[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-tao-train-oneformer":3,"mdc-2kym1v-key":31,"related-repo-nvidia-tao-train-oneformer":3481,"related-org-nvidia-tao-train-oneformer":3585},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":20,"repoUrl":21,"updatedAt":22,"license":23,"forks":24,"topics":25,"repo":26,"sourceUrl":29,"mdContent":30},"tao-train-oneformer","train OneFormer models for image segmentation","OneFormer for universal image segmentation. Unifies panoptic, instance, and semantic segmentation with a single architecture using task-conditioned queries. Use when training, evaluating, exporting, quantizing, or running inference for a TAO OneFormer model. Trigger phrases include \"train OneFormer\", \"universal segmentation\", \"task-conditioned segmentation\", \"panoptic \u002F instance \u002F semantic in one model\".",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,17],{"name":13,"slug":14,"type":15},"Deep Learning","deep-learning","tag",{"name":9,"slug":8,"type":15},{"name":18,"slug":19,"type":15},"Computer Vision","computer-vision",2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-14T05:30:22.895491","Apache-2.0",281,[],{"repoUrl":21,"stars":20,"forks":24,"topics":27,"description":28},[],"AI agent skills published by NVIDIA","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Ftao-train-oneformer","---\nname: tao-train-oneformer\ndescription: OneFormer for universal image segmentation. Unifies panoptic, instance, and semantic segmentation with a\n  single architecture using task-conditioned queries. Use when training, evaluating, exporting, quantizing, or running\n  inference for a TAO OneFormer model. Trigger phrases include \"train OneFormer\", \"universal segmentation\",\n  \"task-conditioned segmentation\", \"panoptic \u002F instance \u002F semantic in one model\".\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- segmentation\n---\n\n# OneFormer\n\nOneFormer for universal image segmentation. Unifies panoptic, instance, and semantic segmentation with a single architecture using task-conditioned queries.\n\nSet train.pretrained_backbone and\u002For train.pretrained_model.\n\nFor TAO Deploy TensorRT actions (`gen_trt_engine`, TensorRT `evaluate`, and TensorRT `inference`), read `references\u002Ftao-deploy-oneformer.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:** segmentation\n- **Formats:** coco_panoptic, coco\n- **Monitoring metric:** mIoU\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| evaluate | dataset.train.images | train_datasets | images.tar.gz | No |\n| evaluate | dataset.label_map | train_datasets | label_map.json | No |\n| evaluate | dataset.train.annotations | train_datasets | annotations.json | No |\n| evaluate | dataset.train.panoptic | train_datasets | images_panoptic.tar.gz | No |\n| evaluate | dataset.val.images | eval_dataset | images.tar.gz | No |\n| evaluate | dataset.val.annotations | eval_dataset | annotations.json | No |\n| evaluate | dataset.val.panoptic | eval_dataset | images_panoptic.tar.gz | No |\n| evaluate | dataset.test.images | eval_dataset | images.tar.gz | No |\n| evaluate | dataset.test.annotations | eval_dataset | annotations.json | No |\n| evaluate | dataset.test.panoptic | eval_dataset | images_panoptic.tar.gz | No |\n| inference | dataset.train.images | train_datasets | images.tar.gz | No |\n| inference | dataset.label_map | train_datasets | label_map.json | No |\n| inference | dataset.train.annotations | train_datasets | annotations.json | No |\n| inference | dataset.train.panoptic | train_datasets | images_panoptic.tar.gz | No |\n| inference | dataset.val.images | eval_dataset | images.tar.gz | No |\n| inference | dataset.val.annotations | eval_dataset | annotations.json | No |\n| inference | dataset.val.panoptic | eval_dataset | images_panoptic.tar.gz | No |\n| inference | dataset.test.images | inference_dataset | images.tar.gz | No |\n| quantize | dataset.train.images | train_datasets | images.tar.gz | No |\n| quantize | dataset.train.annotations | train_datasets | annotations.json | No |\n| quantize | dataset.label_map | train_datasets | label_map.json | No |\n| quantize | dataset.train.panoptic | train_datasets | images_panoptic.tar.gz | No |\n| quantize | dataset.val.images | eval_dataset | images.tar.gz | No |\n| quantize | dataset.val.annotations | eval_dataset | annotations.json | No |\n| quantize | dataset.val.panoptic | eval_dataset | images_panoptic.tar.gz | No |\n| quantize | dataset.test.images | eval_dataset | images.tar.gz | No |\n| quantize | dataset.quant_calibration_dataset.images_dir | calibration_dataset | images.tar.gz | No |\n| train | dataset.train.images | train_datasets | images.tar.gz | No |\n| train | dataset.train.annotations | train_datasets | annotations.json | No |\n| train | dataset.label_map | train_datasets | label_map.json | No |\n| train | dataset.train.panoptic | train_datasets | images_panoptic.tar.gz | No |\n| train | dataset.val.images | eval_dataset | images.tar.gz | No |\n| train | dataset.val.annotations | eval_dataset | annotations.json | No |\n| train | dataset.val.panoptic | eval_dataset | images_panoptic.tar.gz | No |\n| train | dataset.test.images | eval_dataset | images.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\"\nS3_EVAL = \"s3:\u002F\u002Fbucket\u002Fdata\u002Feval\"\nS3_INFERENCE = \"s3:\u002F\u002Fbucket\u002Fdata\u002Finference\"\nS3_CALIBRATION = \"s3:\u002F\u002Fbucket\u002Fdata\u002Fcalibration\"\n```\n\n**train (mandatory data sources):**\n```python\n{\n    \"train.num_gpus\": 1,\n    \"train.num_epochs\": 10,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"model.sem_seg_head.num_classes\": 133,\n    \"dataset.contiguous_id\": True,\n    \"train.precision\": \"32\",\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\n{\n    \"evaluate.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"model.sem_seg_head.num_classes\": 133,\n    \"dataset.contiguous_id\": True,\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.test.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.test.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n}\n```\n\n**export:**\n```python\n{\n    \"export.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"model.sem_seg_head.num_classes\": 133,\n    \"model.export\": True,\n    \"export.onnx_file\": \"\u002Fresults\u002Foneformer_export_640.onnx\",\n}\n```\n\n**inference (mandatory data sources):**\n```python\n{\n    \"inference.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_INFERENCE}\u002Fimages.tar.gz\",\n    \"inference.images_dir\": f\"{S3_INFERENCE}\u002Fimages.tar.gz\",\n}\n```\n\n**quantize (mandatory data sources):**\n```python\n{\n    \"quantize.model_path\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_CALIBRATION}\u002Fimages.tar.gz\",\n}\n```\n\n## Checkpoint Selection\n\nOneFormer training writes epoch-step checkpoints such as\n`model_epoch_000_step_00017.pth` and may also write a\n`oneformer_model_latest.pth` symlink. For checkpoint-dependent actions, use the\nmodel-skill or SDK parent-model resolver and pass the exact selected checkpoint\npath into `evaluate.checkpoint`, `inference.checkpoint`, `export.checkpoint`,\n`quantize.model_path`, or `train.resume_training_checkpoint_path`. Do not pick\nthe `oneformer_model_latest.pth` symlink by name unless the user explicitly asks\nfor latest checkpoint behavior. If the resolver reports a best checkpoint, use\nthat best checkpoint for evaluation\u002Fexport\u002Finference; if the user asks for a\nspecific epoch or step, use the matching epoch-step checkpoint.\n## Eval Dataset\n\nOptional. Val data configured alongside train in the dataset config.\n\n## Important Parameters\n\n- **model.sem_seg_head.num_classes**: Number of segmentation class indices available to the head. Default 133 for COCO panoptic data when `dataset.contiguous_id: True` remaps raw category ids through the label map. Do not shrink this to a global workflow class count unless the label map and annotations have actually been reduced to that class set.\n- **model.one_former.hidden_dim**: Keep at 256 for local smoke runs unless\n  the text encoder width is changed in lock-step. Reducing hidden_dim alone\n  causes a text feature\u002Fcontext dimension mismatch during training.\n- **model.backbone.name**: Default D2SwinTransformer (Swin-based). embed_dim=192, depths=[2,2,18,2] by default.\n- **train.num_epochs**: Default 50 — significantly higher than most TAO models. OneFormer needs more epochs for convergence.\n- **train.optim.lr**: Learning rate. Default 1e-5. Lower than Mask2Former's 2e-4.\n- **model.task_toggling**: Enable\u002Fdisable specific tasks: semantic_on, instance_on, panoptic_on.\n- **export.task**: Export task mode. Options: semantic, instance, panoptic. Default semantic. Export input defaults to 640x640.\n- **inference.mode**: Inference mode. Options: semantic, instance, panoptic. Default semantic. image_size defaults to [1024, 1024].\n- **evaluate.iou_per_class**: Report per-class IoU in evaluation. Default True.\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- Uses explicit `DDPStrategy` with `find_unused_parameters=True`, `gradient_as_bucket_view=True`, `process_group_backend=\"nccl\"`\n- `sync_batchnorm` is always enabled\n- No fsdp support — DDP only\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 4 GPU(s). 24GB+ (A100 recommended) VRAM per GPU. OneFormer is memory-intensive like Mask2Former. batch_size=1 is the default. Multi-GPU needed for reasonable training speed, especially with 50 epochs.\n\n## Error Patterns\n\n**CUDA out of memory**: batch_size is already 1. Reduce image resolution or use a smaller Swin configuration.\n\n**Extracted S3 tarball points one level too high**: For local Docker runs,\n`images.tar.gz` and `images_panoptic.tar.gz` may extract wrapper directories\nsuch as `images\u002F` and `images_panoptic\u002F`. Set `dataset.*.images`,\n`dataset.*.panoptic`, `inference.images_dir`, and quantization calibration\npaths to the actual folder containing image or panoptic files, not the wrapper\ndirectory. A one-level-too-high path fails with `FileNotFoundError` for the\nfirst annotation image even though recursive file counts look correct.\n\n**default_specs missing results_dir**: The CLI `default_specs` subtask ignores\n`-e` experiment specs for `results_dir`; pass a Hydra-style override instead:\n`oneformer default_specs results_dir=\u002Fpath\u002Fto\u002Fdefault_specs`.\n\n**Invalid Lightning precision `fp32`**: Use `train.precision: \"32\"` in\ntrain\u002FAutoML\u002Fevaluate\u002Finference specs. The current Lightning stack rejects the\nlegacy `fp32` string.\n\n**PyTorch 2.6 checkpoint load failure on downstream actions**: Current\nOneFormer checkpoints include 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,\nquantize, or resume job env vars so Lightning can load the full checkpoint.\nDo not use this env var for untrusted checkpoints.\n\n**CUDA device-side assert in matcher\u002Fclass cost**: If training fails in\n`oneformer\u002Futils\u002Fmatcher.py` while indexing `out_prob[:, tgt_ids]`, compare\nthe effective target ids with `model.sem_seg_head.num_classes`. The packaged\nCOCO panoptic sample has 133 compact classes after `dataset.contiguous_id:\nTrue` remapping, so use `model.sem_seg_head.num_classes: 133` even when a\nbroader validation workflow passes a smaller generic `num_classes` value.\nOnly use a smaller class count when the label map and annotations are reduced\nto that exact contiguous class set.\n\n**Inference returns PASS with no predictions**: OneFormer prediction reads\n`inference.images_dir`, not `dataset.test.images`. Declare and populate\n`inference.images_dir` with the image folder or tarball for every inference\nrun. `dataset.test.images` may still be useful for shared dataset context, but\nit does not drive the PyTorch predict dataloader.\n\n**Export output path pre-created as a directory**: Do not declare\n`export.onnx_file` as a file output. The OneFormer exporter asserts that the\nONNX path does not already exist, while the local runner pre-creates declared\noutput paths. Set `export.onnx_file` explicitly in the spec to a non-existing\nfile path under the mounted results tree. Keep the default 640x640 export\nshape for smoke validation; very small export shapes can trigger PyTorch ONNX\nshape-inference failures.\n\n**Quantize cannot find the training label map from an AutoML checkpoint**:\nOneFormer Lightning checkpoints retain train-time absolute dataset paths in\ntheir saved hparams. When running downstream actions from an AutoML child\ncheckpoint, keep the parent AutoML job directory accessible at its original\n`\u002Fresults\u002F\u003Cjob_id>` path inside the action container in addition to passing the\nresolved checkpoint path. Otherwise quantize can fail while loading checkpoint\nhparams even when the current spec includes a valid `dataset.label_map`.\n\n**Slow training**: 50 default epochs with batch_size=1 is slow on single GPU. Use multi-GPU distributed training.\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 `oneformer.config.json`:\n\n| Action | Spec Field | Inference Function | Meaning |\n|---|---|---|---|\n| evaluate | `encryption_key` | `key` | encryption key |\n| evaluate | `evaluate.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| evaluate | `evaluate.trt_engine` | `parent_model` | model file inferred from the parent job results folder |\n| evaluate | `results_dir` | `output_dir` | current job results directory |\n| export | `encryption_key` | `key` | encryption key |\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| gen_trt_engine | `encryption_key` | `key` | encryption key |\n| gen_trt_engine | `gen_trt_engine.onnx_file` | `parent_model` | model file inferred from the parent job results folder |\n| gen_trt_engine | `gen_trt_engine.trt_engine` | `create_engine_file` | output TensorRT engine path |\n| gen_trt_engine | `results_dir` | `output_dir` | current job results directory |\n| inference | `encryption_key` | `key` | encryption key |\n| inference | `inference.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| inference | `inference.trt_engine` | `parent_model` | model file inferred from the parent job results folder |\n| inference | `results_dir` | `output_dir` | current job results directory |\n| quantize | `encryption_key` | `key` | encryption key |\n| quantize | `quantize.model_path` | `parent_model` | model file inferred from the parent job results folder |\n| quantize | `results_dir` | `output_dir` | current job results directory |\n| train | `encryption_key` | `key` | encryption key |\n| train | `results_dir` | `output_dir` | current job results directory |\n| train | `train.pretrained_backbone` | `{'link': 'https:\u002F\u002Fgithub.com\u002FSwinTransformer\u002Fstorage\u002Freleases\u002Fdownload\u002Fv1.0.8\u002Fswin_tiny_patch4_window7_224_22k.pth', 'destination_path': '\u002Fptm\u002Fmask2former\u002Fswin_tiny_patch4_window7_224_22k\u002Fswin_tiny_patch4_window7_224_22k.pth'}` | {'link': 'https:\u002F\u002Fgithub.com\u002FSwinTransformer\u002Fstorage\u002Freleases\u002Fdownload\u002Fv1.0.8\u002Fswin_tiny_patch4_window7_224_22k.pth', 'destination_path': '\u002Fptm\u002Fmask2former\u002Fswin_tiny_patch4_window7_224_22k\u002Fswin_tiny_patch4_window7_224_22k.pth'} |\n| train | `train.pretrained_model` | `ptm_if_no_resume_model` | PTM when no resume checkpoint exists |\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-oneformer](references\u002Ftao-deploy-oneformer.md)\n",{"data":32,"body":40},{"name":4,"description":6,"license":23,"compatibility":33,"metadata":34,"allowed-tools":37,"tags":38},"Requires docker + nvidia-container-toolkit.",{"version":35,"author":36},"0.1.0","NVIDIA Corporation","Read Bash",[39],"segmentation",{"type":41,"children":42},"root",[43,52,58,63,117,124,217,223,332,364,370,406,413,1283,1289,1309,1358,1366,1522,1530,1645,1653,1705,1713,1807,1815,1908,1914,1980,1986,1991,1997,2111,2117,2134,2228,2280,2325,2331,2336,2342,2352,2420,2461,2492,2510,2566,2604,2629,2653,2663,2669,2698,2711,3423,3458,3464,3475],{"type":44,"tag":45,"props":46,"children":48},"element","h1",{"id":47},"oneformer",[49],{"type":50,"value":51},"text","OneFormer",{"type":44,"tag":53,"props":54,"children":55},"p",{},[56],{"type":50,"value":57},"OneFormer for universal image segmentation. Unifies panoptic, instance, and semantic segmentation with a single architecture using task-conditioned queries.",{"type":44,"tag":53,"props":59,"children":60},{},[61],{"type":50,"value":62},"Set train.pretrained_backbone and\u002For train.pretrained_model.",{"type":44,"tag":53,"props":64,"children":65},{},[66,68,75,77,83,85,91,93,99,101,107,109,115],{"type":50,"value":67},"For TAO Deploy TensorRT actions (",{"type":44,"tag":69,"props":70,"children":72},"code",{"className":71},[],[73],{"type":50,"value":74},"gen_trt_engine",{"type":50,"value":76},", TensorRT ",{"type":44,"tag":69,"props":78,"children":80},{"className":79},[],[81],{"type":50,"value":82},"evaluate",{"type":50,"value":84},", and TensorRT ",{"type":44,"tag":69,"props":86,"children":88},{"className":87},[],[89],{"type":50,"value":90},"inference",{"type":50,"value":92},"), read ",{"type":44,"tag":69,"props":94,"children":96},{"className":95},[],[97],{"type":50,"value":98},"references\u002Ftao-deploy-oneformer.md",{"type":50,"value":100}," first. Deploy spec templates live in this skill's ",{"type":44,"tag":69,"props":102,"children":104},{"className":103},[],[105],{"type":50,"value":106},"references\u002F",{"type":50,"value":108}," folder with the ",{"type":44,"tag":69,"props":110,"children":112},{"className":111},[],[113],{"type":50,"value":114},"spec_template_deploy_*.yaml",{"type":50,"value":116}," prefix.",{"type":44,"tag":118,"props":119,"children":121},"h2",{"id":120},"dataclass-schemas",[122],{"type":50,"value":123},"Dataclass Schemas",{"type":44,"tag":53,"props":125,"children":126},{},[127,129,135,137,143,145,151,153,159,161,167,169,175,177,183,185,191,193,199,201,207,209,215],{"type":50,"value":128},"Generated TAO Core schemas are packaged in ",{"type":44,"tag":69,"props":130,"children":132},{"className":131},[],[133],{"type":50,"value":134},"schemas\u002F\u003Caction>.schema.json",{"type":50,"value":136},", with ",{"type":44,"tag":69,"props":138,"children":140},{"className":139},[],[141],{"type":50,"value":142},"schemas\u002Fmanifest.json",{"type":50,"value":144}," listing available actions. Each generated schema also emits ",{"type":44,"tag":69,"props":146,"children":148},{"className":147},[],[149],{"type":50,"value":150},"references\u002Fspec_template_\u003Caction>.yaml",{"type":50,"value":152}," from the schema top-level ",{"type":44,"tag":69,"props":154,"children":156},{"className":155},[],[157],{"type":50,"value":158},"default",{"type":50,"value":160}," field. AutoML enablement is declared at the model layer in ",{"type":44,"tag":69,"props":162,"children":164},{"className":163},[],[165],{"type":50,"value":166},"references\u002Fskill_info.yaml",{"type":50,"value":168}," via ",{"type":44,"tag":69,"props":170,"children":172},{"className":171},[],[173],{"type":50,"value":174},"automl_enabled",{"type":50,"value":176},". Runnable AutoML still requires ",{"type":44,"tag":69,"props":178,"children":180},{"className":179},[],[181],{"type":50,"value":182},"schemas\u002Ftrain.schema.json",{"type":50,"value":184}," and ",{"type":44,"tag":69,"props":186,"children":188},{"className":187},[],[189],{"type":50,"value":190},"references\u002Fspec_template_train.yaml",{"type":50,"value":192}," to exist and parse. Use the packaged train schema for ",{"type":44,"tag":69,"props":194,"children":196},{"className":195},[],[197],{"type":50,"value":198},"automl_default_parameters",{"type":50,"value":200},", ",{"type":44,"tag":69,"props":202,"children":204},{"className":203},[],[205],{"type":50,"value":206},"automl_disabled_parameters",{"type":50,"value":208},", defaults, min\u002Fmax bounds, enums, option weights, math conditions, dependencies, and popular parameters. Do not expect ",{"type":44,"tag":69,"props":210,"children":212},{"className":211},[],[213],{"type":50,"value":214},"~\u002Ftao-core",{"type":50,"value":216}," at runtime; maintainers regenerate schemas\u002Ftemplates before packaging the skill bank.",{"type":44,"tag":118,"props":218,"children":220},{"id":219},"train-action-policy",[221],{"type":50,"value":222},"Train Action Policy",{"type":44,"tag":53,"props":224,"children":225},{},[226,228,233,235,241,243,249,251,257,259,265,267,273,275,280,281,287,289,294,295,300,302,308,310,316,318,323,325,330],{"type":50,"value":227},"This model is AutoML-enabled at the model layer. Before handling any train-stage request, read ",{"type":44,"tag":69,"props":229,"children":231},{"className":230},[],[232],{"type":50,"value":166},{"type":50,"value":234}," and resolve the run override from either an explicit ",{"type":44,"tag":69,"props":236,"children":238},{"className":237},[],[239],{"type":50,"value":240},"automl_policy",{"type":50,"value":242}," value or the user's workflow request. Use ",{"type":44,"tag":69,"props":244,"children":246},{"className":245},[],[247],{"type":50,"value":248},"automl_policy: on",{"type":50,"value":250}," by default and only expose ",{"type":44,"tag":69,"props":252,"children":254},{"className":253},[],[255],{"type":50,"value":256},"on",{"type":50,"value":258}," \u002F ",{"type":44,"tag":69,"props":260,"children":262},{"className":261},[],[263],{"type":50,"value":264},"off",{"type":50,"value":266}," in new launch prompts. Treat phrases like \"turn off AutoML\", \"disable AutoML\", \"no HPO\", or \"plain training\" as ",{"type":44,"tag":69,"props":268,"children":270},{"className":269},[],[271],{"type":50,"value":272},"automl_policy: off",{"type":50,"value":274}," for this run only. When ",{"type":44,"tag":69,"props":276,"children":278},{"className":277},[],[279],{"type":50,"value":248},{"type":50,"value":200},{"type":44,"tag":69,"props":282,"children":284},{"className":283},[],[285],{"type":50,"value":286},"automl_enabled: true",{"type":50,"value":288},", and both ",{"type":44,"tag":69,"props":290,"children":292},{"className":291},[],[293],{"type":50,"value":182},{"type":50,"value":184},{"type":44,"tag":69,"props":296,"children":298},{"className":297},[],[299],{"type":50,"value":190},{"type":50,"value":301}," are packaged, route the train action through ",{"type":44,"tag":69,"props":303,"children":305},{"className":304},[],[306],{"type":50,"value":307},"tao-skill-bank:tao-run-automl",{"type":50,"value":309}," by default with this model's ",{"type":44,"tag":69,"props":311,"children":313},{"className":312},[],[314],{"type":50,"value":315},"skill_dir",{"type":50,"value":317},". Preserve workflow\u002Fapplication overrides for datasets, specs, output directories, GPU\u002Fplatform settings, parent checkpoints, and ",{"type":44,"tag":69,"props":319,"children":321},{"className":320},[],[322],{"type":50,"value":240},{"type":50,"value":324},". Use direct model training only when ",{"type":44,"tag":69,"props":326,"children":328},{"className":327},[],[329],{"type":50,"value":272},{"type":50,"value":331}," 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.",{"type":44,"tag":53,"props":333,"children":334},{},[335,337,342,343,348,349,355,357,362],{"type":50,"value":336},"Non-train actions such as ",{"type":44,"tag":69,"props":338,"children":340},{"className":339},[],[341],{"type":50,"value":82},{"type":50,"value":200},{"type":44,"tag":69,"props":344,"children":346},{"className":345},[],[347],{"type":50,"value":90},{"type":50,"value":200},{"type":44,"tag":69,"props":350,"children":352},{"className":351},[],[353],{"type":50,"value":354},"export",{"type":50,"value":356},", and deploy flows stay in this model skill. The per-run ",{"type":44,"tag":69,"props":358,"children":360},{"className":359},[],[361],{"type":50,"value":240},{"type":50,"value":363}," override does not change model metadata.",{"type":44,"tag":118,"props":365,"children":367},{"id":366},"training-requirements",[368],{"type":50,"value":369},"Training Requirements",{"type":44,"tag":371,"props":372,"children":373},"ul",{},[374,386,396],{"type":44,"tag":375,"props":376,"children":377},"li",{},[378,384],{"type":44,"tag":379,"props":380,"children":381},"strong",{},[382],{"type":50,"value":383},"Dataset type:",{"type":50,"value":385}," segmentation",{"type":44,"tag":375,"props":387,"children":388},{},[389,394],{"type":44,"tag":379,"props":390,"children":391},{},[392],{"type":50,"value":393},"Formats:",{"type":50,"value":395}," coco_panoptic, coco",{"type":44,"tag":375,"props":397,"children":398},{},[399,404],{"type":44,"tag":379,"props":400,"children":401},{},[402],{"type":50,"value":403},"Monitoring metric:",{"type":50,"value":405}," mIoU",{"type":44,"tag":407,"props":408,"children":410},"h3",{"id":409},"per-action-dataset-requirements",[411],{"type":50,"value":412},"Per-Action Dataset Requirements",{"type":44,"tag":414,"props":415,"children":416},"table",{},[417,451],{"type":44,"tag":418,"props":419,"children":420},"thead",{},[421],{"type":44,"tag":422,"props":423,"children":424},"tr",{},[425,431,436,441,446],{"type":44,"tag":426,"props":427,"children":428},"th",{},[429],{"type":50,"value":430},"Action",{"type":44,"tag":426,"props":432,"children":433},{},[434],{"type":50,"value":435},"Spec Key",{"type":44,"tag":426,"props":437,"children":438},{},[439],{"type":50,"value":440},"Source",{"type":44,"tag":426,"props":442,"children":443},{},[444],{"type":50,"value":445},"Files",{"type":44,"tag":426,"props":447,"children":448},{},[449],{"type":50,"value":450},"List?",{"type":44,"tag":452,"props":453,"children":454},"tbody",{},[455,483,508,533,558,583,607,631,655,679,703,726,749,772,795,818,841,864,888,912,935,958,981,1004,1027,1050,1073,1098,1122,1145,1168,1191,1214,1237,1260],{"type":44,"tag":422,"props":456,"children":457},{},[458,463,468,473,478],{"type":44,"tag":459,"props":460,"children":461},"td",{},[462],{"type":50,"value":82},{"type":44,"tag":459,"props":464,"children":465},{},[466],{"type":50,"value":467},"dataset.train.images",{"type":44,"tag":459,"props":469,"children":470},{},[471],{"type":50,"value":472},"train_datasets",{"type":44,"tag":459,"props":474,"children":475},{},[476],{"type":50,"value":477},"images.tar.gz",{"type":44,"tag":459,"props":479,"children":480},{},[481],{"type":50,"value":482},"No",{"type":44,"tag":422,"props":484,"children":485},{},[486,490,495,499,504],{"type":44,"tag":459,"props":487,"children":488},{},[489],{"type":50,"value":82},{"type":44,"tag":459,"props":491,"children":492},{},[493],{"type":50,"value":494},"dataset.label_map",{"type":44,"tag":459,"props":496,"children":497},{},[498],{"type":50,"value":472},{"type":44,"tag":459,"props":500,"children":501},{},[502],{"type":50,"value":503},"label_map.json",{"type":44,"tag":459,"props":505,"children":506},{},[507],{"type":50,"value":482},{"type":44,"tag":422,"props":509,"children":510},{},[511,515,520,524,529],{"type":44,"tag":459,"props":512,"children":513},{},[514],{"type":50,"value":82},{"type":44,"tag":459,"props":516,"children":517},{},[518],{"type":50,"value":519},"dataset.train.annotations",{"type":44,"tag":459,"props":521,"children":522},{},[523],{"type":50,"value":472},{"type":44,"tag":459,"props":525,"children":526},{},[527],{"type":50,"value":528},"annotations.json",{"type":44,"tag":459,"props":530,"children":531},{},[532],{"type":50,"value":482},{"type":44,"tag":422,"props":534,"children":535},{},[536,540,545,549,554],{"type":44,"tag":459,"props":537,"children":538},{},[539],{"type":50,"value":82},{"type":44,"tag":459,"props":541,"children":542},{},[543],{"type":50,"value":544},"dataset.train.panoptic",{"type":44,"tag":459,"props":546,"children":547},{},[548],{"type":50,"value":472},{"type":44,"tag":459,"props":550,"children":551},{},[552],{"type":50,"value":553},"images_panoptic.tar.gz",{"type":44,"tag":459,"props":555,"children":556},{},[557],{"type":50,"value":482},{"type":44,"tag":422,"props":559,"children":560},{},[561,565,570,575,579],{"type":44,"tag":459,"props":562,"children":563},{},[564],{"type":50,"value":82},{"type":44,"tag":459,"props":566,"children":567},{},[568],{"type":50,"value":569},"dataset.val.images",{"type":44,"tag":459,"props":571,"children":572},{},[573],{"type":50,"value":574},"eval_dataset",{"type":44,"tag":459,"props":576,"children":577},{},[578],{"type":50,"value":477},{"type":44,"tag":459,"props":580,"children":581},{},[582],{"type":50,"value":482},{"type":44,"tag":422,"props":584,"children":585},{},[586,590,595,599,603],{"type":44,"tag":459,"props":587,"children":588},{},[589],{"type":50,"value":82},{"type":44,"tag":459,"props":591,"children":592},{},[593],{"type":50,"value":594},"dataset.val.annotations",{"type":44,"tag":459,"props":596,"children":597},{},[598],{"type":50,"value":574},{"type":44,"tag":459,"props":600,"children":601},{},[602],{"type":50,"value":528},{"type":44,"tag":459,"props":604,"children":605},{},[606],{"type":50,"value":482},{"type":44,"tag":422,"props":608,"children":609},{},[610,614,619,623,627],{"type":44,"tag":459,"props":611,"children":612},{},[613],{"type":50,"value":82},{"type":44,"tag":459,"props":615,"children":616},{},[617],{"type":50,"value":618},"dataset.val.panoptic",{"type":44,"tag":459,"props":620,"children":621},{},[622],{"type":50,"value":574},{"type":44,"tag":459,"props":624,"children":625},{},[626],{"type":50,"value":553},{"type":44,"tag":459,"props":628,"children":629},{},[630],{"type":50,"value":482},{"type":44,"tag":422,"props":632,"children":633},{},[634,638,643,647,651],{"type":44,"tag":459,"props":635,"children":636},{},[637],{"type":50,"value":82},{"type":44,"tag":459,"props":639,"children":640},{},[641],{"type":50,"value":642},"dataset.test.images",{"type":44,"tag":459,"props":644,"children":645},{},[646],{"type":50,"value":574},{"type":44,"tag":459,"props":648,"children":649},{},[650],{"type":50,"value":477},{"type":44,"tag":459,"props":652,"children":653},{},[654],{"type":50,"value":482},{"type":44,"tag":422,"props":656,"children":657},{},[658,662,667,671,675],{"type":44,"tag":459,"props":659,"children":660},{},[661],{"type":50,"value":82},{"type":44,"tag":459,"props":663,"children":664},{},[665],{"type":50,"value":666},"dataset.test.annotations",{"type":44,"tag":459,"props":668,"children":669},{},[670],{"type":50,"value":574},{"type":44,"tag":459,"props":672,"children":673},{},[674],{"type":50,"value":528},{"type":44,"tag":459,"props":676,"children":677},{},[678],{"type":50,"value":482},{"type":44,"tag":422,"props":680,"children":681},{},[682,686,691,695,699],{"type":44,"tag":459,"props":683,"children":684},{},[685],{"type":50,"value":82},{"type":44,"tag":459,"props":687,"children":688},{},[689],{"type":50,"value":690},"dataset.test.panoptic",{"type":44,"tag":459,"props":692,"children":693},{},[694],{"type":50,"value":574},{"type":44,"tag":459,"props":696,"children":697},{},[698],{"type":50,"value":553},{"type":44,"tag":459,"props":700,"children":701},{},[702],{"type":50,"value":482},{"type":44,"tag":422,"props":704,"children":705},{},[706,710,714,718,722],{"type":44,"tag":459,"props":707,"children":708},{},[709],{"type":50,"value":90},{"type":44,"tag":459,"props":711,"children":712},{},[713],{"type":50,"value":467},{"type":44,"tag":459,"props":715,"children":716},{},[717],{"type":50,"value":472},{"type":44,"tag":459,"props":719,"children":720},{},[721],{"type":50,"value":477},{"type":44,"tag":459,"props":723,"children":724},{},[725],{"type":50,"value":482},{"type":44,"tag":422,"props":727,"children":728},{},[729,733,737,741,745],{"type":44,"tag":459,"props":730,"children":731},{},[732],{"type":50,"value":90},{"type":44,"tag":459,"props":734,"children":735},{},[736],{"type":50,"value":494},{"type":44,"tag":459,"props":738,"children":739},{},[740],{"type":50,"value":472},{"type":44,"tag":459,"props":742,"children":743},{},[744],{"type":50,"value":503},{"type":44,"tag":459,"props":746,"children":747},{},[748],{"type":50,"value":482},{"type":44,"tag":422,"props":750,"children":751},{},[752,756,760,764,768],{"type":44,"tag":459,"props":753,"children":754},{},[755],{"type":50,"value":90},{"type":44,"tag":459,"props":757,"children":758},{},[759],{"type":50,"value":519},{"type":44,"tag":459,"props":761,"children":762},{},[763],{"type":50,"value":472},{"type":44,"tag":459,"props":765,"children":766},{},[767],{"type":50,"value":528},{"type":44,"tag":459,"props":769,"children":770},{},[771],{"type":50,"value":482},{"type":44,"tag":422,"props":773,"children":774},{},[775,779,783,787,791],{"type":44,"tag":459,"props":776,"children":777},{},[778],{"type":50,"value":90},{"type":44,"tag":459,"props":780,"children":781},{},[782],{"type":50,"value":544},{"type":44,"tag":459,"props":784,"children":785},{},[786],{"type":50,"value":472},{"type":44,"tag":459,"props":788,"children":789},{},[790],{"type":50,"value":553},{"type":44,"tag":459,"props":792,"children":793},{},[794],{"type":50,"value":482},{"type":44,"tag":422,"props":796,"children":797},{},[798,802,806,810,814],{"type":44,"tag":459,"props":799,"children":800},{},[801],{"type":50,"value":90},{"type":44,"tag":459,"props":803,"children":804},{},[805],{"type":50,"value":569},{"type":44,"tag":459,"props":807,"children":808},{},[809],{"type":50,"value":574},{"type":44,"tag":459,"props":811,"children":812},{},[813],{"type":50,"value":477},{"type":44,"tag":459,"props":815,"children":816},{},[817],{"type":50,"value":482},{"type":44,"tag":422,"props":819,"children":820},{},[821,825,829,833,837],{"type":44,"tag":459,"props":822,"children":823},{},[824],{"type":50,"value":90},{"type":44,"tag":459,"props":826,"children":827},{},[828],{"type":50,"value":594},{"type":44,"tag":459,"props":830,"children":831},{},[832],{"type":50,"value":574},{"type":44,"tag":459,"props":834,"children":835},{},[836],{"type":50,"value":528},{"type":44,"tag":459,"props":838,"children":839},{},[840],{"type":50,"value":482},{"type":44,"tag":422,"props":842,"children":843},{},[844,848,852,856,860],{"type":44,"tag":459,"props":845,"children":846},{},[847],{"type":50,"value":90},{"type":44,"tag":459,"props":849,"children":850},{},[851],{"type":50,"value":618},{"type":44,"tag":459,"props":853,"children":854},{},[855],{"type":50,"value":574},{"type":44,"tag":459,"props":857,"children":858},{},[859],{"type":50,"value":553},{"type":44,"tag":459,"props":861,"children":862},{},[863],{"type":50,"value":482},{"type":44,"tag":422,"props":865,"children":866},{},[867,871,875,880,884],{"type":44,"tag":459,"props":868,"children":869},{},[870],{"type":50,"value":90},{"type":44,"tag":459,"props":872,"children":873},{},[874],{"type":50,"value":642},{"type":44,"tag":459,"props":876,"children":877},{},[878],{"type":50,"value":879},"inference_dataset",{"type":44,"tag":459,"props":881,"children":882},{},[883],{"type":50,"value":477},{"type":44,"tag":459,"props":885,"children":886},{},[887],{"type":50,"value":482},{"type":44,"tag":422,"props":889,"children":890},{},[891,896,900,904,908],{"type":44,"tag":459,"props":892,"children":893},{},[894],{"type":50,"value":895},"quantize",{"type":44,"tag":459,"props":897,"children":898},{},[899],{"type":50,"value":467},{"type":44,"tag":459,"props":901,"children":902},{},[903],{"type":50,"value":472},{"type":44,"tag":459,"props":905,"children":906},{},[907],{"type":50,"value":477},{"type":44,"tag":459,"props":909,"children":910},{},[911],{"type":50,"value":482},{"type":44,"tag":422,"props":913,"children":914},{},[915,919,923,927,931],{"type":44,"tag":459,"props":916,"children":917},{},[918],{"type":50,"value":895},{"type":44,"tag":459,"props":920,"children":921},{},[922],{"type":50,"value":519},{"type":44,"tag":459,"props":924,"children":925},{},[926],{"type":50,"value":472},{"type":44,"tag":459,"props":928,"children":929},{},[930],{"type":50,"value":528},{"type":44,"tag":459,"props":932,"children":933},{},[934],{"type":50,"value":482},{"type":44,"tag":422,"props":936,"children":937},{},[938,942,946,950,954],{"type":44,"tag":459,"props":939,"children":940},{},[941],{"type":50,"value":895},{"type":44,"tag":459,"props":943,"children":944},{},[945],{"type":50,"value":494},{"type":44,"tag":459,"props":947,"children":948},{},[949],{"type":50,"value":472},{"type":44,"tag":459,"props":951,"children":952},{},[953],{"type":50,"value":503},{"type":44,"tag":459,"props":955,"children":956},{},[957],{"type":50,"value":482},{"type":44,"tag":422,"props":959,"children":960},{},[961,965,969,973,977],{"type":44,"tag":459,"props":962,"children":963},{},[964],{"type":50,"value":895},{"type":44,"tag":459,"props":966,"children":967},{},[968],{"type":50,"value":544},{"type":44,"tag":459,"props":970,"children":971},{},[972],{"type":50,"value":472},{"type":44,"tag":459,"props":974,"children":975},{},[976],{"type":50,"value":553},{"type":44,"tag":459,"props":978,"children":979},{},[980],{"type":50,"value":482},{"type":44,"tag":422,"props":982,"children":983},{},[984,988,992,996,1000],{"type":44,"tag":459,"props":985,"children":986},{},[987],{"type":50,"value":895},{"type":44,"tag":459,"props":989,"children":990},{},[991],{"type":50,"value":569},{"type":44,"tag":459,"props":993,"children":994},{},[995],{"type":50,"value":574},{"type":44,"tag":459,"props":997,"children":998},{},[999],{"type":50,"value":477},{"type":44,"tag":459,"props":1001,"children":1002},{},[1003],{"type":50,"value":482},{"type":44,"tag":422,"props":1005,"children":1006},{},[1007,1011,1015,1019,1023],{"type":44,"tag":459,"props":1008,"children":1009},{},[1010],{"type":50,"value":895},{"type":44,"tag":459,"props":1012,"children":1013},{},[1014],{"type":50,"value":594},{"type":44,"tag":459,"props":1016,"children":1017},{},[1018],{"type":50,"value":574},{"type":44,"tag":459,"props":1020,"children":1021},{},[1022],{"type":50,"value":528},{"type":44,"tag":459,"props":1024,"children":1025},{},[1026],{"type":50,"value":482},{"type":44,"tag":422,"props":1028,"children":1029},{},[1030,1034,1038,1042,1046],{"type":44,"tag":459,"props":1031,"children":1032},{},[1033],{"type":50,"value":895},{"type":44,"tag":459,"props":1035,"children":1036},{},[1037],{"type":50,"value":618},{"type":44,"tag":459,"props":1039,"children":1040},{},[1041],{"type":50,"value":574},{"type":44,"tag":459,"props":1043,"children":1044},{},[1045],{"type":50,"value":553},{"type":44,"tag":459,"props":1047,"children":1048},{},[1049],{"type":50,"value":482},{"type":44,"tag":422,"props":1051,"children":1052},{},[1053,1057,1061,1065,1069],{"type":44,"tag":459,"props":1054,"children":1055},{},[1056],{"type":50,"value":895},{"type":44,"tag":459,"props":1058,"children":1059},{},[1060],{"type":50,"value":642},{"type":44,"tag":459,"props":1062,"children":1063},{},[1064],{"type":50,"value":574},{"type":44,"tag":459,"props":1066,"children":1067},{},[1068],{"type":50,"value":477},{"type":44,"tag":459,"props":1070,"children":1071},{},[1072],{"type":50,"value":482},{"type":44,"tag":422,"props":1074,"children":1075},{},[1076,1080,1085,1090,1094],{"type":44,"tag":459,"props":1077,"children":1078},{},[1079],{"type":50,"value":895},{"type":44,"tag":459,"props":1081,"children":1082},{},[1083],{"type":50,"value":1084},"dataset.quant_calibration_dataset.images_dir",{"type":44,"tag":459,"props":1086,"children":1087},{},[1088],{"type":50,"value":1089},"calibration_dataset",{"type":44,"tag":459,"props":1091,"children":1092},{},[1093],{"type":50,"value":477},{"type":44,"tag":459,"props":1095,"children":1096},{},[1097],{"type":50,"value":482},{"type":44,"tag":422,"props":1099,"children":1100},{},[1101,1106,1110,1114,1118],{"type":44,"tag":459,"props":1102,"children":1103},{},[1104],{"type":50,"value":1105},"train",{"type":44,"tag":459,"props":1107,"children":1108},{},[1109],{"type":50,"value":467},{"type":44,"tag":459,"props":1111,"children":1112},{},[1113],{"type":50,"value":472},{"type":44,"tag":459,"props":1115,"children":1116},{},[1117],{"type":50,"value":477},{"type":44,"tag":459,"props":1119,"children":1120},{},[1121],{"type":50,"value":482},{"type":44,"tag":422,"props":1123,"children":1124},{},[1125,1129,1133,1137,1141],{"type":44,"tag":459,"props":1126,"children":1127},{},[1128],{"type":50,"value":1105},{"type":44,"tag":459,"props":1130,"children":1131},{},[1132],{"type":50,"value":519},{"type":44,"tag":459,"props":1134,"children":1135},{},[1136],{"type":50,"value":472},{"type":44,"tag":459,"props":1138,"children":1139},{},[1140],{"type":50,"value":528},{"type":44,"tag":459,"props":1142,"children":1143},{},[1144],{"type":50,"value":482},{"type":44,"tag":422,"props":1146,"children":1147},{},[1148,1152,1156,1160,1164],{"type":44,"tag":459,"props":1149,"children":1150},{},[1151],{"type":50,"value":1105},{"type":44,"tag":459,"props":1153,"children":1154},{},[1155],{"type":50,"value":494},{"type":44,"tag":459,"props":1157,"children":1158},{},[1159],{"type":50,"value":472},{"type":44,"tag":459,"props":1161,"children":1162},{},[1163],{"type":50,"value":503},{"type":44,"tag":459,"props":1165,"children":1166},{},[1167],{"type":50,"value":482},{"type":44,"tag":422,"props":1169,"children":1170},{},[1171,1175,1179,1183,1187],{"type":44,"tag":459,"props":1172,"children":1173},{},[1174],{"type":50,"value":1105},{"type":44,"tag":459,"props":1176,"children":1177},{},[1178],{"type":50,"value":544},{"type":44,"tag":459,"props":1180,"children":1181},{},[1182],{"type":50,"value":472},{"type":44,"tag":459,"props":1184,"children":1185},{},[1186],{"type":50,"value":553},{"type":44,"tag":459,"props":1188,"children":1189},{},[1190],{"type":50,"value":482},{"type":44,"tag":422,"props":1192,"children":1193},{},[1194,1198,1202,1206,1210],{"type":44,"tag":459,"props":1195,"children":1196},{},[1197],{"type":50,"value":1105},{"type":44,"tag":459,"props":1199,"children":1200},{},[1201],{"type":50,"value":569},{"type":44,"tag":459,"props":1203,"children":1204},{},[1205],{"type":50,"value":574},{"type":44,"tag":459,"props":1207,"children":1208},{},[1209],{"type":50,"value":477},{"type":44,"tag":459,"props":1211,"children":1212},{},[1213],{"type":50,"value":482},{"type":44,"tag":422,"props":1215,"children":1216},{},[1217,1221,1225,1229,1233],{"type":44,"tag":459,"props":1218,"children":1219},{},[1220],{"type":50,"value":1105},{"type":44,"tag":459,"props":1222,"children":1223},{},[1224],{"type":50,"value":594},{"type":44,"tag":459,"props":1226,"children":1227},{},[1228],{"type":50,"value":574},{"type":44,"tag":459,"props":1230,"children":1231},{},[1232],{"type":50,"value":528},{"type":44,"tag":459,"props":1234,"children":1235},{},[1236],{"type":50,"value":482},{"type":44,"tag":422,"props":1238,"children":1239},{},[1240,1244,1248,1252,1256],{"type":44,"tag":459,"props":1241,"children":1242},{},[1243],{"type":50,"value":1105},{"type":44,"tag":459,"props":1245,"children":1246},{},[1247],{"type":50,"value":618},{"type":44,"tag":459,"props":1249,"children":1250},{},[1251],{"type":50,"value":574},{"type":44,"tag":459,"props":1253,"children":1254},{},[1255],{"type":50,"value":553},{"type":44,"tag":459,"props":1257,"children":1258},{},[1259],{"type":50,"value":482},{"type":44,"tag":422,"props":1261,"children":1262},{},[1263,1267,1271,1275,1279],{"type":44,"tag":459,"props":1264,"children":1265},{},[1266],{"type":50,"value":1105},{"type":44,"tag":459,"props":1268,"children":1269},{},[1270],{"type":50,"value":642},{"type":44,"tag":459,"props":1272,"children":1273},{},[1274],{"type":50,"value":574},{"type":44,"tag":459,"props":1276,"children":1277},{},[1278],{"type":50,"value":477},{"type":44,"tag":459,"props":1280,"children":1281},{},[1282],{"type":50,"value":482},{"type":44,"tag":407,"props":1284,"children":1286},{"id":1285},"typical-spec-overrides",[1287],{"type":50,"value":1288},"Typical Spec Overrides",{"type":44,"tag":53,"props":1290,"children":1291},{},[1292,1294,1299,1301,1307],{"type":50,"value":1293},"Data source overrides are ",{"type":44,"tag":379,"props":1295,"children":1296},{},[1297],{"type":50,"value":1298},"mandatory for every action",{"type":50,"value":1300}," — the agent MUST construct data source paths from the Per-Action Dataset Requirements table above and include them in ",{"type":44,"tag":69,"props":1302,"children":1304},{"className":1303},[],[1305],{"type":50,"value":1306},"spec_overrides",{"type":50,"value":1308},".",{"type":44,"tag":1310,"props":1311,"children":1316},"pre",{"className":1312,"code":1313,"language":1314,"meta":1315,"style":1315},"language-python shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","S3_TRAIN = \"s3:\u002F\u002Fbucket\u002Fdata\u002Ftrain\"\nS3_EVAL = \"s3:\u002F\u002Fbucket\u002Fdata\u002Feval\"\nS3_INFERENCE = \"s3:\u002F\u002Fbucket\u002Fdata\u002Finference\"\nS3_CALIBRATION = \"s3:\u002F\u002Fbucket\u002Fdata\u002Fcalibration\"\n","python","",[1317],{"type":44,"tag":69,"props":1318,"children":1319},{"__ignoreMap":1315},[1320,1331,1340,1349],{"type":44,"tag":1321,"props":1322,"children":1325},"span",{"class":1323,"line":1324},"line",1,[1326],{"type":44,"tag":1321,"props":1327,"children":1328},{},[1329],{"type":50,"value":1330},"S3_TRAIN = \"s3:\u002F\u002Fbucket\u002Fdata\u002Ftrain\"\n",{"type":44,"tag":1321,"props":1332,"children":1334},{"class":1323,"line":1333},2,[1335],{"type":44,"tag":1321,"props":1336,"children":1337},{},[1338],{"type":50,"value":1339},"S3_EVAL = \"s3:\u002F\u002Fbucket\u002Fdata\u002Feval\"\n",{"type":44,"tag":1321,"props":1341,"children":1343},{"class":1323,"line":1342},3,[1344],{"type":44,"tag":1321,"props":1345,"children":1346},{},[1347],{"type":50,"value":1348},"S3_INFERENCE = \"s3:\u002F\u002Fbucket\u002Fdata\u002Finference\"\n",{"type":44,"tag":1321,"props":1350,"children":1352},{"class":1323,"line":1351},4,[1353],{"type":44,"tag":1321,"props":1354,"children":1355},{},[1356],{"type":50,"value":1357},"S3_CALIBRATION = \"s3:\u002F\u002Fbucket\u002Fdata\u002Fcalibration\"\n",{"type":44,"tag":53,"props":1359,"children":1360},{},[1361],{"type":44,"tag":379,"props":1362,"children":1363},{},[1364],{"type":50,"value":1365},"train (mandatory data sources):",{"type":44,"tag":1310,"props":1367,"children":1369},{"className":1312,"code":1368,"language":1314,"meta":1315,"style":1315},"{\n    \"train.num_gpus\": 1,\n    \"train.num_epochs\": 10,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"model.sem_seg_head.num_classes\": 133,\n    \"dataset.contiguous_id\": True,\n    \"train.precision\": \"32\",\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n}\n",[1370],{"type":44,"tag":69,"props":1371,"children":1372},{"__ignoreMap":1315},[1373,1381,1389,1397,1405,1414,1423,1432,1441,1450,1459,1468,1477,1486,1495,1504,1513],{"type":44,"tag":1321,"props":1374,"children":1375},{"class":1323,"line":1324},[1376],{"type":44,"tag":1321,"props":1377,"children":1378},{},[1379],{"type":50,"value":1380},"{\n",{"type":44,"tag":1321,"props":1382,"children":1383},{"class":1323,"line":1333},[1384],{"type":44,"tag":1321,"props":1385,"children":1386},{},[1387],{"type":50,"value":1388},"    \"train.num_gpus\": 1,\n",{"type":44,"tag":1321,"props":1390,"children":1391},{"class":1323,"line":1342},[1392],{"type":44,"tag":1321,"props":1393,"children":1394},{},[1395],{"type":50,"value":1396},"    \"train.num_epochs\": 10,\n",{"type":44,"tag":1321,"props":1398,"children":1399},{"class":1323,"line":1351},[1400],{"type":44,"tag":1321,"props":1401,"children":1402},{},[1403],{"type":50,"value":1404},"    \"train.checkpoint_interval\": 10,\n",{"type":44,"tag":1321,"props":1406,"children":1408},{"class":1323,"line":1407},5,[1409],{"type":44,"tag":1321,"props":1410,"children":1411},{},[1412],{"type":50,"value":1413},"    \"train.validation_interval\": 10,\n",{"type":44,"tag":1321,"props":1415,"children":1417},{"class":1323,"line":1416},6,[1418],{"type":44,"tag":1321,"props":1419,"children":1420},{},[1421],{"type":50,"value":1422},"    \"model.sem_seg_head.num_classes\": 133,\n",{"type":44,"tag":1321,"props":1424,"children":1426},{"class":1323,"line":1425},7,[1427],{"type":44,"tag":1321,"props":1428,"children":1429},{},[1430],{"type":50,"value":1431},"    \"dataset.contiguous_id\": True,\n",{"type":44,"tag":1321,"props":1433,"children":1435},{"class":1323,"line":1434},8,[1436],{"type":44,"tag":1321,"props":1437,"children":1438},{},[1439],{"type":50,"value":1440},"    \"train.precision\": \"32\",\n",{"type":44,"tag":1321,"props":1442,"children":1444},{"class":1323,"line":1443},9,[1445],{"type":44,"tag":1321,"props":1446,"children":1447},{},[1448],{"type":50,"value":1449},"    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n",{"type":44,"tag":1321,"props":1451,"children":1453},{"class":1323,"line":1452},10,[1454],{"type":44,"tag":1321,"props":1455,"children":1456},{},[1457],{"type":50,"value":1458},"    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n",{"type":44,"tag":1321,"props":1460,"children":1462},{"class":1323,"line":1461},11,[1463],{"type":44,"tag":1321,"props":1464,"children":1465},{},[1466],{"type":50,"value":1467},"    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n",{"type":44,"tag":1321,"props":1469,"children":1471},{"class":1323,"line":1470},12,[1472],{"type":44,"tag":1321,"props":1473,"children":1474},{},[1475],{"type":50,"value":1476},"    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n",{"type":44,"tag":1321,"props":1478,"children":1480},{"class":1323,"line":1479},13,[1481],{"type":44,"tag":1321,"props":1482,"children":1483},{},[1484],{"type":50,"value":1485},"    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n",{"type":44,"tag":1321,"props":1487,"children":1489},{"class":1323,"line":1488},14,[1490],{"type":44,"tag":1321,"props":1491,"children":1492},{},[1493],{"type":50,"value":1494},"    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n",{"type":44,"tag":1321,"props":1496,"children":1498},{"class":1323,"line":1497},15,[1499],{"type":44,"tag":1321,"props":1500,"children":1501},{},[1502],{"type":50,"value":1503},"    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n",{"type":44,"tag":1321,"props":1505,"children":1507},{"class":1323,"line":1506},16,[1508],{"type":44,"tag":1321,"props":1509,"children":1510},{},[1511],{"type":50,"value":1512},"    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n",{"type":44,"tag":1321,"props":1514,"children":1516},{"class":1323,"line":1515},17,[1517],{"type":44,"tag":1321,"props":1518,"children":1519},{},[1520],{"type":50,"value":1521},"}\n",{"type":44,"tag":53,"props":1523,"children":1524},{},[1525],{"type":44,"tag":379,"props":1526,"children":1527},{},[1528],{"type":50,"value":1529},"evaluate (mandatory data sources):",{"type":44,"tag":1310,"props":1531,"children":1533},{"className":1312,"code":1532,"language":1314,"meta":1315,"style":1315},"{\n    \"evaluate.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"model.sem_seg_head.num_classes\": 133,\n    \"dataset.contiguous_id\": True,\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.test.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.test.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n}\n",[1534],{"type":44,"tag":69,"props":1535,"children":1536},{"__ignoreMap":1315},[1537,1544,1552,1559,1566,1573,1580,1587,1594,1601,1608,1615,1622,1630,1638],{"type":44,"tag":1321,"props":1538,"children":1539},{"class":1323,"line":1324},[1540],{"type":44,"tag":1321,"props":1541,"children":1542},{},[1543],{"type":50,"value":1380},{"type":44,"tag":1321,"props":1545,"children":1546},{"class":1323,"line":1333},[1547],{"type":44,"tag":1321,"props":1548,"children":1549},{},[1550],{"type":50,"value":1551},"    \"evaluate.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n",{"type":44,"tag":1321,"props":1553,"children":1554},{"class":1323,"line":1342},[1555],{"type":44,"tag":1321,"props":1556,"children":1557},{},[1558],{"type":50,"value":1422},{"type":44,"tag":1321,"props":1560,"children":1561},{"class":1323,"line":1351},[1562],{"type":44,"tag":1321,"props":1563,"children":1564},{},[1565],{"type":50,"value":1431},{"type":44,"tag":1321,"props":1567,"children":1568},{"class":1323,"line":1407},[1569],{"type":44,"tag":1321,"props":1570,"children":1571},{},[1572],{"type":50,"value":1449},{"type":44,"tag":1321,"props":1574,"children":1575},{"class":1323,"line":1416},[1576],{"type":44,"tag":1321,"props":1577,"children":1578},{},[1579],{"type":50,"value":1467},{"type":44,"tag":1321,"props":1581,"children":1582},{"class":1323,"line":1425},[1583],{"type":44,"tag":1321,"props":1584,"children":1585},{},[1586],{"type":50,"value":1458},{"type":44,"tag":1321,"props":1588,"children":1589},{"class":1323,"line":1434},[1590],{"type":44,"tag":1321,"props":1591,"children":1592},{},[1593],{"type":50,"value":1476},{"type":44,"tag":1321,"props":1595,"children":1596},{"class":1323,"line":1443},[1597],{"type":44,"tag":1321,"props":1598,"children":1599},{},[1600],{"type":50,"value":1485},{"type":44,"tag":1321,"props":1602,"children":1603},{"class":1323,"line":1452},[1604],{"type":44,"tag":1321,"props":1605,"children":1606},{},[1607],{"type":50,"value":1494},{"type":44,"tag":1321,"props":1609,"children":1610},{"class":1323,"line":1461},[1611],{"type":44,"tag":1321,"props":1612,"children":1613},{},[1614],{"type":50,"value":1503},{"type":44,"tag":1321,"props":1616,"children":1617},{"class":1323,"line":1470},[1618],{"type":44,"tag":1321,"props":1619,"children":1620},{},[1621],{"type":50,"value":1512},{"type":44,"tag":1321,"props":1623,"children":1624},{"class":1323,"line":1479},[1625],{"type":44,"tag":1321,"props":1626,"children":1627},{},[1628],{"type":50,"value":1629},"    \"dataset.test.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n",{"type":44,"tag":1321,"props":1631,"children":1632},{"class":1323,"line":1488},[1633],{"type":44,"tag":1321,"props":1634,"children":1635},{},[1636],{"type":50,"value":1637},"    \"dataset.test.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n",{"type":44,"tag":1321,"props":1639,"children":1640},{"class":1323,"line":1497},[1641],{"type":44,"tag":1321,"props":1642,"children":1643},{},[1644],{"type":50,"value":1521},{"type":44,"tag":53,"props":1646,"children":1647},{},[1648],{"type":44,"tag":379,"props":1649,"children":1650},{},[1651],{"type":50,"value":1652},"export:",{"type":44,"tag":1310,"props":1654,"children":1656},{"className":1312,"code":1655,"language":1314,"meta":1315,"style":1315},"{\n    \"export.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"model.sem_seg_head.num_classes\": 133,\n    \"model.export\": True,\n    \"export.onnx_file\": \"\u002Fresults\u002Foneformer_export_640.onnx\",\n}\n",[1657],{"type":44,"tag":69,"props":1658,"children":1659},{"__ignoreMap":1315},[1660,1667,1675,1682,1690,1698],{"type":44,"tag":1321,"props":1661,"children":1662},{"class":1323,"line":1324},[1663],{"type":44,"tag":1321,"props":1664,"children":1665},{},[1666],{"type":50,"value":1380},{"type":44,"tag":1321,"props":1668,"children":1669},{"class":1323,"line":1333},[1670],{"type":44,"tag":1321,"props":1671,"children":1672},{},[1673],{"type":50,"value":1674},"    \"export.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n",{"type":44,"tag":1321,"props":1676,"children":1677},{"class":1323,"line":1342},[1678],{"type":44,"tag":1321,"props":1679,"children":1680},{},[1681],{"type":50,"value":1422},{"type":44,"tag":1321,"props":1683,"children":1684},{"class":1323,"line":1351},[1685],{"type":44,"tag":1321,"props":1686,"children":1687},{},[1688],{"type":50,"value":1689},"    \"model.export\": True,\n",{"type":44,"tag":1321,"props":1691,"children":1692},{"class":1323,"line":1407},[1693],{"type":44,"tag":1321,"props":1694,"children":1695},{},[1696],{"type":50,"value":1697},"    \"export.onnx_file\": \"\u002Fresults\u002Foneformer_export_640.onnx\",\n",{"type":44,"tag":1321,"props":1699,"children":1700},{"class":1323,"line":1416},[1701],{"type":44,"tag":1321,"props":1702,"children":1703},{},[1704],{"type":50,"value":1521},{"type":44,"tag":53,"props":1706,"children":1707},{},[1708],{"type":44,"tag":379,"props":1709,"children":1710},{},[1711],{"type":50,"value":1712},"inference (mandatory data sources):",{"type":44,"tag":1310,"props":1714,"children":1716},{"className":1312,"code":1715,"language":1314,"meta":1315,"style":1315},"{\n    \"inference.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_INFERENCE}\u002Fimages.tar.gz\",\n    \"inference.images_dir\": f\"{S3_INFERENCE}\u002Fimages.tar.gz\",\n}\n",[1717],{"type":44,"tag":69,"props":1718,"children":1719},{"__ignoreMap":1315},[1720,1727,1735,1742,1749,1756,1763,1770,1777,1784,1792,1800],{"type":44,"tag":1321,"props":1721,"children":1722},{"class":1323,"line":1324},[1723],{"type":44,"tag":1321,"props":1724,"children":1725},{},[1726],{"type":50,"value":1380},{"type":44,"tag":1321,"props":1728,"children":1729},{"class":1323,"line":1333},[1730],{"type":44,"tag":1321,"props":1731,"children":1732},{},[1733],{"type":50,"value":1734},"    \"inference.checkpoint\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n",{"type":44,"tag":1321,"props":1736,"children":1737},{"class":1323,"line":1342},[1738],{"type":44,"tag":1321,"props":1739,"children":1740},{},[1741],{"type":50,"value":1449},{"type":44,"tag":1321,"props":1743,"children":1744},{"class":1323,"line":1351},[1745],{"type":44,"tag":1321,"props":1746,"children":1747},{},[1748],{"type":50,"value":1467},{"type":44,"tag":1321,"props":1750,"children":1751},{"class":1323,"line":1407},[1752],{"type":44,"tag":1321,"props":1753,"children":1754},{},[1755],{"type":50,"value":1458},{"type":44,"tag":1321,"props":1757,"children":1758},{"class":1323,"line":1416},[1759],{"type":44,"tag":1321,"props":1760,"children":1761},{},[1762],{"type":50,"value":1476},{"type":44,"tag":1321,"props":1764,"children":1765},{"class":1323,"line":1425},[1766],{"type":44,"tag":1321,"props":1767,"children":1768},{},[1769],{"type":50,"value":1485},{"type":44,"tag":1321,"props":1771,"children":1772},{"class":1323,"line":1434},[1773],{"type":44,"tag":1321,"props":1774,"children":1775},{},[1776],{"type":50,"value":1494},{"type":44,"tag":1321,"props":1778,"children":1779},{"class":1323,"line":1443},[1780],{"type":44,"tag":1321,"props":1781,"children":1782},{},[1783],{"type":50,"value":1503},{"type":44,"tag":1321,"props":1785,"children":1786},{"class":1323,"line":1452},[1787],{"type":44,"tag":1321,"props":1788,"children":1789},{},[1790],{"type":50,"value":1791},"    \"dataset.test.images\": f\"{S3_INFERENCE}\u002Fimages.tar.gz\",\n",{"type":44,"tag":1321,"props":1793,"children":1794},{"class":1323,"line":1461},[1795],{"type":44,"tag":1321,"props":1796,"children":1797},{},[1798],{"type":50,"value":1799},"    \"inference.images_dir\": f\"{S3_INFERENCE}\u002Fimages.tar.gz\",\n",{"type":44,"tag":1321,"props":1801,"children":1802},{"class":1323,"line":1470},[1803],{"type":44,"tag":1321,"props":1804,"children":1805},{},[1806],{"type":50,"value":1521},{"type":44,"tag":53,"props":1808,"children":1809},{},[1810],{"type":44,"tag":379,"props":1811,"children":1812},{},[1813],{"type":50,"value":1814},"quantize (mandatory data sources):",{"type":44,"tag":1310,"props":1816,"children":1818},{"className":1312,"code":1817,"language":1314,"meta":1315,"style":1315},"{\n    \"quantize.model_path\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n    \"dataset.train.images\": f\"{S3_TRAIN}\u002Fimages.tar.gz\",\n    \"dataset.train.annotations\": f\"{S3_TRAIN}\u002Fannotations.json\",\n    \"dataset.label_map\": f\"{S3_TRAIN}\u002Flabel_map.json\",\n    \"dataset.train.panoptic\": f\"{S3_TRAIN}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.val.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.val.annotations\": f\"{S3_EVAL}\u002Fannotations.json\",\n    \"dataset.val.panoptic\": f\"{S3_EVAL}\u002Fimages_panoptic.tar.gz\",\n    \"dataset.test.images\": f\"{S3_EVAL}\u002Fimages.tar.gz\",\n    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_CALIBRATION}\u002Fimages.tar.gz\",\n}\n",[1819],{"type":44,"tag":69,"props":1820,"children":1821},{"__ignoreMap":1315},[1822,1829,1837,1844,1851,1858,1865,1872,1879,1886,1893,1901],{"type":44,"tag":1321,"props":1823,"children":1824},{"class":1323,"line":1324},[1825],{"type":44,"tag":1321,"props":1826,"children":1827},{},[1828],{"type":50,"value":1380},{"type":44,"tag":1321,"props":1830,"children":1831},{"class":1323,"line":1333},[1832],{"type":44,"tag":1321,"props":1833,"children":1834},{},[1835],{"type":50,"value":1836},"    \"quantize.model_path\": \"\u003Cselected train\u002FAutoML checkpoint>\",\n",{"type":44,"tag":1321,"props":1838,"children":1839},{"class":1323,"line":1342},[1840],{"type":44,"tag":1321,"props":1841,"children":1842},{},[1843],{"type":50,"value":1449},{"type":44,"tag":1321,"props":1845,"children":1846},{"class":1323,"line":1351},[1847],{"type":44,"tag":1321,"props":1848,"children":1849},{},[1850],{"type":50,"value":1458},{"type":44,"tag":1321,"props":1852,"children":1853},{"class":1323,"line":1407},[1854],{"type":44,"tag":1321,"props":1855,"children":1856},{},[1857],{"type":50,"value":1467},{"type":44,"tag":1321,"props":1859,"children":1860},{"class":1323,"line":1416},[1861],{"type":44,"tag":1321,"props":1862,"children":1863},{},[1864],{"type":50,"value":1476},{"type":44,"tag":1321,"props":1866,"children":1867},{"class":1323,"line":1425},[1868],{"type":44,"tag":1321,"props":1869,"children":1870},{},[1871],{"type":50,"value":1485},{"type":44,"tag":1321,"props":1873,"children":1874},{"class":1323,"line":1434},[1875],{"type":44,"tag":1321,"props":1876,"children":1877},{},[1878],{"type":50,"value":1494},{"type":44,"tag":1321,"props":1880,"children":1881},{"class":1323,"line":1443},[1882],{"type":44,"tag":1321,"props":1883,"children":1884},{},[1885],{"type":50,"value":1503},{"type":44,"tag":1321,"props":1887,"children":1888},{"class":1323,"line":1452},[1889],{"type":44,"tag":1321,"props":1890,"children":1891},{},[1892],{"type":50,"value":1512},{"type":44,"tag":1321,"props":1894,"children":1895},{"class":1323,"line":1461},[1896],{"type":44,"tag":1321,"props":1897,"children":1898},{},[1899],{"type":50,"value":1900},"    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_CALIBRATION}\u002Fimages.tar.gz\",\n",{"type":44,"tag":1321,"props":1902,"children":1903},{"class":1323,"line":1470},[1904],{"type":44,"tag":1321,"props":1905,"children":1906},{},[1907],{"type":50,"value":1521},{"type":44,"tag":118,"props":1909,"children":1911},{"id":1910},"checkpoint-selection",[1912],{"type":50,"value":1913},"Checkpoint Selection",{"type":44,"tag":53,"props":1915,"children":1916},{},[1917,1919,1925,1927,1933,1935,1941,1942,1948,1949,1955,1957,1963,1965,1971,1973,1978],{"type":50,"value":1918},"OneFormer training writes epoch-step checkpoints such as\n",{"type":44,"tag":69,"props":1920,"children":1922},{"className":1921},[],[1923],{"type":50,"value":1924},"model_epoch_000_step_00017.pth",{"type":50,"value":1926}," and may also write a\n",{"type":44,"tag":69,"props":1928,"children":1930},{"className":1929},[],[1931],{"type":50,"value":1932},"oneformer_model_latest.pth",{"type":50,"value":1934}," symlink. For checkpoint-dependent actions, use the\nmodel-skill or SDK parent-model resolver and pass the exact selected checkpoint\npath into ",{"type":44,"tag":69,"props":1936,"children":1938},{"className":1937},[],[1939],{"type":50,"value":1940},"evaluate.checkpoint",{"type":50,"value":200},{"type":44,"tag":69,"props":1943,"children":1945},{"className":1944},[],[1946],{"type":50,"value":1947},"inference.checkpoint",{"type":50,"value":200},{"type":44,"tag":69,"props":1950,"children":1952},{"className":1951},[],[1953],{"type":50,"value":1954},"export.checkpoint",{"type":50,"value":1956},",\n",{"type":44,"tag":69,"props":1958,"children":1960},{"className":1959},[],[1961],{"type":50,"value":1962},"quantize.model_path",{"type":50,"value":1964},", or ",{"type":44,"tag":69,"props":1966,"children":1968},{"className":1967},[],[1969],{"type":50,"value":1970},"train.resume_training_checkpoint_path",{"type":50,"value":1972},". Do not pick\nthe ",{"type":44,"tag":69,"props":1974,"children":1976},{"className":1975},[],[1977],{"type":50,"value":1932},{"type":50,"value":1979}," symlink by name unless the user explicitly asks\nfor latest checkpoint behavior. If the resolver reports a best checkpoint, use\nthat best checkpoint for evaluation\u002Fexport\u002Finference; if the user asks for a\nspecific epoch or step, use the matching epoch-step checkpoint.",{"type":44,"tag":118,"props":1981,"children":1983},{"id":1982},"eval-dataset",[1984],{"type":50,"value":1985},"Eval Dataset",{"type":44,"tag":53,"props":1987,"children":1988},{},[1989],{"type":50,"value":1990},"Optional. Val data configured alongside train in the dataset config.",{"type":44,"tag":118,"props":1992,"children":1994},{"id":1993},"important-parameters",[1995],{"type":50,"value":1996},"Important Parameters",{"type":44,"tag":371,"props":1998,"children":1999},{},[2000,2018,2028,2045,2055,2065,2075,2085,2101],{"type":44,"tag":375,"props":2001,"children":2002},{},[2003,2008,2010,2016],{"type":44,"tag":379,"props":2004,"children":2005},{},[2006],{"type":50,"value":2007},"model.sem_seg_head.num_classes",{"type":50,"value":2009},": Number of segmentation class indices available to the head. Default 133 for COCO panoptic data when ",{"type":44,"tag":69,"props":2011,"children":2013},{"className":2012},[],[2014],{"type":50,"value":2015},"dataset.contiguous_id: True",{"type":50,"value":2017}," remaps raw category ids through the label map. Do not shrink this to a global workflow class count unless the label map and annotations have actually been reduced to that class set.",{"type":44,"tag":375,"props":2019,"children":2020},{},[2021,2026],{"type":44,"tag":379,"props":2022,"children":2023},{},[2024],{"type":50,"value":2025},"model.one_former.hidden_dim",{"type":50,"value":2027},": Keep at 256 for local smoke runs unless\nthe text encoder width is changed in lock-step. Reducing hidden_dim alone\ncauses a text feature\u002Fcontext dimension mismatch during training.",{"type":44,"tag":375,"props":2029,"children":2030},{},[2031,2036,2038,2043],{"type":44,"tag":379,"props":2032,"children":2033},{},[2034],{"type":50,"value":2035},"model.backbone.name",{"type":50,"value":2037},": Default D2SwinTransformer (Swin-based). embed_dim=192, depths=",{"type":44,"tag":1321,"props":2039,"children":2040},{},[2041],{"type":50,"value":2042},"2,2,18,2",{"type":50,"value":2044}," by default.",{"type":44,"tag":375,"props":2046,"children":2047},{},[2048,2053],{"type":44,"tag":379,"props":2049,"children":2050},{},[2051],{"type":50,"value":2052},"train.num_epochs",{"type":50,"value":2054},": Default 50 — significantly higher than most TAO models. OneFormer needs more epochs for convergence.",{"type":44,"tag":375,"props":2056,"children":2057},{},[2058,2063],{"type":44,"tag":379,"props":2059,"children":2060},{},[2061],{"type":50,"value":2062},"train.optim.lr",{"type":50,"value":2064},": Learning rate. Default 1e-5. Lower than Mask2Former's 2e-4.",{"type":44,"tag":375,"props":2066,"children":2067},{},[2068,2073],{"type":44,"tag":379,"props":2069,"children":2070},{},[2071],{"type":50,"value":2072},"model.task_toggling",{"type":50,"value":2074},": Enable\u002Fdisable specific tasks: semantic_on, instance_on, panoptic_on.",{"type":44,"tag":375,"props":2076,"children":2077},{},[2078,2083],{"type":44,"tag":379,"props":2079,"children":2080},{},[2081],{"type":50,"value":2082},"export.task",{"type":50,"value":2084},": Export task mode. Options: semantic, instance, panoptic. Default semantic. Export input defaults to 640x640.",{"type":44,"tag":375,"props":2086,"children":2087},{},[2088,2093,2095,2100],{"type":44,"tag":379,"props":2089,"children":2090},{},[2091],{"type":50,"value":2092},"inference.mode",{"type":50,"value":2094},": Inference mode. Options: semantic, instance, panoptic. Default semantic. image_size defaults to ",{"type":44,"tag":1321,"props":2096,"children":2097},{},[2098],{"type":50,"value":2099},"1024, 1024",{"type":50,"value":1308},{"type":44,"tag":375,"props":2102,"children":2103},{},[2104,2109],{"type":44,"tag":379,"props":2105,"children":2106},{},[2107],{"type":50,"value":2108},"evaluate.iou_per_class",{"type":50,"value":2110},": Report per-class IoU in evaluation. Default True.",{"type":44,"tag":118,"props":2112,"children":2114},{"id":2113},"multi-gpu-multi-node",[2115],{"type":50,"value":2116},"Multi-GPU \u002F Multi-Node",{"type":44,"tag":53,"props":2118,"children":2119},{},[2120,2125,2127,2132],{"type":44,"tag":379,"props":2121,"children":2122},{},[2123],{"type":50,"value":2124},"Launch method:",{"type":50,"value":2126}," Lightning-managed (single ",{"type":44,"tag":69,"props":2128,"children":2130},{"className":2129},[],[2131],{"type":50,"value":1314},{"type":50,"value":2133}," process, Lightning spawns workers).",{"type":44,"tag":414,"props":2135,"children":2136},{},[2137,2157],{"type":44,"tag":418,"props":2138,"children":2139},{},[2140],{"type":44,"tag":422,"props":2141,"children":2142},{},[2143,2147,2152],{"type":44,"tag":426,"props":2144,"children":2145},{},[2146],{"type":50,"value":435},{"type":44,"tag":426,"props":2148,"children":2149},{},[2150],{"type":50,"value":2151},"Description",{"type":44,"tag":426,"props":2153,"children":2154},{},[2155],{"type":50,"value":2156},"Default",{"type":44,"tag":452,"props":2158,"children":2159},{},[2160,2182,2207],{"type":44,"tag":422,"props":2161,"children":2162},{},[2163,2172,2177],{"type":44,"tag":459,"props":2164,"children":2165},{},[2166],{"type":44,"tag":69,"props":2167,"children":2169},{"className":2168},[],[2170],{"type":50,"value":2171},"train.num_gpus",{"type":44,"tag":459,"props":2173,"children":2174},{},[2175],{"type":50,"value":2176},"Number of GPUs",{"type":44,"tag":459,"props":2178,"children":2179},{},[2180],{"type":50,"value":2181},"1",{"type":44,"tag":422,"props":2183,"children":2184},{},[2185,2194,2199],{"type":44,"tag":459,"props":2186,"children":2187},{},[2188],{"type":44,"tag":69,"props":2189,"children":2191},{"className":2190},[],[2192],{"type":50,"value":2193},"train.gpu_ids",{"type":44,"tag":459,"props":2195,"children":2196},{},[2197],{"type":50,"value":2198},"GPU device indices",{"type":44,"tag":459,"props":2200,"children":2201},{},[2202],{"type":44,"tag":1321,"props":2203,"children":2204},{},[2205],{"type":50,"value":2206},"0",{"type":44,"tag":422,"props":2208,"children":2209},{},[2210,2219,2224],{"type":44,"tag":459,"props":2211,"children":2212},{},[2213],{"type":44,"tag":69,"props":2214,"children":2216},{"className":2215},[],[2217],{"type":50,"value":2218},"train.num_nodes",{"type":44,"tag":459,"props":2220,"children":2221},{},[2222],{"type":50,"value":2223},"Number of nodes",{"type":44,"tag":459,"props":2225,"children":2226},{},[2227],{"type":50,"value":2181},{"type":44,"tag":371,"props":2229,"children":2230},{},[2231,2264,2275],{"type":44,"tag":375,"props":2232,"children":2233},{},[2234,2236,2242,2244,2250,2251,2257,2258],{"type":50,"value":2235},"Uses explicit ",{"type":44,"tag":69,"props":2237,"children":2239},{"className":2238},[],[2240],{"type":50,"value":2241},"DDPStrategy",{"type":50,"value":2243}," with ",{"type":44,"tag":69,"props":2245,"children":2247},{"className":2246},[],[2248],{"type":50,"value":2249},"find_unused_parameters=True",{"type":50,"value":200},{"type":44,"tag":69,"props":2252,"children":2254},{"className":2253},[],[2255],{"type":50,"value":2256},"gradient_as_bucket_view=True",{"type":50,"value":200},{"type":44,"tag":69,"props":2259,"children":2261},{"className":2260},[],[2262],{"type":50,"value":2263},"process_group_backend=\"nccl\"",{"type":44,"tag":375,"props":2265,"children":2266},{},[2267,2273],{"type":44,"tag":69,"props":2268,"children":2270},{"className":2269},[],[2271],{"type":50,"value":2272},"sync_batchnorm",{"type":50,"value":2274}," is always enabled",{"type":44,"tag":375,"props":2276,"children":2277},{},[2278],{"type":50,"value":2279},"No fsdp support — DDP only",{"type":44,"tag":53,"props":2281,"children":2282},{},[2283,2288,2290,2296,2297,2303,2304,2310,2311,2317,2318,2324],{"type":44,"tag":379,"props":2284,"children":2285},{},[2286],{"type":50,"value":2287},"Multi-node env vars",{"type":50,"value":2289}," (set by orchestrator): ",{"type":44,"tag":69,"props":2291,"children":2293},{"className":2292},[],[2294],{"type":50,"value":2295},"WORLD_SIZE",{"type":50,"value":200},{"type":44,"tag":69,"props":2298,"children":2300},{"className":2299},[],[2301],{"type":50,"value":2302},"NODE_RANK",{"type":50,"value":200},{"type":44,"tag":69,"props":2305,"children":2307},{"className":2306},[],[2308],{"type":50,"value":2309},"MASTER_ADDR",{"type":50,"value":200},{"type":44,"tag":69,"props":2312,"children":2314},{"className":2313},[],[2315],{"type":50,"value":2316},"MASTER_PORT",{"type":50,"value":200},{"type":44,"tag":69,"props":2319,"children":2321},{"className":2320},[],[2322],{"type":50,"value":2323},"NUM_GPU_PER_NODE",{"type":50,"value":1308},{"type":44,"tag":118,"props":2326,"children":2328},{"id":2327},"hardware",[2329],{"type":50,"value":2330},"Hardware",{"type":44,"tag":53,"props":2332,"children":2333},{},[2334],{"type":50,"value":2335},"Minimum 2 GPU(s), recommended 4 GPU(s). 24GB+ (A100 recommended) VRAM per GPU. OneFormer is memory-intensive like Mask2Former. batch_size=1 is the default. Multi-GPU needed for reasonable training speed, especially with 50 epochs.",{"type":44,"tag":118,"props":2337,"children":2339},{"id":2338},"error-patterns",[2340],{"type":50,"value":2341},"Error Patterns",{"type":44,"tag":53,"props":2343,"children":2344},{},[2345,2350],{"type":44,"tag":379,"props":2346,"children":2347},{},[2348],{"type":50,"value":2349},"CUDA out of memory",{"type":50,"value":2351},": batch_size is already 1. Reduce image resolution or use a smaller Swin configuration.",{"type":44,"tag":53,"props":2353,"children":2354},{},[2355,2360,2362,2367,2368,2373,2375,2381,2382,2388,2390,2396,2397,2403,2404,2410,2412,2418],{"type":44,"tag":379,"props":2356,"children":2357},{},[2358],{"type":50,"value":2359},"Extracted S3 tarball points one level too high",{"type":50,"value":2361},": For local Docker runs,\n",{"type":44,"tag":69,"props":2363,"children":2365},{"className":2364},[],[2366],{"type":50,"value":477},{"type":50,"value":184},{"type":44,"tag":69,"props":2369,"children":2371},{"className":2370},[],[2372],{"type":50,"value":553},{"type":50,"value":2374}," may extract wrapper directories\nsuch as ",{"type":44,"tag":69,"props":2376,"children":2378},{"className":2377},[],[2379],{"type":50,"value":2380},"images\u002F",{"type":50,"value":184},{"type":44,"tag":69,"props":2383,"children":2385},{"className":2384},[],[2386],{"type":50,"value":2387},"images_panoptic\u002F",{"type":50,"value":2389},". Set ",{"type":44,"tag":69,"props":2391,"children":2393},{"className":2392},[],[2394],{"type":50,"value":2395},"dataset.*.images",{"type":50,"value":1956},{"type":44,"tag":69,"props":2398,"children":2400},{"className":2399},[],[2401],{"type":50,"value":2402},"dataset.*.panoptic",{"type":50,"value":200},{"type":44,"tag":69,"props":2405,"children":2407},{"className":2406},[],[2408],{"type":50,"value":2409},"inference.images_dir",{"type":50,"value":2411},", and quantization calibration\npaths to the actual folder containing image or panoptic files, not the wrapper\ndirectory. A one-level-too-high path fails with ",{"type":44,"tag":69,"props":2413,"children":2415},{"className":2414},[],[2416],{"type":50,"value":2417},"FileNotFoundError",{"type":50,"value":2419}," for the\nfirst annotation image even though recursive file counts look correct.",{"type":44,"tag":53,"props":2421,"children":2422},{},[2423,2428,2430,2436,2438,2444,2446,2452,2454,2460],{"type":44,"tag":379,"props":2424,"children":2425},{},[2426],{"type":50,"value":2427},"default_specs missing results_dir",{"type":50,"value":2429},": The CLI ",{"type":44,"tag":69,"props":2431,"children":2433},{"className":2432},[],[2434],{"type":50,"value":2435},"default_specs",{"type":50,"value":2437}," subtask ignores\n",{"type":44,"tag":69,"props":2439,"children":2441},{"className":2440},[],[2442],{"type":50,"value":2443},"-e",{"type":50,"value":2445}," experiment specs for ",{"type":44,"tag":69,"props":2447,"children":2449},{"className":2448},[],[2450],{"type":50,"value":2451},"results_dir",{"type":50,"value":2453},"; pass a Hydra-style override instead:\n",{"type":44,"tag":69,"props":2455,"children":2457},{"className":2456},[],[2458],{"type":50,"value":2459},"oneformer default_specs results_dir=\u002Fpath\u002Fto\u002Fdefault_specs",{"type":50,"value":1308},{"type":44,"tag":53,"props":2462,"children":2463},{},[2464,2475,2477,2483,2485,2490],{"type":44,"tag":379,"props":2465,"children":2466},{},[2467,2469],{"type":50,"value":2468},"Invalid Lightning precision ",{"type":44,"tag":69,"props":2470,"children":2472},{"className":2471},[],[2473],{"type":50,"value":2474},"fp32",{"type":50,"value":2476},": Use ",{"type":44,"tag":69,"props":2478,"children":2480},{"className":2479},[],[2481],{"type":50,"value":2482},"train.precision: \"32\"",{"type":50,"value":2484}," in\ntrain\u002FAutoML\u002Fevaluate\u002Finference specs. The current Lightning stack rejects the\nlegacy ",{"type":44,"tag":69,"props":2486,"children":2488},{"className":2487},[],[2489],{"type":50,"value":2474},{"type":50,"value":2491}," string.",{"type":44,"tag":53,"props":2493,"children":2494},{},[2495,2500,2502,2508],{"type":44,"tag":379,"props":2496,"children":2497},{},[2498],{"type":50,"value":2499},"PyTorch 2.6 checkpoint load failure on downstream actions",{"type":50,"value":2501},": Current\nOneFormer checkpoints include OmegaConf objects. For checkpoints produced by\nthe same trusted TAO train\u002FAutoML workflow, set\n",{"type":44,"tag":69,"props":2503,"children":2505},{"className":2504},[],[2506],{"type":50,"value":2507},"TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1",{"type":50,"value":2509}," in downstream evaluate, inference, export,\nquantize, or resume job env vars so Lightning can load the full checkpoint.\nDo not use this env var for untrusted checkpoints.",{"type":44,"tag":53,"props":2511,"children":2512},{},[2513,2518,2520,2526,2528,2534,2536,2541,2543,2548,2550,2556,2558,2564],{"type":44,"tag":379,"props":2514,"children":2515},{},[2516],{"type":50,"value":2517},"CUDA device-side assert in matcher\u002Fclass cost",{"type":50,"value":2519},": If training fails in\n",{"type":44,"tag":69,"props":2521,"children":2523},{"className":2522},[],[2524],{"type":50,"value":2525},"oneformer\u002Futils\u002Fmatcher.py",{"type":50,"value":2527}," while indexing ",{"type":44,"tag":69,"props":2529,"children":2531},{"className":2530},[],[2532],{"type":50,"value":2533},"out_prob[:, tgt_ids]",{"type":50,"value":2535},", compare\nthe effective target ids with ",{"type":44,"tag":69,"props":2537,"children":2539},{"className":2538},[],[2540],{"type":50,"value":2007},{"type":50,"value":2542},". The packaged\nCOCO panoptic sample has 133 compact classes after ",{"type":44,"tag":69,"props":2544,"children":2546},{"className":2545},[],[2547],{"type":50,"value":2015},{"type":50,"value":2549}," remapping, so use ",{"type":44,"tag":69,"props":2551,"children":2553},{"className":2552},[],[2554],{"type":50,"value":2555},"model.sem_seg_head.num_classes: 133",{"type":50,"value":2557}," even when a\nbroader validation workflow passes a smaller generic ",{"type":44,"tag":69,"props":2559,"children":2561},{"className":2560},[],[2562],{"type":50,"value":2563},"num_classes",{"type":50,"value":2565}," value.\nOnly use a smaller class count when the label map and annotations are reduced\nto that exact contiguous class set.",{"type":44,"tag":53,"props":2567,"children":2568},{},[2569,2574,2576,2581,2583,2588,2590,2595,2597,2602],{"type":44,"tag":379,"props":2570,"children":2571},{},[2572],{"type":50,"value":2573},"Inference returns PASS with no predictions",{"type":50,"value":2575},": OneFormer prediction reads\n",{"type":44,"tag":69,"props":2577,"children":2579},{"className":2578},[],[2580],{"type":50,"value":2409},{"type":50,"value":2582},", not ",{"type":44,"tag":69,"props":2584,"children":2586},{"className":2585},[],[2587],{"type":50,"value":642},{"type":50,"value":2589},". Declare and populate\n",{"type":44,"tag":69,"props":2591,"children":2593},{"className":2592},[],[2594],{"type":50,"value":2409},{"type":50,"value":2596}," with the image folder or tarball for every inference\nrun. ",{"type":44,"tag":69,"props":2598,"children":2600},{"className":2599},[],[2601],{"type":50,"value":642},{"type":50,"value":2603}," may still be useful for shared dataset context, but\nit does not drive the PyTorch predict dataloader.",{"type":44,"tag":53,"props":2605,"children":2606},{},[2607,2612,2614,2620,2622,2627],{"type":44,"tag":379,"props":2608,"children":2609},{},[2610],{"type":50,"value":2611},"Export output path pre-created as a directory",{"type":50,"value":2613},": Do not declare\n",{"type":44,"tag":69,"props":2615,"children":2617},{"className":2616},[],[2618],{"type":50,"value":2619},"export.onnx_file",{"type":50,"value":2621}," as a file output. The OneFormer exporter asserts that the\nONNX path does not already exist, while the local runner pre-creates declared\noutput paths. Set ",{"type":44,"tag":69,"props":2623,"children":2625},{"className":2624},[],[2626],{"type":50,"value":2619},{"type":50,"value":2628}," explicitly in the spec to a non-existing\nfile path under the mounted results tree. Keep the default 640x640 export\nshape for smoke validation; very small export shapes can trigger PyTorch ONNX\nshape-inference failures.",{"type":44,"tag":53,"props":2630,"children":2631},{},[2632,2637,2639,2645,2647,2652],{"type":44,"tag":379,"props":2633,"children":2634},{},[2635],{"type":50,"value":2636},"Quantize cannot find the training label map from an AutoML checkpoint",{"type":50,"value":2638},":\nOneFormer Lightning checkpoints retain train-time absolute dataset paths in\ntheir saved hparams. When running downstream actions from an AutoML child\ncheckpoint, keep the parent AutoML job directory accessible at its original\n",{"type":44,"tag":69,"props":2640,"children":2642},{"className":2641},[],[2643],{"type":50,"value":2644},"\u002Fresults\u002F\u003Cjob_id>",{"type":50,"value":2646}," path inside the action container in addition to passing the\nresolved checkpoint path. Otherwise quantize can fail while loading checkpoint\nhparams even when the current spec includes a valid ",{"type":44,"tag":69,"props":2648,"children":2650},{"className":2649},[],[2651],{"type":50,"value":494},{"type":50,"value":1308},{"type":44,"tag":53,"props":2654,"children":2655},{},[2656,2661],{"type":44,"tag":379,"props":2657,"children":2658},{},[2659],{"type":50,"value":2660},"Slow training",{"type":50,"value":2662},": 50 default epochs with batch_size=1 is slow on single GPU. Use multi-GPU distributed training.",{"type":44,"tag":118,"props":2664,"children":2666},{"id":2665},"spec-param-parent-model-inference",[2667],{"type":50,"value":2668},"Spec Param \u002F Parent Model Inference",{"type":44,"tag":53,"props":2670,"children":2671},{},[2672,2674,2680,2682,2688,2690,2696],{"type":50,"value":2673},"Model-specific inference mappings belong in this MD file, not in ",{"type":44,"tag":69,"props":2675,"children":2677},{"className":2676},[],[2678],{"type":50,"value":2679},"config.json",{"type":50,"value":2681},". Generated runners should read this section and apply the mappings with SDK helpers before ",{"type":44,"tag":69,"props":2683,"children":2685},{"className":2684},[],[2686],{"type":50,"value":2687},"create_job()",{"type":50,"value":2689},". This mirrors the old microservices ",{"type":44,"tag":69,"props":2691,"children":2693},{"className":2692},[],[2694],{"type":50,"value":2695},"infer_params.py",{"type":50,"value":2697}," flow.",{"type":44,"tag":53,"props":2699,"children":2700},{},[2701,2703,2709],{"type":50,"value":2702},"Inference mappings from TAO Core ",{"type":44,"tag":69,"props":2704,"children":2706},{"className":2705},[],[2707],{"type":50,"value":2708},"oneformer.config.json",{"type":50,"value":2710},":",{"type":44,"tag":414,"props":2712,"children":2713},{},[2714,2739],{"type":44,"tag":418,"props":2715,"children":2716},{},[2717],{"type":44,"tag":422,"props":2718,"children":2719},{},[2720,2724,2729,2734],{"type":44,"tag":426,"props":2721,"children":2722},{},[2723],{"type":50,"value":430},{"type":44,"tag":426,"props":2725,"children":2726},{},[2727],{"type":50,"value":2728},"Spec Field",{"type":44,"tag":426,"props":2730,"children":2731},{},[2732],{"type":50,"value":2733},"Inference Function",{"type":44,"tag":426,"props":2735,"children":2736},{},[2737],{"type":50,"value":2738},"Meaning",{"type":44,"tag":452,"props":2740,"children":2741},{},[2742,2772,2801,2829,2858,2885,2912,2941,2968,2995,3023,3053,3080,3107,3134,3162,3189,3216,3243,3270,3297,3324,3364,3394],{"type":44,"tag":422,"props":2743,"children":2744},{},[2745,2749,2758,2767],{"type":44,"tag":459,"props":2746,"children":2747},{},[2748],{"type":50,"value":82},{"type":44,"tag":459,"props":2750,"children":2751},{},[2752],{"type":44,"tag":69,"props":2753,"children":2755},{"className":2754},[],[2756],{"type":50,"value":2757},"encryption_key",{"type":44,"tag":459,"props":2759,"children":2760},{},[2761],{"type":44,"tag":69,"props":2762,"children":2764},{"className":2763},[],[2765],{"type":50,"value":2766},"key",{"type":44,"tag":459,"props":2768,"children":2769},{},[2770],{"type":50,"value":2771},"encryption key",{"type":44,"tag":422,"props":2773,"children":2774},{},[2775,2779,2787,2796],{"type":44,"tag":459,"props":2776,"children":2777},{},[2778],{"type":50,"value":82},{"type":44,"tag":459,"props":2780,"children":2781},{},[2782],{"type":44,"tag":69,"props":2783,"children":2785},{"className":2784},[],[2786],{"type":50,"value":1940},{"type":44,"tag":459,"props":2788,"children":2789},{},[2790],{"type":44,"tag":69,"props":2791,"children":2793},{"className":2792},[],[2794],{"type":50,"value":2795},"parent_model",{"type":44,"tag":459,"props":2797,"children":2798},{},[2799],{"type":50,"value":2800},"model file inferred from the parent job results folder",{"type":44,"tag":422,"props":2802,"children":2803},{},[2804,2808,2817,2825],{"type":44,"tag":459,"props":2805,"children":2806},{},[2807],{"type":50,"value":82},{"type":44,"tag":459,"props":2809,"children":2810},{},[2811],{"type":44,"tag":69,"props":2812,"children":2814},{"className":2813},[],[2815],{"type":50,"value":2816},"evaluate.trt_engine",{"type":44,"tag":459,"props":2818,"children":2819},{},[2820],{"type":44,"tag":69,"props":2821,"children":2823},{"className":2822},[],[2824],{"type":50,"value":2795},{"type":44,"tag":459,"props":2826,"children":2827},{},[2828],{"type":50,"value":2800},{"type":44,"tag":422,"props":2830,"children":2831},{},[2832,2836,2844,2853],{"type":44,"tag":459,"props":2833,"children":2834},{},[2835],{"type":50,"value":82},{"type":44,"tag":459,"props":2837,"children":2838},{},[2839],{"type":44,"tag":69,"props":2840,"children":2842},{"className":2841},[],[2843],{"type":50,"value":2451},{"type":44,"tag":459,"props":2845,"children":2846},{},[2847],{"type":44,"tag":69,"props":2848,"children":2850},{"className":2849},[],[2851],{"type":50,"value":2852},"output_dir",{"type":44,"tag":459,"props":2854,"children":2855},{},[2856],{"type":50,"value":2857},"current job results directory",{"type":44,"tag":422,"props":2859,"children":2860},{},[2861,2865,2873,2881],{"type":44,"tag":459,"props":2862,"children":2863},{},[2864],{"type":50,"value":354},{"type":44,"tag":459,"props":2866,"children":2867},{},[2868],{"type":44,"tag":69,"props":2869,"children":2871},{"className":2870},[],[2872],{"type":50,"value":2757},{"type":44,"tag":459,"props":2874,"children":2875},{},[2876],{"type":44,"tag":69,"props":2877,"children":2879},{"className":2878},[],[2880],{"type":50,"value":2766},{"type":44,"tag":459,"props":2882,"children":2883},{},[2884],{"type":50,"value":2771},{"type":44,"tag":422,"props":2886,"children":2887},{},[2888,2892,2900,2908],{"type":44,"tag":459,"props":2889,"children":2890},{},[2891],{"type":50,"value":354},{"type":44,"tag":459,"props":2893,"children":2894},{},[2895],{"type":44,"tag":69,"props":2896,"children":2898},{"className":2897},[],[2899],{"type":50,"value":1954},{"type":44,"tag":459,"props":2901,"children":2902},{},[2903],{"type":44,"tag":69,"props":2904,"children":2906},{"className":2905},[],[2907],{"type":50,"value":2795},{"type":44,"tag":459,"props":2909,"children":2910},{},[2911],{"type":50,"value":2800},{"type":44,"tag":422,"props":2913,"children":2914},{},[2915,2919,2927,2936],{"type":44,"tag":459,"props":2916,"children":2917},{},[2918],{"type":50,"value":354},{"type":44,"tag":459,"props":2920,"children":2921},{},[2922],{"type":44,"tag":69,"props":2923,"children":2925},{"className":2924},[],[2926],{"type":50,"value":2619},{"type":44,"tag":459,"props":2928,"children":2929},{},[2930],{"type":44,"tag":69,"props":2931,"children":2933},{"className":2932},[],[2934],{"type":50,"value":2935},"create_onnx_file",{"type":44,"tag":459,"props":2937,"children":2938},{},[2939],{"type":50,"value":2940},"output ONNX path",{"type":44,"tag":422,"props":2942,"children":2943},{},[2944,2948,2956,2964],{"type":44,"tag":459,"props":2945,"children":2946},{},[2947],{"type":50,"value":354},{"type":44,"tag":459,"props":2949,"children":2950},{},[2951],{"type":44,"tag":69,"props":2952,"children":2954},{"className":2953},[],[2955],{"type":50,"value":2451},{"type":44,"tag":459,"props":2957,"children":2958},{},[2959],{"type":44,"tag":69,"props":2960,"children":2962},{"className":2961},[],[2963],{"type":50,"value":2852},{"type":44,"tag":459,"props":2965,"children":2966},{},[2967],{"type":50,"value":2857},{"type":44,"tag":422,"props":2969,"children":2970},{},[2971,2975,2983,2991],{"type":44,"tag":459,"props":2972,"children":2973},{},[2974],{"type":50,"value":74},{"type":44,"tag":459,"props":2976,"children":2977},{},[2978],{"type":44,"tag":69,"props":2979,"children":2981},{"className":2980},[],[2982],{"type":50,"value":2757},{"type":44,"tag":459,"props":2984,"children":2985},{},[2986],{"type":44,"tag":69,"props":2987,"children":2989},{"className":2988},[],[2990],{"type":50,"value":2766},{"type":44,"tag":459,"props":2992,"children":2993},{},[2994],{"type":50,"value":2771},{"type":44,"tag":422,"props":2996,"children":2997},{},[2998,3002,3011,3019],{"type":44,"tag":459,"props":2999,"children":3000},{},[3001],{"type":50,"value":74},{"type":44,"tag":459,"props":3003,"children":3004},{},[3005],{"type":44,"tag":69,"props":3006,"children":3008},{"className":3007},[],[3009],{"type":50,"value":3010},"gen_trt_engine.onnx_file",{"type":44,"tag":459,"props":3012,"children":3013},{},[3014],{"type":44,"tag":69,"props":3015,"children":3017},{"className":3016},[],[3018],{"type":50,"value":2795},{"type":44,"tag":459,"props":3020,"children":3021},{},[3022],{"type":50,"value":2800},{"type":44,"tag":422,"props":3024,"children":3025},{},[3026,3030,3039,3048],{"type":44,"tag":459,"props":3027,"children":3028},{},[3029],{"type":50,"value":74},{"type":44,"tag":459,"props":3031,"children":3032},{},[3033],{"type":44,"tag":69,"props":3034,"children":3036},{"className":3035},[],[3037],{"type":50,"value":3038},"gen_trt_engine.trt_engine",{"type":44,"tag":459,"props":3040,"children":3041},{},[3042],{"type":44,"tag":69,"props":3043,"children":3045},{"className":3044},[],[3046],{"type":50,"value":3047},"create_engine_file",{"type":44,"tag":459,"props":3049,"children":3050},{},[3051],{"type":50,"value":3052},"output TensorRT engine path",{"type":44,"tag":422,"props":3054,"children":3055},{},[3056,3060,3068,3076],{"type":44,"tag":459,"props":3057,"children":3058},{},[3059],{"type":50,"value":74},{"type":44,"tag":459,"props":3061,"children":3062},{},[3063],{"type":44,"tag":69,"props":3064,"children":3066},{"className":3065},[],[3067],{"type":50,"value":2451},{"type":44,"tag":459,"props":3069,"children":3070},{},[3071],{"type":44,"tag":69,"props":3072,"children":3074},{"className":3073},[],[3075],{"type":50,"value":2852},{"type":44,"tag":459,"props":3077,"children":3078},{},[3079],{"type":50,"value":2857},{"type":44,"tag":422,"props":3081,"children":3082},{},[3083,3087,3095,3103],{"type":44,"tag":459,"props":3084,"children":3085},{},[3086],{"type":50,"value":90},{"type":44,"tag":459,"props":3088,"children":3089},{},[3090],{"type":44,"tag":69,"props":3091,"children":3093},{"className":3092},[],[3094],{"type":50,"value":2757},{"type":44,"tag":459,"props":3096,"children":3097},{},[3098],{"type":44,"tag":69,"props":3099,"children":3101},{"className":3100},[],[3102],{"type":50,"value":2766},{"type":44,"tag":459,"props":3104,"children":3105},{},[3106],{"type":50,"value":2771},{"type":44,"tag":422,"props":3108,"children":3109},{},[3110,3114,3122,3130],{"type":44,"tag":459,"props":3111,"children":3112},{},[3113],{"type":50,"value":90},{"type":44,"tag":459,"props":3115,"children":3116},{},[3117],{"type":44,"tag":69,"props":3118,"children":3120},{"className":3119},[],[3121],{"type":50,"value":1947},{"type":44,"tag":459,"props":3123,"children":3124},{},[3125],{"type":44,"tag":69,"props":3126,"children":3128},{"className":3127},[],[3129],{"type":50,"value":2795},{"type":44,"tag":459,"props":3131,"children":3132},{},[3133],{"type":50,"value":2800},{"type":44,"tag":422,"props":3135,"children":3136},{},[3137,3141,3150,3158],{"type":44,"tag":459,"props":3138,"children":3139},{},[3140],{"type":50,"value":90},{"type":44,"tag":459,"props":3142,"children":3143},{},[3144],{"type":44,"tag":69,"props":3145,"children":3147},{"className":3146},[],[3148],{"type":50,"value":3149},"inference.trt_engine",{"type":44,"tag":459,"props":3151,"children":3152},{},[3153],{"type":44,"tag":69,"props":3154,"children":3156},{"className":3155},[],[3157],{"type":50,"value":2795},{"type":44,"tag":459,"props":3159,"children":3160},{},[3161],{"type":50,"value":2800},{"type":44,"tag":422,"props":3163,"children":3164},{},[3165,3169,3177,3185],{"type":44,"tag":459,"props":3166,"children":3167},{},[3168],{"type":50,"value":90},{"type":44,"tag":459,"props":3170,"children":3171},{},[3172],{"type":44,"tag":69,"props":3173,"children":3175},{"className":3174},[],[3176],{"type":50,"value":2451},{"type":44,"tag":459,"props":3178,"children":3179},{},[3180],{"type":44,"tag":69,"props":3181,"children":3183},{"className":3182},[],[3184],{"type":50,"value":2852},{"type":44,"tag":459,"props":3186,"children":3187},{},[3188],{"type":50,"value":2857},{"type":44,"tag":422,"props":3190,"children":3191},{},[3192,3196,3204,3212],{"type":44,"tag":459,"props":3193,"children":3194},{},[3195],{"type":50,"value":895},{"type":44,"tag":459,"props":3197,"children":3198},{},[3199],{"type":44,"tag":69,"props":3200,"children":3202},{"className":3201},[],[3203],{"type":50,"value":2757},{"type":44,"tag":459,"props":3205,"children":3206},{},[3207],{"type":44,"tag":69,"props":3208,"children":3210},{"className":3209},[],[3211],{"type":50,"value":2766},{"type":44,"tag":459,"props":3213,"children":3214},{},[3215],{"type":50,"value":2771},{"type":44,"tag":422,"props":3217,"children":3218},{},[3219,3223,3231,3239],{"type":44,"tag":459,"props":3220,"children":3221},{},[3222],{"type":50,"value":895},{"type":44,"tag":459,"props":3224,"children":3225},{},[3226],{"type":44,"tag":69,"props":3227,"children":3229},{"className":3228},[],[3230],{"type":50,"value":1962},{"type":44,"tag":459,"props":3232,"children":3233},{},[3234],{"type":44,"tag":69,"props":3235,"children":3237},{"className":3236},[],[3238],{"type":50,"value":2795},{"type":44,"tag":459,"props":3240,"children":3241},{},[3242],{"type":50,"value":2800},{"type":44,"tag":422,"props":3244,"children":3245},{},[3246,3250,3258,3266],{"type":44,"tag":459,"props":3247,"children":3248},{},[3249],{"type":50,"value":895},{"type":44,"tag":459,"props":3251,"children":3252},{},[3253],{"type":44,"tag":69,"props":3254,"children":3256},{"className":3255},[],[3257],{"type":50,"value":2451},{"type":44,"tag":459,"props":3259,"children":3260},{},[3261],{"type":44,"tag":69,"props":3262,"children":3264},{"className":3263},[],[3265],{"type":50,"value":2852},{"type":44,"tag":459,"props":3267,"children":3268},{},[3269],{"type":50,"value":2857},{"type":44,"tag":422,"props":3271,"children":3272},{},[3273,3277,3285,3293],{"type":44,"tag":459,"props":3274,"children":3275},{},[3276],{"type":50,"value":1105},{"type":44,"tag":459,"props":3278,"children":3279},{},[3280],{"type":44,"tag":69,"props":3281,"children":3283},{"className":3282},[],[3284],{"type":50,"value":2757},{"type":44,"tag":459,"props":3286,"children":3287},{},[3288],{"type":44,"tag":69,"props":3289,"children":3291},{"className":3290},[],[3292],{"type":50,"value":2766},{"type":44,"tag":459,"props":3294,"children":3295},{},[3296],{"type":50,"value":2771},{"type":44,"tag":422,"props":3298,"children":3299},{},[3300,3304,3312,3320],{"type":44,"tag":459,"props":3301,"children":3302},{},[3303],{"type":50,"value":1105},{"type":44,"tag":459,"props":3305,"children":3306},{},[3307],{"type":44,"tag":69,"props":3308,"children":3310},{"className":3309},[],[3311],{"type":50,"value":2451},{"type":44,"tag":459,"props":3313,"children":3314},{},[3315],{"type":44,"tag":69,"props":3316,"children":3318},{"className":3317},[],[3319],{"type":50,"value":2852},{"type":44,"tag":459,"props":3321,"children":3322},{},[3323],{"type":50,"value":2857},{"type":44,"tag":422,"props":3325,"children":3326},{},[3327,3331,3340,3349],{"type":44,"tag":459,"props":3328,"children":3329},{},[3330],{"type":50,"value":1105},{"type":44,"tag":459,"props":3332,"children":3333},{},[3334],{"type":44,"tag":69,"props":3335,"children":3337},{"className":3336},[],[3338],{"type":50,"value":3339},"train.pretrained_backbone",{"type":44,"tag":459,"props":3341,"children":3342},{},[3343],{"type":44,"tag":69,"props":3344,"children":3346},{"className":3345},[],[3347],{"type":50,"value":3348},"{'link': 'https:\u002F\u002Fgithub.com\u002FSwinTransformer\u002Fstorage\u002Freleases\u002Fdownload\u002Fv1.0.8\u002Fswin_tiny_patch4_window7_224_22k.pth', 'destination_path': '\u002Fptm\u002Fmask2former\u002Fswin_tiny_patch4_window7_224_22k\u002Fswin_tiny_patch4_window7_224_22k.pth'}",{"type":44,"tag":459,"props":3350,"children":3351},{},[3352,3354,3362],{"type":50,"value":3353},"{'link': '",{"type":44,"tag":3355,"props":3356,"children":3360},"a",{"href":3357,"rel":3358},"https:\u002F\u002Fgithub.com\u002FSwinTransformer\u002Fstorage\u002Freleases\u002Fdownload\u002Fv1.0.8\u002Fswin_tiny_patch4_window7_224_22k.pth",[3359],"nofollow",[3361],{"type":50,"value":3357},{"type":50,"value":3363},"', 'destination_path': '\u002Fptm\u002Fmask2former\u002Fswin_tiny_patch4_window7_224_22k\u002Fswin_tiny_patch4_window7_224_22k.pth'}",{"type":44,"tag":422,"props":3365,"children":3366},{},[3367,3371,3380,3389],{"type":44,"tag":459,"props":3368,"children":3369},{},[3370],{"type":50,"value":1105},{"type":44,"tag":459,"props":3372,"children":3373},{},[3374],{"type":44,"tag":69,"props":3375,"children":3377},{"className":3376},[],[3378],{"type":50,"value":3379},"train.pretrained_model",{"type":44,"tag":459,"props":3381,"children":3382},{},[3383],{"type":44,"tag":69,"props":3384,"children":3386},{"className":3385},[],[3387],{"type":50,"value":3388},"ptm_if_no_resume_model",{"type":44,"tag":459,"props":3390,"children":3391},{},[3392],{"type":50,"value":3393},"PTM when no resume checkpoint exists",{"type":44,"tag":422,"props":3395,"children":3396},{},[3397,3401,3409,3418],{"type":44,"tag":459,"props":3398,"children":3399},{},[3400],{"type":50,"value":1105},{"type":44,"tag":459,"props":3402,"children":3403},{},[3404],{"type":44,"tag":69,"props":3405,"children":3407},{"className":3406},[],[3408],{"type":50,"value":1970},{"type":44,"tag":459,"props":3410,"children":3411},{},[3412],{"type":44,"tag":69,"props":3413,"children":3415},{"className":3414},[],[3416],{"type":50,"value":3417},"resume_model",{"type":44,"tag":459,"props":3419,"children":3420},{},[3421],{"type":50,"value":3422},"model file inferred from the current job results folder",{"type":44,"tag":53,"props":3424,"children":3425},{},[3426,3428,3433,3435,3441,3443,3449,3451,3456],{"type":50,"value":3427},"For ",{"type":44,"tag":69,"props":3429,"children":3431},{"className":3430},[],[3432],{"type":50,"value":2795},{"type":50,"value":3434}," or ",{"type":44,"tag":69,"props":3436,"children":3438},{"className":3437},[],[3439],{"type":50,"value":3440},"parent_model_folder",{"type":50,"value":3442},", pass the upstream train\u002Fexport\u002FAutoML child job id as ",{"type":44,"tag":69,"props":3444,"children":3446},{"className":3445},[],[3447],{"type":50,"value":3448},"parent_job_id",{"type":50,"value":3450},". 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 ",{"type":44,"tag":69,"props":3452,"children":3454},{"className":3453},[],[3455],{"type":50,"value":2679},{"type":50,"value":3457}," and do not patch generated runner scripts to guess checkpoint paths.",{"type":44,"tag":118,"props":3459,"children":3461},{"id":3460},"deployment",[3462],{"type":50,"value":3463},"Deployment",{"type":44,"tag":371,"props":3465,"children":3466},{},[3467],{"type":44,"tag":375,"props":3468,"children":3469},{},[3470],{"type":44,"tag":3355,"props":3471,"children":3472},{"href":98},[3473],{"type":50,"value":3474},"tao-deploy-oneformer",{"type":44,"tag":3476,"props":3477,"children":3478},"style",{},[3479],{"type":50,"value":3480},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":3482,"total":3584},[3483,3500,3512,3526,3538,3555,3570],{"slug":3484,"name":3484,"fn":3485,"description":3486,"org":3487,"tags":3488,"stars":20,"repoUrl":21,"updatedAt":3499},"accelerated-computing-cudf","accelerate data processing with cuDF","Official NVIDIA-authored guidance for NVIDIA cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby, CSV\u002FParquet I\u002FO, nullable semantics, and multi-GPU DataFrame workloads.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3489,3492,3495,3496],{"name":3490,"slug":3491,"type":15},"Data Analysis","data-analysis",{"name":3493,"slug":3494,"type":15},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":15},{"name":3497,"slug":3498,"type":15},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":3501,"name":3501,"fn":3502,"description":3503,"org":3504,"tags":3505,"stars":20,"repoUrl":21,"updatedAt":3511},"aiq-deploy","deploy and manage NVIDIA AI-Q infrastructure","Use when asked to install, deploy, run, validate, troubleshoot, or stop NVIDIA AI-Q Blueprint infrastructure.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3506,3507,3510],{"name":3463,"slug":3460,"type":15},{"name":3508,"slug":3509,"type":15},"Infrastructure","infrastructure",{"name":9,"slug":8,"type":15},"2026-07-14T05:29:06.667109",{"slug":3513,"name":3513,"fn":3514,"description":3515,"org":3516,"tags":3517,"stars":20,"repoUrl":21,"updatedAt":3525},"aiq-research","conduct deep research with AI-Q","Use when asked to run deep research or AI-Q research through a reachable NVIDIA AI-Q Blueprint backend.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3518,3521,3522],{"name":3519,"slug":3520,"type":15},"Agents","agents",{"name":9,"slug":8,"type":15},{"name":3523,"slug":3524,"type":15},"Research","research","2026-07-14T05:28:06.816956",{"slug":3527,"name":3527,"fn":3528,"description":3529,"org":3530,"tags":3531,"stars":20,"repoUrl":21,"updatedAt":3537},"amc-run-sample-calibration","run AMC sample dataset calibration","Run end-to-end calibration on the shipped sample dataset (sdg_08_2_sample_data_010926.zip) against a running AMC microservice. Use when user says 'test sample dataset', 'run sample calibration', 'verify AMC install', or 'launch and test'.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3532,3533,3534],{"name":3490,"slug":3491,"type":15},{"name":9,"slug":8,"type":15},{"name":3535,"slug":3536,"type":15},"Testing","testing","2026-07-17T05:29:03.913266",{"slug":3539,"name":3539,"fn":3540,"description":3541,"org":3542,"tags":3543,"stars":20,"repoUrl":21,"updatedAt":3554},"amc-run-video-calibration","calibrate video datasets with AutoMagicCalib","Calibrate a new dataset from pre-recorded video files via the AutoMagicCalib REST API. Use when user has local MP4s and says 'calibrate my videos', 'run AMC on these videos', or similar. For RTSP\u002Flive streams, use amc-run-rtsp-calibration instead.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3544,3547,3550,3551],{"name":3545,"slug":3546,"type":15},"Automation","automation",{"name":3548,"slug":3549,"type":15},"Imaging","imaging",{"name":9,"slug":8,"type":15},{"name":3552,"slug":3553,"type":15},"Video","video","2026-07-17T05:28:53.905004",{"slug":3556,"name":3556,"fn":3557,"description":3558,"org":3559,"tags":3560,"stars":20,"repoUrl":21,"updatedAt":3569},"amc-setup-calibration-stack","deploy AutoMagicCalib microservice with Docker","Launch AutoMagicCalib microservice and web UI from NGC release images via Docker Compose. Use when user says 'deploy auto calibration', 'launch auto calibration', 'launch AMC', 'start MS+UI', or 'set up auto-magic-calib'. Requires NGC API key.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3561,3562,3565,3566],{"name":3463,"slug":3460,"type":15},{"name":3563,"slug":3564,"type":15},"Docker","docker",{"name":9,"slug":8,"type":15},{"name":3567,"slug":3568,"type":15},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":3571,"name":3571,"fn":3572,"description":3573,"org":3574,"tags":3575,"stars":20,"repoUrl":21,"updatedAt":3583},"cudaq-guide","develop quantum applications with CUDA-Q","CUDA-Q onboarding guide for installation, test programs, GPU simulation, QPU hardware, and quantum applications.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3576,3577,3580],{"name":9,"slug":8,"type":15},{"name":3578,"slug":3579,"type":15},"Quantum Computing","quantum-computing",{"name":3581,"slug":3582,"type":15},"Simulation","simulation","2026-07-14T05:26:58.898253",305,{"items":3586,"total":3736},[3587,3605,3620,3631,3643,3657,3670,3684,3695,3704,3718,3727],{"slug":3588,"name":3588,"fn":3589,"description":3590,"org":3591,"tags":3592,"stars":3602,"repoUrl":3603,"updatedAt":3604},"nemoclaw-user-guide","retrieve NemoClaw documentation and configuration","Guides human users' AI agents to the NemoClaw docs MCP server and canonical Fern documentation in Markdown form. Use when users ask how to install, configure, operate, troubleshoot, secure, or learn NemoClaw with an AI coding assistant. Trigger keywords - nemoclaw docs, use nemoclaw with ai agent, nemoclaw mcp docs, nemoclaw install help, nemoclaw quickstart, nemoclaw markdown docs, llms.txt, agent skills.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3593,3596,3599],{"name":3594,"slug":3595,"type":15},"Documentation","documentation",{"name":3597,"slug":3598,"type":15},"MCP","mcp",{"name":3600,"slug":3601,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":3606,"name":3606,"fn":3607,"description":3608,"org":3609,"tags":3610,"stars":3617,"repoUrl":3618,"updatedAt":3619},"mcore-build-and-dependency","manage Megatron-LM development environments","Container-based dev environment setup and dependency management for Megatron-LM. Covers acquiring and launching the CI container, uv package management, and updating uv.lock.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3611,3614,3615],{"name":3612,"slug":3613,"type":15},"Containers","containers",{"name":3463,"slug":3460,"type":15},{"name":3616,"slug":1314,"type":15},"Python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":3621,"name":3621,"fn":3622,"description":3623,"org":3624,"tags":3625,"stars":3617,"repoUrl":3618,"updatedAt":3630},"mcore-bump-base-image","update NVIDIA PyTorch base images","Bump the NVIDIA PyTorch base image (`nvcr.io\u002Fnvidia\u002Fpytorch:YY.MM-py3`) used by Megatron-LM CI. Covers the two pin sites (GitHub CI in `docker\u002F.ngc_version.dev` and GitLab CI in `.gitlab\u002Fstages\u002F01.build.yml`), the post-bump CI loop (re-run functional tests, refresh golden values, mark broken tests), and the gotchas that bit PRs",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3626,3629],{"name":3627,"slug":3628,"type":15},"CI\u002FCD","ci-cd",{"name":3463,"slug":3460,"type":15},"2026-07-14T05:25:59.97109",{"slug":3632,"name":3632,"fn":3633,"description":3634,"org":3635,"tags":3636,"stars":3617,"repoUrl":3618,"updatedAt":3642},"mcore-cicd","manage CI\u002FCD pipelines for Megatron-LM","CI\u002FCD reference for Megatron-LM. Covers CI pipeline structure, PR scope labels, triggering internal GitLab CI (which force-pushes the current branch to a pull-request\u002FBRANCH ref — always dry-run and verify the destination first; never run against shared or protected branches), and CI failure investigation.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3637,3638,3639],{"name":3627,"slug":3628,"type":15},{"name":3463,"slug":3460,"type":15},{"name":3640,"slug":3641,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":3644,"name":3644,"fn":3645,"description":3646,"org":3647,"tags":3648,"stars":3617,"repoUrl":3618,"updatedAt":3656},"mcore-create-issue","investigate CI failures and create issues","Investigate a failing GitHub Actions run or job and create a GitHub issue for the failure.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3649,3652,3653],{"name":3650,"slug":3651,"type":15},"Debugging","debugging",{"name":3640,"slug":3641,"type":15},{"name":3654,"slug":3655,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":3658,"name":3658,"fn":3659,"description":3660,"org":3661,"tags":3662,"stars":3617,"repoUrl":3618,"updatedAt":3669},"mcore-linting-and-formatting","lint and format Megatron-LM code","Linting and formatting for Megatron-LM. Covers running autoformat.sh, tools (ruff, black, isort, pylint, mypy), and code style rules.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3663,3666],{"name":3664,"slug":3665,"type":15},"Best Practices","best-practices",{"name":3667,"slug":3668,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":3671,"name":3671,"fn":3672,"description":3673,"org":3674,"tags":3675,"stars":3617,"repoUrl":3618,"updatedAt":3683},"mcore-migrate-gpt-to-hybrid","migrate Megatron-LM models to HybridModel","Migration guide for moving Megatron Core GPTModel checkpoints, model providers, training commands, and layer mappings to HybridModel.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3676,3679,3682],{"name":3677,"slug":3678,"type":15},"Machine Learning","machine-learning",{"name":3680,"slug":3681,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":3685,"name":3685,"fn":3686,"description":3687,"org":3688,"tags":3689,"stars":3617,"repoUrl":3618,"updatedAt":3694},"mcore-onboard-gb200-1node-tests","onboard functional tests for GB200","Onboard 1-node GitHub MR functional tests for GB200 from existing mr-scoped 2-node tests.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3690,3693],{"name":3691,"slug":3692,"type":15},"QA","qa",{"name":3535,"slug":3536,"type":15},"2026-07-14T05:25:53.673039",{"slug":3696,"name":3696,"fn":3697,"description":3698,"org":3699,"tags":3700,"stars":3617,"repoUrl":3618,"updatedAt":3703},"mcore-run-on-slurm","launch distributed training jobs on SLURM","How to launch distributed Megatron-LM training jobs on a SLURM cluster. Covers a minimal sbatch skeleton, environment-variable setup for torch.distributed.run, CUDA_DEVICE_MAX_CONNECTIONS rules across hardware and parallelism modes, container conventions, monitoring, and per-rank failure diagnosis.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3701,3702],{"name":3463,"slug":3460,"type":15},{"name":3508,"slug":3509,"type":15},"2026-07-14T05:25:49.362534",{"slug":3705,"name":3705,"fn":3706,"description":3707,"org":3708,"tags":3709,"stars":3617,"repoUrl":3618,"updatedAt":3717},"mcore-split-pr","split pull requests to reduce review load","Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3710,3713,3714],{"name":3711,"slug":3712,"type":15},"Code Review","code-review",{"name":3640,"slug":3641,"type":15},{"name":3715,"slug":3716,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":3719,"name":3719,"fn":3720,"description":3721,"org":3722,"tags":3723,"stars":3617,"repoUrl":3618,"updatedAt":3726},"mcore-testing","run and manage Megatron-LM tests","Test system for Megatron-LM. Covers test layout, recipe YAML structure, adding and running unit and functional tests, golden values, marker filters, and CI parity.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3724,3725],{"name":3691,"slug":3692,"type":15},{"name":3535,"slug":3536,"type":15},"2026-07-14T05:25:54.928983",{"slug":3728,"name":3728,"fn":3729,"description":3730,"org":3731,"tags":3732,"stars":3617,"repoUrl":3618,"updatedAt":3735},"nightly-sync","manage nightly main-to-dev sync workflows","Domain knowledge for the nightly main-to-dev sync workflow. Covers merge strategy, CI architecture, failure investigation, and known issues.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[3733,3734],{"name":3545,"slug":3546,"type":15},{"name":3627,"slug":3628,"type":15},"2026-07-30T05:29:03.275638",496]