跳到正文
cccskills
免费在 GitHub 公开

deepctr

Use this DeepCTR repo skill for CTR/recommender feature columns, Keras models, sequence/session models, multitask models, and legacy TensorFlow Estimator workflows.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill
查看安装方法

包含的文件（32）

  • SKILL.md2.5 KB
  • references/installation-and-compatibility.md2.2 KB
  • references/repo-provenance.md1.3 KB
  • references/repo-routing-metadata.json318 B
  • references/troubleshooting.md3.0 KB
  • scripts/check_deepctr_env.py3.6 KB
  • sub-skills/data-and-feature-columns/references/data-formats.md7.1 KB
  • sub-skills/data-and-feature-columns/references/feature-columns.md12.4 KB
  • sub-skills/data-and-feature-columns/references/troubleshooting.md5.4 KB
  • sub-skills/data-and-feature-columns/scripts/validate_feature_spec.py32.9 KB
  • sub-skills/data-and-feature-columns/SKILL.md4.4 KB
  • sub-skills/estimator-workflows/references/api-reference.md8.9 KB
  • sub-skills/estimator-workflows/references/tfrecord-and-pandas.md9.1 KB
  • sub-skills/estimator-workflows/references/troubleshooting.md9.3 KB
  • sub-skills/estimator-workflows/scripts/check_estimator_runtime.py14.9 KB
  • sub-skills/estimator-workflows/SKILL.md3.2 KB
  • sub-skills/keras-model-workflows/references/api-reference.md11.4 KB
  • sub-skills/keras-model-workflows/references/model-catalog.md5.8 KB
  • sub-skills/keras-model-workflows/references/troubleshooting.md9.5 KB
  • sub-skills/keras-model-workflows/references/workflows.md10.8 KB
  • sub-skills/keras-model-workflows/scripts/keras_tiny_ctr_smoke.py6.8 KB
  • sub-skills/keras-model-workflows/SKILL.md3.5 KB
  • sub-skills/multitask-models/references/api-reference.md4.4 KB
  • sub-skills/multitask-models/references/troubleshooting.md2.9 KB
  • sub-skills/multitask-models/references/workflows.md4.7 KB
  • sub-skills/multitask-models/scripts/multitask_tiny_smoke.py6.3 KB
  • sub-skills/multitask-models/SKILL.md1.3 KB
  • sub-skills/sequence-models/references/api-reference.md5.1 KB
  • sub-skills/sequence-models/references/sequence-inputs.md6.2 KB
  • sub-skills/sequence-models/references/troubleshooting.md6.9 KB
  • sub-skills/sequence-models/scripts/sequence_tiny_smoke.py8.4 KB
  • sub-skills/sequence-models/SKILL.md3.6 KB

SKILL.md（原文）

安装前可以查看交给 AI 代理的指令内容。

DeepCTR

Use this repo skill when the task mentions DeepCTR, click-through-rate prediction, recommender models, sparse/dense feature columns, DIN/BST/DIEN/DSIN, multitask heads, or the legacy TensorFlow Estimator surface.

Install and verify

DeepCTR does not install TensorFlow for you. Install a TensorFlow build that matches your Python and platform, then install DeepCTR:

python -m pip install "numpy<2" "tensorflow<2.21"
python -m pip install deepctr

If you are using a TensorFlow 2.20-style stack with legacy Keras requirements, install the matching tf-keras package and set TF_USE_LEGACY_KERAS=1 before running DeepCTR. See references/installation-and-compatibility.md for compatibility notes.

Quick environment check:

python scripts/check_deepctr_env.py --json

Route by need

  • data-and-feature-columns for SparseFeat, DenseFeat, VarLenSparseFeat, hashing, vocabulary paths, and input schema validation.
  • keras-model-workflows for ordinary CTR/regression models, compile/fit/predict/save/load, and model selection.
  • sequence-models for DIN, BST, DIEN, DSIN, hist_/sess_ naming, and sequence/session debugging.
  • multitask-models for SharedBottom, ESMM, MMOE, PLE, and multi-output target packing.
  • estimator-workflows for tf.estimator, TFRecord/Pandas input functions, and runtime gating.

Read first

Common starting point

If the user has not specified a model family yet, start with data-and-feature-columns when the problem is input shaping, otherwise keras-model-workflows for single-output model choice.

What this skill does not do

  • It does not tell you to open the original repository's examples or tests at runtime.
  • It does not require a GPU unless the user explicitly asks for GPU or multi-GPU execution.
  • It does not require Estimator workflows when tf.estimator is unavailable; in that case, use the Keras routes.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护，适用 Google 的隐私权政策和服务条款。

同一仓库的技能

3d-resnets-pytorch

免费

Routes 3D ResNets PyTorch video action-recognition workflows across training, inference, and data preparation.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill3382026年9月3日 更新

3ddfa

免费

Guide 3DDFA Python inference, geometry rendering, training/evaluation, and optional C++ ONNX workflows for 3D dense face alignment.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill3382026年9月3日 更新

3ddfa-v2

免费

Routes 3DDFA_V2 face-alignment setup, still-image demos, video tracking, and ONNX benchmarking workflows.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill3382026年9月3日 更新

ab3dmot

免费

Operate AB3DMOT 3D multi-object tracking workflows for KITTI and nuScenes data, tracking, evaluation, and visualization.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill3382026年9月3日 更新

accelerate

免费

Use Hugging Face Accelerate for PyTorch training-loop migration, distributed launch/configuration, DeepSpeed/FSDP/TPU backend setup, big-model inference/offload, checkpointing, tracking, and troubleshooting.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill3382026年9月3日 更新

acme

免费

Route Acme reinforcement-learning framework tasks across core loops, replay/data, JAX agents, and TensorFlow/Sonnet agents.

中文简介正在准备中。当前显示原文说明。

VectorSpaceLab/AREX-Skill3382026年9月3日 更新

查看 VectorSpaceLab 的全部技能

报告此技能的问题