Tao Train Mask Grounding Dino

Mask Grounding 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).

Mask Grounding DINO for grounded instance segmentation. Extends Grounding DINO with mask prediction head for open-set segmentation guided by text prompts.

Set train.pretrained_model_path for full model weights.

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

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 for an action requires schemas/<action>.schema.json and references/spec_template_<action>.yaml to exist and parse. Use the packaged selected-action 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

• Dataset type: segmentation
• Formats: odvg, coco, coco_raw
• Monitoring metric: val_loss
• AutoML metric contract: Use val_loss emitted during training and minimize it. Standalone evaluation metrics are checkpoint-validation KPIs, not the recommendation-selection objective.
• Evaluation checkpoint metrics: [segm] test_mAP50 and [bbox] test_mAP50 when the checkpoint produces class predictions. A very short smoke-trained checkpoint can complete evaluation successfully with an empty KPI dictionary; verify the evaluation action and result artifact in that case. Train-stage AutoML selection remains based on val_loss.

Per-Action Dataset Requirements

Typical Spec Overrides

Data source overrides are mandatory for every action — the agent MUST construct data source paths from the Per-Action Dataset Requirements table above and include them in spec_overrides.

S3_TRAIN = "s3://bucket/data/train"
S3_EVAL = "s3://bucket/data/eval"

train (mandatory data sources):

{
    "train.num_gpus": 1,
    "train.num_epochs": 10,
    "train.checkpoint_interval": 10,
    "train.validation_interval": 10,
    "dataset.val_data_sources.data_type": "OD",
    "model.num_region_queries": 100,
    "dataset.train_data_sources": [{"image_dir": f"{S3_TRAIN}/images.tar.gz", "json_file": f"{S3_TRAIN}/annotations_odvg.jsonl", "label_map": f"{S3_TRAIN}/annotations_odvg_labelmap.json"}],
    "dataset.val_data_sources": {"image_dir": f"{S3_EVAL}/images.tar.gz", "json_file": f"{S3_EVAL}/annotations.json"},
}

evaluate (mandatory data sources):

{
    "evaluate.checkpoint": "<selected train/AutoML checkpoint>",
    "dataset.test_data_sources.data_type": "OD",
    "dataset.test_data_sources": {"image_dir": f"{S3_EVAL}/images.tar.gz", "json_file": f"{S3_EVAL}/annotations.json"},
}

inference (mandatory data sources):

{
    "inference.checkpoint": "<selected train/AutoML checkpoint>",
    "dataset.infer_data_sources.data_type": "OD",
    "dataset.infer_data_sources": {"image_dir": f"{S3_EVAL}/images.tar.gz", "captions": ["person", "bicycle", "car"]},
}

quantize (mandatory data sources):

{
    "dataset.train_data_sources": [{"image_dir": f"{S3_TRAIN}/images.tar.gz", "json_file": f"{S3_TRAIN}/annotations_odvg.jsonl", "label_map": f"{S3_TRAIN}/annotations_odvg_labelmap.json"}],
    "dataset.val_data_sources": {"image_dir": f"{S3_EVAL}/images.tar.gz", "json_file": f"{S3_EVAL}/annotations.json"},
    "dataset.quant_calibration_data_sources": {"image_dir": f"{S3_TRAIN}/images.tar.gz", "json_file": f"{S3_TRAIN}/annotations_odvg.jsonl", "label_map": f"{S3_TRAIN}/annotations_odvg_labelmap.json"},
}

Eval Dataset

Optional. Validation uses COCO-format annotations even when training uses ODVG.

Important Parameters

• model.backbone: Default swin_tiny_224_1k. Same backbone options as Grounding DINO.
• train.optim.lr: Learning rate. Default 2e-4. lr_backbone 2e-5. Reuses GDINOTrainExpConfig — same training setup as Grounding DINO.
• model.num_queries: Object queries. Default 900.
• model.enc_layers / model.dec_layers: Keep both at 6 for train/AutoML runs. The mask head asserts six decoder outputs during validation, so copying Grounding DINO smoke overrides that reduce transformer layers causes an immediate failure.
• AutoML metric note: Use metric="val_loss" with direction="minimize" for train-stage AutoML. The packaged train loop logs validation loss scalars; it does not emit [bbox] val_mAP@50 during the train job. Standalone evaluation emits [segm] test_mAP50 and [bbox] test_mAP50 only when predictions survive its thresholds; do not substitute either conditional test metric for the training loss used to rank AutoML recommendations.
• model.has_mask: Enables mask prediction head. Default True. Adds mask/dice/rela loss coefficients.
• model.num_region_queries: Number of region queries for mask prediction. Default 100.
• model.loss_types: Loss components. Default [labels, boxes, masks]. Includes mask_loss_coef, dice_loss_coef, rela_loss_coef.
• evaluate.ioi_threshold: IoI threshold for mask evaluation. Default 0.5.
• evaluate.nms_threshold: NMS threshold. Default 0.2.
• evaluate.text_threshold: Text matching threshold. Default 0.3.
• dataset.has_mask: Dataset includes mask annotations. Default True. val_data_sources default data_type is "VG".