Skill

Official

Pass

Audit score 90

tao-train-dino

nvidia/skills

Transformer-based DINO detector for 2D object detection with denoising training and distillation support.

What is tao-train-dino?

DINO (DETR with Improved DeNoising Anchor Boxes) is a transformer-based 2D object detector with denoising training, multi-scale features, and optional distillation. Use it to train, evaluate, export, distill, quantize, or run inference on object detection models in COCO format.

Train DINO detectors with denoising anchor boxes and multi-scale feature extraction
Evaluate models using COCO mAP and mAP50 metrics
Export trained models to TensorRT engines for deployment
Distill knowledge from teacher to student DINO models
Quantize models for optimized inference
Run inference on images using trained or pretrained checkpoints

How to install tao-train-dino

npx skills add https://github.com/nvidia/skills --skill tao-train-dino

Prerequisites

Docker and nvidia-container-toolkit installed
TAO skill bank initialized (run tao-setup skill if needed)
COCO-format training and validation datasets with annotations.json
num_classes value matching your dataset's category count

Claude Code

Cursor

Windsurf

Cline

How to use tao-train-dino

1.Read the relevant reference guide (dino-data-specs.md for data, dino-actions-errors.md for parameters)
2.Prepare your COCO-format dataset with train and validation splits
3.Create a spec file using the template from references/spec_template_<action>.yaml
4.Run the desired action (train, evaluate, export, distill, quantize, or inference) with your spec
5.Monitor training metrics (val_mAP50 for quick checks, val_mAP for COCO benchmarks)

Use cases

Good for

Training a custom object detector on your own COCO-formatted dataset
Evaluating detector performance across validation splits
Exporting a trained model to TensorRT for production deployment
Distilling a large DINO model into a smaller one for edge devices
Quantizing a detector for reduced memory and latency requirements

Who it's for

Computer vision engineers training object detectors
ML practitioners working with COCO-format datasets
Teams deploying detectors to production with TensorRT
Researchers experimenting with transformer-based detection architectures

tao-train-dino FAQ

What datasets does DINO support?

DINO accepts COCO and coco_raw formats. Both training and validation dataset URIs are required; if you lack a separate eval split, reuse the training URI.

What is the default num_classes value?

The default is 91 (COCO dataset). Set num_classes to max(category_id) + 1 for your dataset to avoid CUDA errors.

Does DINO support AutoML/hyperparameter optimization?

Yes, DINO is AutoML-enabled. By default, train requests route through tao-run-automl. Use automl_policy: off to disable AutoML for a single run.

How do I deploy a DINO model to production?

Export your trained model using the export action, then follow the TensorRT workflow in references/tao-deploy-dino.md to generate a TensorRT engine.

What metrics should I use for evaluation?

Use test_mAP50 for quick operational checks and AutoML scoring. Use test_mAP for COCO/paper-style benchmark comparisons.

Full instructions (SKILL.md)

Source of truth, from nvidia/skills.

name: tao-train-dino description: DINO (DETR with Improved DeNoising Anchor Boxes) for 2D object detection. Transformer-based detector with denoising training, multi-scale features, and optional distillation support. Use when training, evaluating, exporting, distilling, quantizing, or running inference for a TAO DINO detector. Trigger phrases include "train DINO", "DETR object detection", "TAO 2D detection", "DINO with distillation". license: Apache-2.0 compatibility: Requires docker + nvidia-container-toolkit. metadata: tags:

object
detection version: "0.1.0" author: NVIDIA Corporation allowed-tools: Read Bash

DINO

Standalone install? If this session was not initialized by the TAO skill bank plugin, run the tao-setup skill first (host preflight, credentials, cross-skill discovery).

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/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
AutoML metric contract: for the default evaluation-backed workflow, use test_mAP50 with maximize direction. Use test_mAP only when the user explicitly requests COCO/paper-style mAP.
Training monitoring metrics: val_mAP50 (logged as Validation mAP50) for quick operational checks; val_mAP for COCO/paper-style benchmark comparisons.
Evaluate action metrics: test_mAP50 for AP50 and test_mAP for COCO mAP. AutoML workflows that score recommendations with the standalone evaluate action must use the corresponding test_* KPI.

Required datasets — MUST resolve both:

Dataset	Required	Why
Train dataset URI	Yes	Training data (COCO format)
Validation dataset URI	Yes — ALWAYS	DINO 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:

Train dataset URI — S3 path to COCO-format training data
Validation dataset URI — S3 path to COCO-format val data (can be same as train)
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

Default values: 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. For training-log-only scoring, use val_mAP50 (extracted from Validation mAP50) or val_mAP. When each recommendation is scored through the standalone evaluate action, use test_mAP50 or test_mAP with direction="maximize".

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, 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. For parent_model/parent_model_folder, pass the upstream train/export/AutoML child job id as the parent job id; list the parent result folder, filter checkpoint artifacts, and select the resolved model.

See references/dino-tuning-multigpu.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.