torch-rechub 0.0.1__py3-none-any.whl → 0.0.4__py3-none-any.whl

torch_rechub-0.0.4.dist-info/METADATA

@@ -0,0 +1,391 @@
+Metadata-Version: 2.4
+Name: torch-rechub
+Version: 0.0.4
+Summary: A Pytorch Toolbox for Recommendation Models, Easy-to-use and Easy-to-extend.
+Project-URL: Homepage, https://github.com/datawhalechina/torch-rechub
+Project-URL: Documentation, https://www.torch-rechub.com
+Project-URL: Repository, https://github.com/datawhalechina/torch-rechub.git
+Project-URL: Issues, https://github.com/datawhalechina/torch-rechub/issues
+Author-email: rechub team <morningsky@tju.edu.cn>
+License: MIT
+License-File: LICENSE
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.9
+Requires-Dist: accelerate>=1.0.1
+Requires-Dist: numpy>=1.19.0
+Requires-Dist: pandas>=1.2.0
+Requires-Dist: scikit-learn>=0.24.0
+Requires-Dist: torch>=1.10.0
+Requires-Dist: tqdm>=4.60.0
+Requires-Dist: transformers>=4.46.3
+Provides-Extra: dev
+Requires-Dist: bandit>=1.7.0; extra == 'dev'
+Requires-Dist: flake8>=3.8.0; extra == 'dev'
+Requires-Dist: isort==5.13.2; extra == 'dev'
+Requires-Dist: mypy>=0.800; extra == 'dev'
+Requires-Dist: pre-commit>=2.20.0; extra == 'dev'
+Requires-Dist: pytest-cov>=2.0; extra == 'dev'
+Requires-Dist: pytest>=6.0; extra == 'dev'
+Requires-Dist: toml>=0.10.2; extra == 'dev'
+Requires-Dist: yapf==0.43.0; extra == 'dev'
+Provides-Extra: onnx
+Requires-Dist: onnx>=1.12.0; extra == 'onnx'
+Requires-Dist: onnxruntime>=1.12.0; extra == 'onnx'
+Description-Content-Type: text/markdown
+
+# 🔥 Torch-RecHub - 轻量、高效、易用的 PyTorch 推荐系统框架
+
+> 🚀 **30+ 主流推荐模型** | 🎯 **开箱即用** | 📦 **一键部署 ONNX** | 🤖 **支持生成式推荐 (HSTU/HLLM)**
+
+[![许可证](https://img.shields.io/badge/license-MIT-blue?style=for-the-badge)](LICENSE)
+![GitHub Repo stars](https://img.shields.io/github/stars/datawhalechina/torch-rechub?style=for-the-badge)
+![GitHub forks](https://img.shields.io/github/forks/datawhalechina/torch-rechub?style=for-the-badge)
+![GitHub issues](https://img.shields.io/github/issues/datawhalechina/torch-rechub?style=for-the-badge)
+[![Python 版本](https://img.shields.io/badge/python-3.9%2B-orange?style=for-the-badge)](https://www.python.org/)
+[![PyTorch 版本](https://img.shields.io/badge/pytorch-1.7%2B-orange?style=for-the-badge)](https://pytorch.org/)
+[![annoy 版本](https://img.shields.io/badge/annoy-1.17%2B-orange?style=for-the-badge)](https://github.com/spotify/annoy)
+[![pandas 版本](https://img.shields.io/badge/pandas-1.2%2B-orange?style=for-the-badge)](https://pandas.pydata.org/)
+[![numpy 版本](https://img.shields.io/badge/numpy-1.19%2B-orange?style=for-the-badge)](https://numpy.org/)
+[![scikit-learn 版本](https://img.shields.io/badge/scikit_learn-0.23%2B-orange?style=for-the-badge)](https://scikit-learn.org/)
+[![torch-rechub 版本](https://img.shields.io/badge/torch_rechub-0.0.3%2B-orange?style=for-the-badge)](https://pypi.org/project/torch-rechub/)
+
+[English](README_en.md) | 简体中文
+
+**在线文档：** https://datawhalechina.github.io/torch-rechub/ （英文）| https://datawhalechina.github.io/torch-rechub/zh/ （简体中文）
+
+**Torch-RecHub** —— **10 行代码实现工业级推荐系统**。30+ 主流模型开箱即用，支持一键 ONNX 部署，让你专注于业务而非工程。
+
+![Torch-RecHub 横幅](docs/public/img/banner.png)
+
+## 🎯 为什么选择 Torch-RecHub？
+
+| 特性 | Torch-RecHub | 其他框架 |
+|------|-------------|---------|
+| 代码行数 | **10行** 完成训练+评估+部署 | 100+ 行 |
+| 模型覆盖 | **30+** 主流模型 | 有限 |
+| 生成式推荐 | ✅ HSTU/HLLM (Meta 2024) | ❌ |
+| ONNX 一键导出 | ✅ 内置支持 | 需手动适配 |
+| 学习曲线 | 极低 | 陡峭 |
+
+## ✨ 特性
+
+* **模块化设计:** 易于添加新的模型、数据集和评估指标。
+* **基于 PyTorch:** 利用 PyTorch 的动态图和 GPU 加速能力。
+* **丰富的模型库:** 涵盖 **30+** 经典和前沿推荐算法（召回、排序、多任务、生成式推荐等）。
+* **标准化流程:** 提供统一的数据加载、训练和评估流程。
+* **易于配置:** 通过配置文件或命令行参数轻松调整实验设置。
+* **可复现性:** 旨在确保实验结果的可复现性。
+* **ONNX 导出:** 支持将训练好的模型导出为 ONNX 格式，便于部署到生产环境。
+* **其他特性:** 例如，支持负采样、多任务学习等。
+
+## 📖 目录
+
+- [🔥 Torch-RecHub - 轻量、高效、易用的 PyTorch 推荐系统框架](#-torch-rechub---轻量高效易用的-pytorch-推荐系统框架)
+  - [🎯 为什么选择 Torch-RecHub？](#-为什么选择-torch-rechub)
+  - [✨ 特性](#-特性)
+  - [📖 目录](#-目录)
+  - [🔧 安装](#-安装)
+    - [环境要求](#环境要求)
+    - [安装步骤](#安装步骤)
+  - [🚀 快速开始](#-快速开始)
+  - [📂 项目结构](#-项目结构)
+  - [💡 支持的模型](#-支持的模型)
+  - [📊 支持的数据集](#-支持的数据集)
+  - [🧪 示例](#-示例)
+    - [精排（CTR预测）](#精排ctr预测)
+    - [多任务排序](#多任务排序)
+    - [召回模型](#召回模型)
+  - [👨‍💻‍ 贡献者](#-贡献者)
+  - [🤝 贡献指南](#-贡献指南)
+  - [📜 许可证](#-许可证)
+  - [📚 引用](#-引用)
+  - [📫 联系方式](#-联系方式)
+  - [⭐️ 项目 star 历史](#️-项目-star-历史)
+
+## 🔧 安装
+
+### 环境要求
+
+* Python 3.9+
+* PyTorch 1.7+ (建议使用支持 CUDA 的版本以获得 GPU 加速)
+* NumPy
+* Pandas
+* SciPy
+* Scikit-learn
+
+### 安装步骤
+
+**稳定版（推荐用户使用）：**
+```bash
+pip install torch-rechub
+```
+
+**最新版：**
+```bash
+# 首先安装 uv（如果尚未安装）
+pip install uv
+
+# 克隆并安装
+git clone https://github.com/datawhalechina/torch-rechub.git
+cd torch-rechub
+uv sync
+```
+
+
+
+## 🚀 快速开始
+
+以下是一个简单的示例，展示如何在 MovieLens 数据集上训练模型（例如 DSSM）：
+
+```bash
+# 克隆仓库（如果使用最新版）
+git clone https://github.com/datawhalechina/torch-rechub.git
+cd torch-rechub
+uv sync
+
+# 运行示例
+python examples/matching/run_ml_dssm.py
+
+# 或使用自定义参数：
+python examples/matching/run_ml_dssm.py --model_name dssm --device 'cuda:0' --learning_rate 0.001 --epoch 50 --batch_size 4096 --weight_decay 0.0001 --save_dir 'saved/dssm_ml-100k'
+```
+
+训练完成后，模型文件将保存在 `saved/dssm_ml-100k` 目录下（或你配置的其他目录）。
+
+## 📂 项目结构
+
+```
+torch-rechub/             # 根目录
+├── README.md             # 项目文档
+├── pyproject.toml        # 项目配置和依赖
+├── torch_rechub/         # 核心代码库
+│   ├── basic/            # 基础组件
+│   │   ├── activation.py # 激活函数
+│   │   ├── features.py   # 特征工程
+│   │   ├── layers.py     # 神经网络层
+│   │   ├── loss_func.py  # 损失函数
+│   │   └── metric.py     # 评估指标
+│   ├── models/           # 推荐模型实现
+│   │   ├── matching/     # 召回模型（DSSM/MIND/GRU4Rec等）
+│   │   ├── ranking/      # 排序模型（WideDeep/DeepFM/DIN等）
+│   │   └── multi_task/   # 多任务模型（MMoE/ESMM等）
+│   ├── trainers/         # 训练框架
+│   │   ├── ctr_trainer.py    # CTR预测训练器
+│   │   ├── match_trainer.py  # 召回模型训练器
+│   │   └── mtl_trainer.py    # 多任务学习训练器
+│   └── utils/            # 工具函数
+│       ├── data.py       # 数据处理工具
+│       ├── match.py      # 召回工具
+│       ├── mtl.py        # 多任务工具
+│       └── onnx_export.py # ONNX 导出工具
+├── examples/             # 示例脚本
+│   ├── matching/         # 召回任务示例
+│   ├── ranking/          # 排序任务示例
+│   └── generative/       # 生成式推荐示例（HSTU、HLLM 等）
+├── docs/                 # 文档（VitePress，多语言）
+├── tutorials/            # Jupyter教程
+├── tests/                # 单元测试
+├── config/               # 配置文件
+└── scripts/              # 工具脚本
+```
+
+## 💡 支持的模型
+
+本框架目前支持 **30+** 主流推荐模型：
+
+### 排序模型 (Ranking Models) - 13个
+
+| 模型 | 论文 | 简介 |
+|------|------|------|
+| **DeepFM** | [IJCAI 2017](https://arxiv.org/abs/1703.04247) | FM + Deep 联合训练 |
+| **Wide&Deep** | [DLRS 2016](https://arxiv.org/abs/1606.07792) | 记忆 + 泛化能力结合 |
+| **DCN** | [KDD 2017](https://arxiv.org/abs/1708.05123) | 显式特征交叉网络 |
+| **DCN-v2** | [WWW 2021](https://arxiv.org/abs/2008.13535) | 增强版交叉网络 |
+| **DIN** | [KDD 2018](https://arxiv.org/abs/1706.06978) | 注意力机制捕捉用户兴趣 |
+| **DIEN** | [AAAI 2019](https://arxiv.org/abs/1809.03672) | 兴趣演化建模 |
+| **BST** | [DLP-KDD 2019](https://arxiv.org/abs/1905.06874) | Transformer 序列建模 |
+| **AFM** | [IJCAI 2017](https://arxiv.org/abs/1708.04617) | 注意力因子分解机 |
+| **AutoInt** | [CIKM 2019](https://arxiv.org/abs/1810.11921) | 自动特征交互学习 |
+| **FiBiNET** | [RecSys 2019](https://arxiv.org/abs/1905.09433) | 特征重要性 + 双线性交互 |
+| **DeepFFM** | [RecSys 2019](https://arxiv.org/abs/1611.00144) | 场感知因子分解机 |
+| **EDCN** | [KDD 2021](https://arxiv.org/abs/2106.03032) | 增强型交叉网络 |
+
+### 召回模型 (Matching Models) - 12个
+
+| 模型 | 论文 | 简介 |
+|------|------|------|
+| **DSSM** | [CIKM 2013](https://posenhuang.github.io/papers/cikm2013_DSSM_fullversion.pdf) | 经典双塔召回模型 |
+| **YoutubeDNN** | [RecSys 2016](https://dl.acm.org/doi/10.1145/2959100.2959190) | YouTube 深度召回 |
+| **YoutubeSBC** | [RecSys 2019](https://dl.acm.org/doi/10.1145/3298689.3346997) | 采样偏差校正版本 |
+| **MIND** | [CIKM 2019](https://arxiv.org/abs/1904.08030) | 多兴趣动态路由 |
+| **SINE** | [WSDM 2021](https://arxiv.org/abs/2103.06920) | 稀疏兴趣网络 |
+| **GRU4Rec** | [ICLR 2016](https://arxiv.org/abs/1511.06939) | GRU 序列推荐 |
+| **SASRec** | [ICDM 2018](https://arxiv.org/abs/1808.09781) | 自注意力序列推荐 |
+| **NARM** | [CIKM 2017](https://arxiv.org/abs/1711.04725) | 神经注意力会话推荐 |
+| **STAMP** | [KDD 2018](https://dl.acm.org/doi/10.1145/3219819.3219895) | 短期注意力记忆优先 |
+| **ComiRec** | [KDD 2020](https://arxiv.org/abs/2005.09347) | 可控多兴趣推荐 |
+
+### 多任务模型 (Multi-Task Models) - 5个
+
+| 模型 | 论文 | 简介 |
+|------|------|------|
+| **ESMM** | [SIGIR 2018](https://arxiv.org/abs/1804.07931) | 全空间多任务建模 |
+| **MMoE** | [KDD 2018](https://dl.acm.org/doi/10.1145/3219819.3220007) | 多门控专家混合 |
+| **PLE** | [RecSys 2020](https://dl.acm.org/doi/10.1145/3383313.3412236) | 渐进式分层提取 |
+| **AITM** | [KDD 2021](https://arxiv.org/abs/2105.08489) | 自适应信息迁移 |
+| **SharedBottom** | - | 经典多任务共享底层 |
+
+### 生成式推荐 (Generative Recommendation) - 2个
+
+| 模型 | 论文 | 简介 |
+|------|------|------|
+| **HSTU** | [Meta 2024](https://arxiv.org/abs/2402.17152) | 层级序列转换单元，支撑 Meta 万亿参数推荐系统 |
+| **HLLM** | [2024](https://arxiv.org/abs/2409.12740) | 层级大语言模型推荐，融合 LLM 语义理解能力 |
+
+## 📊 支持的数据集
+
+框架内置了对以下常见数据集格式的支持或提供了处理脚本：
+
+* **MovieLens**
+* **Amazon**
+* **Criteo**
+* **Avazu** 
+* **Census-Income**
+* **BookCrossing**
+* **Ali-ccp**
+* **Yidian**
+* ... 
+
+我们期望的数据格式通常是包含以下字段的交互文件：
+- 用户 ID
+- 物品 ID 
+- 评分（可选）
+- 时间戳（可选）
+
+具体格式要求请参考 `tutorials` 目录下的示例代码。
+
+你可以方便地集成你自己的数据集，只需确保它符合框架要求的数据格式，或编写自定义的数据加载器。
+
+
+## 🧪 示例
+
+所有模型使用案例参考 `/examples`
+
+
+### 精排（CTR预测）
+
+```python
+from torch_rechub.models.ranking import DeepFM
+from torch_rechub.trainers import CTRTrainer
+from torch_rechub.utils.data import DataGenerator
+
+dg = DataGenerator(x, y)
+train_dataloader, val_dataloader, test_dataloader = dg.generate_dataloader(split_ratio=[0.7, 0.1], batch_size=256)
+
+model = DeepFM(deep_features=deep_features, fm_features=fm_features, mlp_params={"dims": [256, 128], "dropout": 0.2, "activation": "relu"})
+
+ctr_trainer = CTRTrainer(model)
+ctr_trainer.fit(train_dataloader, val_dataloader)
+auc = ctr_trainer.evaluate(ctr_trainer.model, test_dataloader)
+ctr_trainer.export_onnx("deepfm.onnx")
+```
+
+### 多任务排序
+
+```python
+from torch_rechub.models.multi_task import SharedBottom, ESMM, MMOE, PLE, AITM
+from torch_rechub.trainers import MTLTrainer
+
+task_types = ["classification", "classification"] 
+model = MMOE(features, task_types, 8, expert_params={"dims": [32,16]}, tower_params_list=[{"dims": [32, 16]}, {"dims": [32, 16]}])
+
+mtl_trainer = MTLTrainer(model)
+mtl_trainer.fit(train_dataloader, val_dataloader)
+auc = ctr_trainer.evaluate(ctr_trainer.model, test_dataloader)
+mtl_trainer.export_onnx("mmoe.onnx")
+```
+
+### 召回模型
+
+```python
+from torch_rechub.models.matching import DSSM
+from torch_rechub.trainers import MatchTrainer
+from torch_rechub.utils.data import MatchDataGenerator
+
+dg = MatchDataGenerator(x, y)
+train_dl, test_dl, item_dl = dg.generate_dataloader(test_user, all_item, batch_size=256)
+
+model = DSSM(user_features, item_features, temperature=0.02,
+             user_params={
+                 "dims": [256, 128, 64],
+                 "activation": 'prelu',  
+             },
+             item_params={
+                 "dims": [256, 128, 64],
+                 "activation": 'prelu', 
+             })
+
+match_trainer = MatchTrainer(model)
+match_trainer.fit(train_dl)
+match_trainer.export_onnx("dssm.onnx")
+# 双塔模型可分别导出用户塔和物品塔: 
+# match_trainer.export_onnx("user_tower.onnx", mode="user")
+# match_trainer.export_onnx("dssm_item.onnx", tower="item")
+```
+
+## 👨‍💻‍ 贡献者
+
+感谢所有的贡献者！
+
+![GitHub contributors](https://img.shields.io/github/contributors/datawhalechina/torch-rechub?color=32A9C3&labelColor=1B3C4A&logo=contributorcovenant)
+
+[![contributors](https://contrib.rocks/image?repo=datawhalechina/torch-rechub)](https://github.com/datawhalechina/torch-rechub/graphs/contributors)
+
+## 🤝 贡献指南
+
+我们欢迎各种形式的贡献！请查看 [CONTRIBUTING.md](CONTRIBUTING.md) 了解详细的贡献指南。
+
+我们也欢迎通过 [Issues](https://github.com/datawhalechina/torch-rechub/issues) 报告 Bug 或提出功能建议。
+
+## 📜 许可证
+
+本项目采用 [MIT 许可证](LICENSE)。
+
+## 📚 引用
+
+如果你在研究或工作中使用了本框架，请考虑引用：
+
+```bibtex
+@misc{torch_rechub,
+    title = {Torch-RecHub},
+    author = {Datawhale},
+    year = {2022},
+    publisher = {GitHub},
+    journal = {GitHub repository},
+    howpublished = {\url{https://github.com/datawhalechina/torch-rechub}},
+    note = {A PyTorch-based recommender system framework providing easy-to-use and extensible solutions}
+}
+```
+
+## 📫 联系方式
+
+* **项目负责人:** [1985312383](https://github.com/1985312383) 
+* [**GitHub Disscussions**](https://github.com/datawhalechina/torch-rechub/discussions)
+
+## ⭐️ 项目 star 历史
+
+[![Star History Chart](https://api.star-history.com/svg?repos=datawhalechina/torch-rechub&type=Date)](https://www.star-history.com/#datawhalechina/torch-rechub&Date)
+
+---
+
+*最后更新: [2025-12-04]*

torch_rechub-0.0.4.dist-info/RECORD

@@ -0,0 +1,62 @@
+torch_rechub/__init__.py,sha256=XUwV85oz-uIokuE9qj3nmbUQg3EY8dZcDMohlob3suw,245
+torch_rechub/basic/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+torch_rechub/basic/activation.py,sha256=hIZDCe7cAgV3bX2UnvUrkO8pQs4iXxkQGD0J4GejbVg,1600
+torch_rechub/basic/callback.py,sha256=ZeiDSDQAZUKmyK1AyGJCnqEJ66vwfwlX5lOyu6-h2G0,946
+torch_rechub/basic/features.py,sha256=TLHR5EaNvIbKyKd730Qt8OlLpV0Km91nv2TMnq0HObk,3562
+torch_rechub/basic/initializers.py,sha256=V6hprXvRexcw3vrYsf8Qp-F52fp8uzPMpa1CvkHofy8,3196
+torch_rechub/basic/layers.py,sha256=URWk78dlffMOAhDVDhOhugcr4nmwEa192AI1diktC-4,39653
+torch_rechub/basic/loss_func.py,sha256=6bjljqpiuUP6O8-wUbGd8FSvflY5Dp_DV_57OuQVMz4,7969
+torch_rechub/basic/metaoptimizer.py,sha256=y-oT4MV3vXnSQ5Zd_ZEHP1KClITEi3kbZa6RKjlkYw8,3093
+torch_rechub/basic/metric.py,sha256=9JsaJJGvT6VRvsLoM2Y171CZxESsjYTofD3qnMI-bPM,8443
+torch_rechub/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+torch_rechub/models/generative/__init__.py,sha256=TsCdVIhOcalQwqKZKjEuNbHKyIjyclapKGNwYfFR7TM,135
+torch_rechub/models/generative/hllm.py,sha256=6Vrp5Bh0fTFHCn7C-3EqzOyc7UunOyEY9TzAKGHrW-8,9669
+torch_rechub/models/generative/hstu.py,sha256=M1ByAWHxrkvmwaXPNdGFrbAQYyYJswnABTY3jCEceyg,7846
+torch_rechub/models/matching/__init__.py,sha256=fjWOzJB8loPGy8rJMG-6G-NUISp7k3sD_1FdsKGw1as,471
+torch_rechub/models/matching/comirec.py,sha256=8KB5rg2FWlZaG73CBI7_J8-J-XpjTPnblwh5OsbtAbc,9439
+torch_rechub/models/matching/dssm.py,sha256=1Q0JYpt1h_7NWlLN5a_RbCoUSubZwpYTVEXccSn44eg,3003
+torch_rechub/models/matching/dssm_facebook.py,sha256=n3MS7FT_kyJSDnVTlPCv_nPJ0MHCtMgJRUPDRh7jBLM,3508
+torch_rechub/models/matching/dssm_senet.py,sha256=_E-xEh44XvOaBHP8XdSRkFsTvajhovxlYyCt3H9P61c,4052
+torch_rechub/models/matching/gru4rec.py,sha256=cJtYCkFyg3cPYkOy_YeXRAsTev0cBPiicrj68xJup9k,3932
+torch_rechub/models/matching/mind.py,sha256=NIUeqWhrnZeiFDMNFvXfMx1GMBMaCZnc6nxNZCJpwSE,4933
+torch_rechub/models/matching/narm.py,sha256=2dlTuan9AFrku53WJlBbTwgLlfOHsas3-JBFGxEz7oE,3167
+torch_rechub/models/matching/sasrec.py,sha256=QDfKrFl-aduWg6rY3R13RrdpMiApVugDmtEsWJulgzg,5534
+torch_rechub/models/matching/sine.py,sha256=sUTUHbnewdSBd51epDIp9j-B1guKkhm6eM-KkZ3oS3Q,6746
+torch_rechub/models/matching/stamp.py,sha256=DBVM3iCoQTBKwO7oKHg5SCCDXqTuRJ4Ko1n7StgEovA,3308
+torch_rechub/models/matching/youtube_dnn.py,sha256=EQV_GoEs2Hxwg1U3Dj7-lWkEejEqGmtZ7D9CgfknQdA,3368
+torch_rechub/models/matching/youtube_sbc.py,sha256=paw9uRnbNw_-EaFpRogy7rB4vhw4KN0Qf8BfQylTj4I,4757
+torch_rechub/models/multi_task/__init__.py,sha256=5N8aJ32fzxniDm4d-AeNSi81CFWyBhjoSaK3OC-XCkY,189
+torch_rechub/models/multi_task/aitm.py,sha256=hlG4opauSmM4SNZBPqogPZKNCPheGCVx--JSEWXeIJ8,3355
+torch_rechub/models/multi_task/esmm.py,sha256=y6Gv_mWRw7lcZm-wjw2OHJVwvDHFfvkVSnJvQGO6kUk,2742
+torch_rechub/models/multi_task/mmoe.py,sha256=yprs9P5vL6C4mf3lHb0uyjyJisYVny5UZ6Q9MGVfn-0,2831
+torch_rechub/models/multi_task/ple.py,sha256=5QieTfL-jhHMzuSOlmgXc0dGARJzpLT2sHHdMedjtpM,6448
+torch_rechub/models/multi_task/shared_bottom.py,sha256=XH32YOGsb8M97ggcm6RPY2k3T5g9GgeZsJ2NjwpZoko,1866
+torch_rechub/models/ranking/__init__.py,sha256=VWhbC4918N1hDiUIj7shHp-N7oWOYYpLc4siBtMUV0w,447
+torch_rechub/models/ranking/afm.py,sha256=VbTa3HSyYmpTMdSOr54ex6HOn_uA9tYtmlwX7IKDhUs,2267
+torch_rechub/models/ranking/autoint.py,sha256=VQ8BrPYaSXZzDcSvQ0spcxl5rSqNL-gqemH7UKiLUDk,3880
+torch_rechub/models/ranking/bst.py,sha256=b9cZUnFV51Fw4UAOuPJpog9pHMb2cuGr-yNAvqK7Xpg,3632
+torch_rechub/models/ranking/dcn.py,sha256=ErWAF9MlKsZj_aF1DQQa1cHkobMyo7J5CTreUptRzhw,1302
+torch_rechub/models/ranking/dcn_v2.py,sha256=09mx4LP8D1CVCpFZpJmjCWUm6OzJ9lXgyAvSyaaxzQo,2832
+torch_rechub/models/ranking/deepffm.py,sha256=XYYnpwsJ028fyNcH1MWGPWQVrlonO5aiZO7bpC25PnE,6044
+torch_rechub/models/ranking/deepfm.py,sha256=5yKLrdLPuD3NigGL3bKnG5HS3kqCuRw7gAIkXl6qY9Q,1783
+torch_rechub/models/ranking/dien.py,sha256=2jaPluJf0K_ctHM3MTJ042bXxGLI8KphOTkKVScKUAg,8817
+torch_rechub/models/ranking/din.py,sha256=HsOCEErea3KwEiyWw4M_aX_LMC_-Sqs1C_zeRLKLV_c,4542
+torch_rechub/models/ranking/edcn.py,sha256=6f_S8I6Ir16kCIU54R4EfumWfUFOND5KDKUPHMgsVU0,4997
+torch_rechub/models/ranking/fibinet.py,sha256=fmEJ9WkO8Mn0RtK_8aRHlnQFh_jMBPO0zODoHZPWmDA,2234
+torch_rechub/models/ranking/widedeep.py,sha256=eciRvWRBHLlctabLLS5NB7k3MnqrWXCBdpflOU6jMB0,1636
+torch_rechub/trainers/__init__.py,sha256=NSa2DqgfE1HGDyj40YgrbtUrfBHBxNBpw57XtaAB_jE,148
+torch_rechub/trainers/ctr_trainer.py,sha256=RDUXkn7GwLzs3f0kWZwGDNCpqiMeGXo7R6ezFeZdPg8,9075
+torch_rechub/trainers/match_trainer.py,sha256=xox5eaPKjSgErJQpbSr29sbyGs1p2sFaKEjxACE6uMI,11276
+torch_rechub/trainers/matching.md,sha256=vIBQ3UMmVpUpyk38rrkelFwm_wXVXqMOuqzYZ4M8bzw,30
+torch_rechub/trainers/mtl_trainer.py,sha256=tC4c2KIc-H8Wvj4qCzcW6TyfMLRPJyfQvTaN0dDePFg,12598
+torch_rechub/trainers/seq_trainer.py,sha256=lXKRx7XbZ3iJuqp_f05vw_jkn8X5j8HmH6Nr-typiIU,12043
+torch_rechub/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+torch_rechub/utils/data.py,sha256=vzLAAVt6dujg_vbGhQewiJc0l6JzwzdcM_9EjoOz898,19882
+torch_rechub/utils/hstu_utils.py,sha256=qLON_pJDC-kDyQn1PoN_HaHi5xTNCwZPgJeV51Z61Lc,6207
+torch_rechub/utils/match.py,sha256=l9qDwJGHPP9gOQTMYoqGVdWrlhDx1F1-8UnQwDWrEyk,18143
+torch_rechub/utils/mtl.py,sha256=AxU05ezizCuLdbPuCg1ZXE0WAStzuxaS5Sc3nwMCBpI,5737
+torch_rechub/utils/onnx_export.py,sha256=uRcAD4uZ3eIQbM-DPhdc0bkaPaslNsOYny6BOeLVBfU,13660
+torch_rechub-0.0.4.dist-info/METADATA,sha256=SNm71v_YOfculnc13p266bD_8yLo0U_16F_aJQPDvYo,16149
+torch_rechub-0.0.4.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+torch_rechub-0.0.4.dist-info/licenses/LICENSE,sha256=V7ietiX9G_84HtgEbxDgxClniqXGm2t5q8WM4AHGTu0,1066
+torch_rechub-0.0.4.dist-info/RECORD,,

{torch_rechub-0.0.1.dist-info → torch_rechub-0.0.4.dist-info}/WHEEL

@@ -1,5 +1,4 @@
 Wheel-Version: 1.0
-Generator: bdist_wheel (0.37.1)
+Generator: hatchling 1.27.0
 Root-Is-Purelib: true
 Tag: py3-none-any
-

{torch_rechub-0.0.1.dist-info → torch_rechub-0.0.4.dist-info/licenses}/LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2022 Mincai Lai
+Copyright (c) 2022 Datawhale
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

torch_rechub/basic/utils.py

@@ -1,168 +0,0 @@
-import random
-import torch
-import numpy as np
-import pandas as pd
-from torch.utils.data import Dataset, DataLoader, random_split
-from sklearn.metrics import roc_auc_score, mean_squared_error
-
-
-class TorchDataset(Dataset):
-
-    def __init__(self, x, y):
-        super(TorchDataset, self).__init__()
-        self.x = x
-        self.y = y
-
-    def __getitem__(self, index):
-        return {k: v[index] for k, v in self.x.items()}, self.y[index]
-
-    def __len__(self):
-        return len(self.y)
-
-
-class DataGenerator(object):
-
-    def __init__(self, x, y):
-        super(DataGenerator, self).__init__()
-        self.dataset = TorchDataset(x, y)
-        self.length = len(self.dataset)
-
-    def generate_dataloader(self, x_val=None, y_val=None, x_test=None, y_test=None, split_ratio=None, batch_size=16, num_workers=8):
-        if split_ratio != None:
-            train_length = int(self.length * split_ratio[0])
-            val_length = int(self.length * split_ratio[1])
-            test_length = self.length - train_length - val_length
-            print("the samples of train : val : test are  %d : %d : %d" % (train_length, val_length, test_length))
-            train_dataset, val_dataset, test_dataset = random_split(self.dataset, (train_length, val_length, test_length))
-        else:
-            train_dataset = self.dataset
-            val_dataset = TorchDataset(x_val, y_val)
-            test_dataset = TorchDataset(x_test, y_test)
-
-        train_dataloader = DataLoader(train_dataset, batch_size=batch_size, num_workers=num_workers)
-        val_dataloader = DataLoader(val_dataset, batch_size=batch_size, num_workers=num_workers)
-        test_dataloader = DataLoader(test_dataset, batch_size=batch_size, num_workers=num_workers)
-        return train_dataloader, val_dataloader, test_dataloader
-
-
-class PredictDataset(Dataset):
-
-    def __init__(self, x):
-        super(TorchDataset, self).__init__()
-        self.x = x
-
-    def __getitem__(self, index):
-        return {k: v[index] for k, v in self.x.items()}
-
-    def __len__(self):
-        return len(self.x[list(self.x.keys())[0]])
-
-
-def get_auto_embedding_dim(num_classes):
-    """ Calculate the dim of embedding vector according to number of classes in the category
-    emb_dim = [6 * (num_classes)^(1/4)]
-    reference: Deep & Cross Network for Ad Click Predictions.(ADKDD'17)
-    
-    Args:
-        num_classes: number of classes in the category
-    
-    Returns:
-        the dim of embedding vector
-    """
-    return np.floor(6 * np.pow(num_classes, 0.26))
-
-
-def get_loss_func(task_type="classification"):
-    if task_type == "classification":
-        return torch.nn.BCELoss()
-    elif task_type == "regression":
-        return torch.nn.MSELoss()
-    else:
-        raise ValueError("task_type must be classification or regression")
-
-
-def get_metric_func(task_type="classification"):
-    if task_type == "classification":
-        return roc_auc_score
-    elif task_type == "regression":
-        return mean_squared_error
-    else:
-        raise ValueError("task_type must be classification or regression")
-
-
-def create_seq_features(data, max_len=50, drop_short=3, shuffle=True):
-    """Build a sequence of user's history by time.
-    
-    Args:
-        data (pd.DataFrame): must contain keys: `user_id, item_id, cate_id, time`.
-        max_len (int): the max length of a user history sequence.
-        drop_short (int): remove some inactive user who's sequence length < drop_short.
-        shuffle (bool): shuffle data if true.
-    
-    Returns: 
-        train (pd.DataFrame): target item will be each item before last two items.
-        val (pd.DataFrame): target item is the second to last item of user's history sequence.
-        test (pd.DataFrame): target item is the last item of user's history sequence.
-    """
-    n_users, n_items, n_cates = data["user_id"].max(), data["item_id"].max(), data["cate_id"].max()
-    # 0 to be used as the symbol for padding
-    data = data.astype('int32')
-    data['item_id'] = data['item_id'].apply(lambda x: x + 1)
-    data['cate_id'] = data['cate_id'].apply(lambda x: x + 1)
-
-    item_cate_map = data[['item_id', 'cate_id']]
-    item2cate_dict = item_cate_map.set_index(['item_id'])['cate_id'].to_dict()
-
-    data = data.sort_values(['user_id', 'time']).groupby('user_id').agg(click_hist_list=('item_id', list), cate_hist_hist=('cate_id', list)).reset_index()
-
-    # Sliding window to construct negative samples
-    train_data, val_data, test_data = [], [], []
-    for item in data.itertuples():
-        if len(item[2]) < drop_short:
-            continue
-        click_hist_list = item[2][:max_len]
-        cate_hist_list = item[3][:max_len]
-
-        def neg_sample():
-            neg = click_hist_list[0]
-            while neg in click_hist_list:
-                neg = random.randint(1, n_items)
-            return neg
-
-        neg_list = [neg_sample() for _ in range(len(click_hist_list))]
-        hist_list = []
-        cate_list = []
-        for i in range(1, len(click_hist_list)):
-            hist_list.append(click_hist_list[i - 1])
-            cate_list.append(cate_hist_list[i - 1])
-            hist_list_pad = hist_list + [0] * (max_len - len(hist_list))
-            cate_list_pad = cate_list + [0] * (max_len - len(cate_list))
-            if i == len(click_hist_list) - 1:
-                test_data.append([hist_list_pad, cate_list_pad, click_hist_list[i], cate_hist_list[i], 1])
-                test_data.append([hist_list_pad, cate_list_pad, neg_list[i], item2cate_dict[neg_list[i]], 0])
-            if i == len(click_hist_list) - 2:
-                val_data.append([hist_list_pad, cate_list_pad, click_hist_list[i], cate_hist_list[i], 1])
-                val_data.append([hist_list_pad, cate_list_pad, neg_list[i], item2cate_dict[neg_list[i]], 0])
-            else:
-                train_data.append([hist_list_pad, cate_list_pad, click_hist_list[i], cate_hist_list[i], 1])
-                train_data.append([hist_list_pad, cate_list_pad, neg_list[i], item2cate_dict[neg_list[i]], 0])
-
-    # shuffle
-    if shuffle:
-        random.shuffle(train_data)
-        random.shuffle(val_data)
-        random.shuffle(test_data)
-
-    col_name = ['history_item', 'history_cate', 'target_item', 'target_cate', 'label']
-    train = pd.DataFrame(train_data, columns=col_name)
-    val = pd.DataFrame(val_data, columns=col_name)
-    test = pd.DataFrame(test_data, columns=col_name)
-
-    return train, val, test
-
-
-def df_to_input_dict(data):
-    data_dict = data.to_dict('list')
-    for key in data.keys():
-        data_dict[key] = np.array(data_dict[key])
-    return data_dict