executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
Explorer les Skills
Productivité · SKILL.md
multitask-models
multitask-modelsDeepCTR multitask workflows for SharedBottom, ESMM, MMOE, and PLE.
Independent GitHub snapshot · Automated classification · Original documentation governs
Documentation de la Skill
GitHubname: multitask-models description: "DeepCTR multitask workflows for SharedBottom, ESMM, MMOE, and PLE." disable-model-invocation: true metadata: disco-role: operating license: Apache 2.0
Multitask Models
Use this sub-skill for DeepCTR multi-output learning tasks that need shared feature inputs and task-specific heads.
Covers
- SharedBottom, ESMM, MMOE, and PLE
- task name / task type alignment
- binary vs regression validation where supported
- shared feature reuse from one
dnn_feature_columnslist - compile / fit / evaluate / predict for multi-output models
- tiny synthetic smoke checks via the bundled script
Route elsewhere
- Single-output CTR or regression model selection:
keras-model-workflows - Feature column construction details:
data-and-feature-columns - Estimator workflows:
estimator-workflows
Read first
- API reference: constructor signatures, task validation, and output name order
- Workflows: two-label preprocessing, compile/fit/evaluate/predict patterns, and model selection
- Troubleshooting: task-list, string sparse, loss-order, and output-shape failures
Smoke check
- Tiny multitask smoke: synthetic checks for SharedBottom, ESMM, MMOE, and PLE