[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AABYSA4TDKXZB3S5V4J9MF":3,"skill-same:01M7AABYSA4TDKXZB3S5V4J9MF":141},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":34,"commitSha":35,"skillDirectory":36,"forksCount":37,"repositoryHtmlUrl":38,"repositoryHomepageUrl":39,"ownerHtmlUrl":40,"ownerBlogUrl":41,"ownerDescription":42,"updatedAt":43,"install":44,"resources":57},"01M7AABYSA4TDKXZB3S5V4J9MF","tao-train-sparse4d",null,"Sparse4D for multi-camera temporal 3D object detection and tracking. Uses sparse queries with deformable attention across camera views and time for end-to-end 3D perception, with an instance bank for temporal tracking. Use when training, evaluating, exporting, quantizing, or running inference for a TAO Sparse4D model. Trigger phrases include \"train Sparse4D\", \"multi-camera 3D detection\", \"temporal 3D tracker\", \"sparse query 3D perception\".","themusebook\u002Fskills","themusebook","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F349927247?v=4","01M7A5C1RV5S3V5FF6ZACVDNRN","musebook","01M7A5C1SFC3KRZAXNN60XJ7KM",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: tao-train-sparse4d\ndescription: Sparse4D for multi-camera temporal 3D object detection and tracking. Uses sparse queries with deformable\n  attention across camera views and time for end-to-end 3D perception, with an instance bank for temporal tracking. Use when\n  training, evaluating, exporting, quantizing, or running inference for a TAO Sparse4D model. Trigger phrases include\n  \"train Sparse4D\", \"multi-camera 3D detection\", \"temporal 3D tracker\", \"sparse query 3D perception\".\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- temporal\n- 3d\n- detection\n- tracking\n---\n\n# Sparse4D\n\nSparse4D for multi-camera temporal 3D object detection and tracking. Uses sparse queries with deformable attention across camera views and time for end-to-end 3D perception. Includes instance bank for temporal tracking.\n\nUse a pretrained ResNet-101 backbone when one is available by setting\n`train.pretrained_model_path`. For local smoke validation, Sparse4D training\ncan run with an empty `train.pretrained_model_path`, but production runs should\nstill use a compatible PTM.\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:** sparse4d\n- **Formats:** ovpkl\n- **Monitoring metric:** val_mAP\n- Current TAO Sparse4D training emits this value in status\u002Flogs as\n  `img_bbox_NuScenes\u002FmAP` and `mAP`; AutoML metric\n  extractors should treat those emitted keys as aliases for `val_mAP`.\n  Multi-fidelity AutoML algorithms such as Hyperband, ASHA, and BOHB may\n  promote a checkpoint to a resume job that completes without emitting a fresh\n  `val_mAP` alias. In that case, compare AutoML's carried metric to the source\n  rung job that emitted `img_bbox_NuScenes\u002FmAP` or `mAP`, while still verifying\n  that the promoted job resumed from the explicit epoch\u002Fstep checkpoint,\n  produced a real checkpoint, and is usable for evaluate\u002Finference.\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| dataset_convert | aicity.root | id |  | No |\n| evaluate | dataset.data_root | eval_dataset | (from convert job, spec: aicity.split) | No |\n| evaluate | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| evaluate | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| evaluate | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| evaluate | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | No |\n| export | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| inference | dataset.data_root | inference_dataset | (from convert job, spec: aicity.split) | No |\n| inference | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| inference | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| inference | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| inference | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.data_root | train_datasets | (from convert job, spec: aicity.split) | No |\n| quantize | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| quantize | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.quant_calibration_dataset.images_dir | train_datasets |  | No |\n| train | dataset.data_root | train_datasets | (from convert job, spec: aicity.split) | No |\n| train | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| train | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| train | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| train | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | 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\"\nS3_EVAL = \"s3:\u002F\u002Fbucket\u002Fdata\u002Feval\"\nCONVERTED_SCENE = \"\u003Cscene-from-converter>\"  # e.g. \"subsetscene+bev-sensor-random-0\"\n```\n\n**train (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"train.num_epochs\": 30,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"train.num_gpus\": 1,\n    \"dataset.sequences.split_num\": 90,\n    \"dataset.train_dataset.sequences_split_num\": 90,\n    \"dataset.data_root\": f\"{S3_TRAIN}\u002Ftrain\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"dataset.data_root\": f\"{S3_EVAL}\u002Fval\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n}\n```\n\n**export (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n}\n```\n\n**inference (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"dataset.data_root\": f\"{S3_EVAL}\u002Ftest\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n}\n```\n\n**quantize (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"dataset.data_root\": f\"{S3_TRAIN}\u002Ftrain\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_TRAIN}\",\n}\n```\n\nSee `references\u002Flocal_docker_conversion.md` for local-docker conversion roots and mounts, H5 depth-path normalization, converted annotation filenames, smoke-run `max_num_cams`\u002Fanchor contracts for export compatibility, and converted-artifact verification before train\u002Fevaluate\u002Finference.\n\n## Eval Dataset\n\nOptional. Val\u002Ftest splits configured via dataset ann_file paths.\n\n## Important Parameters\n\n- **model.backbone**: Backbone. Default resnet_101.\n- **model.neck.out_channels**: FPN output channels. Default 256. num_outs=4.\n- **model.input_shape**: Input image shape [W, H]. Default [1408, 512].\n- **model.head.num_output**: Number of detection output queries. Default 300.\n- **model.head.num_decoder**: Number of decoder layers. Default 6.\n- **model.head.temporal**: Enable temporal reasoning. Default True.\n- **model.head.instance_bank.num_anchor**: Instance bank anchors. Default 900.\n- **model.head.instance_bank.num_temp_instances**: Temporal instance count. Default 600.\n- **model.depth_branch.loss_weight**: Depth supervision loss weight. Default 0.2.\n- **dataset.batch_size**: Per-GPU batch size. Default 2.\n- **dataset.num_frames**: Sequence length. Default 200.\n- **dataset.classes**: Detection classes. Default [person, gr1_t2, agility_digit, nova_carter]. num_ids=70 for tracking.\n- **train.optim.lr**: Learning rate. Default 5e-5. img_backbone lr_mult=0.2.\n- **train.lr_scheduler**: Cosine scheduler with linear warmup (500 iters, ratio 0.333).\n- **train.grad_clip.max_norm**: Gradient clipping. Default 25.\n- **train.precision**: Options: bf16, fp16, fp32. Default bf16.\n- **evaluate.metrics**: Eval metrics. Default [\"detection\"]. Optional tracking evaluation.\n- **evaluate.tracking.enabled**: Enable tracking evaluation. tracking_threshold=0.2.\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| `train.num_nodes` | Number of nodes | 1 |\n\n- Multi-GPU strategy: `ddp_find_unused_parameters_true` (no fsdp support)\n- `sync_batchnorm` is always enabled (True)\n- Iterations per epoch computed as: `num_frames * num_bev_groups \u002F (num_nodes * num_gpus * batch_size)`\n- **Scaling:** When increasing GPUs, effective batch size grows and iterations-per-epoch shrinks proportionally\n\n**Multi-node env vars** (set by orchestrator): `WORLD_SIZE`, `NODE_RANK`, `MASTER_ADDR`, `MASTER_PORT`, `NUM_GPU_PER_NODE`.\n\n## Hardware\n\nMinimum 2 GPU(s), recommended 8 GPU(s). 40GB+ (A100 recommended) VRAM per GPU. Multi-camera temporal model is memory intensive. bf16 required for practical training. Multi-GPU strongly recommended. Instance bank requires substantial memory for temporal reasoning.\n\n## Error Patterns\n\n**dataset_convert required**: Must run dataset_convert first to produce annotation pickles and anchor_init.npy.\n\n**dataset_convert container\u002Fcommand**: Sparse4D conversion is an AICity to\nOVPKL annotations conversion. Launch `dataset_convert` with the action-level\n`tao_toolkit.data_services` image and `annotations convert -e {config_path}`;\ndo not use the PyTorch `sparse4d` CLI for conversion. Train\u002Fevaluate\u002Fexport\u002F\ninference still use the model-level PyTorch image.\n\n**Stable raw-data path**: The AICity to OVPKL converter writes image paths into\nthe generated pickle files. Keep `aicity.root` at `\u002Fdata\u002Faicity_root` during\nconversion, then point `dataset.data_root` at the split folder, for example\n`\u002Fdata\u002Faicity_root\u002Ftrain` for training or `\u002Fdata\u002Faicity_root\u002Fval` for\nevaluation. This preserves the converter's absolute RGB paths and relative\ndepth paths.\n\n**H5 depth tuple mismatch**: If training fails with an H5 path error where the\ntrainer tries to open a camera directory such as\n`\u002Fdata\u002Faicity_root\u002Ftrain\u002F\u003Cscene>\u002FCamera`, run\n`models\u002Fsparse4d\u002Fscripts\u002Fnormalize_depth_paths.py --data-root \u003Chost-aicity-root>\u002Ftrain \u003Cconverted-ann-dir>`\nafter `dataset_convert` and before train\u002Fevaluate\u002Finference. The helper rewrites\nconverted `depth_map_path` tuples to point at\n`\u003Cscene>\u002Fdepth_maps\u002F\u003Ccamera>.h5` with the H5 dataset key basename.\n\n**Missing anchor file**: Set model.head.instance_bank.anchor to the anchor_init.npy path from dataset_convert results.\n\n**Temporal OOM**: Reduce dataset.num_frames or dataset.batch_size if running out of memory during temporal training.\n\n**Quantize image compatibility**: The model-skill wiring should pass\n`quantize.model_path` through the parent-model resolver, and checkpoint handoff\nshould select the exact epoch\u002Fstep checkpoint just like evaluate, inference,\nexport, and resume. TorchAO checkpoint quantization passes in the\n`validation-fixes-20260525` PyT image and writes\n`quantized_model_torchao.pth`. Older 7.0.0-rc PyT images may fail inside the\nSparse4D quantize entrypoint or lack ONNX quantization dependencies; do not\nremove or skip the advertised `quantize` action if that occurs. Report the\ncontainer\u002Fimage failure and keep the exact checkpoint path visible.\n\n## Spec Param \u002F Parent Model Inference\n\nSee `references\u002Fspec_param_inference.md` for the model-specific inference mappings from TAO Core `sparse4d.config.json` (the per-action spec-field to inference-function table) and the `parent_model`\u002F`parent_job_id` checkpoint-resolution rules that generated runners apply with SDK helpers before `create_job()`.\n","# Sparse4D\n\nSparse4D for multi-camera temporal 3D object detection and tracking. Uses sparse queries with deformable attention across camera views and time for end-to-end 3D perception. Includes instance bank for temporal tracking.\n\nUse a pretrained ResNet-101 backbone when one is available by setting\n`train.pretrained_model_path`. For local smoke validation, Sparse4D training\ncan run with an empty `train.pretrained_model_path`, but production runs should\nstill use a compatible PTM.\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:** sparse4d\n- **Formats:** ovpkl\n- **Monitoring metric:** val_mAP\n- Current TAO Sparse4D training emits this value in status\u002Flogs as\n  `img_bbox_NuScenes\u002FmAP` and `mAP`; AutoML metric\n  extractors should treat those emitted keys as aliases for `val_mAP`.\n  Multi-fidelity AutoML algorithms such as Hyperband, ASHA, and BOHB may\n  promote a checkpoint to a resume job that completes without emitting a fresh\n  `val_mAP` alias. In that case, compare AutoML's carried metric to the source\n  rung job that emitted `img_bbox_NuScenes\u002FmAP` or `mAP`, while still verifying\n  that the promoted job resumed from the explicit epoch\u002Fstep checkpoint,\n  produced a real checkpoint, and is usable for evaluate\u002Finference.\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| dataset_convert | aicity.root | id |  | No |\n| evaluate | dataset.data_root | eval_dataset | (from convert job, spec: aicity.split) | No |\n| evaluate | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| evaluate | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| evaluate | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| evaluate | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | No |\n| export | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| inference | dataset.data_root | inference_dataset | (from convert job, spec: aicity.split) | No |\n| inference | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| inference | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| inference | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| inference | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.data_root | train_datasets | (from convert job, spec: aicity.split) | No |\n| quantize | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| quantize | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | No |\n| quantize | dataset.quant_calibration_dataset.images_dir | train_datasets |  | No |\n| train | dataset.data_root | train_datasets | (from convert job, spec: aicity.split) | No |\n| train | model.head.instance_bank.anchor | train_datasets | \u002Fresults\u002F{dataset_convert_job_id}\u002Fanchor_init.npy | No |\n| train | dataset.train_dataset.ann_file | train_datasets | (from convert job, spec: aicity.split) | No |\n| train | dataset.val_dataset.ann_file | eval_dataset | (from convert job, spec: aicity.split) | No |\n| train | dataset.test_dataset.ann_file | inference_dataset | (from convert job, spec: aicity.split) | 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\"\nS3_EVAL = \"s3:\u002F\u002Fbucket\u002Fdata\u002Feval\"\nCONVERTED_SCENE = \"\u003Cscene-from-converter>\"  # e.g. \"subsetscene+bev-sensor-random-0\"\n```\n\n**train (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"train.num_epochs\": 30,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"train.num_gpus\": 1,\n    \"dataset.sequences.split_num\": 90,\n    \"dataset.train_dataset.sequences_split_num\": 90,\n    \"dataset.data_root\": f\"{S3_TRAIN}\u002Ftrain\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"dataset.data_root\": f\"{S3_EVAL}\u002Fval\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n}\n```\n\n**export (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n}\n```\n\n**inference (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"dataset.data_root\": f\"{S3_EVAL}\u002Ftest\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n}\n```\n\n**quantize (mandatory data sources):**\n```python\nCONVERTED = \"s3:\u002F\u002Fbucket\u002Fresults\u002F\u003Cdataset_convert_job_id>\"\n{\n    \"dataset.data_root\": f\"{S3_TRAIN}\u002Ftrain\",\n    \"model.head.instance_bank.anchor\": f\"{CONVERTED}\u002Fanchor_init.npy\",\n    \"dataset.train_dataset.ann_file\": f\"{CONVERTED}\u002Ftrain\u002F{CONVERTED_SCENE}_infos_train.pkl\",\n    \"dataset.val_dataset.ann_file\": f\"{CONVERTED}\u002Fval\u002F{CONVERTED_SCENE}_infos_val.pkl\",\n    \"dataset.test_dataset.ann_file\": f\"{CONVERTED}\u002Ftest\u002F{CONVERTED_SCENE}_infos_test.pkl\",\n    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_TRAIN}\",\n}\n```\n\nSee `references\u002Flocal_docker_conversion.md` for local-docker conversion roots and mounts, H5 depth-path normalization, converted annotation filenames, smoke-run `max_num_cams`\u002Fanchor contracts for export compatibility, and converted-artifact verification before train\u002Fevaluate\u002Finference.\n\n## Eval Dataset\n\nOptional. Val\u002Ftest splits configured via dataset ann_file paths.\n\n## Important Parameters\n\n- **model.backbone**: Backbone. Default resnet_101.\n- **model.neck.out_channels**: FPN output channels. Default 256. num_outs=4.\n- **model.input_shape**: Input image shape [W, H]. Default [1408, 512].\n- **model.head.num_output**: Number of detection output queries. Default 300.\n- **model.head.num_decoder**: Number of decoder layers. Default 6.\n- **model.head.temporal**: Enable temporal reasoning. Default True.\n- **model.head.instance_bank.num_anchor**: Instance bank anchors. Default 900.\n- **model.head.instance_bank.num_temp_instances**: Temporal instance count. Default 600.\n- **model.depth_branch.loss_weight**: Depth supervision loss weight. Default 0.2.\n- **dataset.batch_size**: Per-GPU batch size. Default 2.\n- **dataset.num_frames**: Sequence length. Default 200.\n- **dataset.classes**: Detection classes. Default [person, gr1_t2, agility_digit, nova_carter]. num_ids=70 for tracking.\n- **train.optim.lr**: Learning rate. Default 5e-5. img_backbone lr_mult=0.2.\n- **train.lr_scheduler**: Cosine scheduler with linear warmup (500 iters, ratio 0.333).\n- **train.grad_clip.max_norm**: Gradient clipping. Default 25.\n- **train.precision**: Options: bf16, fp16, fp32. Default bf16.\n- **evaluate.metrics**: Eval metrics. Default [\"detection\"]. Optional tracking evaluation.\n- **evaluate.tracking.enabled**: Enable tracking evaluation. tracking_threshold=0.2.\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| `train.num_nodes` | Number of nodes | 1 |\n\n- Multi-GPU strategy: `ddp_find_unused_parameters_true` (no fsdp support)\n- `sync_batchnorm` is always enabled (True)\n- Iterations per epoch computed as: `num_frames * num_bev_groups \u002F (num_nodes * num_gpus * batch_size)`\n- **Scaling:** When increasing GPUs, effective batch size grows and iterations-per-epoch shrinks proportionally\n\n**Multi-node env vars** (set by orchestrator): `WORLD_SIZE`, `NODE_RANK`, `MASTER_ADDR`, `MASTER_PORT`, `NUM_GPU_PER_NODE`.\n\n## Hardware\n\nMinimum 2 GPU(s), recommended 8 GPU(s). 40GB+ (A100 recommended) VRAM per GPU. Multi-camera temporal model is memory intensive. bf16 required for practical training. Multi-GPU strongly recommended. Instance bank requires substantial memory for temporal reasoning.\n\n## Error Patterns\n\n**dataset_convert required**: Must run dataset_convert first to produce annotation pickles and anchor_init.npy.\n\n**dataset_convert container\u002Fcommand**: Sparse4D conversion is an AICity to\nOVPKL annotations conversion. Launch `dataset_convert` with the action-level\n`tao_toolkit.data_services` image and `annotations convert -e {config_path}`;\ndo not use the PyTorch `sparse4d` CLI for conversion. Train\u002Fevaluate\u002Fexport\u002F\ninference still use the model-level PyTorch image.\n\n**Stable raw-data path**: The AICity to OVPKL converter writes image paths into\nthe generated pickle files. Keep `aicity.root` at `\u002Fdata\u002Faicity_root` during\nconversion, then point `dataset.data_root` at the split folder, for example\n`\u002Fdata\u002Faicity_root\u002Ftrain` for training or `\u002Fdata\u002Faicity_root\u002Fval` for\nevaluation. This preserves the converter's absolute RGB paths and relative\ndepth paths.\n\n**H5 depth tuple mismatch**: If training fails with an H5 path error where the\ntrainer tries to open a camera directory such as\n`\u002Fdata\u002Faicity_root\u002Ftrain\u002F\u003Cscene>\u002FCamera`, run\n`models\u002Fsparse4d\u002Fscripts\u002Fnormalize_depth_paths.py --data-root \u003Chost-aicity-root>\u002Ftrain \u003Cconverted-ann-dir>`\nafter `dataset_convert` and before train\u002Fevaluate\u002Finference. The helper rewrites\nconverted `depth_map_path` tuples to point at\n`\u003Cscene>\u002Fdepth_maps\u002F\u003Ccamera>.h5` with the H5 dataset key basename.\n\n**Missing anchor file**: Set model.head.instance_bank.anchor to the anchor_init.npy path from dataset_convert results.\n\n**Temporal OOM**: Reduce dataset.num_frames or dataset.batch_size if running out of memory during temporal training.\n\n**Quantize image compatibility**: The model-skill wiring should pass\n`quantize.model_path` through the parent-model resolver, and checkpoint handoff\nshould select the exact epoch\u002Fstep checkpoint just like evaluate, inference,\nexport, and resume. TorchAO checkpoint quantization passes in the\n`validation-fixes-20260525` PyT image and writes\n`quantized_model_torchao.pth`. Older 7.0.0-rc PyT images may fail inside the\nSparse4D quantize entrypoint or lack ONNX quantization dependencies; do not\nremove or skip the advertised `quantize` action if that occurs. Report the\ncontainer\u002Fimage failure and keep the exact checkpoint path visible.\n\n## Spec Param \u002F Parent Model Inference\n\nSee `references\u002Fspec_param_inference.md` for the model-specific inference mappings from TAO Core `sparse4d.config.json` (the per-action spec-field to inference-function table) and the `parent_model`\u002F`parent_job_id` checkpoint-resolution rules that generated runners apply with SDK helpers before `create_job()`.\n",{"name":5,"tags":24,"license":15,"metadata":29,"description":7,"allowed-tools":32,"compatibility":33},[25,26,27,28],"temporal","3d","detection","tracking",{"author":30,"version":31},"NVIDIA Corporation","0.1.0","Read Bash","Requires docker + nvidia-container-toolkit.","main","1c523aba5cdf867a79f078847075492d7778092a","skills\u002Fnvidia\u002Ftao-train-sparse4d",1,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills","https:\u002F\u002Fskills.x402.wtf","https:\u002F\u002Fgithub.com\u002Fthemusebook","https:\u002F\u002Fmusebook.trade","The official Github of musebook","2026-11-02 23:42:26",{"source":8,"skillName":5,"commands":45,"prompt":55,"zipAvailable":56},[46,49,52],{"tool":47,"command":48},"npx","npx skills add themusebook\u002Fskills --skill tao-train-sparse4d",{"tool":50,"command":51},"bunx","bunx skills add themusebook\u002Fskills --skill tao-train-sparse4d",{"tool":53,"command":54},"pnpm","pnpm dlx skills add themusebook\u002Fskills --skill tao-train-sparse4d","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002FSKILL.md",true,[58,64,68,72,76,80,84,88,92,96,100,103,107,111,115,119,123,127,131,135],{"path":59,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":62,"sourceHtmlUrl":63},"BENCHMARK.md","REFERENCE","参考文档",4062,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002FBENCHMARK.md",{"path":65,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":66,"sourceHtmlUrl":67},"evals\u002Fevals.json",804,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fevals\u002Fevals.json",{"path":69,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":70,"sourceHtmlUrl":71},"references\u002Flocal_docker_conversion.md",3739,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Flocal_docker_conversion.md",{"path":73,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":74,"sourceHtmlUrl":75},"references\u002Fskill_info.yaml",7625,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fskill_info.yaml",{"path":77,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":78,"sourceHtmlUrl":79},"references\u002Fspec_param_inference.md",2256,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_param_inference.md",{"path":81,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":82,"sourceHtmlUrl":83},"references\u002Fspec_template_dataset_convert.yaml",1302,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_template_dataset_convert.yaml",{"path":85,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":86,"sourceHtmlUrl":87},"references\u002Fspec_template_evaluate.yaml",7130,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_template_evaluate.yaml",{"path":89,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":90,"sourceHtmlUrl":91},"references\u002Fspec_template_export.yaml",7146,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_template_export.yaml",{"path":93,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":94,"sourceHtmlUrl":95},"references\u002Fspec_template_inference.yaml",7163,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_template_inference.yaml",{"path":97,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":98,"sourceHtmlUrl":99},"references\u002Fspec_template_quantize.yaml",6930,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_template_quantize.yaml",{"path":101,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":98,"sourceHtmlUrl":102},"references\u002Fspec_template_train.yaml","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Freferences\u002Fspec_template_train.yaml",{"path":104,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":105,"sourceHtmlUrl":106},"schemas\u002Fdataset_convert.schema.json",11148,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Fdataset_convert.schema.json",{"path":108,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":109,"sourceHtmlUrl":110},"schemas\u002Fevaluate.schema.json",113837,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Fevaluate.schema.json",{"path":112,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":113,"sourceHtmlUrl":114},"schemas\u002Fexport.schema.json",113416,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Fexport.schema.json",{"path":116,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":117,"sourceHtmlUrl":118},"schemas\u002Finference.schema.json",113971,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Finference.schema.json",{"path":120,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":121,"sourceHtmlUrl":122},"schemas\u002Fmanifest.json",16273,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Fmanifest.json",{"path":124,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":125,"sourceHtmlUrl":126},"schemas\u002Fquantize.schema.json",109870,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Fquantize.schema.json",{"path":128,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":129,"sourceHtmlUrl":130},"schemas\u002Ftrain.schema.json",109864,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fschemas\u002Ftrain.schema.json",{"path":132,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":133,"sourceHtmlUrl":134},"skill-card.md",3923,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fskill-card.md",{"path":136,"resourceType":137,"resourceTypeDesc":138,"sizeBytes":139,"sourceHtmlUrl":140},"skill.oms.sig","OTHER","其他",8549,"https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-sparse4d\u002Fskill.oms.sig",[142,151,158,165,172,179],{"skillNo":143,"name":144,"nameZh":6,"description":145,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":146,"licenseFileUrl":147,"sourceHtmlUrl":148,"categories":149,"keywords":150,"publishedAt":19},"01M7AB34SPKVSDRRG009PM1CZH","youtube-clipper","YouTube 视频智能剪辑工具。下载视频和字幕，AI 分析生成精细章节（几分钟级别）， 用户选择片段后自动剪辑、翻译字幕为中英双语、烧录字幕到视频，并生成总结文案。 使用场景：当用户需要剪辑 YouTube 视频、生成短视频片段、制作双语字幕版本时。 关键词：视频剪辑、YouTube、字幕翻译、双语字幕、视频下载、clip video","MIT","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fyoutube-clipper\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fyoutube-clipper\u002FSKILL.md",[],[],{"skillNo":152,"name":153,"nameZh":6,"description":154,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":155,"categories":156,"keywords":157,"publishedAt":19},"01M7AB18BH6N7KR7939HRNWVMQ","viem-integration","Integrate EVM blockchains using viem. Use when user says \"read blockchain data\", \"send transaction\", \"interact with smart contract\", \"connect to Ethereum\", \"use viem\", \"use wagmi\", \"wallet integration\", \"viem setup\", or mentions blockchain\u002FEVM development with TypeScript.","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fviem-integration\u002FSKILL.md",[],[],{"skillNo":159,"name":160,"nameZh":6,"description":161,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":162,"categories":163,"keywords":164,"publishedAt":19},"01M7AB12GHGP48WCNXDQQ9222H","v4-security-foundations","Security-first Uniswap v4 hook development. Use when user mentions \"v4 hooks\", \"hook security\", \"PoolManager\", \"beforeSwap\", \"afterSwap\", or asks about V4 hook best practices, vulnerabilities, or audit requirements.","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fv4-security-foundations\u002FSKILL.md",[],[],{"skillNo":166,"name":167,"nameZh":6,"description":168,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":169,"categories":170,"keywords":171,"publishedAt":19},"01M7AB0XM3PSR7V3Q2ZRR0Q0FB","v4-sdk-integration","App-layer SDK guide for building swap and liquidity experiences directly with the Uniswap v4 SDK. Use when user asks about \"v4 sdk\", \"uniswap v4\", \"v4 swap\", \"v4 liquidity\", \"PoolManager\", \"V4Planner\", \"StateView\", \"PositionManager\", \"pool state\", \"v4 position\", \"uniswap sdk\", or when building swap\u002Fliquidity UX directly with SDKs rather than via the Trading API.","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fv4-sdk-integration\u002FSKILL.md",[],[],{"skillNo":173,"name":174,"nameZh":6,"description":175,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":176,"categories":177,"keywords":178,"publishedAt":19},"01M7AB0SGA93ASA588DJW4P9WT","v4-hook-generator","Generate Uniswap v4 hook contracts via OpenZeppelin MCP. Use when building custom swap logic, async swaps, hook-owned liquidity, custom curves, dynamic fees, MEV protection, limit orders, or oracle hooks.","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fv4-hook-generator\u002FSKILL.md",[],[],{"skillNo":180,"name":181,"nameZh":6,"description":182,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":146,"licenseFileUrl":6,"sourceHtmlUrl":183,"categories":184,"keywords":185,"publishedAt":19},"01M7AAZYH75H4CS03WSW2EZ8DM","testing","For testing with Light Protocol programs and clients on localnet, devnet, and mainnet validation.","https:\u002F\u002Fgithub.com\u002Fthemusebook\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftesting\u002FSKILL.md",[],[]]