evaluation metrics, distilled or quantized models, ONNX files, and TensorRT detection engines.

  • Fine-tuned DINO checkpoints
  • ONNX and TensorRT detection engines
  • mAP50 evaluation metrics

By the numbers

  • Default num_classes 91 for COCO object detection
  • Default training uses 10 epochs and batch_size 4
  • Skill metadata version 0.1.0 with AutoML HPO support

Files

SKILL.mdMarkdownGitHub ↗

DINO

DINO (DETR with Improved DeNoising Anchor Boxes) for 2D object detection. Transformer-based detector with denoising training, multi-scale features, and optional distillation support.

Uses pretrained backbone weights (e.g. ResNet-50 ImageNet). Set model.pretrained_backbone_path for backbone-only or train.pretrained_model_path for full model.

When To Use

Train, evaluate, export, distill, quantize, or run inference for a TAO DINO 2D object detector.

For TAO Deploy TensorRT actions (gen_trt_engine, TensorRT evaluate, and TensorRT inference), read references/tao-deploy-dino.md first. Deploy spec templates live in this skill's references/ folder with the spec_template_deploy_*.yaml prefix.

Reference Map

  • references/dino-data-specs.md — dataset contracts, per-action dataset requirements, per-action spec-override examples (train, evaluate, export, deploy/gen_trt_engine, inference, quantize, distill), data-source arrays, checkpoint inference, and dataset layout.
  • references/dino-actions-errors.md — important parameters, default values, evaluate/export defaults, hardware, and the full error-pattern catalog.
  • references/dino-tuning-multigpu.md — full AutoML/HPO notes (metrics, hyperparameters, extractor) and multi-GPU spec consistency.
  • references/dino-automl-sdk.md — AutoML metrics, SDK orchestration internals, data-source gap, and spec-param/parent-model inference.
  • references/tao-deploy-dino.md — TensorRT deploy workflow.
  • references/detailed-guide.md — map to the detailed model guide.

Dataclass Schemas

Generated TAO Core schemas are packaged in schemas/<action>.schema.json, with schemas/manifest.json listing available actions. Each generated schema also emits references/spec_template_<action>.yaml from the schema top-level default field. AutoML enablement is declared at the model layer in references/skill_info.yaml via automl_enabled. Runnable AutoML still requires schemas/train.schema.json and references/spec_template_train.yaml to exist and parse. Use the packaged train schema for automl_default_parameters, automl_disabled_parameters, defaults, min/max bounds, enums, option weights, math conditions, dependencies, and popular parameters. Do not expect ~/tao-core at runtime; maintainers regenerate schemas/templates before packaging the skill bank.

Train Action Policy

This model is AutoML-enabled at the model layer. Before handling any train-stage request, read references/skill_info.yaml and resolve the run override from either an explicit automl_policy value or the user's workflow request. Use automl_policy: on by default and only expose on / off in new launch prompts. Treat phrases like "turn off AutoML", "disable AutoML", "no HPO", or "plain training" as automl_policy: off for this run only. When automl_policy: on, automl_enabled: true, and both schemas/train.schema.json and references/spec_template_train.yaml are packaged, route the train action through tao-skill-bank:tao-run-automl by default with this model's skill_dir. Preserve workflow/application overrides for datasets, specs, output directories, GPU/platform settings, parent checkpoints, and automl_policy. Use direct model training only when automl_policy: off or the packaged train schema/template is missing; in the missing-schema case, report that AutoML is enabled but not runnable for this model until schemas are generated.

Non-train actions such as evaluate, inference, export, and deploy flows stay in this model skill. The per-run automl_policy override does not change model metadata.

Training Requirements

The agent MUST read this section before generating any training or AutoML script for DINO.

  • Dataset type: object_detection
  • Formats: coco, coco_raw
  • Accepted dataset intents: training, evaluation, testing, calibration
  • Monitoring metric: mAP50 for quick operational checks; val_mAP for

COCO/paper-style benchmark comparisons.

Required datasets — MUST resolve both:

DatasetRequiredWhy
Train dataset URIYesTraining data (COCO format)
Validation dataset URIYes — ALWAYSDINO unconditionally builds a val dataloader. Omitting val_data_sources causes FileNotFoundError at startup regardless of the metric or workflow. If the user has no separate eval split, reuse the train URI.

Required inputs before generating any training spec:

1. Train dataset URI — S3 path to COCO-format training data 2. Validation dataset URI — S3 path to COCO-format val data (can be same as train) 3. `num_classes` — How many object classes? Default 91 (COCO). Must be >= max(category_id) + 1. Too low causes CUDA error: device-side assert triggered.

Resolve these from the user request or the default profile below. Prompt only for values that are still missing after applying the profile rules.

Bankable local default profile for DINO AutoML smoke runs:

Use this profile only when the user asks to run DINO AutoML and does not provide dataset or class-count inputs. This profile is intentionally small and local to this skill bank; it is for smoke/iteration runs, not a production benchmark. Do not search previous runners, logs, session state, shell history, or the home directory to recover these values.

DINO_AUTOML_PROFILE = {
    "train_dataset_uri": "s3://nvcf-storage-handling/data/tao_od_synthetic_subset_train_no_convert",
    "validation_dataset_uri": "s3://nvcf-storage-handling/data/tao_od_synthetic_subset_val_no_convert",
    "object_classes": 4,
    "dataset_num_classes": 5,
    "image_archive": "images.tar.gz",
    "annotation_file": "annotations.json",
    "max_recommendations": 10,
    "train_num_epochs": 10,
    "train_checkpoint_interval": 10,
    "train_validation_interval": 1,
    "train_num_gpus": 1,
}

If the user supplies any dataset URI or class-count value, prefer the user value and ask for any remaining required DINO value. Do not partially mix a user's custom dataset with this profile's class count unless the user confirms it.

Do not prompt for image layout for the standard DINO dataset. The standard TAO DINO dataset artifact is images.tar.gz plus annotations.json. Use images.tar.gz in the remote image_dir spec override. The SDK downloads the archive and rewrites the runtime spec to the extracted folder named after the archive stem (images.tar.gz -> images). Only deviate if the user explicitly provides a different image artifact name.

Core Workflow

DINO supports train, evaluate, export, distill, quantize, and inference. Data-source overrides are mandatory for every action — DINO's config.json has empty data_sources because the runner cannot auto-resolve array-of-objects spec keys. The agent MUST construct data source paths and include them in spec_overrides.

See references/dino-data-specs.md for the per-action dataset requirements table, the standard dataset artifact (images.tar.gz + annotations.json) and runtime folder rewrite rules, and the complete per-action spec_overrides examples for train, evaluate, export, deploy/gen_trt_engine, inference, quantize, and distill — including checkpoint inference via parent_model, the results_dir/train/ checkpoint location, and the distillation FAN-teacher / student rules.

Important Parameters And Defaults

Key defaults: num_epochs=10, batch_size=4, learning_rate=2e-4, lr_backbone=2e-5, num_classes=91, backbone=resnet_50.

  • dataset.num_classes: Default 91 (COCO). Must be >= max(category_id) + 1. Too low causes CUDA error: device-side assert triggered. Set as <num_classes> + 1 in spec overrides.
  • num_epochs: default 10 (quick iteration); real datasets typically need 30-50+ epochs for good mAP.

See references/dino-actions-errors.md for the full parameter list (backbone options, train.optim.lr/lr_steps, model.num_queries, batch_size), default values, evaluate defaults, export defaults (input 960x544, opset 17, TRT data types, workspace 1024 MB), and hardware requirements.

Multi-GPU And AutoML / HPO

When increasing train.num_gpus, also set train.gpu_ids to the same visible device range, or distributed startup can be inconsistent.

AutoML runs training — all Training Requirements above apply. For no-input local smoke runs, use DINO_AUTOML_PROFILE. Recommended metric is mAP50 (val_mAP for benchmark comparisons) with direction="maximize" and a custom metric_extractor.

See references/dino-tuning-multigpu.md for the full multi-GPU spec-consistency rule (8-GPU example, NCCL timeout note) and the full AutoML/HPO notes (metric selection, metric_extractor, recommended hyperparameters, weight_decay behavior, dense-dataset resume guidance). See references/dino-automl-sdk.md for AutoML metric extractor code, SDK orchestration internals, and parent-model inference mappings.

Error Patterns

Common failures include CUDA OOM (reduce batch_size), missing val_data_sources (FileNotFoundError at startup — always supply val), num_classes too low (CUDA device-side assert), and the parent dino gen_trt_engine / dino convert PyT-CLI restrictions.

See references/dino-actions-errors.md for the complete error-pattern catalog with diagnostics and fixes.

Spec Param / Parent Model Inference

Model-specific inference mappings belong in this MD file, not in config.json. Generated runners read the mappings and apply them with SDK helpers before create_job(). For parent_model/parent_model_folder, pass the upstream train/export/AutoML child job id as parent_job_id; the SDK lists the parent result folder, filters checkpoint artifacts, and returns the selected model.

See references/dino-automl-sdk.md for the full inference-mapping table (per action: parent_model, key, output_dir, ptm_if_no_resume_model, resume_model, create_onnx_file) and the TensorRT-mapping note. TensorRT mappings live in the deploy workflow, not the PyT model skill.

Optional: running via the TAO SDK

When running DINO through the TAO SDK (script_runner orchestration, S3 I/O wrapping, AutoML), skills read references/skill_info.yaml for input and spec-param mappings. See references/dino-automl-sdk.md for SDK orchestration internals, including the data-sources gap and the [0]-indexed inputs declarations. Skip this when running locally with docker run.

Deployment

  • tao-deploy-dino

Related skills

microsoftMicrosoft FoundryDeploy, evaluate, and continuously improve Microsoft Foundry agents from a single agent interface.490k1.3klllllllamaAi Research ReproductionOrchestrate trustworthy, auditable reproduction of deep learning repositories directly from their READMEs.176k511lllllllamaRun TrainSafely execute selected deep learning training commands with standardized evidence capture.176k511lllllllamaExplore RunSafely run isolated exploratory experiments with clear recording and conservative selection before committing changes.176k511lllllllamaPaper Context ResolverFetch precise reproduction-critical details like dataset splits, preprocessing steps, or evaluation protocols from the original academic paper when the repo README leav141k511lllllllamaRepo Intake And PlanScan unfamiliar AI research repositories and receive a minimal, trustworthy reproduction target before investing significant time.140k511

How it compares

Choose tao-train-dino over grounding-dino skills when detection uses a fixed COCO class vocabulary rather than open-vocabulary text prompts.

FAQ

What dataset format does tao-train-dino require?

tao-train-dino requires object_detection datasets in COCO or coco_raw format with images.tar.gz plus annotations.json. Both train and validation dataset URIs are mandatory because DINO unconditionally builds a validation dataloader at startup regardless of monitoring metric choic

What are the default DINO training hyperparameters?

tao-train-dino defaults include num_epochs 10, batch_size 4, learning_rate 2e-4, lr_backbone 2e-5, and num_classes 91 for COCO. Production datasets typically need 30-50+ epochs, and num_classes must be at least max category_id plus one.

Does tao-train-dino support AutoML hyperparameter search?

tao-train-dino is AutoML-enabled at the model layer and routes train actions through tao-run-automl by default when automl_policy is on. The recommended monitoring metric is mAP50 with maximize direction, using packaged schemas/train.schema.json for hyperparameter bounds.

Data Science & MLpipelinesetl
Browse by category
AI & AgentsFrontend & DesignBackend & DataDev ToolsAutomationTesting & ReviewSecurityMonitoring & CloudDocs & Planning
Design & frontend
Design systemsAnimation & motionAccessibility (a11y)CSS & TailwindUI designUXDesign to codeshadcn/uiLanding pagesBest design skills
Most searched skills
find-skillscavemanfrontend-designbrainstormingsystematic-debuggingstop-slopemil-design-engui-ux-pro-maxkarpathy-guidelinesweb-design-guidelinesreact-best-practicesgrill-meimprovemicrosoft-foundryagent-browser
Browse all listings
All agent skillsAll MCP serversAll marketplacesAll plugins

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.