[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79FX3511T2F7GDF836HDZQR":3,"skill-same:01M79FX3511T2F7GDF836HDZQR":162},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":32,"commitSha":33,"skillDirectory":34,"forksCount":14,"repositoryHtmlUrl":35,"repositoryHomepageUrl":6,"ownerHtmlUrl":36,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":37,"install":38,"resources":51},"01M79FX3511T2F7GDF836HDZQR","tao-train-ocrnet",null,"OCRNet for scene text recognition. Recognizes text content from cropped text-region images and supports CTC and attention-based decoders. Use when training, evaluating, exporting, pruning, quantizing, retraining, or running inference for a TAO OCRNet model. Trigger phrases include \"train OCRNet\", \"scene text recognition\", \"OCR cropped text\", \"CTC \u002F attention text decoder\".","Elogic360\u002Fsoftware-factory","Elogic360","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F156535519?v=4","01M74NH3F0RB13PWB7JZDE2V80","elogic360","01M74NH3FPC65082MDZVMFHDM3",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: tao-train-ocrnet\ndescription: OCRNet for scene text recognition. Recognizes text content from cropped text-region images and supports CTC\n  and attention-based decoders. Use when training, evaluating, exporting, pruning, quantizing, retraining, or running\n  inference for a TAO OCRNet model. Trigger phrases include \"train OCRNet\", \"scene text recognition\", \"OCR cropped text\",\n  \"CTC \u002F attention text decoder\".\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- text\n- recognition\n---\n\n# OCRNet\n\nOCRNet for scene text recognition. Recognizes text content from cropped text region images. Supports CTC and attention-based decoders.\n\nSet train.pretrained_model_path for pretrained OCR weights.\n\nFor TAO Deploy TensorRT actions (`gen_trt_engine`, TensorRT `evaluate`, and TensorRT `inference`), read `references\u002Ftao-deploy-ocrnet.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. Treat phrases like \"turn off AutoML\", \"disable AutoML\", \"no HPO\", or \"plain training\" as `automl_policy: off` for this run only; otherwise default to `auto`. When `automl_policy: auto`, `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:** ocrnet\n- **Formats:** default\n- **Monitoring metric:** val_acc\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| dataset_convert | dataset_convert.input_img_dir | id |  | No |\n| dataset_convert | dataset_convert.gt_file | id |  | No |\n| evaluate | dataset.character_list_file | eval_dataset | character_list | No |\n| evaluate | evaluate.test_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| export | dataset.character_list_file | eval_dataset | character_list | No |\n| gen_trt_engine | gen_trt_engine.tensorrt.calibration.cal_image_dir | calibration_dataset |  | Yes |\n| inference | dataset.character_list_file | eval_dataset | character_list | No |\n| inference | inference.inference_dataset_dir | eval_dataset | test.tar.gz | No |\n| prune | dataset.character_list_file | eval_dataset | character_list | No |\n| quantize | dataset.train_dataset_dir | train_datasets | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | Yes |\n| quantize | dataset.val_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| quantize | dataset.character_list_file | eval_dataset | character_list | No |\n| quantize | dataset.quant_calibration_dataset.images_dir | train_datasets | train.tar.gz | No |\n| retrain | dataset.train_dataset_dir | train_datasets | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | Yes |\n| retrain | dataset.val_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| retrain | dataset.character_list_file | eval_dataset | character_list | No |\n| train | dataset.train_dataset_dir | train_datasets | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | Yes |\n| train | dataset.val_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| train | dataset.character_list_file | eval_dataset | character_list | 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\"\n```\n\n**train (mandatory data sources):**\n```python\n{\n    \"train.num_epochs\": 30,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"train.num_gpus\": 1,\n    \"dataset.batch_size\": 16,\n    \"dataset.train_dataset_dir\": [f\"{S3_TRAIN}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\"],\n    \"dataset.val_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n\n**gen_trt_engine (mandatory data sources):**\n```python\n{\n    \"gen_trt_engine.tensorrt.data_type\": \"fp16\",\n    \"gen_trt_engine.tensorrt.calibration.cal_image_dir\": [f\"{S3_TRAIN}\"],\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n    \"evaluate.test_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n}\n```\n\n**export (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n\n**inference (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n    \"inference.inference_dataset_dir\": f\"{S3_EVAL}\u002Ftest.tar.gz\",\n}\n```\n\n**prune (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n\n**quantize (mandatory data sources):**\n```python\n{\n    \"dataset.train_dataset_dir\": [f\"{S3_TRAIN}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\"],\n    \"dataset.val_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_TRAIN}\u002Ftrain.tar.gz\",\n}\n```\n\n**retrain (mandatory data sources):**\n```python\n{\n    \"dataset.train_dataset_dir\": [f\"{S3_TRAIN}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\"],\n    \"dataset.val_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n## Eval Dataset\n\nOptional. Test data provided as separate tarball.\n\n## Important Parameters\n\n- **dataset.character_list_file**: Path to character list defining the supported character set. This determines the output vocabulary size.\n- **model.backbone**: Default ResNet.\n- **model.prediction**: Decoder type. CTC or Attn (attention-based).\n- **train.optim.lr**: Learning rate. Default 1.0 (Adadelta optimizer). High default is specific to Adadelta.\n- **dataset.batch_size**: Per-GPU batch size. Default 16.\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.distributed_strategy` | Strategy name | `auto` |\n\n- Strategy: `auto` for single-GPU, reads `train.distributed_strategy` from config when multi-GPU\n- No explicit `num_nodes` in train script — single-node oriented\n- Lightweight model, single GPU typically sufficient\n\n## Hardware\n\nMinimum 1 GPU(s), recommended 1 GPU(s). 8GB+ VRAM per GPU. OCR text recognition is lightweight. Single GPU is typically sufficient.\n\n## Error Patterns\n\n**dataset_convert required**: If using raw images + gt files, run dataset_convert first to produce LMDB format.\n\n**Character list mismatch**: All characters in training data must be present in the character_list file.\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 `ocrnet.config.json`:\n\n| Action | Spec Field | Inference Function | Meaning |\n|---|---|---|---|\n| dataset_convert | `results_dir` | `output_dir` | current job results directory |\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 | `model.pruned_graph_path` | `pruned_model` | parent pruned model |\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.tensorrt.calibration.cal_cache_file` | `create_cal_cache` | calibration cache path |\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 | `model.pruned_graph_path` | `pruned_model` | parent pruned model |\n| inference | `results_dir` | `output_dir` | current job results directory |\n| prune | `encryption_key` | `key` | encryption key |\n| prune | `prune.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| prune | `prune.pruned_file` | `create_pth_file` | output PTH path |\n| prune | `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| retrain | `encryption_key` | `key` | encryption key |\n| retrain | `model.pruned_graph_path` | `parent_model` | model file inferred from the parent job results folder |\n| retrain | `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_model_path` | `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-ocrnet](references\u002Ftao-deploy-ocrnet.md) — OCRNet deploy workflow for TensorRT engine generation, TensorRT evaluation, and TensorRT inference using TAO Deploy.\n","# OCRNet\n\nOCRNet for scene text recognition. Recognizes text content from cropped text region images. Supports CTC and attention-based decoders.\n\nSet train.pretrained_model_path for pretrained OCR weights.\n\nFor TAO Deploy TensorRT actions (`gen_trt_engine`, TensorRT `evaluate`, and TensorRT `inference`), read `references\u002Ftao-deploy-ocrnet.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. Treat phrases like \"turn off AutoML\", \"disable AutoML\", \"no HPO\", or \"plain training\" as `automl_policy: off` for this run only; otherwise default to `auto`. When `automl_policy: auto`, `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:** ocrnet\n- **Formats:** default\n- **Monitoring metric:** val_acc\n\n### Per-Action Dataset Requirements\n\n| Action | Spec Key | Source | Files | List? |\n|---|---|---|---|---|\n| dataset_convert | dataset_convert.input_img_dir | id |  | No |\n| dataset_convert | dataset_convert.gt_file | id |  | No |\n| evaluate | dataset.character_list_file | eval_dataset | character_list | No |\n| evaluate | evaluate.test_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| export | dataset.character_list_file | eval_dataset | character_list | No |\n| gen_trt_engine | gen_trt_engine.tensorrt.calibration.cal_image_dir | calibration_dataset |  | Yes |\n| inference | dataset.character_list_file | eval_dataset | character_list | No |\n| inference | inference.inference_dataset_dir | eval_dataset | test.tar.gz | No |\n| prune | dataset.character_list_file | eval_dataset | character_list | No |\n| quantize | dataset.train_dataset_dir | train_datasets | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | Yes |\n| quantize | dataset.val_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| quantize | dataset.character_list_file | eval_dataset | character_list | No |\n| quantize | dataset.quant_calibration_dataset.images_dir | train_datasets | train.tar.gz | No |\n| retrain | dataset.train_dataset_dir | train_datasets | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | Yes |\n| retrain | dataset.val_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| retrain | dataset.character_list_file | eval_dataset | character_list | No |\n| train | dataset.train_dataset_dir | train_datasets | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | Yes |\n| train | dataset.val_dataset_dir | eval_dataset | results\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb | No |\n| train | dataset.character_list_file | eval_dataset | character_list | 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\"\n```\n\n**train (mandatory data sources):**\n```python\n{\n    \"train.num_epochs\": 30,\n    \"train.checkpoint_interval\": 10,\n    \"train.validation_interval\": 10,\n    \"train.num_gpus\": 1,\n    \"dataset.batch_size\": 16,\n    \"dataset.train_dataset_dir\": [f\"{S3_TRAIN}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\"],\n    \"dataset.val_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n\n**gen_trt_engine (mandatory data sources):**\n```python\n{\n    \"gen_trt_engine.tensorrt.data_type\": \"fp16\",\n    \"gen_trt_engine.tensorrt.calibration.cal_image_dir\": [f\"{S3_TRAIN}\"],\n}\n```\n\n**evaluate (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n    \"evaluate.test_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n}\n```\n\n**export (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n\n**inference (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n    \"inference.inference_dataset_dir\": f\"{S3_EVAL}\u002Ftest.tar.gz\",\n}\n```\n\n**prune (mandatory data sources):**\n```python\n{\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n\n**quantize (mandatory data sources):**\n```python\n{\n    \"dataset.train_dataset_dir\": [f\"{S3_TRAIN}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\"],\n    \"dataset.val_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n    \"dataset.quant_calibration_dataset.images_dir\": f\"{S3_TRAIN}\u002Ftrain.tar.gz\",\n}\n```\n\n**retrain (mandatory data sources):**\n```python\n{\n    \"dataset.train_dataset_dir\": [f\"{S3_TRAIN}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\"],\n    \"dataset.val_dataset_dir\": f\"{S3_EVAL}\u002Fresults\u002F{dataset_convert_job_id}\u002Fdataset_convert\u002Flmdb\",\n    \"dataset.character_list_file\": f\"{S3_EVAL}\u002Fcharacter_list\",\n}\n```\n## Eval Dataset\n\nOptional. Test data provided as separate tarball.\n\n## Important Parameters\n\n- **dataset.character_list_file**: Path to character list defining the supported character set. This determines the output vocabulary size.\n- **model.backbone**: Default ResNet.\n- **model.prediction**: Decoder type. CTC or Attn (attention-based).\n- **train.optim.lr**: Learning rate. Default 1.0 (Adadelta optimizer). High default is specific to Adadelta.\n- **dataset.batch_size**: Per-GPU batch size. Default 16.\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.distributed_strategy` | Strategy name | `auto` |\n\n- Strategy: `auto` for single-GPU, reads `train.distributed_strategy` from config when multi-GPU\n- No explicit `num_nodes` in train script — single-node oriented\n- Lightweight model, single GPU typically sufficient\n\n## Hardware\n\nMinimum 1 GPU(s), recommended 1 GPU(s). 8GB+ VRAM per GPU. OCR text recognition is lightweight. Single GPU is typically sufficient.\n\n## Error Patterns\n\n**dataset_convert required**: If using raw images + gt files, run dataset_convert first to produce LMDB format.\n\n**Character list mismatch**: All characters in training data must be present in the character_list file.\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 `ocrnet.config.json`:\n\n| Action | Spec Field | Inference Function | Meaning |\n|---|---|---|---|\n| dataset_convert | `results_dir` | `output_dir` | current job results directory |\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 | `model.pruned_graph_path` | `pruned_model` | parent pruned model |\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.tensorrt.calibration.cal_cache_file` | `create_cal_cache` | calibration cache path |\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 | `model.pruned_graph_path` | `pruned_model` | parent pruned model |\n| inference | `results_dir` | `output_dir` | current job results directory |\n| prune | `encryption_key` | `key` | encryption key |\n| prune | `prune.checkpoint` | `parent_model` | model file inferred from the parent job results folder |\n| prune | `prune.pruned_file` | `create_pth_file` | output PTH path |\n| prune | `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| retrain | `encryption_key` | `key` | encryption key |\n| retrain | `model.pruned_graph_path` | `parent_model` | model file inferred from the parent job results folder |\n| retrain | `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_model_path` | `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-ocrnet](references\u002Ftao-deploy-ocrnet.md) — OCRNet deploy workflow for TensorRT engine generation, TensorRT evaluation, and TensorRT inference using TAO Deploy.\n",{"name":5,"tags":24,"license":15,"metadata":27,"description":7,"allowed-tools":30,"compatibility":31},[25,26],"text","recognition",{"author":28,"version":29},"NVIDIA Corporation","0.1.0","Read Bash","Requires docker + nvidia-container-toolkit.","main","c475fc64e94e803b50d14a791dfee8853fbb2c5c","skills\u002Fnvidia\u002Ftao-train-ocrnet","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory","https:\u002F\u002Fgithub.com\u002FElogic360","2026-11-08 22:21:22",{"source":8,"skillName":5,"commands":39,"prompt":49,"zipAvailable":50},[40,43,46],{"tool":41,"command":42},"npx","npx skills add Elogic360\u002Fsoftware-factory --skill tao-train-ocrnet",{"tool":44,"command":45},"bunx","bunx skills add Elogic360\u002Fsoftware-factory --skill tao-train-ocrnet",{"tool":47,"command":48},"pnpm","pnpm dlx skills add Elogic360\u002Fsoftware-factory --skill tao-train-ocrnet","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002FSKILL.md",true,[52,58,62,66,70,74,78,82,86,90,94,98,101,104,108,112,116,120,124,128,132,136,140,144,148,152,156],{"path":53,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":56,"sourceHtmlUrl":57},"BENCHMARK.md","REFERENCE","参考文档",4233,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002FBENCHMARK.md",{"path":59,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":60,"sourceHtmlUrl":61},"evals\u002Fevals.json",740,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fevals\u002Fevals.json",{"path":63,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Fskill_info.yaml",2306,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fskill_info.yaml",{"path":67,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Fspec_template_dataset_convert.yaml",1817,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_dataset_convert.yaml",{"path":71,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":72,"sourceHtmlUrl":73},"references\u002Fspec_template_deploy_experiment.yaml",1369,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_deploy_experiment.yaml",{"path":75,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":76,"sourceHtmlUrl":77},"references\u002Fspec_template_evaluate.yaml",1960,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_evaluate.yaml",{"path":79,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":80,"sourceHtmlUrl":81},"references\u002Fspec_template_export.yaml",1818,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_export.yaml",{"path":83,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":84,"sourceHtmlUrl":85},"references\u002Fspec_template_gen_trt_engine.yaml",2141,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_gen_trt_engine.yaml",{"path":87,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":88,"sourceHtmlUrl":89},"references\u002Fspec_template_inference.yaml",1939,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_inference.yaml",{"path":91,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":92,"sourceHtmlUrl":93},"references\u002Fspec_template_prune.yaml",1912,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_prune.yaml",{"path":95,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":96,"sourceHtmlUrl":97},"references\u002Fspec_template_quantize.yaml",1746,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_quantize.yaml",{"path":99,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":96,"sourceHtmlUrl":100},"references\u002Fspec_template_retrain.yaml","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_retrain.yaml",{"path":102,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":96,"sourceHtmlUrl":103},"references\u002Fspec_template_train.yaml","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Fspec_template_train.yaml",{"path":105,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":106,"sourceHtmlUrl":107},"references\u002Ftao-deploy-ocrnet.md",4860,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Ftao-deploy-ocrnet.md",{"path":109,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":110,"sourceHtmlUrl":111},"references\u002Ftao-deploy-ocrnet.skill_info.yaml",2187,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Freferences\u002Ftao-deploy-ocrnet.skill_info.yaml",{"path":113,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":114,"sourceHtmlUrl":115},"schemas\u002Fdataset_convert.schema.json",29632,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fdataset_convert.schema.json",{"path":117,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":118,"sourceHtmlUrl":119},"schemas\u002Fevaluate.schema.json",32095,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fevaluate.schema.json",{"path":121,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":122,"sourceHtmlUrl":123},"schemas\u002Fexport.schema.json",29707,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fexport.schema.json",{"path":125,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":126,"sourceHtmlUrl":127},"schemas\u002Fgen_trt_engine.schema.json",37154,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fgen_trt_engine.schema.json",{"path":129,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":130,"sourceHtmlUrl":131},"schemas\u002Finference.schema.json",31868,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Finference.schema.json",{"path":133,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":134,"sourceHtmlUrl":135},"schemas\u002Fmanifest.json",20327,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fmanifest.json",{"path":137,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":138,"sourceHtmlUrl":139},"schemas\u002Fprune.schema.json",31670,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fprune.schema.json",{"path":141,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":142,"sourceHtmlUrl":143},"schemas\u002Fquantize.schema.json",28711,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fquantize.schema.json",{"path":145,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":146,"sourceHtmlUrl":147},"schemas\u002Fretrain.schema.json",28709,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Fretrain.schema.json",{"path":149,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":150,"sourceHtmlUrl":151},"schemas\u002Ftrain.schema.json",28705,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fschemas\u002Ftrain.schema.json",{"path":153,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":154,"sourceHtmlUrl":155},"skill-card.md",3825,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fskill-card.md",{"path":157,"resourceType":158,"resourceTypeDesc":159,"sizeBytes":160,"sourceHtmlUrl":161},"skill.oms.sig","OTHER","其他",10309,"https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Ftao-train-ocrnet\u002Fskill.oms.sig",[163,170,177,184,191,198],{"skillNo":164,"name":165,"nameZh":6,"description":166,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":167,"categories":168,"keywords":169,"publishedAt":19},"01M79G206MT3J6S0GY3C45P2KV","vss-summarize-video","Use to summarize a recorded video via the LVS summarization microservice (HITL-gated) with a VLM fallback. Not for report generation or live RTSP captioning.","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Fvss-summarize-video\u002FSKILL.md",[],[],{"skillNo":171,"name":172,"nameZh":6,"description":173,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":174,"categories":175,"keywords":176,"publishedAt":19},"01M79G1WN63PGRE8CKD47WEBQ5","vss-setup-video-analytics-api","Use to deploy the vss-video-analytics-api REST service standalone (config-source, data-log bind, Elasticsearch, optional Kafka). Not for full warehouse deploy.","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Fvss-setup-video-analytics-api\u002FSKILL.md",[],[],{"skillNo":178,"name":179,"nameZh":6,"description":180,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":181,"categories":182,"keywords":183,"publishedAt":19},"01M79G1S9F7VY3EF6H53PRP5AN","vss-setup-behavior-analytics","Use to deploy the vss-behavior-analytics service standalone (entrypoint, config-source, optional calibration). Not for the full warehouse deploy.","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Fvss-setup-behavior-analytics\u002FSKILL.md",[],[],{"skillNo":185,"name":186,"nameZh":6,"description":187,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":188,"categories":189,"keywords":190,"publishedAt":19},"01M79G1N1S1Y27PFZPYMNK1MTQ","vss-search-archive","Use this skill to run top-level VSS fusion search on archived video, or to ingest video files \u002F RTSP streams for search. Do NOT use for ad-hoc visual Q&A (use vss-ask-video), live captioning (use vss-deploy-dense-captioning), or video summarization and reports (use vss-summarize-video).","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Fvss-search-archive\u002FSKILL.md",[],[],{"skillNo":192,"name":193,"nameZh":6,"description":194,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":195,"categories":196,"keywords":197,"publishedAt":19},"01M79G1H6WN0Q6C0VCAXCZX6B7","vss-query-analytics","Use this skill when reading video-analytics metrics, incidents, alerts, and sensor data via the VA-MCP server (port 9901). Not for live VLM or incident-range narrative reports.","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Fvss-query-analytics\u002FSKILL.md",[],[],{"skillNo":199,"name":200,"nameZh":6,"description":201,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":202,"categories":203,"keywords":204,"publishedAt":19},"01M79G1DYPNXQS7QHZ0AAF5FX5","vss-manage-video-io-storage","Use to call the VIOS REST API (sensor list, timelines, clip extraction, snapshots, add\u002Fdelete sensors and streams). Not for VLM inference or search.","https:\u002F\u002Fgithub.com\u002FElogic360\u002Fsoftware-factory\u002Fblob\u002Fmain\u002Fskills\u002Fnvidia\u002Fvss-manage-video-io-storage\u002FSKILL.md",[],[]]