[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78JFJB9KV4VTRF604CZG2CC":3,"skill-same:01M78JFJB9KV4VTRF604CZG2CC":141},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":32,"commitSha":33,"skillDirectory":34,"forksCount":13,"repositoryHtmlUrl":35,"repositoryHomepageUrl":36,"ownerHtmlUrl":37,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":38,"install":39,"resources":52},"01M78JFJB9KV4VTRF604CZG2CC","tao-train-metric-learning-recognition",null,"Metric-learning recognition (ml-recog) for fine-grained visual recognition. Learns embeddings for retrieval-based matching (e.g., retail product recognition) using triplet \u002F contrastive losses. Use when training, evaluating, exporting, or running inference for a TAO metric-learning recognition model. Trigger phrases include \"train metric learning\", \"ml-recog\", \"retrieval embeddings\", \"triplet loss recognition\", \"fine-grained matching\".","matheussluzz\u002FZeuZ-Agent","matheussluzz","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F209255758?v=4","01M78HKHH1D1ASC9KC9BG583DH","01M78HKHHP6W7XRXQM2SBCAM3R",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: tao-train-metric-learning-recognition\ndescription: Metric-learning recognition (ml-recog) for fine-grained visual recognition. Learns embeddings for\n  retrieval-based matching (e.g., retail product recognition) using triplet \u002F contrastive losses. Use when training,\n  evaluating, exporting, or running inference for a TAO metric-learning recognition model. Trigger phrases include\n  \"train metric learning\", \"ml-recog\", \"retrieval embeddings\", \"triplet loss recognition\", \"fine-grained matching\".\nlicense: Apache-2.0\ncompatibility: Requires docker + nvidia-container-toolkit.\nmetadata:\n  version: \"0.1.0\"\n  author: NVIDIA Corporation\nallowed-tools: Read Bash\ntags:\n- metric\n- learning\n- recognition\n---\n\n# ML Recog\n\nMetric learning recognition for fine-grained visual recognition. Learns embeddings for retrieval-based matching (e.g., retail product recognition). Uses triplet\u002Fcontrastive losses.\n\nSet model.pretrained_model_path for pretrained backbone.\n\nFor TAO Deploy TensorRT actions (`gen_trt_engine`, TensorRT `evaluate`, and TensorRT `inference`), read `references\u002Ftao-deploy-metric-learning-recognition.md` first. Deploy spec templates live in this skill's `references\u002F` folder with the `spec_template_deploy_*.yaml` prefix.\n\n## Dataclass Schemas\n\nGenerated TAO Core schemas are packaged in `schemas\u002F\u003Caction>.schema.json`, with `schemas\u002Fmanifest.json` listing available actions. Each generated schema also emits `references\u002Fspec_template_\u003Caction>.yaml` from the schema top-level `default` field. AutoML enablement is declared at the model layer in `references\u002Fskill_info.yaml` via `automl_enabled`. Runnable AutoML still requires `schemas\u002Ftrain.schema.json` and `references\u002Fspec_template_train.yaml` to exist and parse. Use the packaged train schema for `automl_default_parameters`, `automl_disabled_parameters`, defaults, min\u002Fmax bounds, enums, option weights, math conditions, dependencies, and popular parameters. Do not expect `~\u002Ftao-core` at runtime; maintainers regenerate schemas\u002Ftemplates before packaging the skill bank.\n\n## Train Action Policy\n\nThis model is AutoML-enabled at the model layer. Before handling any train-stage request, read `references\u002Fskill_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` \u002F `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\u002Ftrain.schema.json` and `references\u002Fspec_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\u002Fapplication overrides for datasets, specs, output directories, GPU\u002Fplatform settings, parent checkpoints, and `automl_policy`. Use direct model training only when `automl_policy: off` or the packaged train schema\u002Ftemplate is missing; in the missing-schema case, report that AutoML is enabled but not runnable for this model until schemas are generated.\n\nNon-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.\n\n## Training Requirements\n\n- **Dataset type:** ml_recog\n- **Formats:** default\n- **Monitoring metric:** val Precision at Rank 1\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| evaluate | dataset.val_dataset | train_datasets | reference: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz, query: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz | No |\n| inference | dataset.val_dataset | train_datasets | reference: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz, query:  | No |\n| inference | inference.input_path | train_datasets | metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz | No |\n| train | dataset.train_dataset | train_datasets | metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Ftrain.tar.gz | No |\n| train | dataset.val_dataset | train_datasets | reference: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Freference.tar.gz, query: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Fval.tar.gz | No |\n\n### Typical Spec Overrides\n\nData 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`.\n\n```python\nS3_TRAIN = \"s3:\u002F\u002Fbucket\u002Fdata\u002Ftrain\"\n```\n\n**train (mandatory data sources):**\n```python\n{\n    \"train.num_epochs\": 30,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"train.num_gpus\": 1,\n    \"dataset.train_dataset\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Ftrain.tar.gz\",\n    \"dataset.val_dataset\": {\"reference\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Freference.tar.gz\", \"query\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Fval.tar.gz\"},\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\n{\n    \"evaluate.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.val_dataset\": {\"reference\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz\", \"query\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz\"},\n}\n```\n\n**inference (mandatory data sources):**\n```python\n{\n    \"inference.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.val_dataset\": {\"reference\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz\"},\n    \"inference.input_path\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz\",\n}\n```\n## Eval Dataset\n\nRequired. Evaluation requires reference and query datasets for retrieval metrics.\n\n## Important Parameters\n\n- **model.backbone**: Default resnet_50. Options: resnet_50, resnet_101, fan_small, fan_base, fan_large, fan_tiny, nvdinov2_vit_large_legacy.\n- **model.feat_dim**: Embedding dimension. Default 256. Output feature vector size for similarity matching.\n- **train.batch_size**: Per-GPU batch size. Default 4. `val_batch_size` also 4. For training and AutoML search, `train.batch_size` must be divisible by `dataset.num_instance`.\n- **dataset.num_instance**: Instances per identity in a batch (P\u002FK sampling). Default 4. Controls how many images of the same class appear together. If using a custom AutoML range for `train.batch_size`, use explicit options that are multiples of this value.\n- **train.optim.trunk.base_lr**: Learning rate for the trunk (backbone). Default 3.5e-4 (Adam).\n- **train.optim.embedder.base_lr**: Learning rate for the embedding head. Default 3.5e-4.\n- **train.optim.triplet_loss_margin**: Margin for triplet loss. Default 0.3. smooth_loss=True by default.\n- **train.optim.miner_function_margin**: Hard mining margin. Default 0.1. Controls pair mining difficulty.\n- **train.optim.steps**: LR decay steps. Default [40, 70] with gamma=0.1.\n- **dataset.train_dataset**: Path to training images organized in class folders.\n- **dataset.val_dataset**: Dict with 'reference' and 'query' keys pointing to ImageNet-format directories for retrieval evaluation.\n\n## Multi-GPU \u002F Multi-Node\n\n**Launch method:** Lightning-managed (single `python` process, Lightning spawns workers).\n\n| Spec Key | Description | Default |\n|----------|-------------|---------|\n| `train.num_gpus` | Number of GPUs | 1 |\n| `train.gpu_ids` | GPU device indices | [0] |\n\n- Strategy: `auto` (Lightning picks best strategy automatically)\n- No explicit `num_nodes` or `distributed_strategy` config — single-node oriented\n\n## Hardware\n\nMinimum 1 GPU(s), recommended 2 GPU(s). 16GB+ VRAM per GPU. Metric learning benefits from larger batch sizes for better triplet sampling but is otherwise moderate on memory.\n\n## Error Patterns\n\n**Reference\u002Fquery mismatch**: Ensure reference and query datasets share compatible class namespaces for evaluation.\n\n**PyTorch 2.6 checkpoint load failure on checkpoint actions**: Current TAO\nML-Recog checkpoints may contain OmegaConf objects. For checkpoints produced by\nthe same trusted TAO train\u002FAutoML workflow, set\n`TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1` in downstream evaluate, inference, export,\nor resume\u002Fretrain job env vars so Lightning can load the full checkpoint. Do not\nuse this env var for untrusted checkpoints.\n\n## Spec Param \u002F Parent Model Inference\n\nModel-specific inference mappings belong in this MD file, not in `config.json`. Generated runners should read this section and apply the mappings with SDK helpers before `create_job()`. This mirrors the old microservices `infer_params.py` flow.\n\nInference mappings from TAO Core `ml_recog.config.json`:\n\n| Action | Spec Field | Inference Function | Meaning |\n|---|---|---|---|\n| evaluate | `evaluate.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| evaluate | `results_dir` | `output_dir` | current job results directory |\n| export | `export.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| export | `export.onnx_file` | `create_onnx_file` | output ONNX path |\n| export | `results_dir` | `output_dir` | current job results directory |\n| inference | `inference.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| inference | `results_dir` | `output_dir` | current job results directory |\n| train | `model.pretrained_model_path` | `ptm_if_no_resume_model` | PTM when no resume checkpoint exists |\n| train | `results_dir` | `output_dir` | current job results directory |\n| train | `train.resume_training_checkpoint_path` | `resume_model` | model file inferred from the current job results folder |\n\nFor `parent_model` or `parent_model_folder`, pass the upstream train\u002Fexport\u002FAutoML child job id as `parent_job_id`. The SDK lists the parent result folder, filters checkpoint artifacts, and returns the selected model file or folder. Do not add these mappings back to `config.json` and do not patch generated runner scripts to guess checkpoint paths.\n\n## Deployment\n\n- [tao-deploy-metric-learning-recognition](references\u002Ftao-deploy-metric-learning-recognition.md)\n","# ML Recog\n\nMetric learning recognition for fine-grained visual recognition. Learns embeddings for retrieval-based matching (e.g., retail product recognition). Uses triplet\u002Fcontrastive losses.\n\nSet model.pretrained_model_path for pretrained backbone.\n\nFor TAO Deploy TensorRT actions (`gen_trt_engine`, TensorRT `evaluate`, and TensorRT `inference`), read `references\u002Ftao-deploy-metric-learning-recognition.md` first. Deploy spec templates live in this skill's `references\u002F` folder with the `spec_template_deploy_*.yaml` prefix.\n\n## Dataclass Schemas\n\nGenerated TAO Core schemas are packaged in `schemas\u002F\u003Caction>.schema.json`, with `schemas\u002Fmanifest.json` listing available actions. Each generated schema also emits `references\u002Fspec_template_\u003Caction>.yaml` from the schema top-level `default` field. AutoML enablement is declared at the model layer in `references\u002Fskill_info.yaml` via `automl_enabled`. Runnable AutoML still requires `schemas\u002Ftrain.schema.json` and `references\u002Fspec_template_train.yaml` to exist and parse. Use the packaged train schema for `automl_default_parameters`, `automl_disabled_parameters`, defaults, min\u002Fmax bounds, enums, option weights, math conditions, dependencies, and popular parameters. Do not expect `~\u002Ftao-core` at runtime; maintainers regenerate schemas\u002Ftemplates before packaging the skill bank.\n\n## Train Action Policy\n\nThis model is AutoML-enabled at the model layer. Before handling any train-stage request, read `references\u002Fskill_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` \u002F `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\u002Ftrain.schema.json` and `references\u002Fspec_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\u002Fapplication overrides for datasets, specs, output directories, GPU\u002Fplatform settings, parent checkpoints, and `automl_policy`. Use direct model training only when `automl_policy: off` or the packaged train schema\u002Ftemplate is missing; in the missing-schema case, report that AutoML is enabled but not runnable for this model until schemas are generated.\n\nNon-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.\n\n## Training Requirements\n\n- **Dataset type:** ml_recog\n- **Formats:** default\n- **Monitoring metric:** val Precision at Rank 1\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| evaluate | dataset.val_dataset | train_datasets | reference: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz, query: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz | No |\n| inference | dataset.val_dataset | train_datasets | reference: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz, query:  | No |\n| inference | inference.input_path | train_datasets | metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz | No |\n| train | dataset.train_dataset | train_datasets | metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Ftrain.tar.gz | No |\n| train | dataset.val_dataset | train_datasets | reference: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Freference.tar.gz, query: metric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Fval.tar.gz | No |\n\n### Typical Spec Overrides\n\nData 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`.\n\n```python\nS3_TRAIN = \"s3:\u002F\u002Fbucket\u002Fdata\u002Ftrain\"\n```\n\n**train (mandatory data sources):**\n```python\n{\n    \"train.num_epochs\": 30,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"train.num_gpus\": 1,\n    \"dataset.train_dataset\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Ftrain.tar.gz\",\n    \"dataset.val_dataset\": {\"reference\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Freference.tar.gz\", \"query\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Fknown_classes\u002Fval.tar.gz\"},\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\n{\n    \"evaluate.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.val_dataset\": {\"reference\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz\", \"query\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz\"},\n}\n```\n\n**inference (mandatory data sources):**\n```python\n{\n    \"inference.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.val_dataset\": {\"reference\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Freference.tar.gz\"},\n    \"inference.input_path\": f\"{S3_TRAIN}\u002Fmetric_learning_recognition\u002Fretail-product-checkout-dataset_classification_demo\u002Funknown_classes\u002Ftest.tar.gz\",\n}\n```\n## Eval Dataset\n\nRequired. Evaluation requires reference and query datasets for retrieval metrics.\n\n## Important Parameters\n\n- **model.backbone**: Default resnet_50. Options: resnet_50, resnet_101, fan_small, fan_base, fan_large, fan_tiny, nvdinov2_vit_large_legacy.\n- **model.feat_dim**: Embedding dimension. Default 256. Output feature vector size for similarity matching.\n- **train.batch_size**: Per-GPU batch size. Default 4. `val_batch_size` also 4. For training and AutoML search, `train.batch_size` must be divisible by `dataset.num_instance`.\n- **dataset.num_instance**: Instances per identity in a batch (P\u002FK sampling). Default 4. Controls how many images of the same class appear together. If using a custom AutoML range for `train.batch_size`, use explicit options that are multiples of this value.\n- **train.optim.trunk.base_lr**: Learning rate for the trunk (backbone). Default 3.5e-4 (Adam).\n- **train.optim.embedder.base_lr**: Learning rate for the embedding head. Default 3.5e-4.\n- **train.optim.triplet_loss_margin**: Margin for triplet loss. Default 0.3. smooth_loss=True by default.\n- **train.optim.miner_function_margin**: Hard mining margin. Default 0.1. Controls pair mining difficulty.\n- **train.optim.steps**: LR decay steps. Default [40, 70] with gamma=0.1.\n- **dataset.train_dataset**: Path to training images organized in class folders.\n- **dataset.val_dataset**: Dict with 'reference' and 'query' keys pointing to ImageNet-format directories for retrieval evaluation.\n\n## Multi-GPU \u002F Multi-Node\n\n**Launch method:** Lightning-managed (single `python` process, Lightning spawns workers).\n\n| Spec Key | Description | Default |\n|----------|-------------|---------|\n| `train.num_gpus` | Number of GPUs | 1 |\n| `train.gpu_ids` | GPU device indices | [0] |\n\n- Strategy: `auto` (Lightning picks best strategy automatically)\n- No explicit `num_nodes` or `distributed_strategy` config — single-node oriented\n\n## Hardware\n\nMinimum 1 GPU(s), recommended 2 GPU(s). 16GB+ VRAM per GPU. Metric learning benefits from larger batch sizes for better triplet sampling but is otherwise moderate on memory.\n\n## Error Patterns\n\n**Reference\u002Fquery mismatch**: Ensure reference and query datasets share compatible class namespaces for evaluation.\n\n**PyTorch 2.6 checkpoint load failure on checkpoint actions**: Current TAO\nML-Recog checkpoints may contain OmegaConf objects. For checkpoints produced by\nthe same trusted TAO train\u002FAutoML workflow, set\n`TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1` in downstream evaluate, inference, export,\nor resume\u002Fretrain job env vars so Lightning can load the full checkpoint. Do not\nuse this env var for untrusted checkpoints.\n\n## Spec Param \u002F Parent Model Inference\n\nModel-specific inference mappings belong in this MD file, not in `config.json`. Generated runners should read this section and apply the mappings with SDK helpers before `create_job()`. This mirrors the old microservices `infer_params.py` flow.\n\nInference mappings from TAO Core `ml_recog.config.json`:\n\n| Action | Spec Field | Inference Function | Meaning |\n|---|---|---|---|\n| evaluate | `evaluate.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| evaluate | `results_dir` | `output_dir` | current job results directory |\n| export | `export.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| export | `export.onnx_file` | `create_onnx_file` | output ONNX path |\n| export | `results_dir` | `output_dir` | current job results directory |\n| inference | `inference.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| inference | `results_dir` | `output_dir` | current job results directory |\n| train | `model.pretrained_model_path` | `ptm_if_no_resume_model` | PTM when no resume checkpoint exists |\n| train | `results_dir` | `output_dir` | current job results directory |\n| train | `train.resume_training_checkpoint_path` | `resume_model` | model file inferred from the current job results folder |\n\nFor `parent_model` or `parent_model_folder`, pass the upstream train\u002Fexport\u002FAutoML child job id as `parent_job_id`. The SDK lists the parent result folder, filters checkpoint artifacts, and returns the selected model file or folder. Do not add these mappings back to `config.json` and do not patch generated runner scripts to guess checkpoint paths.\n\n## Deployment\n\n- [tao-deploy-metric-learning-recognition](references\u002Ftao-deploy-metric-learning-recognition.md)\n",{"name":5,"tags":23,"license":14,"metadata":27,"description":7,"allowed-tools":30,"compatibility":31},[24,25,26],"metric","learning","recognition",{"author":28,"version":29},"NVIDIA Corporation","0.1.0","Read Bash","Requires docker + nvidia-container-toolkit.","main","fc4f2c7759ae400c66cd38656db70ead882a0d08","catalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent#readme","https:\u002F\u002Fgithub.com\u002Fmatheussluzz","2026-08-27 05:21:51",{"source":8,"skillName":5,"commands":40,"prompt":50,"zipAvailable":51},[41,44,47],{"tool":42,"command":43},"npx","npx skills add matheussluzz\u002FZeuZ-Agent --skill tao-train-metric-learning-recognition",{"tool":45,"command":46},"bunx","bunx skills add matheussluzz\u002FZeuZ-Agent --skill tao-train-metric-learning-recognition",{"tool":48,"command":49},"pnpm","pnpm dlx skills add matheussluzz\u002FZeuZ-Agent --skill tao-train-metric-learning-recognition","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002FSKILL.md",true,[53,59,63,67,71,75,79,83,87,91,95,99,103,107,111,115,119,123,127,131,137],{"path":54,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":57,"sourceHtmlUrl":58},"BENCHMARK.md","REFERENCE","参考文档",4192,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002FBENCHMARK.md",{"path":60,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":61,"sourceHtmlUrl":62},"references\u002Fskill_info.yaml",4828,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fskill_info.yaml",{"path":64,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":65,"sourceHtmlUrl":66},"references\u002Fspec_template_deploy_evaluate.yaml",169,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_deploy_evaluate.yaml",{"path":68,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":69,"sourceHtmlUrl":70},"references\u002Fspec_template_deploy_gen_trt_engine.yaml",582,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_deploy_gen_trt_engine.yaml",{"path":72,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":73,"sourceHtmlUrl":74},"references\u002Fspec_template_deploy_inference.yaml",308,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_deploy_inference.yaml",{"path":76,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":77,"sourceHtmlUrl":78},"references\u002Fspec_template_evaluate.yaml",1856,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_evaluate.yaml",{"path":80,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":81,"sourceHtmlUrl":82},"references\u002Fspec_template_export.yaml",1776,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_export.yaml",{"path":84,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":85,"sourceHtmlUrl":86},"references\u002Fspec_template_gen_trt_engine.yaml",2082,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_gen_trt_engine.yaml",{"path":88,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":89,"sourceHtmlUrl":90},"references\u002Fspec_template_inference.yaml",1887,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_inference.yaml",{"path":92,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":93,"sourceHtmlUrl":94},"references\u002Fspec_template_train.yaml",1687,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Fspec_template_train.yaml",{"path":96,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":97,"sourceHtmlUrl":98},"references\u002Ftao-deploy-metric-learning-recognition.md",5877,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Ftao-deploy-metric-learning-recognition.md",{"path":100,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":101,"sourceHtmlUrl":102},"references\u002Ftao-deploy-metric-learning-recognition.skill_info.yaml",4086,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Freferences\u002Ftao-deploy-metric-learning-recognition.skill_info.yaml",{"path":104,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":105,"sourceHtmlUrl":106},"schemas\u002Fevaluate.schema.json",36414,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fschemas\u002Fevaluate.schema.json",{"path":108,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":109,"sourceHtmlUrl":110},"schemas\u002Fexport.schema.json",35477,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fschemas\u002Fexport.schema.json",{"path":112,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":113,"sourceHtmlUrl":114},"schemas\u002Fgen_trt_engine.schema.json",41865,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fschemas\u002Fgen_trt_engine.schema.json",{"path":116,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":117,"sourceHtmlUrl":118},"schemas\u002Finference.schema.json",36791,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fschemas\u002Finference.schema.json",{"path":120,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":121,"sourceHtmlUrl":122},"schemas\u002Fmanifest.json",10538,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fschemas\u002Fmanifest.json",{"path":124,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":125,"sourceHtmlUrl":126},"schemas\u002Ftrain.schema.json",33292,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fschemas\u002Ftrain.schema.json",{"path":128,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":129,"sourceHtmlUrl":130},"skill-card.md",3889,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fskill-card.md",{"path":132,"resourceType":133,"resourceTypeDesc":134,"sizeBytes":135,"sourceHtmlUrl":136},"skill.oms.sig","OTHER","其他",8937,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fskill.oms.sig",{"path":138,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":139,"sourceHtmlUrl":140},"zeuz.manifest.yaml",172,"https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fcatalog\u002Fbundles\u002Fnvidia\u002Ftao-train-metric-learning-recognition\u002Fzeuz.manifest.yaml",[142,150,157,164,171,178],{"skillNo":143,"name":144,"nameZh":6,"description":145,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":147,"categories":148,"keywords":149,"publishedAt":18},"01M78JQ2R0QR8H2Z6662RFX804","gamma","Fixture skill gamma for discovery benchmarks.","MIT","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fskill-registry-mini\u002Fskills\u002Fgamma\u002FSKILL.md",[],[],{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":156,"publishedAt":18},"01M78JQ1KCVZ15APZG3FFF29K0","beta","Fixture skill beta for discovery benchmarks.","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fskill-registry-mini\u002Fskills\u002Fbeta\u002FSKILL.md",[],[],{"skillNo":158,"name":159,"nameZh":6,"description":160,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":161,"categories":162,"keywords":163,"publishedAt":18},"01M78JQ0FJTK4Q9JKZMX9SVSEN","alpha","Fixture skill alpha for discovery benchmarks.","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fskill-registry-mini\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":165,"name":166,"nameZh":6,"description":167,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":168,"categories":169,"keywords":170,"publishedAt":18},"01M78JPYVYPY7R8QS10KADQPWE","writing-for-agents","Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fskills\u002Fwriting-for-agents\u002FSKILL.md",[],[],{"skillNo":172,"name":173,"nameZh":6,"description":174,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":177,"publishedAt":18},"01M78JPVYHB88RF5ATZVGX84RR","wizard","Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fskills\u002Fwizard\u002FSKILL.md",[],[],{"skillNo":179,"name":180,"nameZh":6,"description":181,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":182,"categories":183,"keywords":184,"publishedAt":18},"01M78JPSSWNTZW89DWGJTX7716","wayfinder","Plan a huge chunk of work (more than one agent session can hold) as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.","https:\u002F\u002Fgithub.com\u002Fmatheussluzz\u002FZeuZ-Agent\u002Fblob\u002Fmain\u002Fskills\u002Fwayfinder\u002FSKILL.md",[],[]]