DeepSklearn

All You Need Is DeepSklearn.

An industrial-grade PyTorch platform for modern deep learning algorithms

What is DeepSklearn?

DeepSklearn is an industrial-grade PyTorch platform that makes modern deep learning algorithms easy to build, train, evaluate, and deploy in a clean, sklearn-style workflow.

It combines a modular deep learning framework with an interactive Streamlit training dashboard, allowing users to configure datasets, select models, tune hyperparameters, start training, and monitor logs directly from a UI. With DeepSklearn, even beginners can launch deep learning training jobs without writing complex boilerplate training code.

The project is designed to support a growing algorithm zoo, including recommendation models, recall models, multi-task learning models, sequence models, and general neural network architectures. DeepSklearn emphasizes practical engineering, clean abstractions, reusable components, and production-inspired training pipelines.

training dashboard

deepsklearn demo

Models

Here is the collection of deep learning and recommendation algorithms supported by this project, categorized into five core domains: generative, multi-task learning, non-sequence, recall, and sequence models.

1. Non-Sequence Models

Model	Full Name	Paper	Status
LR	Logistic Regression	[Cox, 1958] The Regression Analysis of Binary Sequences	✅
FM	Factorization Machines	[Rendle, 2010] Factorization Machines	✅
DNN	Deep Neural Network	-	✅
Wide & Deep	Wide & Deep Learning	[Cheng et al., 2016] Wide & Deep Learning for Recommender Systems	✅
DeepFM	Deep Factorization Machine	[Guo et al., 2017] DeepFM: A Factorization-Machine based Neural Network for CTR Prediction	✅
DCN	Deep & Cross Network	[Wang et al., 2017] Deep & Cross Network for Ad Click Predictions	✅
DCN-V2	Deep & Cross Network V2	[Wang et al., 2021] DCN V2: Improved Deep & Cross Network and Practical Lessons for Web-scale Learning to Rank Systems	✅
PNN	Product-based Neural Network	[Qu et al., 2016] Product-based Neural Networks for User Response Prediction	✅
AutoInt	Automatic Feature Interaction Learning via Self-Attentive Neural Networks	[Song et al., 2019] AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks	✅
AFM	Attentional Factorization Machine	[Xiao et al., 2017] Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks	✅
NFM	Neural Factorization Machine	[He & Chua, 2017] Neural Factorization Machines for Sparse Predictive Analytics	✅
DLRM	Deep Learning Recommendation Model	[Naumov et al., 2019] Deep Learning Recommendation Model for Personalization and Recommendation Systems	✅

2. Sequence (User Behavior Sequence Models)

Model	Full Name	Paper	Status
Average Pooling	Average Pooling for Sequence Modeling	Common sequence aggregation baseline; no single canonical recommendation paper	✅
BST	Behavior Sequence Transformer	[Chen et al., 2019] Behavior Sequence Transformer for E-commerce Recommendation in Alibaba	✅
DIN	Deep Interest Network	[Zhou et al., 2018] Deep Interest Network for Click-Through Rate Prediction	✅

3. Multi-Task Learning

Model	Full Name	Paper	Status
Shared Bottom	Shared-Bottom Multi-Task Learning	[Caruana, 1997] Multitask Learning ([DOI][1])	✅
MMoE	Multi-gate Mixture-of-Experts	[Ma et al., 2018] Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts ([ACM Digital Library][2])	✅
ESMM	Entire Space Multi-Task Model	[Ma et al., 2018] Entire Space Multi-Task Model: An Effective Approach for Estimating Post-Click Conversion Rate ([arXiv][3])	✅
PLE	Progressive Layered Extraction	[Tang et al., 2020] Progressive Layered Extraction: A Novel Multi-Task Learning Model for Personalized Recommendations ([Papers with Code][4])	✅

4. Generative models

Model	Full Name	Paper	Status
GRU4Rec	Gated Recurrent Unit for Recommendation	[Hidasi et al., 2016] Session-based Recommendations with Recurrent Neural Networks ([arXiv][1])	✅
SASRec	Self-Attentive Sequential Recommendation	[Kang & McAuley, 2018] Self-Attentive Sequential Recommendation ([arXiv][2])	✅
BERT4Rec	Bidirectional Encoder Representations from Transformers for Recommendation	[Sun et al., 2019] BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer ([arXiv][3])	✅
TIGER	Transformer Index for Generative Recommenders	[Rajput et al., 2023] Recommender Systems with Generative Retrieval ([arXiv][4])	✅
HSTU	Hierarchical Sequential Transduction Units	[Zhai et al., 2024] Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations ([arXiv][5])	✅

5. Recall

Model	Full Name	Paper	Status
Item2Vec	Neural Item Embedding for Collaborative Filtering	[Barkan & Koenigstein, 2016] Item2Vec: Neural Item Embedding for Collaborative Filtering	✅
ItemCF	Item-Based Collaborative Filtering	[Sarwar et al., 2001] Item-Based Collaborative Filtering Recommendation Algorithms	✅
Swing	Swing Item-to-Item Collaborative Filtering	Alibaba, Swing Recommendation Algorithm	✅
Two-Tower	Two-Tower Neural Network	[Yang et al., 2020] Mixed Negative Sampling for Learning Two-Tower Neural Networks in Recommendations	✅
YouTubeDNN	Deep Neural Network for YouTube Recommendations	[Covington et al., 2016] Deep Neural Networks for YouTube Recommendations	✅

Contact

For questions, suggestions, or collaboration, please contact: Email: guopan.jin@outlook.com

Citation

If you find deepsklearn useful in your research or projects, please consider citing it:

@software{jin2026deepsklearn,
  author = {Jin, Guopan},
  title  = {{DeepSklearn}},
  year   = {2026},
  url    = {https://github.com/guopanjin/deepsklearn}
}

Metadata

Release files for deepsklearn 1.0.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for deepsklearn 1.0.1
File	Size	Uploaded	Provenance
deepsklearn-1.0.1.tar.gz	10.8 kB	Jul 20, 2026	Without provenance	Details

Built distribution (wheel)

Table of built distributions (wheels) for deepsklearn 1.0.1
File	Interpreter	ABI	Platform	Reset
deepsklearn-1.0.1-py3-none-any.whl 5.5 kB Jul 20, 2026 Without provenance	Python 3	none	any	Details

Total release size: 16.3 kB

Release files / deepsklearn-1.0.1.tar.gz

Download URL	deepsklearn-1.0.1.tar.gz
Size	10.8 kB
Tags	Source
SHA-256 checksum How to use checksums	`0eb5d156b30df7eabbab2503bef96bc1780b46fd1ce1e5f528e0ea9631a071f0`
BLAKE2b-256 checksum How to use checksums	`4aafa6a90fdbc186b9addf3b8dc083616c25699fd8f11edbb49ff84c6a31d6cd`
Upload date	Jul 20, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.9.18`

Release files / deepsklearn-1.0.1-py3-none-any.whl

Download URL	deepsklearn-1.0.1-py3-none-any.whl
Size	5.5 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`2b4bb1c283a93bd0ae822eab7c3c0279f93f1b0131fa0c311256bb834ff334c7`
BLAKE2b-256 checksum How to use checksums	`bf404643e407824256f6a04705f87c68974ff126c36d0cf643c19797ef857a80`
Upload date	Jul 20, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.9.18`

Security for deepsklearn 1.0.1

Report project as malware

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.