├── .github └── workflows │ └── test-basic.yml ├── .gitignore ├── LICENSE ├── README.md ├── awesome_autodl ├── __init__.py ├── bins │ ├── __init__.py │ ├── list_email.py │ ├── list_email_FIE_202203.py │ ├── show_infos.py │ └── statistics.py ├── data_cls │ ├── __init__.py │ ├── abbrv.py │ └── paper.py ├── raw_data │ ├── __init__.py │ ├── abbrv.bib │ ├── abbrv.yaml │ └── papers │ │ ├── Automated_Data_Engineering.yaml │ │ ├── Automated_Deployment.yaml │ │ ├── Automated_Maintenance.yaml │ │ ├── Automated_Problem_Formulation.yaml │ │ ├── Hyperparameter_Optimization.yaml │ │ └── Neural_Architecture_Search.yaml └── utils │ ├── __init__.py │ ├── check.py │ ├── filter.py │ ├── fix_invalid_email.py │ └── yaml.py ├── setup.py └── tests ├── test_abbrv.py └── test_format.py /.github/workflows/test-basic.yml: -------------------------------------------------------------------------------- 1 | name: Test Black 2 | on: 3 | push: 4 | branches: 5 | - main 6 | pull_request: 7 | branches: 8 | - main 9 | 10 | 11 | jobs: 12 | build: 13 | strategy: 14 | matrix: 15 | os: [ubuntu-18.04, ubuntu-20.04, macos-latest] 16 | python-version: [3.6, 3.7, 3.8, 3.9] 17 | 18 | runs-on: ${{ matrix.os }} 19 | steps: 20 | - uses: actions/checkout@v2 21 | 22 | - name: Set up Python ${{ matrix.python-version }} 23 | uses: actions/setup-python@v2 24 | with: 25 | python-version: ${{ matrix.python-version }} 26 | 27 | - name: Lint with Black 28 | run: | 29 | python -m pip install --upgrade pip 30 | python -m pip install black 31 | python --version 32 | python -m black --version 33 | echo $PWD ; ls 34 | python -m black ./awesome_autodl/bins/* -l 88 --check --diff --verbose 35 | python -m black ./awesome_autodl/utils/* -l 88 --check --diff --verbose 36 | 37 | - name: Install Awesome-AutoDL from source 38 | run: | 39 | python -m pip install . --force 40 | 41 | - name: Run tests with pytest 42 | run: | 43 | python -m pip install pytest 44 | python -m pytest . --durations=0 45 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | .DS_Store 2 | # Byte-compiled / optimized / DLL files 3 | __pycache__/ 4 | *.py[cod] 5 | *$py.class 6 | 7 | # C extensions 8 | *.so 9 | 10 | # Distribution / packaging 11 | .Python 12 | env/ 13 | build/ 14 | develop-eggs/ 15 | dist/ 16 | downloads/ 17 | eggs/ 18 | .eggs/ 19 | lib64/ 20 | parts/ 21 | sdist/ 22 | var/ 23 | *.egg-info/ 24 | .installed.cfg 25 | *.egg 26 | 27 | # PyInstaller 28 | # Usually these files are written by a python script from a template 29 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 30 | *.manifest 31 | *.spec 32 | 33 | # Installer logs 34 | pip-log.txt 35 | pip-delete-this-directory.txt 36 | 37 | # Unit test / coverage reports 38 | htmlcov/ 39 | .tox/ 40 | .coverage 41 | .coverage.* 42 | .cache 43 | nosetests.xml 44 | coverage.xml 45 | *,cover 46 | .hypothesis/ 47 | 48 | # Translations 49 | *.mo 50 | *.pot 51 | 52 | # Django stuff: 53 | *.log 54 | local_settings.py 55 | 56 | # Flask stuff: 57 | instance/ 58 | .webassets-cache 59 | 60 | # Scrapy stuff: 61 | .scrapy 62 | 63 | # Sphinx documentation 64 | docs/_build/ 65 | 66 | # PyBuilder 67 | target/ 68 | 69 | # IPython Notebook 70 | .ipynb_checkpoints 71 | 72 | # pyenv 73 | .python-version 74 | 75 | # celery beat schedule file 76 | celerybeat-schedule 77 | 78 | # dotenv 79 | .env 80 | 81 | # virtualenv 82 | venv/ 83 | ENV/ 84 | 85 | # Spyder project settings 86 | .spyderproject 87 | 88 | # Rope project settings 89 | .ropeproject 90 | 91 | # Pycharm project 92 | .idea 93 | snapshots 94 | *.pytorch 95 | *.tar.bz 96 | data 97 | .*.swp 98 | main_main.py 99 | *.pdf 100 | */*.pdf 101 | 102 | # Device 103 | scripts-nas/.nfs00* 104 | */.nfs00* 105 | *.DS_Store 106 | 107 | # logs and snapshots 108 | output 109 | logs 110 | 111 | # snapshot 112 | a.pth 113 | cal-merge*.sh 114 | GPU-*.sh 115 | cal.sh 116 | aaa 117 | cx.sh 118 | 119 | NAS-Bench-*-v1_0.pth 120 | lib/NAS-Bench-*-v1_0.pth 121 | others/TF 122 | scripts-search/l2s-algos 123 | TEMP-L.sh 124 | 125 | .nfs00* 126 | *.swo 127 | */*.swo 128 | 129 | # Visual Studio Code 130 | .vscode 131 | mlruns* 132 | outputs 133 | 134 | pytest_cache 135 | *.pkl 136 | *.pth 137 | 138 | *.tgz 139 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2018 Xuanyi Dong [GitHub: D-X-Y] 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 |
Awesome AutoDL [![Awesome](https://awesome.re/badge.svg)](https://awesome.re)
2 | 3 | A curated list of automated deep learning related resources. Inspired by [awesome-deep-vision](https://github.com/kjw0612/awesome-deep-vision), [awesome-adversarial-machine-learning](https://github.com/yenchenlin/awesome-adversarial-machine-learning), [awesome-deep-learning-papers](https://github.com/terryum/awesome-deep-learning-papers), and [awesome-architecture-search](https://github.com/markdtw/awesome-architecture-search). 4 | 5 | Please feel free to [pull requests](https://github.com/D-X-Y/Awesome-AutoDL/pulls) or [open an issue](https://github.com/D-X-Y/Awesome-AutoDL/issues) to add papers. 6 | 7 | --- 8 | 9 |
Table of Contents
10 | 11 | - [Awesome Blogs](#awesome-blogs) 12 | - [Awesome AutoDL Libraies](#awesome-autodl-libraies) 13 | - [Awesome Benchmarks](#awesome-benchmarks) 14 | - [Deep Learning-based NAS and HPO](#deep-learning-based-nas-and-hpo) 15 | - [2021 Venues](#2021-venues) 16 | - [2020 Venues](#2020-venues) 17 | - [2019 Venues](#2019-venues) 18 | - [2018 Venues](#2018-venues) 19 | - [2017 Venues](#2017-venues) 20 | - [Previous Venues](#previous-venues) 21 | - [arXiv](#arxiv) 22 | - [Awesome Surveys](#awesome-surveys) 23 | 24 | --- 25 | 26 | # Awesome Blogs 27 | 28 | - [AutoML info](http://automl.chalearn.org/) and [AutoML Freiburg-Hannover](https://www.automl.org/) 29 | - [What’s the deal with Neural Architecture Search?](https://determined.ai/blog/neural-architecture-search/) 30 | - [Google Could AutoML](https://cloud.google.com/vision/automl/docs/beginners-guide) and [PocketFlow](https://pocketflow.github.io/) 31 | - [AutoML Challenge](http://automl.chalearn.org/) and [AutoDL Challenge](https://autodl.chalearn.org/) 32 | - [In Defense of Weight-sharing for Neural Architecture Search: an optimization perspective](https://determined.ai/blog/ws-optimization-for-nas/) 33 | 34 | # Awesome AutoDL Libraies 35 | 36 | - [PyGlove](https://proceedings.neurips.cc/paper/2020/file/012a91467f210472fab4e11359bbfef6-Paper.pdf) 37 | - [NASLib](https://github.com/automl/NASLib) 38 | - [Keras Tuner](https://keras-team.github.io/keras-tuner/) 39 | - [NNI](https://github.com/microsoft/nni) 40 | - [AutoGluon](https://autogluon.mxnet.io/) 41 | - [Auto-PyTorch](https://github.com/automl/Auto-PyTorch) 42 | - [AutoDL-Projects](https://github.com/D-X-Y/AutoDL-Projects) 43 | - [aw_nas](https://github.com/walkerning/aw_nas) 44 | - [Determined](https://github.com/determined-ai/determined) 45 | - [TPOT](https://github.com/EpistasisLab/tpot) 46 | 47 | # Awesome Benchmarks 48 | 49 | | Title | Venue | Code | 50 | |:--------|:--------:|:--------:| 51 | | [NAS-Bench-101: Towards Reproducible Neural Architecture Search](https://arxiv.org/pdf/1902.09635.pdf) | ICML 2019 | [GitHub](https://github.com/google-research/nasbench) | 52 | | [NAS-Bench-201: Extending the Scope of Reproducible Neural Architecture Search](https://openreview.net/forum?id=HJxyZkBKDr) | ICLR 2020 | [Github](https://github.com/D-X-Y/NAS-Bench-201) | 53 | | [NAS-Bench-301 and the Case for Surrogate Benchmarks for Neural Architecture Search](https://arxiv.org/abs/2008.09777) | arXiv 2020 | [GitHub](https://github.com/automl/nasbench301) | 54 | | [NAS-Bench-1Shot1: Benchmarking and Dissecting One-shot Neural Architecture Search](https://arxiv.org/abs/2001.10422) | ICLR 2020 | [GitHub](https://github.com/automl/nasbench-1shot1) | 55 | | [NATS-Bench: Benchmarking NAS Algorithms for Architecture Topology and Size](https://arxiv.org/abs/2009.00437) | TPAMI 2021 | [GitHub](https://github.com/D-X-Y/NATS-Bench) 56 | | [NAS-Bench-ASR: Reproducible Neural Architecture Search for Speech Recognition](https://openreview.net/forum?id=CU0APx9LMaL) | ICLR 2021 | [GitHub](https://github.com/SamsungLabs/nb-asr) | 57 | | [HW-NAS-Bench: Hardware-Aware Neural Architecture Search Benchmark](https://openreview.net/pdf?id=_0kaDkv3dVf) | ICLR 2021 | [GitHub](https://github.com/RICE-EIC/HW-NAS-Bench) | 58 | | [NAS-Bench-NLP: Neural Architecture Search Benchmark for Natural Language Processing](https://arxiv.org/pdf/2006.07116.pdf) | arXiv 2020 | [GitHub](https://github.com/fmsnew/nas-bench-nlp-release) | 59 | | [NAS-Bench-x11 and the Power of Learning Curves](https://arxiv.org/pdf/2111.03602.pdf) | NeurIPS 2021 | [GitHub](https://github.com/automl/nas-bench-x11) | 60 | 61 | # Deep Learning-based NAS and HPO 62 | 63 | | Type | G | RL | EA | PD | Other | 64 | |:------------|:--------------:|:----------------------:|:-----------------------:|:----------------------:|:----------:| 65 | | Explanation | gradient-based | reinforcement learning | evolutionary algorithm | performance prediction | other types | 66 | 67 | ## 2021 Venues 68 | 69 | | Title | Venue | Type | Code | 70 | |:--------|:--------:|:--------:|:--------:| 71 | | [CATE: Computation-aware Neural Architecture Encoding with Transformers](https://arxiv.org/pdf/2102.07108.pdf) | ICML | O | [GitHub](https://github.com/MSU-MLSys-Lab/CATE) | 72 | | [Searching by Generating: Flexible and Efficient One-Shot NAS with Architecture Generator](https://openaccess.thecvf.com/content/CVPR2021/papers/Huang_Searching_by_Generating_Flexible_and_Efficient_One-Shot_NAS_With_Architecture_CVPR_2021_paper.pdf) | CVPR | G | [Github](https://github.com/eric8607242/SGNAS) | 73 | | [Zen-NAS: A Zero-Shot NAS for High-Performance Deep Image Recognition](https://arxiv.org/abs/2102.01063) | ICCV | EA | [Github](https://github.com/idstcv/ZenNAS) | 74 | | [AutoFormer: Searching Transformers for Visual Recognition](https://arxiv.org/pdf/2107.00651.pdf) |ICCV | EA | [GitHub](https://github.com/microsoft/AutoML) 75 | | [LightTrack: Finding Lightweight Neural Networks for Object Tracking via One-Shot Architecture Search](https://arxiv.org/abs/2104.14545) | CVPR | EA | [GitHub](https://github.com/researchmm/LightTrack) | 76 | | [One-Shot Neural Ensemble Architecture Search by Diversity-Guided Search Space Shrinking](https://arxiv.org/abs/2104.00597) | CVPR | EA | [GitHub](https://github.com/researchmm/NEAS) | 77 | | [DARTS-: Robustly Stepping out of Performance Collapse Without Indicators](https://openreview.net/pdf?id=KLH36ELmwIB) | ICLR | G | [GitHub](https://github.com/Meituan-AutoML/DARTS-) | 78 | | [Zero-Cost Proxies for Lightweight NAS](https://openreview.net/pdf?id=0cmMMy8J5q) | ICLR | O | [GitHub](https://github.com/SamsungLabs/zero-cost-nas) | 79 | | [Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective](https://openreview.net/forum?id=Cnon5ezMHtu) | ICLR | - | [GitHub](https://github.com/VITA-Group/TENAS) | 80 | | [DrNAS: Dirichlet Neural Architecture Search](https://openreview.net/forum?id=9FWas6YbmB3) | ICLR | G | [GitHub](https://github.com/xiangning-chen/DrNAS) | 81 | | [Rethinking Architecture Selection in Differentiable NAS](https://openreview.net/forum?id=PKubaeJkw3) | ICLR | O | [GitHub](https://github.com/ruocwang/darts-pt) | 82 | | [Evolving Reinforcement Learning Algorithms](https://openreview.net/forum?id=0XXpJ4OtjW) | ICLR | EA | [GitHub](https://github.com/jcoreyes/evolvingrl) | 83 | | [AutoHAS: Differentiable Hyper-parameter and Architecture Search](https://arxiv.org/pdf/2006.03656.pdf) | ICLR-W | G | - | 84 | | [FBNetV3: Joint Architecture-Recipe Search using Neural Acquisition Function](https://arxiv.org/abs/2006.02049) | CVPR | PD | [github](https://github.com/facebookresearch/mobile-vision/blob/main/mobile_cv/arch/fbnet_v2/fbnet_modeldef_cls_fbnetv3.py) | 85 | 86 | ## 2020 Venues 87 | 88 | | Title | Venue | Type | Code | 89 | |:--------|:--------:|:--------:|:--------:| 90 | | [Cream of the Crop: Distilling Prioritized Paths For One-Shot Neural Architecture Search](https://papers.nips.cc/paper/2020/file/d072677d210ac4c03ba046120f0802ec-Paper.pdf) | NeurIPS | - | [GitHub](https://github.com/microsoft/Cream) | 91 | | [PyGlove: Symbolic Programming for Automated Machine Learning](https://proceedings.neurips.cc/paper/2020/file/012a91467f210472fab4e11359bbfef6-Paper.pdf) | NeurIPS | library | - | 92 | | [Does Unsupervised Architecture Representation Learning Help Neural Architecture Search](https://arxiv.org/abs/2006.06936) | NeurIPS | PD | [GitHub](https://github.com/MSU-MLSys-Lab/arch2vec) | 93 | | [RandAugment: Practical Automated Data Augmentation with a Reduced Search Space](https://arxiv.org/abs/1909.13719) | NeurIPS | | [GitHub](https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet) | 94 | | [Delta-STN: Efficient Bilevel Optimization for Neural Networks using Structured Response Jacobians](https://arxiv.org/pdf/2010.13514.pdf) | NeurIPS | G | [GitHub](https://github.com/pomonam/Self-Tuning-Networks) | 95 | | [A Study on Encodings for Neural Architecture Search](https://arxiv.org/abs/2007.04965) | NeurIPS | | [GitHub](https://github.com/naszilla/naszilla) | 96 | | [AutoBSS: An Efficient Algorithm for Block Stacking Style Search](https://proceedings.neurips.cc/paper/2020/file/747d3443e319a22747fbb873e8b2f9f2-Paper.pdf) | NeurIPS | | | 97 | | [Bridging the Gap between Sample-based and One-shot Neural Architecture Search with BONAS](https://proceedings.neurips.cc/paper/2020/file/13d4635deccc230c944e4ff6e03404b5-Paper.pdf) | NeurIPS | G | [GitHub](https://github.com/haolibai/APS-channel-search) | 98 | | [Interstellar: Searching Recurrent Architecture for Knowledge Graph Embedding](https://proceedings.neurips.cc/paper/2020/file/722caafb4825ef5d8670710fa29087cf-Paper.pdf) | NeurIPS | | | 99 | | [Revisiting Parameter Sharing for Automatic Neural Channel Number Search](https://proceedings.neurips.cc/paper/2020/file/42cd63cb189c30ed03e42ce2c069566c-Paper.pdf) | NeurIPS | | | 100 | | [Learning Search Space Partition for Black-box Optimization using Monte Carlo Tree Search](https://arxiv.org/pdf/2007.00708.pdf) | NeurIPS | MCTS | [GitHub](https://github.com/facebookresearch/LaMCTS) | 101 | | [Neural Architecture Search using Deep Neural Networks and Monte Carlo Tree Search](https://arxiv.org/abs/1805.07440) | AAAI | MCTS | [GitHub](https://github.com/linnanwang/AlphaX-NASBench101) | 102 | | [Representation Sharing for Fast Object Detector Search and Beyond](https://arxiv.org/pdf/2007.12075v4.pdf) | ECCV | G | [GitHub](https://github.com/msight-tech/research-fad) | 103 | | [Are Labels Necessary for Neural Architecture Search?](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123490766.pdf) | ECCV | G | - | 104 | | [Single Path One-Shot Neural Architecture Search with Uniform Sampling](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123610528.pdf) | ECCV | EA | - | 105 | | [Neural Predictor for Neural Architecture Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123740647.pdf) | ECCV | O | - | 106 | | [BigNAS: Scaling Up Neural Architecture Search with Big Single-Stage Models](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123520681.pdf) | ECCV | G | - | 107 | | [BATS: Binary ArchitecTure Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123680307.pdf) | ECCV | - | - | 108 | | [AttentionNAS: Spatiotemporal Attention Cell Search for Video Classification](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123530443.pdf) | ECCV | - | - | 109 | | [Search What You Want: Barrier Panelty NAS for Mixed Precision Quantization](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123540001.pdf) | ECCV | - | - | 110 | | [Angle-based Search Space Shrinking for Neural Architecture Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123630426.pdf) | ECCV | - | - | 111 | | [Anti-Bandit Neural Architecture Search for Model Defense](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123580069.pdf) | ECCV | - | - | 112 | | [TF-NAS: Rethinking Three Search Freedoms of Latency-Constrained Differentiable Neural Architecture Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123600120.pdf) | ECCV | G | [GitHub](https://github.com/AberHu/TF-NAS) | 113 | | [Fair DARTS: Eliminating Unfair Advantages in Differentiable Architecture Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123600460.pdf) | ECCV | G | [GitHub](https://github.com/xiaomi-automl/FairDARTS) | 114 | | [Off-Policy Reinforcement Learning for Efficient and Effective GAN Architecture Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123520171.pdf) | ECCV | RL | - | 115 | | [DA-NAS: Data Adapted Pruning for Efficient Neural Architecture Search](https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123720579.pdf) | ECCV | G | - | 116 | | [Optimizing Millions of Hyperparameters by Implicit Differentiation](https://arxiv.org/abs/1911.02590) | AISTATS | G | - | 117 | | [Evolving Machine Learning Algorithms From Scratch](https://arxiv.org/pdf/2003.03384.pdf) | ICML | EA | - | 118 | | [Stabilizing Differentiable Architecture Search via Perturbation-based Regularization](https://arxiv.org/abs/2002.05283) | ICML | G | [GitHub](https://github.com/xiangning-chen/SmoothDARTS) | 119 | | [NADS: Neural Architecture Distribution Search for Uncertainty Awareness](https://arxiv.org/pdf/2006.06646.pdf) | ICML | - | - | 120 | | [Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data](https://arxiv.org/abs/1912.07768) | ICML | - | - | 121 | | Neural Architecture Search in a Proxy Validation Loss Landscape | ICML | - | - | 122 | | [Hit-Detector: Hierarchical Trinity Architecture Search for Object Detection](https://arxiv.org/pdf/2003.11818v1.pdf) | CVPR | - | [GitHub](https://github.com/ggjy/HitDet.pytorch) | 123 | | [Designing Network Design Spaces](https://arxiv.org/pdf/2003.13678.pdf) | CVPR | - | [GitHub](https://github.com/facebookresearch/pycls) | 124 | | [UNAS: Differentiable Architecture Search Meets Reinforcement Learning](https://openaccess.thecvf.com/content_CVPR_2020/papers/Vahdat_UNAS_Differentiable_Architecture_Search_Meets_Reinforcement_Learning_CVPR_2020_paper.pdf) | CVPR | G/RL | [GitHub](https://github.com/NVlabs/unas) | 125 | | [MiLeNAS: Efficient Neural Architecture Search via Mixed-Level Reformulation](https://arxiv.org/pdf/2003.12238.pdf) | CVPR | G | [GitHub](https://github.com/chaoyanghe/MiLeNAS) | 126 | | [A Semi-Supervised Assessor of Neural Architectures](https://openaccess.thecvf.com/content_CVPR_2020/papers/Tang_A_Semi-Supervised_Assessor_of_Neural_Architectures_CVPR_2020_paper.pdf) | CVPR | PD | - | 127 | | [Binarizing MobileNet via Evolution-based Searching](https://arxiv.org/abs/2005.06305) | CVPR | EA | - | 128 | | [Rethinking Performance Estimation in Neural Architecture Search](https://openaccess.thecvf.com/content_CVPR_2020/papers/Zheng_Rethinking_Performance_Estimation_in_Neural_Architecture_Search_CVPR_2020_paper.pdf) | CVPR | - | [GitHub](https://github.com/zhengxiawu/rethinking_performance_estimation_in_NAS) | 129 | | [APQ: Joint Search for Network Architecture, Pruning and Quantization Policy](http://openaccess.thecvf.com/content_CVPR_2020/papers/Wang_APQ_Joint_Search_for_Network_Architecture_Pruning_and_Quantization_Policy_CVPR_2020_paper.pdf) | CVPR | G | [GitHub](https://github.com/mit-han-lab/apq) | 130 | | [SGAS: Sequential Greedy Architecture Search](http://openaccess.thecvf.com/content_CVPR_2020/papers/Li_SGAS_Sequential_Greedy_Architecture_Search_CVPR_2020_paper.pdf) | CVPR | G | [Github](https://github.com/lightaime/sgas) | 131 | | [Can Weight Sharing Outperform Random Architecture Search? An Investigation With TuNAS](http://openaccess.thecvf.com/content_CVPR_2020/papers/Bender_Can_Weight_Sharing_Outperform_Random_Architecture_Search_An_Investigation_With_CVPR_2020_paper.pdf) | CVPR | RL | - | 132 | | [FBNetV2: Differentiable Neural Architecture Search for Spatial and Channel Dimensions](https://arxiv.org/abs/2004.05565) | CVPR | G | [Github](https://github.com/facebookresearch/mobile-vision) | 133 | | [AdversarialNAS: Adversarial Neural Architecture Search for GANs](https://arxiv.org/pdf/1912.02037.pdf) | CVPR | G | [Github](https://github.com/chengaopro/AdversarialNAS) | 134 | | [When NAS Meets Robustness: In Search of Robust Architectures against Adversarial Attacks](https://arxiv.org/abs/1911.10695) | CVPR | G | [Github](https://github.com/gmh14/RobNets) | 135 | | [Block-wisely Supervised Neural Architecture Search with Knowledge Distillation](https://www.xiaojun.ai/papers/CVPR2020_04676.pdf) | CVPR | G | [Github](https://github.com/changlin31/DNA) | 136 | | [Overcoming Multi-Model Forgetting in One-Shot NAS with Diversity Maximization](https://www.xiaojun.ai/papers/cvpr-2020-zhang.pdf) | CVPR | G | [Github](https://github.com/MiaoZhang0525/NSAS_FOR_CVPR) | 137 | | [Densely Connected Search Space for More Flexible Neural Architecture Search](https://arxiv.org/abs/1906.09607) | CVPR | G | [Github](https://github.com/JaminFong/DenseNAS) | 138 | | [EfficientDet: Scalable and Efficient Object Detection](https://arxiv.org/abs/1911.09070) | CVPR | RL | - | 139 | | [NAS-BENCH-201: Extending the Scope of Reproducible Neural Architecture Search](https://openreview.net/forum?id=HJxyZkBKDr) | ICLR | - | [Github](https://github.com/D-X-Y/AutoDL-Projects) | 140 | | [Understanding Architectures Learnt by Cell-based Neural Architecture Search](https://openreview.net/forum?id=BJxH22EKPS) | ICLR | G | [GitHub](https://github.com/shuyao95/Understanding-NAS) | 141 | | [Evaluating The Search Phase of Neural Architecture Search](https://openreview.net/forum?id=H1loF2NFwr) | ICLR | - | | 142 | | [AtomNAS: Fine-Grained End-to-End Neural Architecture Search](https://openreview.net/forum?id=BylQSxHFwr) | ICLR | | [GitHub](https://github.com/meijieru/AtomNAS) | 143 | | [Fast Neural Network Adaptation via Parameter Remapping and Architecture Search](https://openreview.net/forum?id=rklTmyBKPH) | ICLR | - | [GitHub](https://github.com/JaminFong/FNA) | 144 | | [Once for All: Train One Network and Specialize it for Efficient Deployment](https://openreview.net/forum?id=HylxE1HKwS) | ICLR | G | [GitHub](https://github.com/mit-han-lab/once-for-all) | 145 | | Efficient Transformer for Mobile Applications | ICLR | - | - | 146 | | [PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search](https://arxiv.org/pdf/1907.05737v4.pdf) | ICLR | G | [GitHub](https://github.com/yuhuixu1993/PC-DARTS) | 147 | | Adversarial AutoAugment | ICLR | - | - | 148 | | [NAS evaluation is frustratingly hard](https://arxiv.org/abs/1912.12522) | ICLR | - | [GitHub](https://github.com/antoyang/NAS-Benchmark) | 149 | | [FasterSeg: Searching for Faster Real-time Semantic Segmentation](https://openreview.net/pdf?id=BJgqQ6NYvB) | ICLR | G | [GitHub](https://github.com/TAMU-VITA/FasterSeg) | 150 | | [Computation Reallocation for Object Detection](https://openreview.net/forum?id=SkxLFaNKwB) | ICLR | - | - | 151 | | [Towards Fast Adaptation of Neural Architectures with Meta Learning](https://openreview.net/pdf?id=r1eowANFvr) | ICLR | - | [GitHub](https://github.com/dongzelian/T-NAS) | 152 | | [AssembleNet: Searching for Multi-Stream Neural Connectivity in Video Architectures](https://arxiv.org/pdf/1905.13209v4.pdf) | ICLR | EA | - | 153 | | How to Own the NAS in Your Spare Time | ICLR | - | - | 154 | | [Fast, Accurate and Lightweight Super-Resolution with Neural Architecture Search](https://arxiv.org/pdf/1901.07261.pdf) | ICPR | G | [github](https://github.com/falsr/FALSR) | 155 | 156 | ## 2019 Venues 157 | 158 | | Title | Venue | Type | Code | 159 | |:--------|:--------:|:--------:|:--------:| 160 | | [Self-Tuning Networks: Bilevel Optimization of Hyperparameters using Structured Best-Response Functions](https://arxiv.org/abs/1903.03088) | ICLR | - | - | 161 | | [DATA: Differentiable ArchiTecture Approximation](http://papers.nips.cc/paper/8374-data-differentiable-architecture-approximation) | NeurIPS | - | - | 162 | | [Random Search and Reproducibility for Neural Architecture Search](https://arxiv.org/pdf/1902.07638v3.pdf) | UAI | G | [GitHub](https://github.com/D-X-Y/NAS-Projects/blob/master/scripts-search/algos/RANDOM-NAS.sh) | 163 | | [Improved Differentiable Architecture Search for Language Modeling and Named Entity Recognition](https://www.aclweb.org/anthology/D19-1367.pdf/) | EMNLP | G | - | 164 | | [Continual and Multi-Task Architecture Search](https://www.aclweb.org/anthology/P19-1185.pdf) | ACL | RL | - | 165 | | [Progressive Differentiable Architecture Search: Bridging the Depth Gap Between Search and Evaluation](https://arxiv.org/pdf/1904.12760v1.pdf) | ICCV | G | [GitHub](https://github.com/chenxin061/pdarts) | 166 | | [Multinomial Distribution Learning for Effective Neural Architecture Search](https://arxiv.org/pdf/1905.07529v3.pdf) | ICCV | - | [GitHub](https://github.com/tanglang96/MDENAS) | 167 | | [Searching for MobileNetV3](https://arxiv.org/pdf/1905.02244v5.pdf) | ICCV | EA | - | 168 | | [Multinomial Distribution Learning for Effective Neural Architecture Search](http://openaccess.thecvf.com/content_ICCV_2019/papers/Zheng_Multinomial_Distribution_Learning_for_Effective_Neural_Architecture_Search_ICCV_2019_paper.pdf) | ICCV | - | [GitHub](https://github.com/tanglang96/MDENAS) | 169 | | [Fast and Practical Neural Architecture Search](http://openaccess.thecvf.com/content_ICCV_2019/papers/Cui_Fast_and_Practical_Neural_Architecture_Search_ICCV_2019_paper.pdf) | ICCV | | | 170 | | [Teacher Guided Architecture Search](http://openaccess.thecvf.com/content_ICCV_2019/papers/Bashivan_Teacher_Guided_Architecture_Search_ICCV_2019_paper.pdf) | ICCV | | - | 171 | | [AutoDispNet: Improving Disparity Estimation With AutoML](http://openaccess.thecvf.com/content_ICCV_2019/papers/Saikia_AutoDispNet_Improving_Disparity_Estimation_With_AutoML_ICCV_2019_paper.pdf) | ICCV | G | - | 172 | | [Resource Constrained Neural Network Architecture Search: Will a Submodularity Assumption Help?](http://openaccess.thecvf.com/content_ICCV_2019/papers/Xiong_Resource_Constrained_Neural_Network_Architecture_Search_Will_a_Submodularity_Assumption_ICCV_2019_paper.pdf) | ICCV | EA | - | 173 | | [One-Shot Neural Architecture Search via Self-Evaluated Template Network](https://arxiv.org/abs/1910.05733) | ICCV | G | [Github](https://github.com/D-X-Y/NAS-Projects) | 174 | | [Evolving Space-Time Neural Architectures for Videos](https://arxiv.org/abs/1811.10636) | ICCV | EA | [GitHub](https://sites.google.com/view/evanet-video) | 175 | | [AutoGAN: Neural Architecture Search for Generative Adversarial Networks](https://arxiv.org/pdf/1908.03835.pdf) | ICCV | RL | [github](https://github.com/TAMU-VITA/AutoGAN) | 176 | | [Discovering Neural Wirings](https://arxiv.org/pdf/1906.00586.pdf) | NeurIPS | G | [Github](https://github.com/allenai/dnw) | 177 | | [Towards modular and programmable architecture search](https://arxiv.org/abs/1909.13404) | NeurIPS | [Other](https://github.com/D-X-Y/Awesome-NAS/issues/10) | [Github](https://github.com/negrinho/deep_architect) | 178 | | [Network Pruning via Transformable Architecture Search](https://arxiv.org/abs/1905.09717) | NeurIPS | G | [Github](https://github.com/D-X-Y/NAS-Projects) | 179 | | [Deep Active Learning with a NeuralArchitecture Search](https://arxiv.org/pdf/1811.07579.pdf) | NeurIPS | - | - | 180 | | [DetNAS: Backbone Search for Object Detection](https://arxiv.org/pdf/1903.10979v4.pdf) | NeurIPS | EA | [GitHub](https://github.com/megvii-model/DetNAS) | 181 | | [SpArSe: Sparse Architecture Search for CNNs on Resource-Constrained Microcontrollers](https://arxiv.org/pdf/1905.12107v1.pdf) | NeurIPS | - | - | 182 | | [Efficient Forward Architecture Search](https://arxiv.org/abs/1905.13360) | NeurIPS | G | [Github](https://github.com/microsoft/petridishnn) | 183 | | Efficient Neural ArchitectureTransformation Search in Channel-Level for Object Detection | NeurIPS | G | - | 184 | | [XNAS: Neural Architecture Search with Expert Advice](https://arxiv.org/pdf/1906.08031v1.pdf) | NeurIPS | G | [GitHub](https://github.com/NivNayman/XNAS) | 185 | | [DARTS: Differentiable Architecture Search](https://arxiv.org/abs/1806.09055) | ICLR | G | [github](https://github.com/quark0/darts) | 186 | | [ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware](https://openreview.net/pdf?id=HylVB3AqYm) | ICLR | RL/G | [github](https://github.com/MIT-HAN-LAB/ProxylessNAS) | 187 | | [Graph HyperNetworks for Neural Architecture Search](https://arxiv.org/pdf/1810.05749.pdf) | ICLR | G | - | 188 | | [Learnable Embedding Space for Efficient Neural Architecture Compression](https://openreview.net/forum?id=S1xLN3C9YX) | ICLR | Other | [github](https://github.com/Friedrich1006/ESNAC) | 189 | | [Efficient Multi-Objective Neural Architecture Search via Lamarckian Evolution](https://arxiv.org/abs/1804.09081) | ICLR | EA | - | 190 | | [SNAS: stochastic neural architecture search](https://openreview.net/pdf?id=rylqooRqK7) | ICLR | G | - | 191 | | [NetTailor: Tuning the Architecture, Not Just the Weights](https://arxiv.org/abs/1907.00274) | CVPR | G | [Github](https://github.com/pedro-morgado/nettailor) | 192 | | [Searching for A Robust Neural Architecture in Four GPU Hours](http://xuanyidong.com/publication/gradient-based-diff-sampler/) | CVPR | G | [Github](https://github.com/D-X-Y/NAS-Projects) | 193 | | [ChamNet: Towards Efficient Network Design through Platform-Aware Model Adaptation](http://openaccess.thecvf.com/content_CVPR_2019/papers/Dai_ChamNet_Towards_Efficient_Network_Design_Through_Platform-Aware_Model_Adaptation_CVPR_2019_paper.pdf) | CVPR | - | - | 194 | | [Partial Order Pruning: for Best Speed/Accuracy Trade-off in Neural Architecture Search](https://arxiv.org/pdf/1903.03777.pdf) | CVPR | EA | [github](https://github.com/lixincn2015/Partial-Order-Pruning) | 195 | | [FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search](https://arxiv.org/abs/1812.03443) | CVPR | G | - | 196 | | [RENAS: Reinforced Evolutionary Neural Architecture Search](https://arxiv.org/abs/1808.00193) | CVPR | G | - | 197 | | [Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation](https://arxiv.org/pdf/1901.02985.pdf) | CVPR | G | [GitHub](https://github.com/tensorflow/models/tree/master/research/deeplab) | 198 | | [MnasNet: Platform-Aware Neural Architecture Search for Mobile](https://arxiv.org/abs/1807.11626) | CVPR | RL | [Github](https://github.com/AnjieZheng/MnasNet-PyTorch) | 199 | | [MFAS: Multimodal Fusion Architecture Search](https://arxiv.org/pdf/1903.06496.pdf) | CVPR | EA | - | 200 | | [A Neurobiological Evaluation Metric for Neural Network Model Search](https://arxiv.org/pdf/1805.10726.pdf) | CVPR | Other | - | 201 | | [Fast Neural Architecture Search of Compact Semantic Segmentation Models via Auxiliary Cells](https://arxiv.org/abs/1810.10804) | CVPR | RL | - | 202 | | [Customizable Architecture Search for Semantic Segmentation](https://arxiv.org/pdf/1908.09550v1.pdf) | CVPR | - | - | 203 | | [Regularized Evolution for Image Classifier Architecture Search](https://arxiv.org/pdf/1802.01548.pdf) | AAAI | EA | - | 204 | | AutoAugment: Learning Augmentation Policies from Data | CVPR | RL | - | 205 | | Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules | ICML | EA | - | 206 | | [The Evolved Transformer](https://arxiv.org/pdf/1901.11117.pdf) | ICML | EA | [Github](https://github.com/tensorflow/tensor2tensor/blob/master/tensor2tensor/models/evolved_transformer.py) | 207 | | [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](https://arxiv.org/pdf/1905.11946v5.pdf) | ICML | RL | - | 208 | | [NAS-Bench-101: Towards Reproducible Neural Architecture Search](https://arxiv.org/abs/1902.09635) | ICML | Other | [Github](https://github.com/google-research/nasbench) | 209 | | [On Network Design Spaces for Visual Recognition](https://arxiv.org/abs/1905.13214) | ICCV | G | [Github](https://github.com/facebookresearch/nds) | 210 | 211 | ## 2018 Venues 212 | 213 | | Title | Venue | Type | Code | 214 | |:--------|:--------:|:--------:|:--------:| 215 | | [Towards Automatically-Tuned Deep Neural Networks](https://link.springer.com/content/pdf/10.1007%2F978-3-030-05318-5.pdf) | BOOK | - | [GitHub](https://github.com/automl/Auto-PyTorch) | 216 | | [NetAdapt: Platform-Aware Neural Network Adaptation for Mobile Applications](https://arxiv.org/pdf/1804.03230.pdf) | ECCV | - | [github](https://github.com/denru01/netadapt) | 217 | | [Efficient Architecture Search by Network Transformation](https://arxiv.org/pdf/1707.04873.pdf) | AAAI | RL | [github](https://github.com/han-cai/EAS) | 218 | | [Learning Transferable Architectures for Scalable Image Recognition](http://openaccess.thecvf.com/content_cvpr_2018/papers/Zoph_Learning_Transferable_Architectures_CVPR_2018_paper.pdf) | CVPR | RL | [github](https://github.com/tensorflow/models/tree/master/research/slim/nets/nasnet) | 219 | | [N2N learning: Network to Network Compression via Policy Gradient Reinforcement Learning](https://openreview.net/forum?id=B1hcZZ-AW) | ICLR | RL | - | 220 | | [A Flexible Approach to Automated RNN Architecture Generation](https://openreview.net/forum?id=SkOb1Fl0Z) | ICLR | RL/PD | - | 221 | | [Practical Block-wise Neural Network Architecture Generation](http://openaccess.thecvf.com/content_cvpr_2018/papers/Zhong_Practical_Block-Wise_Neural_CVPR_2018_paper.pdf) | CVPR | RL | - | [Efficient Neural Architecture Search via Parameter Sharing](http://proceedings.mlr.press/v80/pham18a.html) | ICML | RL | [github](https://github.com/melodyguan/enas) | 222 | | [Path-Level Network Transformation for Efficient Architecture Search](https://arxiv.org/abs/1806.02639) | ICML | RL | [github](https://github.com/han-cai/PathLevel-EAS) | 223 | | [Hierarchical Representations for Efficient Architecture Search](https://openreview.net/forum?id=BJQRKzbA-) | ICLR | EA | - | 224 | | [Understanding and Simplifying One-Shot Architecture Search](http://proceedings.mlr.press/v80/bender18a/bender18a.pdf) | ICML | G | - | 225 | | [SMASH: One-Shot Model Architecture Search through HyperNetworks](https://arxiv.org/pdf/1708.05344.pdf) | ICLR | G | [github](https://github.com/ajbrock/SMASH) | 226 | | [Neural Architecture Optimization](https://arxiv.org/pdf/1808.07233.pdf) | NeurIPS | G | [github](https://github.com/renqianluo/NAO) | 227 | | [Searching for efficient multi-scale architectures for dense image prediction](https://papers.nips.cc/paper/8087-searching-for-efficient-multi-scale-architectures-for-dense-image-prediction.pdf) | NeurIPS | Other | - | 228 | | [Progressive Neural Architecture Search](http://openaccess.thecvf.com/content_ECCV_2018/papers/Chenxi_Liu_Progressive_Neural_Architecture_ECCV_2018_paper.pdf) | ECCV | PD | [github](https://github.com/chenxi116/PNASNet) | 229 | | [Neural Architecture Search with Bayesian Optimisation and Optimal Transport](https://arxiv.org/pdf/1802.07191.pdf) | NeurIPS | Other | [github](https://github.com/kirthevasank/nasbot) | 230 | | [Differentiable Neural Network Architecture Search](https://openreview.net/pdf?id=BJ-MRKkwG) | ICLR-W | G | - | 231 | | [Accelerating Neural Architecture Search using Performance Prediction](https://arxiv.org/abs/1705.10823) | ICLR-W | PD | - | 232 | 233 | ## 2017 Venues 234 | 235 | | Title | Venue | Type | Code | 236 | |:--------|:--------:|:--------:|:--------:| 237 | | [Neural Architecture Search with Reinforcement Learning](https://arxiv.org/abs/1611.01578) | ICLR | RL | - | 238 | | [Designing Neural Network Architectures using Reinforcement Learning](https://openreview.net/pdf?id=S1c2cvqee) | ICLR | RL | - | [github](https://github.com/bowenbaker/metaqnn) | 239 | | [Neural Optimizer Search with Reinforcement Learning](http://proceedings.mlr.press/v70/bello17a/bello17a.pdf) | ICML | RL | - | [Large-Scale Evolution of Image Classifiers](https://arxiv.org/pdf/1703.01041.pdf) | ICML | EA | - | 240 | | [Learning Curve Prediction with Bayesian Neural Networks](http://ml.informatik.uni-freiburg.de/papers/17-ICLR-LCNet.pdf) | ICLR | PD | - | 241 | | [Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization](https://arxiv.org/abs/1603.06560) | ICLR | PD | - | 242 | | [Hyperparameter Optimization: A Spectral Approach](https://arxiv.org/abs/1706.00764) | NeurIPS-W | Other | [github](https://github.com/callowbird/Harmonica) | 243 | | Learning to Compose Domain-Specific Transformations for Data Augmentation | NeurIPS | - | - | 244 | 245 | ## Previous Venues 246 | 247 | 2012-2016 248 | 249 | | Title | Venue | Type | Code | 250 | |:--------|:--------:|:--------:|:--------:| 251 | | [Speeding up Automatic Hyperparameter Optimization of Deep Neural Networksby Extrapolation of Learning Curves](http://ml.informatik.uni-freiburg.de/papers/15-IJCAI-Extrapolation_of_Learning_Curves.pdf) | IJCAI | PD | [github](https://github.com/automl/pylearningcurvepredictor) | 252 | 253 | ## arXiv 254 | 255 | | Title | Date | Type | Code | 256 | |:--------|:--------:|:--------:|:--------:| 257 | | [NSGA-NET: A Multi-Objective Genetic Algorithm for Neural Architecture Search](https://arxiv.org/pdf/1810.03522.pdf) | 2018.10 | EA | - | 258 | | [Training Frankenstein’s Creature to Stack: HyperTree Architecture Search](https://arxiv.org/pdf/1810.11714.pdf) | 2018.10 | G | - | 259 | | [Population Based Training of Neural Networks](https://arxiv.org/abs/1711.09846) | 2017.11 | EA | [GitHub](https://github.com/MattKleinsmith/pbt) | 260 | | [EmotionNAS: Two-stream Architecture Search for Speech Emotion Recognition](https://arxiv.org/pdf/2203.13617v1.pdf) | 2022.3 | G | - | 261 | | [U-Boost NAS: Utilization-Boosted Differentiable Neural Architecture Search](https://arxiv.org/pdf/2203.12412) | 2022.3 | G | [Github](https://github.com/yuezuegu/UBoostNAS) | 262 | 263 | # Awesome Surveys 264 | 265 | | Title | Venue | Year | Code | 266 | |:--------|:--------:|:--------:|:--------:| 267 | | [A Comprehensive Survey of Neural Architecture Search: Challenges and Solutions](https://arxiv.org/pdf/2006.02903.pdf) | ACM Computing Surveys | 2021 | - | 268 | | [Automated Machine Learning on Graphs: A Survey](https://arxiv.org/pdf/2103.00742v3.pdf) | ICLR-W | 2021 | [GitHub](https://github.com/THUMNLab/AutoGL) | 269 | | [On Hyperparameter Optimization of Machine Learning Algorithms: Theory and Practice](https://arxiv.org/pdf/2007.15745.pdf) | Neurocomputing | 2020 |[github](https://github.com/LiYangHart/Hyperparameter-Optimization-of-Machine-Learning-Algorithms) | 270 | | [AutonoML: Towards an Integrated Framework for Autonomous Machine Learning](https://arxiv.org/pdf/2012.12600.pdf) | arXiv | 2020 | - | 271 | | [Automated Machine Learning](https://link.springer.com/book/10.1007/978-3-030-05318-5) | Springer Book | 2019 | - | 272 | | [Neural architecture search: A survey](http://www.jmlr.org/papers/volume20/18-598/18-598.pdf) | JMLR | 2019 | - | 273 | | [AutoML: A Survey of the State-of-the-Art](https://arxiv.org/pdf/1908.00709.pdf) | arXiv | 2019 | [GitHub](https://github.com/marsggbo/automl_a_survey_of_state_of_the_art) | 274 | | [A Survey on Neural Architecture Search](https://arxiv.org/pdf/1905.01392.pdf) | arXiv | 2019 | - | 275 | | [Taking human out of learning applications: A survey on automated machine learning](https://arxiv.org/pdf/1810.13306.pdf) | arXiv | 2018 | - | 276 | | [IoT Data Analytics in Dynamic Environments: From An Automated Machine Learning Perspective](https://arxiv.org/pdf/2209.08018.pdf) | Engineering Applications of Artificial Intelligence | 2022 |[github](https://github.com/Western-OC2-Lab/AutoML-Implementation-for-Static-and-Dynamic-Data-Analytics) | 277 | -------------------------------------------------------------------------------- /awesome_autodl/__init__.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2021.09 # 3 | ##################################################################################### 4 | # Automated Deep Learning: Neural Architecture Search Is Not the End, arXiv 2021.12 # 5 | ##################################################################################### 6 | # This package is used to analyze the AutoDL-related papers. More detailed reports # 7 | # can be found in the above paper. # 8 | ##################################################################################### 9 | from pathlib import Path 10 | from collections import OrderedDict 11 | 12 | 13 | def version(): 14 | versions = ["v0.1"] # 2021.09.03 15 | versions = ["v0.2"] # 2021.09.04 16 | versions = ["v0.3"] # 2022.01.17 17 | versions = ["v1.0"] # 2022.01.20 18 | versions = ["v1.1"] # 2022.01.21 19 | versions = ["v1.2"] # 2022.03.20 20 | versions = ["v1.3"] # 2022.03.27 21 | return versions[-1] 22 | 23 | 24 | def autodl_topic2file(): 25 | topic2file = OrderedDict() 26 | topic2file["Automated Problem Formulation"] = "Automated_Problem_Formulation.yaml" 27 | topic2file["Automated Data Engineering"] = "Automated_Data_Engineering.yaml" 28 | topic2file["Neural Architecture Search"] = "Neural_Architecture_Search.yaml" 29 | topic2file["Hyperparameter Optimization"] = "Hyperparameter_Optimization.yaml" 30 | topic2file["Automated Deployment"] = "Automated_Deployment.yaml" 31 | topic2file["Automated Maintenance"] = "Automated_Maintenance.yaml" 32 | return topic2file 33 | 34 | 35 | def root(): 36 | return Path(__file__).parent 37 | 38 | 39 | def get_data_dir(): 40 | return root() / "raw_data" 41 | 42 | 43 | def get_bib_abbrv_file(): 44 | return get_data_dir() / "abbrv.bib" 45 | 46 | 47 | def autodl_topic2path(): 48 | topic2file = autodl_topic2file() 49 | topic2path = OrderedDict() 50 | xdir = get_data_dir() / "papers" 51 | for topic, file_name in topic2file.items(): 52 | topic2path[topic] = xdir / file_name 53 | if not topic2path[topic].exists(): 54 | ValueError(f"Can not find {topic} at {topic2path[topic]}") 55 | return topic2path 56 | 57 | 58 | def autodl_topic2papers(): 59 | from awesome_autodl.utils import load_yaml, dump_yaml 60 | from awesome_autodl.data_cls import AutoDLpaper 61 | 62 | topic2path = autodl_topic2path() 63 | topic2papers = OrderedDict() 64 | for topic, xpath in topic2path.items(): 65 | if not xpath.exists(): 66 | ValueError(f"Can not find {topic} at {xpath}.") 67 | papers = [] 68 | raw_data = load_yaml(xpath) 69 | assert isinstance( 70 | raw_data, (list, tuple) 71 | ), f"invalid type of raw data: {type(raw_data)}" 72 | for per_data in raw_data: 73 | papers.append(AutoDLpaper(per_data)) 74 | topic2papers[topic] = papers 75 | print(f"Load {topic} completed with {len(papers)} papers.") 76 | return topic2papers 77 | 78 | 79 | def get_bib_abbrv_obj(): 80 | from awesome_autodl.data_cls import BibAbbreviations 81 | 82 | xfile = str(get_bib_abbrv_file()) 83 | return BibAbbreviations(xfile) 84 | -------------------------------------------------------------------------------- /awesome_autodl/bins/__init__.py: -------------------------------------------------------------------------------- 1 | ################################################## 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2021 # 3 | ################################################## 4 | -------------------------------------------------------------------------------- /awesome_autodl/bins/list_email.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.01 # 3 | ######################################################################################## 4 | # python -m awesome_autodl.bins.list_email # 5 | ######################################################################################## 6 | import argparse 7 | from collections import OrderedDict, defaultdict 8 | from awesome_autodl import autodl_topic2papers 9 | 10 | 11 | def generate_email(author, counts): 12 | message = ( 13 | f"Dear {author}," 14 | + "\n\n" 15 | + "We at the UTS' Complex Adaptive Systems Lab, part of the Data Science Institute in Sydney, Australia, have recently compiled a comprehensive conceptual review of automated deep learning (AutoDL), see initial version at https://arxiv.org/abs/2112.09245." 16 | + "\n\n" 17 | + "The intent of this review is to establish a broader snapshot of the AutoDL field beyond just neural architecture search (NAS), as well as motivate a discussion on how best to evaluate the quality of AutoDL research." 18 | + "\n\n" 19 | ) 20 | if counts >= 3: 21 | message += f"In compiling this review, we have found more than {counts} of your works (referenced in the review) to be significant for informing some of the content, and we hope you find the paper of interest." 22 | else: 23 | message += "In compiling this review, we have found your work (referenced in the review) to be significant for informing some of the content, and we hope you find the paper of interest." 24 | message += ( 25 | "\n\n" 26 | + "We thus welcome any feedback on the accuracy of our coverage - particularly with regard to your referenced contributions - and invite a broader discussion, if desired." 27 | + "\n\n" 28 | + "If you have moved on from this research area or are otherwise uninterested, we apologise for any inconvenience." 29 | + "\n\n" 30 | + "Best regards," 31 | + "\n" 32 | + "Xuanyi Dong, David J. Kedziora, Kaska Musial and Bogdan Gabrys" 33 | ) 34 | return message 35 | 36 | 37 | if __name__ == "__main__": 38 | parser = argparse.ArgumentParser("Analysis the AutoDL papers.") 39 | parser.add_argument( 40 | "--output_file", 41 | type=str, 42 | help="The path to save the final directory.", 43 | ) 44 | args = parser.parse_args() 45 | 46 | author_email_title = [] 47 | topic2papers = autodl_topic2papers() 48 | for topic, papers in topic2papers.items(): 49 | # print(f'Collect {len(papers)} papers for "{topic}"') 50 | for paper in papers: 51 | if not paper.discussed: 52 | continue 53 | for author, email in paper.author_email.items(): 54 | # print(f"{author:25s}, {email:15s} : {paper.title}") 55 | assert email is not None, "f{paper} has a None value for email" 56 | author_email_title.append((author, email.lower(), paper.title)) 57 | # print(f"There are {len(author_email_title)} items in total.") 58 | 59 | author2email = OrderedDict() 60 | author2counts = defaultdict(lambda: 0) 61 | for author, email, title in author_email_title: 62 | if author in author2email: 63 | assert ( 64 | author2email[author] == email 65 | ), f"{author} : {author2email[author]} vs {email}" 66 | author2email[author] = email 67 | author2counts[author] += 1 68 | # print(f"There are {len(author2email)} unique authors.") 69 | 70 | for author, email in author2email.items(): 71 | message = generate_email(author, author2counts[author]) 72 | print(f"\n\nemail:to:{email}") 73 | print(f"message:\n{message}") 74 | -------------------------------------------------------------------------------- /awesome_autodl/bins/list_email_FIE_202203.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.03 # 3 | ######################################################################################## 4 | # python -m awesome_autodl.bins.list_email_FIE_202203 # 5 | ######################################################################################## 6 | import argparse 7 | from collections import OrderedDict, defaultdict 8 | from awesome_autodl import autodl_topic2papers 9 | from awesome_autodl.utils import email_old_to_new_202203 10 | from awesome_autodl.bins.list_email import generate_email 11 | 12 | 13 | if __name__ == "__main__": 14 | parser = argparse.ArgumentParser("Analysis the AutoDL papers.") 15 | parser.add_argument( 16 | "--output_file", 17 | type=str, 18 | help="The path to save the final directory.", 19 | ) 20 | args = parser.parse_args() 21 | 22 | author_email_title = [] 23 | topic2papers = autodl_topic2papers() 24 | for topic, papers in topic2papers.items(): 25 | # print(f'Collect {len(papers)} papers for "{topic}"') 26 | for paper in papers: 27 | if not paper.discussed: 28 | continue 29 | for author, email in paper.author_email.items(): 30 | # print(f"{author:25s}, {email:15s} : {paper.title}") 31 | assert email is not None, "f{paper} has a None value for email" 32 | author_email_title.append((author, email.lower(), paper.title)) 33 | # print(f"There are {len(author_email_title)} items in total.") 34 | 35 | author2email = OrderedDict() 36 | author2counts = defaultdict(lambda: 0) 37 | for author, email, title in author_email_title: 38 | if author in author2email: 39 | assert ( 40 | author2email[author] == email 41 | ), f"{author} : {author2email[author]} vs {email}" 42 | if ( 43 | email in email_old_to_new_202203 44 | and email_old_to_new_202203[email] is not None 45 | ): 46 | author2email[author] = email_old_to_new_202203[email] 47 | author2counts[author] += 1 48 | print( 49 | f"During fixing the invalid email issue, there are {len(author2email)} unique authors." 50 | ) 51 | 52 | for author, email in author2email.items(): 53 | message = generate_email(author, author2counts[author]) 54 | print(f"\n\nemail:to:{email}") 55 | print(f"message:\n{message}") 56 | -------------------------------------------------------------------------------- /awesome_autodl/bins/show_infos.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2020.08 # 3 | ######################################################################################## 4 | # python -m awesome_autodl.bins.show_infos --root awesome_autodl/raw_data 5 | ######################################################################################## 6 | import argparse 7 | from pathlib import Path 8 | from awesome_autodl.utils import load_yaml 9 | from awesome_autodl.utils import dump_yaml 10 | from awesome_autodl.utils import check_and_sort_by_date 11 | from awesome_autodl.utils import filter_ele_w_value 12 | 13 | 14 | name2file = { 15 | "Automated Problem Formulation": "Automated_Problem_Formulation.yaml", 16 | "Automated Data Engineering": "Automated_Data_Engineering.yaml", 17 | "Neural Architecture Search": "Neural_Architecture_Search.yaml", 18 | "Hyperparameter Optimization": "Hyperparameter_Optimization.yaml", 19 | "Automated Deployment": "Automated_Deployment.yaml", 20 | "Automated Maintenance": "Automated_Maintenance.yaml", 21 | } 22 | 23 | # add abbreviation 24 | name2file["ADE"] = name2file["Automated Data Engineering"] 25 | name2file["NAS"] = name2file["Neural Architecture Search"] 26 | name2file["HPO"] = name2file["Hyperparameter Optimization"] 27 | name2file["AD"] = name2file["Automated Deployment"] 28 | name2file["AM"] = name2file["Automated Maintenance"] 29 | 30 | 31 | def scale(xlist, scale): 32 | for x in xlist: 33 | print(int(x * scale)) 34 | 35 | 36 | def show_nas(root): 37 | abbrv = load_yaml(root / "abbrv.yaml") 38 | topic_path = root / "papers" / name2file["NAS"] 39 | assert topic_path.exists(), f"Did not find {topic_path}" 40 | print(f"Process NAS topic from {topic_path}") 41 | data = load_yaml(topic_path) 42 | data = check_and_sort_by_date(data) 43 | print(f"Find {len(data)} papers for NAS") 44 | # Search Space Analysis 45 | nasnet_like_search_space_papers = filter_ele_w_value(data, "search_space", "NASNet") 46 | print( 47 | f"NASNet-like search space: {len(nasnet_like_search_space_papers)}/{len(data)} : {len(nasnet_like_search_space_papers)*100./len(data):.3f}" 48 | ) 49 | mbconv_based_search_space_papers = filter_ele_w_value( 50 | data, "search_space", "MBConv" 51 | ) 52 | print( 53 | f"MBConv-based search space: {len(mbconv_based_search_space_papers)}/{len(data)} : {len(mbconv_based_search_space_papers)*100./len(data):.3f}" 54 | ) 55 | size_based_search_space_papers = filter_ele_w_value(data, "search_space", "size") 56 | print( 57 | f"Size-related search space: {len(size_based_search_space_papers)}/{len(data)} : {len(size_based_search_space_papers)*100./len(data):.3f}" 58 | ) 59 | ratio_keys = ["NASNet", "MBConv", "size"] 60 | ratios = [len(filter_ele_w_value(data, "search_space", key)) for key in ratio_keys] 61 | ratios = [float(ratio) / len(data) for ratio in ratios] 62 | ratios.append(1 - sum(ratios)) 63 | # Search Strategy 64 | ratio_keys = ["Differential", "RL", "Evolution"] 65 | ratios = [ 66 | len(filter_ele_w_value(data, "search_strategy", key)) for key in ratio_keys 67 | ] 68 | ratios = [float(ratio) / len(data) for ratio in ratios] 69 | ratios.append(1 - sum(ratios)) 70 | # scale(ratios, 500) 71 | 72 | # Efficient candidate evaluation 73 | weight_sharing_papers = ( 74 | filter_ele_w_value(data, "eval_boost", "weight sharing") 75 | + filter_ele_w_value(data, "eval_boost", "HyperNet") 76 | + filter_ele_w_value(data, "eval_boost", "weight i") 77 | + filter_ele_w_value(data, "eval_boost", "Net2Net") 78 | ) 79 | print( 80 | f"Weight sharing: {len(weight_sharing_papers)}/{len(data)} : {len(weight_sharing_papers)*100./len(data):.3f}" 81 | ) 82 | 83 | 84 | def show_all(root): 85 | 86 | name2data = dict() 87 | total = 0 88 | for name, file in name2file.items(): 89 | if len(name) < 10: 90 | topic_path = root / "papers" / file 91 | data = load_yaml(topic_path) 92 | name2data[name] = data 93 | total += len(data) 94 | for name, data in name2data.items(): 95 | print(f"{name:5s}: {len(data)}/{total} = {len(data)*1./total:.3f}") 96 | 97 | 98 | if __name__ == "__main__": 99 | parser = argparse.ArgumentParser("Analysis the AutoDL papers.") 100 | parser.add_argument( 101 | "--root", 102 | type=str, 103 | required=True, 104 | help="The path for the data directory.", 105 | ) 106 | args = parser.parse_args() 107 | root = Path(args.root) 108 | show_all(root) 109 | # show_nas(root) 110 | -------------------------------------------------------------------------------- /awesome_autodl/bins/statistics.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2020.08 # 3 | ######################################################################################## 4 | # python -m awesome_autodl.bins.statistics --topic "ADE" --root awesome_autodl/raw_data 5 | # python -m awesome_autodl.bins.statistics --topic "NAS" --root awesome_autodl/raw_data 6 | # python -m awesome_autodl.bins.statistics --topic "HPO" --root awesome_autodl/raw_data 7 | # python -m awesome_autodl.bins.statistics --topic "AD" --root awesome_autodl/raw_data 8 | # python -m awesome_autodl.bins.statistics --topic "AM" --root awesome_autodl/raw_data 9 | ######################################################################################## 10 | import argparse 11 | from pathlib import Path 12 | from awesome_autodl.utils import load_yaml 13 | from awesome_autodl.utils import dump_yaml 14 | from awesome_autodl.utils import check_and_sort_by_date 15 | 16 | 17 | name2file = { 18 | "Automated Problem Formulation": "Automated_Problem_Formulation.yaml", 19 | "Automated Data Engineering": "Automated_Data_Engineering.yaml", 20 | "Neural Architecture Search": "Neural_Architecture_Search.yaml", 21 | "Hyperparameter Optimization": "Hyperparameter_Optimization.yaml", 22 | "Automated Deployment": "Automated_Deployment.yaml", 23 | "Automated Maintenance": "Automated_Maintenance.yaml", 24 | } 25 | 26 | # add abbreviation 27 | name2file["ADE"] = name2file["Automated Data Engineering"] 28 | name2file["NAS"] = name2file["Neural Architecture Search"] 29 | name2file["HPO"] = name2file["Hyperparameter Optimization"] 30 | name2file["AD"] = name2file["Automated Deployment"] 31 | name2file["AM"] = name2file["Automated Maintenance"] 32 | 33 | 34 | def main(root, topic): 35 | abbrv = load_yaml(root / "abbrv.yaml") 36 | topic_path = root / "papers" / name2file[topic] 37 | assert topic_path.exists(), f"Did not find {topic_path}" 38 | print(f"Process {topic_path}") 39 | data = load_yaml(topic_path) 40 | data = check_and_sort_by_date(data) 41 | print(f"Find {len(data)} papers for {topic}") 42 | dump_yaml(data, path=topic_path) 43 | 44 | 45 | if __name__ == "__main__": 46 | parser = argparse.ArgumentParser("Analysis the AutoDL papers.") 47 | parser.add_argument( 48 | "--root", 49 | type=str, 50 | required=True, 51 | help="The path for the data directory.", 52 | ) 53 | parser.add_argument( 54 | "--topic", 55 | type=str, 56 | required=True, 57 | choices=list(name2file.keys()), 58 | help="Choose the AutoDL sub-topic.", 59 | ) 60 | args = parser.parse_args() 61 | main(Path(args.root), args.topic) 62 | -------------------------------------------------------------------------------- /awesome_autodl/data_cls/__init__.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.01 # 3 | ##################################################### 4 | from awesome_autodl.data_cls.abbrv import BibAbbreviations 5 | from awesome_autodl.data_cls.paper import AutoDLpaper 6 | -------------------------------------------------------------------------------- /awesome_autodl/data_cls/abbrv.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.01 # 3 | ##################################################### 4 | import re 5 | import os 6 | 7 | 8 | class BibAbbreviations: 9 | """A class to maintain the paper venue abbreviation.""" 10 | 11 | def __init__(self, xfile): 12 | xfile = str(xfile) 13 | assert os.path.isfile(xfile) 14 | with open(xfile) as f: 15 | lines = f.readlines() 16 | lines = [x.strip() for x in lines] 17 | 18 | pattern = r'^@STRING\{([A-Z]|[a-z]|\s|_)*=(\s)*"([A-Z]|[a-z]|\s|\(|\)|\/|\,|\:|\-)*"\}$' 19 | self.prog = re.compile(pattern) 20 | self.abbrv2str = dict() 21 | for index, line in enumerate(lines): 22 | if line: 23 | if self.prog.match(line) is None: 24 | raise ValueError(f"Incorrect line [{index}]: {line}") 25 | key, value = self.decode(line) 26 | if key in self.abbrv2str: 27 | raise ValueError(f"Already defined {key}") 28 | self.abbrv2str[key] = value 29 | 30 | def __contains__(self, key): 31 | return key in self.abbrv2str 32 | 33 | def __getitem__(self, key): 34 | return self.abbrv2str[key] 35 | 36 | def __len__(self): 37 | return len(self.abbrv2str) 38 | 39 | def keys(self, sort=False): 40 | keys = list(self.abbrv2str.keys()) 41 | if sort: 42 | keys = sorted(keys) 43 | return keys 44 | 45 | def decode(self, line): 46 | head = "@STRING{" 47 | assert len(line) > len(head) 48 | assert line[: len(head)] == head and line[-1] == "}" 49 | line = line[len(head) : -1] 50 | key, value = line.split("=") 51 | key, value = key.strip(" "), value.strip(" ") 52 | return key, value 53 | 54 | def __repr__(self): 55 | return f"{self.__class__.__name__}(" + f"{len(self)} abbrev pairs)" 56 | -------------------------------------------------------------------------------- /awesome_autodl/data_cls/paper.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.01 # 3 | ##################################################### 4 | from typing import Dict, Text, Any 5 | from typing import Optional 6 | import re 7 | 8 | 9 | class AutoDLpaper: 10 | """A class to maintain the paper attribute.""" 11 | 12 | # required fields, must include 13 | title: Text = None 14 | venue: Text = None 15 | venue_date: Text = None # yyyy.mm 16 | online_date: Text = None # yyyy.mm 17 | contacts: Optional[Dict[Text, Text]] = None # a pair of name and email 18 | 19 | required_fields = ("title", "venue", "venue_date", "online_date", "contacts") 20 | 21 | # none-required fields 22 | search_space: Text = None 23 | search_strategy: Text = None 24 | candidate_evaluation: Text = None 25 | 26 | autodl_aspect_fields = ("search_space", "search_strategy", "candidate_evaluation") 27 | 28 | # non-required fileds 29 | links: Optional[Dict[Text, Text]] = None 30 | 31 | # misc 32 | discussed: bool = None 33 | misc: Text = None 34 | 35 | def __init__(self, data: Dict[Text, Any]): 36 | self.check_raw_data(data) 37 | self.reset_value(data) 38 | 39 | def reset_value(self, data): 40 | # set the basic value 41 | for field in self.required_fields[:-1]: 42 | if not isinstance(data[field], str): 43 | raise TypeError( 44 | f"Expect {field} is str instead of {type(data[field])}." 45 | ) 46 | self.title = data["title"] 47 | self.venue = data["venue"] 48 | date_pattern = re.compile("^([0-9]){4}.([0-9]){2}$") 49 | if date_pattern.match(data["venue_date"]) is None: 50 | raise ValueError(f"Invalid venue date ({data['venue_date']}) :: {data}") 51 | if date_pattern.match(data["online_date"]) is None: 52 | raise ValueError(f"Invalid online date ({data['online_date']}) :: {data}") 53 | self.venue_date = data["venue_date"] 54 | self.online_date = data["online_date"] 55 | 56 | # set the contact information 57 | if data["contacts"] is not None: 58 | assert isinstance(data["contacts"], dict) 59 | self.contacts = data["contacts"] 60 | # TODO(xuanyidong): check the email address 61 | # regex = r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b' 62 | 63 | # set the AutoDL aspect fields 64 | for field in self.autodl_aspect_fields: 65 | if data[field] is None: 66 | continue 67 | assert isinstance(data[field], str) 68 | setattr(self, field, data[field]) 69 | 70 | # set the misc info 71 | if "discussed" in data: 72 | assert isinstance( 73 | data["discussed"], bool 74 | ), f"The discussed field must be a bool instead of a {type(data['discussed'])}" 75 | self.discussed = data["discussed"] 76 | if "misc" in data and data["misc"] is not None: 77 | assert isinstance( 78 | data["misc"], str 79 | ), f"The misc field must be a str instead of a {type(data['discussed'])}" 80 | self.misc = data["misc"] 81 | 82 | def check_raw_data(self, data): 83 | """Check whether the necessary field is included in `data`.""" 84 | if not isinstance(data, dict): 85 | raise TypeError(f"Expect data to be a dict instead of {type(data)}") 86 | for field in self.required_fields: 87 | if field not in data: 88 | raise ValueError(f"Missing {field} in {data}") 89 | for field in self.autodl_aspect_fields: 90 | if field not in data: 91 | raise ValueError( 92 | f"Missing {field} in {data}" 93 | + "Please leave this field as blank if you are not sure about it." 94 | ) 95 | all_fields = ( 96 | list(self.required_fields) 97 | + list(self.autodl_aspect_fields) 98 | + ["discussed", "misc", "links"] 99 | ) 100 | for key in data.keys(): 101 | if key not in all_fields: 102 | raise ValueError(f"Find unexpected field: {key} in {data}") 103 | 104 | @property 105 | def author_email(self): 106 | if self.contacts is None: 107 | return dict() 108 | else: 109 | return self.contacts 110 | 111 | def __repr__(self): 112 | return f"{self.__class__.__name__}({self.title})" 113 | -------------------------------------------------------------------------------- /awesome_autodl/raw_data/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/D-X-Y/Awesome-AutoDL/9fb9fe201d6c2b325ee7df7c12656fe9e2971ce4/awesome_autodl/raw_data/__init__.py -------------------------------------------------------------------------------- /awesome_autodl/raw_data/abbrv.bib: -------------------------------------------------------------------------------- 1 | @STRING{IEEE_J_IP = "IEEE Transactions on Image Processing (TIP)"} 2 | @STRING{IEEE_J_PAMI = "IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)"} 3 | @STRING{nature = "Nature"} 4 | @STRING{IEEE_TSMC = "IEEE Transactions on Systems, Man, and Cybernetics: Systems (TSMC)"} 5 | @STRING{IEEE_J_TASE = "IEEE Transactions on Automation Science and Engineering (TASE)"} 6 | @STRING{acm_comp_sur = "ACM Computing Surveys (CSUR)"} 7 | @STRING{evo_comp = "Evolutionary Computation"} 8 | @STRING{NeurCom = "Neural Computation"} 9 | @STRING{JAIR = "Journal of Artificial Intelligence Research (JAIR)"} 10 | @STRING{IEEE_J_TCAD = "IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD)"} 11 | @STRING{IEEE_NNSP = "IEEE Workshop on Neural Networks for Signal Processing (NNSP)"} 12 | @STRING{NN_ToT = "Neural Networks: Tricks of the Trade"} 13 | @STRING{ISLPED = "IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED)"} 14 | @STRING{TACO = "ACM Transactions on Architecture and Code Optimization (TACO)"} 15 | @STRING{eurosys = "Proceedings of the EuroSys Conference (EuroSys)"} 16 | @STRING{CogSci = "Cognitive Science Society (CogSci)"} 17 | @STRING{AAMAS = "Proceedings of the International Conference on Autonomous Agents and Multiagent Systems (AAMAS)"} 18 | @STRING{Neurocomputing = "Neurocomputing"} 19 | 20 | @STRING{IJCV = "International Journal of Computer Vision (IJCV)"} 21 | @STRING{uai = "The Conference on Uncertainty in Artificial Intelligence (UAI)"} 22 | @STRING{aistats = "The International Conference on Artificial Intelligence and Statistics (AISTATS)"} 23 | @STRING{cvpr = "Proceedings of the IEEE Conference Computer Vision Pattern Recognition (CVPR)"} 24 | @STRING{cvpr_w = "Proceedings of the IEEE Conference Computer Vision Pattern Recognition (CVPR) Workshop"} 25 | @STRING{iclr = "International Conference on Learning Representations (ICLR)"} 26 | @STRING{iclr_w = "International Conference on Learning Representations (ICLR) Workshop"} 27 | @STRING{iccv = "Proceedings of the IEEE International Conference Computer Vision (ICCV)"} 28 | @STRING{iccv_w = "Proceedings of the IEEE International Conference Computer Vision (ICCV) Workshop"} 29 | @STRING{icpr = "Proceedings of the IEEE International Conference Pattern Recognition (ICPR)"} 30 | @STRING{nips = "Proceedings of the International Conference on Neural Information Processing Systems (NeurIPS)"} 31 | @STRING{nips_w = "Proceedings of the International Conference on Neural Information Processing Systems (NeurIPS) Workshop"} 32 | @STRING{eccv = "Proceedings of the European Conference on Computer Vision (ECCV)"} 33 | @STRING{eccv_w = "Proceedings of the European Conference on Computer Vision (ECCV) Workshop"} 34 | @STRING{ecai_w = "The European Conference on Artificial Intelligence (ECAI) Workshop"} 35 | @STRING{ijcai = "International Joint Conferences on Artificial Intelligence (IJCAI)"} 36 | @STRING{miccai = "Proceedings of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI)"} 37 | @STRING{icml = "The International Conference on Machine Learning (ICML)"} 38 | @STRING{icml_w = "The International Conference on Machine Learning (ICML) Workshop"} 39 | @STRING{aaai = "AAAI Conference on Artificial Intelligence (AAAI)"} 40 | @STRING{jmlr = "Journal of Machine Learning Research (JMLR)"} 41 | @STRING{sigkdd = "Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD)"} 42 | @STRING{bmvc = "Proceedings of the British Machine Vision Conference (BMVC)"} 43 | @STRING{www = "Proceedings of the International Conference on World Wide Web (WWW)"} 44 | @STRING{mascots = "IEEE International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS)"} 45 | @STRING{emnlp = "Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP)"} 46 | @STRING{wacv = "Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV)"} 47 | @STRING{dac = "The ACM/ESDA/IEEE Design Automation Conference (DAC)"} 48 | @STRING{acl = "The Annual Meeting of the Association for Computational Linguistics"} 49 | @STRING{lion = "Proceedings of the International Conference on Learning and Intelligent Optimization (LION)"} 50 | @STRING{tkde = "IEEE Transactions on Knowledge and Data Engineering (TKDE)"} 51 | @STRING{mlsys = "The Conference on Machine Learning and Systems (MLSys)"} 52 | @STRING{pricai = "The Pacific Rim International Conferences on Artificial Intelligence (PRICAI)"} 53 | @STRING{icann = "Proceedings of the International Conference on Artificial Neural Networks (ICANN)"} 54 | @STRING{ijcnn = "International Joint Conference on Neural Network (IJCNN)"} 55 | @STRING{iccad = "The IEEE/ACM International Conference on Computer-Aided Design (ICCAD)"} 56 | @STRING{ECML_PKDD = "The Joint European Conference on Machine Learning and Knowledge Discovery in Databases (ECML PKDD)"} 57 | @STRING{osdi = "The USENIX Symposium on Operating Systems Design and Implementation (OSDI)"} 58 | 59 | @STRING{automated_machine_learning = "Automated Machine Learning"} 60 | @STRING{arXiv = "arXiv"} -------------------------------------------------------------------------------- /awesome_autodl/raw_data/abbrv.yaml: -------------------------------------------------------------------------------- 1 | application: 2 | CLS: "classification" 3 | DET: "object detection" 4 | REG: "regression" 5 | SEG: "segmentation" 6 | search_strategy: 7 | RL: "Reinforcement Learning" 8 | Evolution: "Evolutionary Algorithm" 9 | Random: "Random Search" 10 | Differential: "" 11 | BayesOpt: "Bayesian Optimization" 12 | Heuristic: "" 13 | -------------------------------------------------------------------------------- /awesome_autodl/raw_data/papers/Automated_Data_Engineering.yaml: -------------------------------------------------------------------------------- 1 | - title: "AutoAugment: Learning Augmentation Policies from Data" 2 | venue: "cvpr" 3 | venue_date: "2019.06" 4 | online_date: "2018.05" 5 | contacts: {"Ekin D. Cubuk": "cubuk@google.com", "Barret Zoph": "barretzoph@google.com"} 6 | search_space: 7 | search_strategy: 8 | candidate_evaluation: 9 | discussed: true 10 | misc: 11 | - title: "RandAugment: Practical Automated Data Augmentation with a Reduced Search Space" 12 | venue: "nips" 13 | venue_date: "2020.12" 14 | online_date: "2019.09" 15 | contacts: {"Ekin D. Cubuk": "cubuk@google.com", "Barret Zoph": "barretzoph@google.com"} 16 | search_space: 17 | search_strategy: 18 | candidate_evaluation: 19 | discussed: true 20 | misc: 21 | - title: "Teacher Supervises Students How to Learn from Partially Labeled Images for Facial Landmark Detection" 22 | venue: "iccv" 23 | venue_date: "2019.10" 24 | online_date: "2019.08" 25 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 26 | search_space: 27 | search_strategy: 28 | candidate_evaluation: 29 | discussed: true 30 | misc: 31 | - title: "Learning to teach" 32 | venue: "iclr" 33 | venue_date: "2018.04" 34 | online_date: "2018.04" 35 | contacts: {"Yang Fan": "fyabc@mail.ustc.edu.cn", "Fei Tian": "fetia@microsoft.com"} 36 | search_space: 37 | search_strategy: 38 | candidate_evaluation: 39 | discussed: true 40 | misc: 41 | - title: "Automated data cleansing through meta-learning" 42 | venue: "aaai" 43 | venue_date: "2017.02" 44 | online_date: "2017.02" 45 | contacts: {"Ian Gemp": "imgemp@cs.umass.edu"} 46 | search_space: 47 | search_strategy: 48 | candidate_evaluation: 49 | discussed: true 50 | misc: 51 | - title: "Learning Active Learning from Data" 52 | venue: "nips" 53 | venue_date: "2017.12" 54 | online_date: "2017.03" 55 | contacts: {"Ksenia Konyushkova": "ksenia.konyushkova@epfl.ch", "Pascal Fua": "pascal.fua@epfl.ch"} 56 | search_space: 57 | search_strategy: 58 | candidate_evaluation: 59 | discussed: true 60 | misc: 61 | - title: "DADA: Differentiable Automatic Data Augmentation" 62 | venue: "eccv" 63 | venue_date: "2020.08" 64 | online_date: "2020.03" 65 | contacts: {"Yongtao Wang": "wyt@pku.edu.cn"} 66 | search_space: 67 | search_strategy: 68 | candidate_evaluation: 69 | discussed: true 70 | misc: 71 | - title: "Fast autoaugment" 72 | venue: "nips" 73 | venue_date: "2019.12" 74 | online_date: "2019.05" 75 | contacts: {"Sungbin Lim": "sungbin.lim@kakaobrain.com", "Ildoo Kim": "ildoo.kim@kakaobrain.com"} 76 | search_space: 77 | search_strategy: 78 | candidate_evaluation: 79 | discussed: true 80 | misc: 81 | - title: "Automatically Learning Data Augmentation Policies for Dialogue Tasks" 82 | venue: "emnlp" 83 | venue_date: "2019.11" 84 | online_date: "2019.09" 85 | contacts: {"Tong Niu": "tongn@cs.unc.edu", "Mohit Bansal": "mbansal@cs.unc.edu"} 86 | search_space: 87 | search_strategy: 88 | candidate_evaluation: 89 | discussed: true 90 | misc: 91 | - title: "Learning to Reweight Examples for Robust Deep Learning" 92 | venue: "icml" 93 | venue_date: "2018.07" 94 | online_date: "2018.03" 95 | contacts: {"Mengye Ren": "mren@cs.toronto.edu"} 96 | search_space: 97 | search_strategy: 98 | candidate_evaluation: 99 | discussed: true 100 | links: {"paper": "https://arxiv.org/abs/1803.09050", "code": "https://github.com/uber-research/learning-to-reweight-examples"} 101 | misc: 102 | - title: "Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting" 103 | venue: "nips" 104 | venue_date: "2019.12" 105 | online_date: "2019.02" 106 | contacts: {"Jun Shu": "xjtushujun@gmail.com"} 107 | search_space: 108 | search_strategy: 109 | candidate_evaluation: 110 | discussed: true 111 | links: {"paper": "https://arxiv.org/abs/1902.07379", "code": "https://github.com/xjtushujun/meta-weight-net"} 112 | misc: 113 | - title: "Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data" 114 | venue: "icml" 115 | venue_date: "2020.07" 116 | online_date: "2019.12" 117 | contacts: {"Felipe Petroski Such": "fe-lipe.such@gmail.com", "Jeff Clune": "jeffclune@openai.com"} 118 | search_space: 119 | search_strategy: 120 | candidate_evaluation: 121 | discussed: true 122 | links: {"paper": "https://arxiv.org/abs/1912.07768", "code": "https://github.com/uber-research/gtn"} 123 | misc: 124 | - title: "Adaptive Preprocessing for Streaming Data" 125 | venue: "tkde" 126 | venue_date: "2012.07" 127 | online_date: "2012.07" 128 | contacts: {"Indre Zliobaite": "izliobaite@bournemouth.ac.uk", "Bogdan Gabrys": "Bogdan.Gabrys@uts.edu.au"} 129 | search_space: 130 | search_strategy: 131 | candidate_evaluation: 132 | discussed: true 133 | links: {"paper": "https://ieeexplore.ieee.org/document/6247432"} 134 | misc: 135 | - title: "Learning Data Augmentation Strategies for Object Detection" 136 | venue: "eccv" 137 | venue_date: "2020.08" 138 | online_date: "2019.06" 139 | contacts: {"Barret Zoph": "barretzoph@google.com"} 140 | search_space: 141 | search_strategy: 142 | candidate_evaluation: 143 | discussed: true 144 | links: {"paper": "https://arxiv.org/abs/1906.11172", "code": "https://github.com/tensorflow/tpu/tree/master/models/official/detection"} 145 | misc: -------------------------------------------------------------------------------- /awesome_autodl/raw_data/papers/Automated_Deployment.yaml: -------------------------------------------------------------------------------- 1 | - title: "{TVM}: An Automated {End-to-End} Optimizing Compiler for Deep Learning" 2 | venue: "osdi" 3 | venue_date: "2018.10" 4 | online_date: "2018.10" 5 | contacts: {} 6 | search_space: 7 | search_strategy: 8 | candidate_evaluation: 9 | discussed: true 10 | misc: 11 | - title: "DANCE: Differentiable Accelerator/Network Co-Exploration" 12 | venue: "dac" 13 | venue_date: "2021.12" 14 | online_date: "2020.09" 15 | contacts: {"Kanghyun Choi": "kanghyun.choi@yonsei.ac.kr", "Jinho Lee": "leejinho@yonsei.ac.kr"} 16 | search_space: 17 | search_strategy: "differentiable" 18 | candidate_evaluation: 19 | discussed: true 20 | misc: 21 | - title: "Hardware/Software Co-Exploration of Neural Architectures" 22 | venue: "IEEE_J_TCAD" 23 | venue_date: "2020.04" 24 | online_date: "2019.07" 25 | contacts: {"Weiwen Jiang": "wjiang2@nd.edu"} 26 | search_space: 27 | search_strategy: 28 | candidate_evaluation: 29 | discussed: true 30 | misc: 31 | - title: "A Hierarchical Model for Device Placement" 32 | venue: "iclr" 33 | venue_date: "2018.04" 34 | online_date: "2018.04" 35 | contacts: {"Azalia Mirhoseini": "azalia@google.com"} 36 | search_space: 37 | search_strategy: 38 | candidate_evaluation: 39 | discussed: true 40 | misc: 41 | - title: "Placeto: Learning Generalizable Device Placement Algorithms for Distributed Machine Learning" 42 | venue: "nips" 43 | venue_date: "2019.12" 44 | online_date: "2019.06" 45 | contacts: {"Ravichandra Addanki": "addanki@mit.edu"} 46 | search_space: 47 | search_strategy: 48 | candidate_evaluation: 49 | discussed: true 50 | misc: 51 | - title: "A graph placement methodology for fast chip design" 52 | venue: "nature" 53 | venue_date: "2021.06" 54 | online_date: "2021.06" 55 | contacts: {"Azalia Mirhoseini": "azalia@google.com"} 56 | search_space: 57 | search_strategy: 58 | candidate_evaluation: 59 | discussed: true 60 | misc: 61 | - title: "Device placement optimization with reinforcement learning" 62 | venue: "icml" 63 | venue_date: "2017.08" 64 | online_date: "2017.06" 65 | contacts: {"Azalia Mirhoseini": "azalia@google.com"} 66 | search_space: 67 | search_strategy: "RL" 68 | candidate_evaluation: 69 | discussed: true 70 | misc: 71 | - title: "Practical Design Space Exploration" 72 | venue: "mascots" 73 | venue_date: "2019.10" 74 | online_date: "2018.10" 75 | contacts: {"Luigi Nardi": "lnardi@stanford.edu", "Kunle Olukotun": "kunle@stanford.edu"} 76 | search_space: 77 | search_strategy: 78 | candidate_evaluation: 79 | discussed: true 80 | misc: 81 | - title: "Optimus: an efficient dynamic resource scheduler for deep learning clusters" 82 | venue: "eurosys" 83 | venue_date: "2018.04" 84 | online_date: "2018.04" 85 | contacts: {} 86 | search_space: 87 | search_strategy: 88 | candidate_evaluation: 89 | discussed: true 90 | links: {"paper": "https://dl.acm.org/doi/10.1145/3190508.3190517"} 91 | misc: 92 | - title: "A case for efficient accelerator design space exploration via Bayesian optimization" 93 | venue: "ISLPED" 94 | venue_date: "2017.08" 95 | online_date: "2017.08" 96 | contacts: {} 97 | search_space: 98 | search_strategy: "BayesOpt" 99 | candidate_evaluation: 100 | discussed: true 101 | links: {"paper": "https://ieeexplore.ieee.org/document/8009208"} 102 | misc: 103 | - title: "Bayesian Optimization for Efficient Accelerator Synthesis" 104 | venue: "TACO" 105 | venue_date: "2020.12" 106 | online_date: "2020.12" 107 | contacts: {"Atefeh Mehrabi": "atefeh.mehrabi@duke.edu"} 108 | search_space: 109 | search_strategy: "BayesOpt" 110 | candidate_evaluation: 111 | discussed: false 112 | links: {"paper": "https://dl.acm.org/doi/10.1145/3427377"} 113 | misc: 114 | - title: "Co-Exploration of Neural Architectures and Heterogeneous ASIC Accelerator Designs Targeting Multiple Tasks" 115 | venue: "dac" 116 | venue_date: "2020.07" 117 | online_date: "2020.02" 118 | contacts: {"Weiwen Jiang": "wjiang2@nd.edu"} 119 | search_space: 120 | search_strategy: 121 | candidate_evaluation: 122 | discussed: true 123 | links: {"paper": "https://arxiv.org/abs/2002.04116"} 124 | misc: 125 | - title: "A reinforcement learning approach to job-shop scheduling" 126 | venue: "ijcai" 127 | venue_date: "1995.08" 128 | online_date: "1995.08" 129 | contacts: {} 130 | search_space: 131 | search_strategy: 132 | candidate_evaluation: 133 | discussed: true 134 | links: {"paper": "https://dl.acm.org/doi/10.5555/1643031.1643044"} 135 | misc: 136 | - title: "Rethinking Co-design of Neural Architectures and Hardware Accelerators" 137 | venue: "mlsys" 138 | venue_date: "2022.08" 139 | online_date: "2021.02" 140 | contacts: {"Yanqi Zhou": "yanqiz@google.com"} 141 | search_space: 142 | search_strategy: 143 | candidate_evaluation: 144 | discussed: true 145 | links: {"paper": "https://arxiv.org/abs/2102.08619"} 146 | misc: -------------------------------------------------------------------------------- /awesome_autodl/raw_data/papers/Automated_Maintenance.yaml: -------------------------------------------------------------------------------- 1 | - title: "Meta-Learning with Adaptive Hyperparameters" 2 | venue: "nips" 3 | venue_date: "2020.12" 4 | online_date: "2020.11" 5 | contacts: {"Sungyong Baik": "dsybaik@snu.ac.kr", "Kyoung Mu Lee": "kyoungmu@snu.ac.kr"} 6 | search_space: 7 | search_strategy: 8 | candidate_evaluation: 9 | discussed: true 10 | misc: 11 | - title: "Adaptation Strategies for Automated Machine Learning on Evolving Data" 12 | venue: "IEEE_J_PAMI" 13 | venue_date: "2021.03" 14 | online_date: "2020.06" 15 | contacts: {"Bilge Celik": "B.Celik.Aydin@tue.nl"} 16 | search_space: 17 | search_strategy: 18 | candidate_evaluation: 19 | discussed: true 20 | misc: 21 | - title: "Sequential Scenario-Specific Meta Learner for Online Recommendation" 22 | venue: "sigkdd" 23 | venue_date: "2019.08" 24 | online_date: "2019.06" 25 | contacts: {"Zhengxiao Du": "duzx16@mails.tsinghua.edu.cn"} 26 | search_space: 27 | search_strategy: 28 | candidate_evaluation: 29 | discussed: true 30 | misc: 31 | - title: "RL2: Fast Reinforcement Learning via Slow Reinforcement Learning" 32 | venue: "arXiv" 33 | venue_date: "2016.11" 34 | online_date: "2016.11" 35 | contacts: {"Yan Duan": "rocky@covariant.ai"} 36 | search_space: 37 | search_strategy: 38 | candidate_evaluation: 39 | discussed: true 40 | misc: 41 | - title: "DACBench: A Benchmark Library for Dynamic Algorithm Configuration" 42 | venue: "ijcai" 43 | venue_date: "2021.08" 44 | online_date: "2021.05" 45 | contacts: {"Theresa Eimer": "eimer@tnt.uni-hannover.de", "Marius Lindauer": "lindauer@tnt.uni-hannover.de"} 46 | search_space: 47 | search_strategy: 48 | candidate_evaluation: 49 | discussed: true 50 | misc: 51 | - title: "Building Upon Neural Models of How Brains Make Minds" 52 | venue: "IEEE_TSMC" 53 | venue_date: "2020.12" 54 | online_date: "2020.12" 55 | contacts: {} 56 | search_space: 57 | search_strategy: 58 | candidate_evaluation: 59 | discussed: true 60 | misc: 61 | - title: "Completely Derandomized Self-Adaptation in Evolution Strategies" 62 | venue: "evo_comp" 63 | venue_date: "2001.06" 64 | online_date: "2001.06" 65 | contacts: {} 66 | search_space: 67 | search_strategy: 68 | candidate_evaluation: 69 | discussed: true 70 | misc: 71 | - title: "Improving Generalization in Meta Reinforcement Learning using Learned Objectives" 72 | venue: "iclr" 73 | venue_date: "2020.04" 74 | online_date: "2019.10" 75 | contacts: {"Louis Kirsch": "louis@idsia.ch", "Sjoerd van Steenkiste": "sjoerd@idsia.ch", "Jurgen Schmidhuber": "juergen@idsia.ch"} 76 | search_space: 77 | search_strategy: 78 | candidate_evaluation: 79 | discussed: true 80 | misc: 81 | - title: "Towards AutoML in the presence of Drift: first results" 82 | venue: "icml_w" 83 | venue_date: "2018.07" 84 | online_date: "2019.07" 85 | contacts: {"Jorge G. Madrid": "jorgegus.93@gmail.com", "Michele Sebag": "michele.sebag@lri.fr"} 86 | search_space: 87 | search_strategy: 88 | candidate_evaluation: 89 | discussed: true 90 | misc: "The AutoML@ICML 2018 website failed, thus the online date is used as the arxiv date." 91 | - title: "Progressive Neural Networks" 92 | venue: "arXiv" 93 | venue_date: "2016.06" 94 | online_date: "2016.06" 95 | contacts: {"Andrei A. Rusu": "andreirusu@google.com"} 96 | search_space: 97 | search_strategy: 98 | candidate_evaluation: 99 | discussed: true 100 | links: {"paper": "https://arxiv.org/abs/1606.04671"} 101 | misc: 102 | - title: "Memory-based Parameter Adaptation" 103 | venue: "iclr" 104 | venue_date: "2018.05" 105 | online_date: "2018.02" 106 | contacts: {"Pablo Sprechmann": "psprechmann@google.com", "Siddhant M. Jayakumar": "sidmj@google.com"} 107 | search_space: 108 | search_strategy: 109 | candidate_evaluation: 110 | discussed: true 111 | links: {"paper": "https://arxiv.org/abs/1802.10542"} 112 | misc: 113 | - title: "Discovery of Useful Questions as Auxiliary Tasks" 114 | venue: "nips" 115 | venue_date: "2019.12" 116 | online_date: "2019.09" 117 | contacts: {"Vivek Veeriah": "hvveeriah@umich.edu"} 118 | search_space: 119 | search_strategy: 120 | candidate_evaluation: 121 | discussed: true 122 | links: {"paper": "https://arxiv.org/abs/1909.04607"} 123 | misc: 124 | - title: "A Greedy Approach to Adapting the Trace Parameter for Temporal Difference Learning" 125 | venue: "AAMAS" 126 | venue_date: "2016.05" 127 | online_date: "2016.07" 128 | contacts: {"Martha White": "martha@indiana.edu", "Adam White": "adamw@indiana.edu"} 129 | search_space: 130 | search_strategy: 131 | candidate_evaluation: 132 | discussed: true 133 | links: {"paper": "https://arxiv.org/abs/1607.00446"} 134 | misc: 135 | - title: "Meta-Gradient Reinforcement Learning with an Objective Discovered Online" 136 | venue: "nips" 137 | venue_date: "2020.12" 138 | online_date: "2020.07" 139 | contacts: {} 140 | search_space: 141 | search_strategy: 142 | candidate_evaluation: 143 | discussed: true 144 | links: {"paper": "https://arxiv.org/abs/2007.08433"} 145 | misc: 146 | - title: "Meta-Gradient Reinforcement Learning" 147 | venue: "nips" 148 | venue_date: "2018.12" 149 | online_date: "2018.05" 150 | contacts: {} 151 | search_space: 152 | search_strategy: 153 | candidate_evaluation: 154 | discussed: true 155 | links: {"paper": "https://arxiv.org/abs/1805.09801"} 156 | misc: 157 | -------------------------------------------------------------------------------- /awesome_autodl/raw_data/papers/Automated_Problem_Formulation.yaml: -------------------------------------------------------------------------------- 1 | [] 2 | -------------------------------------------------------------------------------- /awesome_autodl/raw_data/papers/Hyperparameter_Optimization.yaml: -------------------------------------------------------------------------------- 1 | - title: "Optuna: A next-generation hyperparameter optimization framework" 2 | venue: "sigkdd" 3 | venue_date: "2019.08" 4 | online_date: "2019.07" 5 | contacts: {"Takuya Akiba": "akiba@preferred.jp"} 6 | search_space: 7 | search_strategy: 8 | candidate_evaluation: 9 | discussed: true 10 | misc: "This is a library/software paper." 11 | - title: "A meta-reinforcement learning approach to optimize parameters and hyper-parameters simultaneously" 12 | venue: "pricai" 13 | venue_date: "2019.08" 14 | online_date: "2019.08" 15 | contacts: {"Abbas Raza Ali": "aali@bournemouth.ac.uk", "Bogdan Gabrys": "bogdan.gabrys@uts.edu.au"} 16 | search_space: 17 | search_strategy: 18 | candidate_evaluation: 19 | discussed: true 20 | misc: 21 | - title: "Learning to learn by gradient descent by gradient descent." 22 | venue: "nips" 23 | venue_date: "2016.12" 24 | online_date: "2016.06" 25 | contacts: {"Marcin Andrychowicz": "marcin.andrychowicz@gmail.com", "Misha Denil": "mdenil@google.com", "Nando de Freitas": "nandodefreitas@google.com"} 26 | search_space: 27 | search_strategy: "differentiable" 28 | candidate_evaluation: 29 | discussed: true 30 | misc: 31 | - title: "Bayesian Optimization of Composite Functions" 32 | venue: "icml" 33 | venue_date: "2019.06" 34 | online_date: "2019.06" 35 | contacts: {"Raul Astudillo": "ra598@cornell.edu", "Peter I. Frazier": "pf98@cornell.edu"} 36 | search_space: 37 | search_strategy: "BayesOpt" 38 | candidate_evaluation: 39 | discussed: true 40 | misc: 41 | - title: "Online Learning Rate Adaptation with Hypergradient Descent" 42 | venue: "iclr" 43 | venue_date: "2018.04" 44 | online_date: "2017.03" 45 | contacts: {"Atılım Gunes Baydin": "gunes@robots.ox.ac.uk", "Robert Cornish": "rcornish@robots.ox.ac.uk", "Frank Wood": "fwood@robots.ox.ac.uk"} 46 | search_space: 47 | search_strategy: "differentiable" 48 | candidate_evaluation: 49 | discussed: true 50 | misc: 51 | - title: "Gradient-Based Optimization of Hyperparameters" 52 | venue: "NeurCom" 53 | venue_date: "2000.08" 54 | online_date: "1999.09" 55 | contacts: {"Yoshua Bengio": "yoshua.bengio@umontreal.ca"} 56 | search_space: 57 | search_strategy: "differentiable" 58 | candidate_evaluation: 59 | discussed: true 60 | misc: 61 | - title: "Algorithms for Hyper-Parameter Optimization" 62 | venue: "nips" 63 | venue_date: "2011.12" 64 | online_date: "2011.12" 65 | contacts: {"James Bergstra": "james.bergstra@uwaterloo.ca", "Balazs Kegl": "balazs.kegl@gmail.com"} 66 | search_space: 67 | search_strategy: 68 | candidate_evaluation: 69 | discussed: true 70 | misc: 71 | - title: "Random search for hyper-parameter optimization" 72 | venue: "jmlr" 73 | venue_date: "2012.02" 74 | online_date: "2012.02" 75 | contacts: {"James Bergstra": "james.bergstra@uwaterloo.ca", "Yoshua Bengio": "yoshua.bengio@umontreal.ca"} 76 | search_space: 77 | search_strategy: "random" 78 | candidate_evaluation: 79 | discussed: true 80 | misc: 81 | - title: "CAVE: Configuration Assessment, Visualization and Evaluation" 82 | venue: "lion" 83 | venue_date: "2018.06" 84 | online_date: "2018.06" 85 | contacts: {"Andre Biedenkapp": "biedenka@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 86 | search_space: 87 | search_strategy: 88 | candidate_evaluation: 89 | discussed: true 90 | misc: 91 | - title: "Speeding up hyper-parameter optimization by extrapolation of learning curves using previous builds" 92 | venue: "ECML_PKDD" 93 | venue_date: "2017.09" 94 | online_date: "2017.09" 95 | contacts: {"Akshay Chandrashekaran": "akshayc@cmu.edu", "Ian R. Lane": "lane@cmu.edu"} 96 | search_space: 97 | search_strategy: 98 | candidate_evaluation: 99 | discussed: true 100 | misc: 101 | - title: "Learning to Learn without Gradient Descent by Gradient Descent" 102 | venue: "icml" 103 | venue_date: "2017.08" 104 | online_date: "2016.11" 105 | contacts: {"Yutian Chen": "yutianc@google.com"} 106 | search_space: 107 | search_strategy: 108 | candidate_evaluation: 109 | discussed: true 110 | misc: 111 | - title: "FBNetV3: Joint Architecture-Recipe Search using Predictor Pretraining" 112 | venue: "cvpr" 113 | venue_date: "2021.06" 114 | online_date: "2020.06" 115 | contacts: {"Xiaoliang Dai": "xiaoliangdai@fb.com", "Alvin Wan": "alvinwan@berkeley.edu"} 116 | search_space: 117 | search_strategy: 118 | candidate_evaluation: 119 | discussed: true 120 | misc: 121 | - title: "Mixed-Variable Bayesian Optimization" 122 | venue: "ijcai" 123 | venue_date: "2020.01" 124 | online_date: "2019.07" 125 | contacts: {"Erik Daxberger": "ead54@cam.ac.uk", "Andreas Krause": "krausea@inf.ethz.ch"} 126 | search_space: 127 | search_strategy: 128 | candidate_evaluation: 129 | discussed: true 130 | misc: 131 | - title: "Speeding up Automatic Hyperparameter Optimization of Deep Neural Networks by Extrapolation of Learning Curves" 132 | venue: "ijcai" 133 | venue_date: "2015.07" 134 | online_date: "2015.07" 135 | contacts: {"Tobias Domhan": "domhant@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 136 | search_space: 137 | search_strategy: 138 | candidate_evaluation: 139 | discussed: true 140 | misc: 141 | - title: "AutoHAS: Efficient Hyperparameter and Architecture Search" 142 | venue: "iclr_w" 143 | venue_date: "2021.05" 144 | online_date: "2020.06" 145 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 146 | search_space: 147 | search_strategy: 148 | candidate_evaluation: "weight sharing" 149 | discussed: true 150 | misc: 151 | - title: "Surrogate Benchmarks for Hyperparameter Optimization" 152 | venue: "ecai_w" 153 | venue_date: "2014.08" 154 | online_date: "2014.08" 155 | contacts: {"Katharina Eggensperger": "eggenspk@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 156 | search_space: 157 | search_strategy: 158 | candidate_evaluation: 159 | discussed: true 160 | misc: 161 | - title: "HPOBench: A Collection of Reproducible Multi-Fidelity Benchmark Problems for HPO" 162 | venue: "nips" 163 | venue_date: "2021.12" 164 | online_date: "2021.09" 165 | contacts: {"Katharina Eggensperger": "eggenspk@cs.uni-freiburg.de", "Neeratyoy Mallik": "mallik@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 166 | search_space: 167 | search_strategy: 168 | candidate_evaluation: 169 | discussed: true 170 | misc: 171 | - title: "AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data" 172 | venue: "icml_w" 173 | venue_date: "2020.07" 174 | online_date: "2020.03" 175 | contacts: {"Nick Erickson": "neerick@amazon.com", "Jonas Mueller": "jonasmue@amazon.com"} 176 | search_space: 177 | search_strategy: 178 | candidate_evaluation: 179 | discussed: true 180 | misc: 181 | - title: "BOHB: Robust and Efficient Hyperparameter Optimization at Scale" 182 | venue: "icml" 183 | venue_date: "2018.07" 184 | online_date: "2018.07" 185 | contacts: {"Stefan Falkner": "sfalkner@informatik.uni-freiburg.de"} 186 | search_space: 187 | search_strategy: 188 | candidate_evaluation: 189 | discussed: true 190 | misc: 191 | - title: "Hyperparameter optimization" 192 | venue: "automated_machine_learning" 193 | venue_date: "2019.12" 194 | online_date: "2019.12" 195 | contacts: {} 196 | search_space: 197 | search_strategy: 198 | candidate_evaluation: 199 | discussed: true 200 | misc: "This is a survey chapter in the book, Automated Machine Learning." 201 | - title: "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks" 202 | venue: "icml" 203 | venue_date: "2017.08" 204 | online_date: "2017.03" 205 | contacts: {"Chelsea Finn": "cbfinn@eecs.berkeley.edu"} 206 | search_space: 207 | search_strategy: 208 | candidate_evaluation: 209 | discussed: true 210 | misc: 211 | - title: "Google Vizier: A Service for Black-Box Optimization" 212 | venue: "sigkdd" 213 | venue_date: "2017.08" 214 | online_date: "2017.06" 215 | contacts: {"Daniel Golovin": "dgg@google.com"} 216 | search_space: 217 | search_strategy: 218 | candidate_evaluation: 219 | discussed: true 220 | misc: "This is a library/software paper." 221 | - title: "A Survey of Methods for Explaining Black Box Models" 222 | venue: "acm_comp_sur" 223 | venue_date: "2018.08" 224 | online_date: "2018.02" 225 | contacts: {"Riccardo Guidotti": "riccardo.guidotti@di.unipi.it"} 226 | search_space: 227 | search_strategy: 228 | candidate_evaluation: 229 | discussed: true 230 | misc: "This is a survey work" 231 | - title: "Learning to Learn Using Gradient Descent" 232 | venue: "icann" 233 | venue_date: "2001.08" 234 | online_date: "2001.08" 235 | contacts: {} 236 | search_space: 237 | search_strategy: 238 | candidate_evaluation: 239 | discussed: true 240 | misc: 241 | - title: "Evolved Policy Gradients" 242 | venue: "nips" 243 | venue_date: "2018.12" 244 | online_date: "2018.02" 245 | contacts: {"Rein Houthooft": "rein.houthooft@openai.com"} 246 | search_space: "RL loss" 247 | search_strategy: "evolution" 248 | candidate_evaluation: 249 | discussed: true 250 | misc: 251 | - title: "Sequential model-based optimization for general algorithm configuration" 252 | venue: "lion" 253 | venue_date: "2011.01" 254 | online_date: "2011.01" 255 | contacts: {"Frank Hutter": "fh@cs.uni-freiburg.de"} 256 | search_space: 257 | search_strategy: 258 | candidate_evaluation: 259 | discussed: true 260 | misc: 261 | - title: "ParamILS: an automatic algorithm configuration framework" 262 | venue: "JAIR" 263 | venue_date: "2009.09" 264 | online_date: "2009.09" 265 | contacts: {"Frank Hutter": "fh@cs.uni-freiburg.de"} 266 | search_space: 267 | search_strategy: 268 | candidate_evaluation: 269 | discussed: true 270 | misc: 271 | - title: "Population Based Training of Neural Networks" 272 | venue: "arXiv" 273 | venue_date: "2017.11" 274 | online_date: "2017.11" 275 | contacts: {} 276 | search_space: 277 | search_strategy: 278 | candidate_evaluation: 279 | discussed: true 280 | misc: 281 | - title: "Non-stochastic Best Arm Identification and Hyperparameter Optimization" 282 | venue: "aistats" 283 | venue_date: "2016.05" 284 | online_date: "2015.02" 285 | contacts: {"Kevin Jamieson": "kjamieson@eecs.berkeley.edu", "Ameet Talwalkar": "talwalkar@cmu.edu"} 286 | search_space: 287 | search_strategy: 288 | candidate_evaluation: 289 | discussed: true 290 | misc: 291 | - title: "Bayesian Optimization with Tree-structured Dependencies" 292 | venue: "icml" 293 | venue_date: "2017.08" 294 | online_date: "2017.08" 295 | contacts: {"Rodolphe Jenatton": "jenat-ton@amazon.de", "Matthias Seeger": "matthias@amazon.de"} 296 | search_space: 297 | search_strategy: "BayesOpt" 298 | candidate_evaluation: 299 | discussed: true 300 | misc: 301 | - title: "Hyp-RL : Hyperparameter Optimization by Reinforcement Learning" 302 | venue: "arXiv" 303 | venue_date: "2019.06" 304 | online_date: "2019.06" 305 | contacts: {"Hadi S. Jomaa": "jomaah@ismll.uni-hildesheim.de", "Lars Schmidt-Thieme": "lars@ismll.uni-hildesheim.de"} 306 | search_space: 307 | search_strategy: 308 | candidate_evaluation: 309 | discussed: true 310 | misc: 311 | - title: "Gaussian Process Bandit Optimisation with Multi-fidelity Evaluations" 312 | venue: "nips" 313 | venue_date: "2016.12" 314 | online_date: "2016.12" 315 | contacts: {"Kirthevasan Kandasamy": "kandasamy@cs.cmu.edu"} 316 | search_space: 317 | search_strategy: 318 | candidate_evaluation: 319 | discussed: true 320 | misc: 321 | - title: "AutonoML: Towards an Integrated Framework for Autonomous Machine Learning" 322 | venue: "arXiv" 323 | venue_date: "2020.12" 324 | online_date: "2020.12" 325 | contacts: {"David Jacob Kedziora": "david.kedziora@uts.edu.au"} 326 | search_space: 327 | search_strategy: 328 | candidate_evaluation: 329 | discussed: true 330 | misc: 331 | - title: "Design and regularization of neural networks: the optimal use of a validation set" 332 | venue: "IEEE_NNSP" 333 | venue_date: "1996.09" 334 | online_date: "1996.09" 335 | contacts: {} 336 | search_space: 337 | search_strategy: 338 | candidate_evaluation: 339 | discussed: true 340 | misc: 341 | - title: "Adaptive Regularization in Neural Network Modeling" 342 | venue: "NN_ToT" 343 | venue_date: "2002.03" 344 | online_date: "2002.03" 345 | contacts: {} 346 | search_space: 347 | search_strategy: 348 | candidate_evaluation: 349 | discussed: true 350 | misc: 351 | - title: "A Generalized Framework for Population Based Training" 352 | venue: "sigkdd" 353 | venue_date: "2019.08" 354 | online_date: "2019.02" 355 | contacts: {"Ang Li": "anglili@google.com"} 356 | search_space: 357 | search_strategy: "PBT" 358 | candidate_evaluation: 359 | discussed: true 360 | misc: 361 | - title: "Learning to Optimize" 362 | venue: "iclr" 363 | venue_date: "2017.04" 364 | online_date: "2016.06" 365 | contacts: {"Ke Li": "ke.li@eecs.berkeley.edu", "Jitendra Malik": "malik@eecs.berkeley.edu"} 366 | search_space: 367 | search_strategy: 368 | candidate_evaluation: 369 | discussed: true 370 | misc: 371 | - title: "Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization" 372 | venue: "jmlr" 373 | venue_date: "2018.04" 374 | online_date: "2016.03" 375 | contacts: {"Lisha Li": "lishal@cs.cmu.edu", "Ameet Talwalkar": "talwalkar@cmu.edu"} 376 | search_space: 377 | search_strategy: 378 | candidate_evaluation: 379 | discussed: true 380 | misc: 381 | - title: "Optimizing Millions of Hyperparameters by Implicit Differentiation" 382 | venue: "aistats" 383 | venue_date: "2020.08" 384 | online_date: "2019.11" 385 | contacts: {"Jonathan Lorraine": "lorraine@cs.toronto.edu", "David Duvenaud": "duvenaud@cs.toronto.edu"} 386 | search_space: 387 | search_strategy: 388 | candidate_evaluation: 389 | discussed: true 390 | misc: 391 | - title: "CMA-ES for Hyperparameter Optimization of Deep Neural Networks" 392 | venue: "iclr_w" 393 | venue_date: "2016.05" 394 | online_date: "2016.04" 395 | contacts: {"Ilya Loshchilov": "ilya@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 396 | search_space: 397 | search_strategy: 398 | candidate_evaluation: 399 | discussed: true 400 | misc: 401 | - title: "Scalable Gradient-Based Tuning of Continuous Regularization Hyperparameters" 402 | venue: "icml" 403 | venue_date: "2016.06" 404 | online_date: "2015.11" 405 | contacts: {"Jelena Luketina": "jelena.luketina@aalto.fi", "Tapani Raiko": "tapani.raiko@aalto.fi"} 406 | search_space: 407 | search_strategy: 408 | candidate_evaluation: 409 | discussed: true 410 | misc: 411 | - title: "Self-Tuning Networks: Bilevel Optimization of Hyperparameters using Structured Best-Response Functions" 412 | venue: "iclr" 413 | venue_date: "2019.05" 414 | online_date: "2019.03" 415 | contacts: {"Matthew MacKay": "mmackay@cs.toronto.edu", "Paul Vicol": "pvicol@cs.toronto.edu"} 416 | search_space: 417 | search_strategy: 418 | candidate_evaluation: 419 | discussed: true 420 | misc: 421 | - title: "Gradient-based Hyperparameter Optimization through Reversible Learning" 422 | venue: "icml" 423 | venue_date: "2015.07" 424 | online_date: "2015.02" 425 | contacts: {"Dougal Maclaurin": "maclaurin@physics.harvard.edu"} 426 | search_space: 427 | search_strategy: 428 | candidate_evaluation: 429 | discussed: true 430 | misc: 431 | - title: "Exploring Opportunistic Meta-knowledge to Reduce Search Spaces for Automated Machine Learning" 432 | venue: "ijcnn" 433 | venue_date: "2021.05" 434 | online_date: "2021.07" 435 | contacts: {"Tien-Dung Nguyen": "TienDung.Nguyen-2@student.uts.edu.au", "Bogdan Gabrys": "bogdan.gabrys@uts.edu.au"} 436 | search_space: 437 | search_strategy: 438 | candidate_evaluation: 439 | discussed: true 440 | misc: 441 | - title: "PABO: Pseudo Agent-Based Multi-Objective Bayesian Hyperparameter Optimization for Efficient Neural Accelerator Design" 442 | venue: "iccad" 443 | venue_date: "2019.11" 444 | online_date: "2019.06" 445 | contacts: {"Maryam Parsa": "mparsa@purdue.edu", "Kaushik Roy": "kaushik@purdue.edu"} 446 | search_space: 447 | search_strategy: 448 | candidate_evaluation: 449 | discussed: true 450 | misc: 451 | - title: "Hyperparameter optimization with approximate gradient" 452 | venue: "icml" 453 | venue_date: "2016.06" 454 | online_date: "2016.02" 455 | contacts: {"Fabian Pedregosa": "f@bianp.net"} 456 | search_space: 457 | search_strategy: 458 | candidate_evaluation: 459 | discussed: true 460 | links: {"paper": "https://arxiv.org/pdf/1602.02355.pdf", "code": "https://github.com/fabianp/hoag"} 461 | misc: 462 | - title: "AutoML-Zero: Evolving Machine Learning Algorithms From Scratch" 463 | venue: "icml" 464 | venue_date: "2020.07" 465 | online_date: "2020.03" 466 | contacts: {"Esteban Real": "ereal@google.com"} 467 | search_space: 468 | search_strategy: "evolution" 469 | candidate_evaluation: 470 | discussed: true 471 | links: {"paper": "https://arxiv.org/abs/2003.03384"} 472 | misc: 473 | - title: "Automatic Composition and Optimization of Multicomponent Predictive Systems With an Extended Auto-WEKA" 474 | venue: "IEEE_J_TASE" 475 | venue_date: "2018.11" 476 | online_date: "2016.12" 477 | contacts: {"Manuel Martin Salvador": "manuel@blinkeye.ai"} 478 | search_space: 479 | search_strategy: 480 | candidate_evaluation: 481 | discussed: true 482 | links: {"paper": "https://arxiv.org/abs/1612.08789"} 483 | misc: 484 | - title: "Truncated Back-propagation for Bilevel Optimization" 485 | venue: "aistats" 486 | venue_date: "2019.04" 487 | online_date: "2018.10" 488 | contacts: {} 489 | search_space: 490 | search_strategy: 491 | candidate_evaluation: 492 | discussed: true 493 | links: {"paper": "https://arxiv.org/abs/1810.10667"} 494 | misc: 495 | - title: "Practical Bayesian Optimization of Machine Learning Algorithms" 496 | venue: "nips" 497 | venue_date: "2012.12" 498 | online_date: "2012.06" 499 | contacts: {"Jasper Snoek": "jasper@cs.toronto.edu", "Ryan P. Adams": "rpa@seas.harvard.edu"} 500 | search_space: 501 | search_strategy: 502 | candidate_evaluation: 503 | discussed: true 504 | links: {"paper": "https://arxiv.org/abs/1206.2944"} 505 | misc: 506 | - title: "Adapting bias by gradient descent" 507 | venue: "aaai" 508 | venue_date: "1992.10" 509 | online_date: "1992.10" 510 | contacts: {"Richard S. Sutton": "sutton@gte.com"} 511 | search_space: 512 | search_strategy: 513 | candidate_evaluation: 514 | discussed: true 515 | links: {"paper": "https://www.aaai.org/Papers/AAAI/1992/AAAI92-027.pdf"} 516 | misc: 517 | - title: "Freeze-Thaw Bayesian Optimization" 518 | venue: "arXiv" 519 | venue_date: "2014.06" 520 | online_date: "2014.06" 521 | contacts: {} 522 | search_space: 523 | search_strategy: "BayesOpt" 524 | candidate_evaluation: 525 | discussed: true 526 | links: {"paper": "https://arxiv.org/abs/1406.3896"} 527 | misc: 528 | - title: "Auto-WEKA: Combined Selection and Hyperparameter Optimization of Classification Algorithms" 529 | venue: "sigkdd" 530 | venue_date: "2013.08" 531 | online_date: "2012.08" 532 | contacts: {"Chris Thornton": "cwthornt@cs.ubc.ca"} 533 | search_space: 534 | search_strategy: 535 | candidate_evaluation: 536 | discussed: true 537 | links: {"paper": "https://arxiv.org/abs/1208.3719", "code": "http://www.cs.ubc.ca/labs/beta/Projects/autoweka"} 538 | misc: 539 | - title: "FBNetV2: Differentiable Neural Architecture Search for Spatial and Channel Dimensions" 540 | venue: "cvpr" 541 | venue_date: "2020.06" 542 | online_date: "2020.04" 543 | contacts: {"Alvin Wan": "alvinwan@berkeley.edu"} 544 | search_space: 545 | search_strategy: 546 | candidate_evaluation: 547 | discussed: true 548 | links: {"paper": "https://arxiv.org/abs/2004.05565", "code": "https://github.com/facebookresearch/mobile-vision"} 549 | misc: 550 | - title: "Learning to reinforcement learn" 551 | venue: "CogSci" 552 | venue_date: "2017.07" 553 | online_date: "2016.11" 554 | contacts: {"JX Wang": "wangjane@google.com"} 555 | search_space: 556 | search_strategy: 557 | candidate_evaluation: 558 | discussed: true 559 | links: {"paper": "https://arxiv.org/abs/1611.05763"} 560 | misc: 561 | - title: "On Hyperparameter Optimization of Machine Learning Algorithms: Theory and Practice" 562 | venue: "Neurocomputing" 563 | venue_date: "2020.11" 564 | online_date: "2020.07" 565 | contacts: {"Li Yang": "lyang339@uwo.ca", "Abdallah Shami": "abdallah.shami@uwo.ca"} 566 | search_space: 567 | search_strategy: 568 | candidate_evaluation: 569 | discussed: true 570 | links: {"paper": "https://arxiv.org/abs/2007.15745", "code": "https://github.com/LiYangHart/Hyperparameter-Optimization-of-Machine-Learning-Algorithms"} 571 | misc: 572 | - title: "Hyper-Parameter Optimization: A Review of Algorithms and Applications" 573 | venue: "arXiv" 574 | venue_date: "2020.03" 575 | online_date: "2020.03" 576 | contacts: {"Tong Yu": "yutong01@inspur.com", "Hong Zhu": "zhuhongbj@inspur.com"} 577 | search_space: 578 | search_strategy: 579 | candidate_evaluation: 580 | discussed: true 581 | links: {"paper": "https://arxiv.org/abs/2003.05689"} 582 | misc: "This is a survey work" 583 | - title: "Towards Automated Deep Learning: Efficient Joint Neural Architecture and Hyperparameter Search" 584 | venue: "icml_w" 585 | venue_date: "2018.07" 586 | online_date: "2018.07" 587 | contacts: {"Arber Zela": "zelaa@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 588 | search_space: 589 | search_strategy: 590 | candidate_evaluation: 591 | discussed: true 592 | links: {"paper": "https://arxiv.org/abs/1807.06906", "code": "https://github.com/arberzela/EfficientNAS"} 593 | misc: -------------------------------------------------------------------------------- /awesome_autodl/raw_data/papers/Neural_Architecture_Search.yaml: -------------------------------------------------------------------------------- 1 | - title: "Zero-Cost Proxies for Lightweight NAS" 2 | venue: "iclr" 3 | venue_date: "2021.05" 4 | online_date: "2021.01" 5 | contacts: {"Mohamed S. Abdelfattah": "mohamed1.a@samsung.com"} 6 | search_space: "RNS" 7 | search_strategy: 8 | candidate_evaluation: "zero-cost proxy" 9 | discussed: true 10 | misc: 11 | - title: "Designing Neural Network Architectures using Reinforcement Learning" 12 | venue: "iclr" 13 | venue_date: "2017.04" 14 | online_date: "2016.11" 15 | contacts: {"Bowen Baker": "bowen@mit.edu", "Otkrist Gupta": "otkrist@mit.edu", "Nikhil Naik": "naik@mit.edu", "Ramesh Raskar": "raskar@mit.edu"} 16 | search_space: 17 | search_strategy: 18 | candidate_evaluation: 19 | discussed: true 20 | misc: 21 | - title: "Accelerating Neural Architecture Search using Performance Prediction" 22 | venue: "iclr_w" 23 | venue_date: "2018.05" 24 | online_date: "2017.05" 25 | contacts: {"Bowen Baker": "bowen@mit.edu", "Otkrist Gupta": "otkrist@mit.edu", "Ramesh Raskar": "raskar@mit.edu", "Nikhil Naik": "naik@mit.edu"} 26 | search_space: 27 | search_strategy: 28 | candidate_evaluation: "predictor" 29 | discussed: true 30 | misc: 31 | - title: "Evolving Memory Cell Structures for Sequence Learning" 32 | venue: "icann" 33 | venue_date: "2009.09" 34 | online_date: "2009.09" 35 | contacts: {"Justin Bayer": "bayer.justin@googlemail.com", "Daan Wierstra": "daan@idsia.ch", "Julian Togelius": "julian@idsia.ch", "Jurgen Schmidhuber": "juergen@idsia.ch"} 36 | search_space: 37 | search_strategy: "evolution" 38 | candidate_evaluation: 39 | discussed: true 40 | misc: 41 | - title: "Understanding and simplifying one-shot architecture search" 42 | venue: "icml" 43 | venue_date: "2018.07" 44 | online_date: "2018.07" 45 | contacts: {"Gabriel Bender": "gbender@google.com"} 46 | search_space: 47 | search_strategy: 48 | candidate_evaluation: 49 | discussed: true 50 | misc: 51 | - title: "Can weight sharing outperform random architecture search? An investigation with TuNAS" 52 | venue: "cvpr" 53 | venue_date: "2020.06" 54 | online_date: "2020.06" 55 | contacts: {"Gabriel Bender": "gbender@google.com", "Hanxiao Liu": "hanxiaol@google.com"} 56 | search_space: "MBS" 57 | search_strategy: "REINFORCE" 58 | candidate_evaluation: "weight sharing" 59 | discussed: true 60 | misc: 61 | - title: "SMASH: one-shot model architecture search through hypernetworks" 62 | venue: "iclr" 63 | venue_date: "2018.05" 64 | online_date: "2017.08" 65 | contacts: {"Andrew Brock": "ajb5@hw.ac.uk", "Theodore Lim": "t.lim@hw.ac.uk", "J.M. Ritchie": "j.m.ritchie@hw.ac.uk"} 66 | search_space: 67 | search_strategy: 68 | candidate_evaluation: "hypernetwork" 69 | discussed: true 70 | misc: 71 | - title: "BATS: Binary architecture search" 72 | venue: "eccv" 73 | venue_date: "2020.08" 74 | online_date: "2020.03" 75 | contacts: {"Adrian Bulat": "adrian@adrianbulat.com", "Brais Martinez": "brais.mart@gmail.com", "Georgios Tzimiropoulos": "g.tzimiropoulos@qmul.ac.uk"} 76 | search_space: 77 | search_strategy: 78 | candidate_evaluation: 79 | discussed: true 80 | misc: 81 | - title: "Once-for-All: Train One Network and Specialize it for Efficient Deployment" 82 | venue: "iclr" 83 | venue_date: "2020.04" 84 | online_date: "2019.08" 85 | contacts: {"Han Cai": "hancai@mit.edu", "Song Han" : "chuangg@mit.edu"} 86 | search_space: 87 | search_strategy: 88 | candidate_evaluation: 89 | discussed: true 90 | misc: 91 | - title: "Path-Level Network Transformation for Efficient Architecture Search" 92 | venue: "icml" 93 | venue_date: "2018.07" 94 | online_date: "2018.06" 95 | contacts: {"Han Cai": "hancai@mit.edu"} 96 | search_space: 97 | search_strategy: 98 | candidate_evaluation: 99 | discussed: true 100 | misc: 101 | - title: "ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware" 102 | venue: "iclr" 103 | venue_date: "2019.05" 104 | online_date: "2018.12" 105 | contacts: {"Han Cai": "hancai@mit.edu", "Ligeng Zhu": "ligeng@mit.edu", "Song Han" : "chuangg@mit.edu"} 106 | search_space: "MBS" 107 | search_strategy: "differentiable, RL" 108 | candidate_evaluation: "weight sharing" 109 | discussed: true 110 | misc: 111 | - title: "Neural architecture search on imagenet in four gpu hours: A theoretically inspired perspective" 112 | venue: "iclr" 113 | venue_date: "2021.05" 114 | online_date: "2021.02" 115 | contacts: {"Wuyang Chen": "wuyang.chen@utexas.edu", "Zhangyang Wang": "atlaswang@utexas.edu"} 116 | search_space: 117 | search_strategy: 118 | candidate_evaluation: "zero-cost proxy" 119 | discussed: true 120 | misc: 121 | - title: "DetNAS: Backbone Search for Object Detection" 122 | venue: "nips" 123 | venue_date: "2019.12" 124 | online_date: "2019.03" 125 | contacts: {"Yukang Chen": "yukang.chen@nlpr.ia.ac.cn", "Xiangyu Zhang": "zhangxiangyu@megvii.com"} 126 | search_space: 127 | search_strategy: 128 | candidate_evaluation: 129 | discussed: true 130 | misc: 131 | - title: "FBNetV3: Joint Architecture-Recipe Search using Predictor Pretraining" 132 | venue: "cvpr" 133 | venue_date: "2021.06" 134 | online_date: "2020.06" 135 | contacts: {"Xiaoliang Dai": "xiaoliangdai@fb.com", "Alvin Wan": "alvinwan@berkeley.edu"} 136 | search_space: 137 | search_strategy: 138 | candidate_evaluation: 139 | discussed: true 140 | misc: 141 | - title: "Bayesian Learning of Neural Network Architectures" 142 | venue: "aistats" 143 | venue_date: "2019.04" 144 | online_date: "2019.01" 145 | contacts: {"Justin Bayer": "bayer.justin@googlemail.com"} 146 | search_space: 147 | search_strategy: "BayesOpt" 148 | candidate_evaluation: 149 | discussed: true 150 | misc: 151 | - title: "More is Less: A More Complicated Network with Less Inference Complexity" 152 | venue: "cvpr" 153 | venue_date: "2017.07" 154 | online_date: "2017.03" 155 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 156 | search_space: 157 | search_strategy: 158 | candidate_evaluation: 159 | discussed: true 160 | misc: 161 | - title: "NATS-Bench: Benchmarking NAS Algorithms for Architecture Topology and Size" 162 | venue: "IEEE_J_PAMI" 163 | venue_date: "2021.01" 164 | online_date: "2020.09" 165 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com", "Bogdan Gabrys": "Bogdan.Gabrys@uts.edu.au"} 166 | search_space: 167 | search_strategy: 168 | candidate_evaluation: 169 | discussed: true 170 | links: {"paper": "https://arxiv.org/abs/2009.00437", "code": "https://github.com/D-X-Y/NATS-Bench"} 171 | misc: 172 | - title: "AutoHAS: Efficient Hyperparameter and Architecture Search" 173 | venue: "iclr_w" 174 | venue_date: "2021.05" 175 | online_date: "2020.06" 176 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 177 | search_space: 178 | search_strategy: 179 | candidate_evaluation: "weight sharing" 180 | discussed: true 181 | links: {"paper": "https://arxiv.org/pdf/2006.03656.pdf"} 182 | misc: 183 | - title: "Network Pruning via Transformable Architecture Search" 184 | venue: "nips" 185 | venue_date: "2019.12" 186 | online_date: "2019.05" 187 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 188 | search_space: "size" 189 | search_strategy: "differentiable" 190 | candidate_evaluation: "weight sharing" 191 | discussed: true 192 | links: {"paper": "https://arxiv.org/abs/1905.09717", "code": "https://github.com/D-X-Y/AutoDL-Projects"} 193 | misc: 194 | - title: "One-Shot Neural Architecture Search via Self-Evaluated Template Network" 195 | venue: "iccv" 196 | venue_date: "2019.10" 197 | online_date: "2019.10" 198 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 199 | search_space: "RNS" 200 | search_strategy: "differentiable" 201 | candidate_evaluation: 202 | discussed: true 203 | misc: 204 | - title: "Searching for A Robust Neural Architecture in Four GPU Hours" 205 | venue: "cvpr" 206 | venue_date: "2019.06" 207 | online_date: "2019.06" 208 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 209 | search_space: "RNS" 210 | search_strategy: "differentiable" 211 | candidate_evaluation: 212 | discussed: true 213 | links: {"paper": "https://arxiv.org/abs/1910.04465", "code": "https://github.com/D-X-Y/AutoDL-Projects"} 214 | misc: 215 | - title: "NAS-bench-201: Extending the scope of reproducible neural architecture search" 216 | venue: "iclr" 217 | venue_date: "2020.04" 218 | online_date: "2020.01" 219 | contacts: {"Xuanyi Dong": "xuanyi.dxy@gmail.com"} 220 | search_space: 221 | search_strategy: 222 | candidate_evaluation: 223 | discussed: true 224 | links: {"paper": "https://arxiv.org/abs/2001.00326", "code": "https://github.com/D-X-Y/NAS-Bench-201"} 225 | misc: 226 | - title: "Neural architecture search: A survey" 227 | venue: "jmlr" 228 | venue_date: "2019.03" 229 | online_date: "2018.08" 230 | contacts: {"Thomas Elsken": "thomas.elsken@de.bosch.com", "Jan Hendrik Metzen": "janhendrik.metzen@de.bosch.com", "Frank Hutter": "fh@cs.uni-freiburg.de"} 231 | search_space: 232 | search_strategy: 233 | candidate_evaluation: 234 | discussed: true 235 | misc: "This is a survey work" 236 | - title: "Densely Connected Search Space for More Flexible Neural Architecture Search" 237 | venue: "cvpr" 238 | venue_date: "2020.06" 239 | online_date: "2019.06" 240 | contacts: {"Jiemin Fang": "jaminfong@hust.edu.cn"} 241 | search_space: 242 | search_strategy: 243 | candidate_evaluation: 244 | discussed: true 245 | misc: 246 | - title: "Spatially Adaptive Computation Time for Residual Networks" 247 | venue: "cvpr" 248 | venue_date: "2017.07" 249 | online_date: "2016.12" 250 | contacts: {"Michael Figurnov": "michael@figurnov.ru", "Maxwell D. Collins": "maxwellcollins@google.com"} 251 | search_space: 252 | search_strategy: 253 | candidate_evaluation: 254 | discussed: true 255 | misc: 256 | - title: "The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks" 257 | venue: "iclr" 258 | venue_date: "2019.05" 259 | online_date: "2018.03" 260 | contacts: {"Jonathan Frankle": "jfrankle@csail.mit.edu", "Michael Carbin": "mcarbin@csail.mit.edu"} 261 | search_space: 262 | search_strategy: 263 | candidate_evaluation: 264 | discussed: true 265 | misc: 266 | - title: "NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection" 267 | venue: "cvpr" 268 | venue_date: "2019.06" 269 | online_date: "2019.04" 270 | contacts: {"Golnaz Ghaisi": "golnazg@google.com"} 271 | search_space: 272 | search_strategy: 273 | candidate_evaluation: 274 | discussed: true 275 | misc: 276 | - title: "Single Path One-Shot Neural Architecture Search with Uniform Sampling" 277 | venue: "eccv" 278 | venue_date: "2020.08" 279 | online_date: "2019.04" 280 | contacts: {"Zichao Guo": "guozichao@megvii.com", "Xiangyu Zhang": "zhangxiangyu@megvii.com"} 281 | search_space: 282 | search_strategy: 283 | candidate_evaluation: 284 | discussed: true 285 | misc: 286 | - title: "Dynamic Neural Networks: A Survey" 287 | venue: "arXiv" 288 | venue_date: "2021.02" 289 | online_date: "2021.02" 290 | contacts: {"Yizeng Han": "hanyz18@mails.tsinghua.edu.cn"} 291 | search_space: 292 | search_strategy: 293 | candidate_evaluation: 294 | discussed: true 295 | misc: "This is a survey work" 296 | - title: "Macro Neural Architecture Search Revisited" 297 | venue: "nips_w" 298 | venue_date: "2018.12" 299 | online_date: "2018.12" 300 | contacts: {"Hanzhang Hu": "hanzhang@cs.cmu.edu"} 301 | search_space: 302 | search_strategy: 303 | candidate_evaluation: 304 | discussed: true 305 | misc: 306 | - title: "Auto-Keras: An Efficient Neural Architecture Search System" 307 | venue: "sigkdd" 308 | venue_date: "2019.08" 309 | online_date: "2018.06" 310 | contacts: {"Haifeng Jin": "jin@tamu.edu", "Qingquan Song": "song_3134@tamu.edu", "Xia Hu": "xiahu@tamu.edu"} 311 | search_space: 312 | search_strategy: 313 | candidate_evaluation: 314 | discussed: true 315 | misc: "This is a library/software paper." 316 | - title: "Neural Architecture Search with Bayesian Optimisation and Optimal Transport" 317 | venue: "nips" 318 | venue_date: "2018.12" 319 | online_date: "2018.02" 320 | contacts: {"Kirthevasan Kandasamy": "kandasamy@cs.cmu.edu", "Eric P Xing": "epxing@cs.cmu.edu"} 321 | search_space: 322 | search_strategy: 323 | candidate_evaluation: 324 | discussed: true 325 | misc: 326 | - title: "Tabular Benchmarks for Joint Architecture and Hyperparameter Optimization" 327 | venue: "arXiv" 328 | venue_date: "2019.05" 329 | online_date: "2019.05" 330 | contacts: {"Aaron Klein": "kleinaa@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 331 | search_space: 332 | search_strategy: 333 | candidate_evaluation: 334 | discussed: true 335 | misc: 336 | - title: "HW-NAS-Bench: Hardware-Aware Neural Architecture Search Benchmark" 337 | venue: "iclr" 338 | venue_date: "2021.05" 339 | online_date: "2021.03" 340 | contacts: {"Chaojian Li": "cl114@rice.edu", "Yingyan Lin": "yingyan.lin@rice.edu"} 341 | search_space: 342 | search_strategy: 343 | candidate_evaluation: 344 | discussed: true 345 | misc: 346 | - title: "Random Search and Reproducibility for Neural Architecture Search" 347 | venue: "uai" 348 | venue_date: "2020.08" 349 | online_date: "2019.02" 350 | contacts: {"Liam Li": "me@liamcli.com", "Ameet Talwalkar": "talwalkar@cmu.edu"} 351 | search_space: 352 | search_strategy: 353 | candidate_evaluation: 354 | discussed: true 355 | misc: 356 | - title: "Best Practices for Scientific Research on Neural Architecture Search" 357 | venue: "jmlr" 358 | venue_date: "2020.11" 359 | online_date: "2019.09" 360 | contacts: {"Marius Lindauer": "lindauer@tnt.uni-hannover.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 361 | search_space: 362 | search_strategy: 363 | candidate_evaluation: 364 | discussed: true 365 | misc: "This is a discussion paper." 366 | - title: "Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation" 367 | venue: "cvpr" 368 | venue_date: "2019.06" 369 | online_date: "2019.01" 370 | contacts: {} 371 | search_space: 372 | search_strategy: 373 | candidate_evaluation: 374 | discussed: true 375 | misc: 376 | - title: "Progressive Neural Architecture Search" 377 | venue: "eccv" 378 | venue_date: "2018.09" 379 | online_date: "2017.12" 380 | contacts: {} 381 | search_space: 382 | search_strategy: 383 | candidate_evaluation: 384 | discussed: true 385 | misc: 386 | - title: "Evolving Normalization-Activation Layers" 387 | venue: "nips" 388 | venue_date: "2020.12" 389 | online_date: "2020.04" 390 | contacts: {"Hanxiao Liu": "hanxiaol@google.com"} 391 | search_space: 392 | search_strategy: "evolution" 393 | candidate_evaluation: 394 | discussed: true 395 | misc: 396 | - title: "DARTS: Differentiable Architecture Search" 397 | venue: "iclr" 398 | venue_date: "2019.05" 399 | online_date: "2018.06" 400 | contacts: {"Hanxiao Liu": "hanxiaol@google.com"} 401 | search_space: 402 | search_strategy: 403 | candidate_evaluation: 404 | discussed: true 405 | misc: 406 | - title: "Neural Architecture Optimization" 407 | venue: "nips" 408 | venue_date: "2018.12" 409 | online_date: "2018.08" 410 | contacts: {"Renqian Luo": "lrq@mail.ustc.edu.cn", "Fei Tian": "fetia@microsoft.com"} 411 | search_space: 412 | search_strategy: 413 | candidate_evaluation: 414 | discussed: true 415 | misc: 416 | - title: "Neural Architecture Search without Training" 417 | venue: "icml" 418 | venue_date: "2021.07" 419 | online_date: "2020.06" 420 | contacts: {"Joseph Mellor": "joe.mellor@ed.ac.uk"} 421 | search_space: 422 | search_strategy: 423 | candidate_evaluation: 424 | discussed: true 425 | misc: 426 | - title: "PyGlove: Symbolic Programming for Automated Machine Learning" 427 | venue: "nips" 428 | venue_date: "2020.12" 429 | online_date: "2020.12" 430 | contacts: {"Daiyi Peng": "daiyip@google.com"} 431 | search_space: 432 | search_strategy: 433 | candidate_evaluation: 434 | discussed: true 435 | misc: "This is a library/software paper." 436 | - title: "Efficient Neural Architecture Search via Parameters Sharing" 437 | venue: "icml" 438 | venue_date: "2018.07" 439 | online_date: "2018.02" 440 | contacts: {"Hieu Pham": "hy-hieu@cmu.edu", "Melody Y. Guan": "mguan@stanford.edu"} 441 | search_space: 442 | search_strategy: 443 | candidate_evaluation: 444 | discussed: true 445 | links: {"paper": "https://arxiv.org/pdf/1802.03268.pdf"} 446 | misc: 447 | - title: "Designing Network Design Spaces" 448 | venue: "cvpr" 449 | venue_date: "2020.06" 450 | online_date: "2020.03" 451 | contacts: {} 452 | search_space: 453 | search_strategy: 454 | candidate_evaluation: 455 | discussed: true 456 | links: {"paper": "https://arxiv.org/pdf/2003.13678.pdf", "code": "https://github.com/facebookresearch/pycls"} 457 | misc: 458 | - title: "Searching for Activation Functions" 459 | venue: "arXiv" 460 | venue_date: "2017.10" 461 | online_date: "2017.10" 462 | contacts: {"Prajit Ramachandran": "prajit@google.com"} 463 | search_space: 464 | search_strategy: 465 | candidate_evaluation: 466 | discussed: true 467 | links: {"paper": "https://arxiv.org/pdf/1710.05941.pdf"} 468 | misc: 469 | - title: "Regularized Evolution for Image Classifier Architecture Search" 470 | venue: "aaai" 471 | venue_date: "2019.01" 472 | online_date: "2018.02" 473 | contacts: {"Esteban Real": "ereal@google.com"} 474 | search_space: 475 | search_strategy: "evolution" 476 | candidate_evaluation: 477 | discussed: true 478 | links: {"paper": "https://arxiv.org/abs/1802.01548"} 479 | misc: 480 | - title: "Large-Scale Evolution of Image Classifiers" 481 | venue: "icml" 482 | venue_date: "2017.08" 483 | online_date: "2017.03" 484 | contacts: {"Esteban Real": "ereal@google.com"} 485 | search_space: 486 | search_strategy: "evolution" 487 | candidate_evaluation: 488 | discussed: true 489 | links: {"paper": "https://arxiv.org/abs/1703.01041"} 490 | misc: 491 | - title: "A Comprehensive Survey of Neural Architecture Search: Challenges and Solutions" 492 | venue: "acm_comp_sur" 493 | venue_date: "2021.05" 494 | online_date: "2020.06" 495 | contacts: {"Pengzhen Ren": "pzhren@foxmail.com"} 496 | search_space: 497 | search_strategy: 498 | candidate_evaluation: 499 | discussed: true 500 | links: {"paper": "https://arxiv.org/abs/2006.02903"} 501 | misc: "This is a survey work" 502 | - title: "Neural Architecture Generator Optimization" 503 | venue: "nips" 504 | venue_date: "2020.12" 505 | online_date: "2020.04" 506 | contacts: {"Binxin Ru": "robin@robots.ox.ac.uk"} 507 | search_space: 508 | search_strategy: 509 | candidate_evaluation: 510 | discussed: true 511 | links: {"paper": "https://arxiv.org/abs/2004.01395", "code": "https://github.com/rubinxin/vega_NAGO"} 512 | misc: 513 | - title: "NAS-Bench-301 and the Case for Surrogate Benchmarks for Neural Architecture Search" 514 | venue: "arXiv" 515 | venue_date: "2020.08" 516 | online_date: "2020.08" 517 | contacts: {"Julien Siems": "siemsj@cs.uni-freiburg.de", "Lucas Zimmer": "zimmerl@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 518 | search_space: 519 | search_strategy: 520 | candidate_evaluation: 521 | discussed: true 522 | links: {"paper": "https://arxiv.org/abs/2008.09777", "code": "https://github.com/automl/nasbench301"} 523 | misc: 524 | - title: "The Evolved Transformer" 525 | venue: "icml" 526 | venue_date: "2019.06" 527 | online_date: "2019.01" 528 | contacts: {"David R. So": "davidso@google.com"} 529 | search_space: "transformer" 530 | search_strategy: "evolution" 531 | candidate_evaluation: 532 | discussed: true 533 | links: {"paper": "https://arxiv.org/abs/1901.11117", "code": "https://github.com/tensorflow/tensor2tensor/blob/master/tensor2tensor/models/evolved_transformer.py"} 534 | misc: 535 | - title: "Primer: Searching for efficient transformers for language modeling" 536 | venue: "nips" 537 | venue_date: "2021.12" 538 | online_date: "2021.09" 539 | contacts: {"David R. So": "davidso@google.com"} 540 | search_space: "transformer" 541 | search_strategy: "evolution" 542 | candidate_evaluation: 543 | discussed: true 544 | links: {"paper": "https://arxiv.org/abs/2109.08668", "code": "https://github.com/google-research/google-research/tree/master/primer"} 545 | misc: 546 | - title: "Searching the Search Space of Vision Transformer" 547 | venue: "nips" 548 | venue_date: "2021.12" 549 | online_date: "2021.11" 550 | contacts: {} 551 | search_space: "transformer" 552 | search_strategy: "evolution" 553 | candidate_evaluation: 554 | discussed: true 555 | links: {"paper": "https://arxiv.org/abs/2111.14725", "code": "https://github.com/microsoft/Cream"} 556 | misc: 557 | - title: "MnasNet: Platform-aware neural architecture search for mobile" 558 | venue: "cvpr" 559 | venue_date: "2019.06" 560 | online_date: "2018.07" 561 | contacts: {"Mingxing Tan": "tanmingxing@google.com"} 562 | search_space: 563 | search_strategy: 564 | candidate_evaluation: 565 | discussed: true 566 | links: {"paper": "https://arxiv.org/abs/1807.11626", "code": "https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet"} 567 | misc: 568 | - title: "EfficientNetV2: Smaller models and faster training" 569 | venue: "icml" 570 | venue_date: "2021.07" 571 | online_date: "2021.04" 572 | contacts: {"Mingxing Tan": "tanmingxing@google.com"} 573 | search_space: 574 | search_strategy: 575 | candidate_evaluation: 576 | discussed: true 577 | links: {"paper": "https://arxiv.org/abs/2104.00298", "code": "https://github.com/google/automl/tree/master/efficientnetv2"} 578 | misc: 579 | - title: "HAQ: Hardware-aware automated quantization with mixed precision" 580 | venue: "cvpr" 581 | venue_date: "2019.06" 582 | online_date: "2018.11" 583 | contacts: {"Kuan Wang": "kuanwang@mit.edu"} 584 | search_space: 585 | search_strategy: 586 | candidate_evaluation: 587 | discussed: true 588 | links: {"paper": "https://arxiv.org/abs/1811.08886"} 589 | misc: 590 | - title: "APQ: Joint Search for Network Architecture, Pruning and Quantization Policy" 591 | venue: "cvpr" 592 | venue_date: "2020.06" 593 | online_date: "2020.06" 594 | contacts: {} 595 | search_space: 596 | search_strategy: 597 | candidate_evaluation: 598 | discussed: true 599 | links: {"paper": "https://arxiv.org/abs/2006.08509", "code": "https://github.com/mit-han-lab/apq"} 600 | misc: 601 | - title: "Neural Predictor for Neural Architecture Search" 602 | venue: "cvpr" 603 | venue_date: "2020.06" 604 | online_date: "2019.12" 605 | contacts: {"Wei Wen": "wei.wen@duke.edu", "Hanxiao Liu": "hanxiaol@google.com"} 606 | search_space: 607 | search_strategy: 608 | candidate_evaluation: 609 | discussed: true 610 | links: {"paper": "https://arxiv.org/abs/1912.00848"} 611 | misc: 612 | - title: "BANANAS: Bayesian Optimization with Neural Architectures for Neural Architecture Search" 613 | venue: "aaai" 614 | venue_date: "2021.02" 615 | online_date: "2019.10" 616 | contacts: {"Colin White": "colin@abacus.ai"} 617 | search_space: 618 | search_strategy: 619 | candidate_evaluation: 620 | discussed: true 621 | links: {"paper": "https://arxiv.org/abs/1910.11858", "code": "https://github.com/naszilla/naszilla"} 622 | misc: 623 | - title: "FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search" 624 | venue: "cvpr" 625 | venue_date: "2019.06" 626 | online_date: "2018.12" 627 | contacts: {"Bichen Wu": "wbc@fb.com"} 628 | search_space: 629 | search_strategy: "differentiable" 630 | candidate_evaluation: 631 | discussed: true 632 | links: {"paper": "https://arxiv.org/abs/1812.03443"} 633 | misc: 634 | - title: "Mixed Precision Quantization of ConvNets via Differentiable Neural Architecture Search" 635 | venue: "arXiv" 636 | venue_date: "2018.12" 637 | online_date: "2018.12" 638 | contacts: {"Bichen Wu": "wbc@fb.com"} 639 | search_space: 640 | search_strategy: "differentiable" 641 | candidate_evaluation: 642 | discussed: true 643 | links: {"paper": "https://arxiv.org/abs/1812.00090"} 644 | misc: 645 | - title: "Exploring Randomly Wired Neural Networks for Image Recognition" 646 | venue: "iccv" 647 | venue_date: "2019.10" 648 | online_date: "2019.04" 649 | contacts: {} 650 | search_space: 651 | search_strategy: 652 | candidate_evaluation: 653 | discussed: true 654 | links: {"paper": "https://arxiv.org/abs/1904.01569"} 655 | misc: 656 | - title: "SNAS: Stochastic Neural Architecture Search" 657 | venue: "iclr" 658 | venue_date: "2019.05" 659 | online_date: "2018.12" 660 | contacts: {"Sirui Xie": "xiesirui@sensetime.com"} 661 | search_space: 662 | search_strategy: 663 | candidate_evaluation: 664 | discussed: true 665 | links: {"paper": "https://arxiv.org/abs/1812.09926", "code": "https://github.com/SNAS-Series/SNAS-Series"} 666 | misc: 667 | - title: "Resource Constrained Neural Network Architecture Search: Will a Submodularity Assumption Help?" 668 | venue: "iccv" 669 | venue_date: "2019.10" 670 | online_date: "2019.04" 671 | contacts: {"Yunyang Xiong": "yxiong43@wisc.edu"} 672 | search_space: 673 | search_strategy: 674 | candidate_evaluation: 675 | discussed: true 676 | links: {"paper": "https://arxiv.org/abs/1904.03786", "code": "https://github.com/yyxiongzju/RCNet"} 677 | misc: 678 | - title: "PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search" 679 | venue: "iclr" 680 | venue_date: "2020.04" 681 | online_date: "2019.07" 682 | contacts: {"Yuhui Xu": "yuhuixu@sjtu.edu.cn"} 683 | search_space: 684 | search_strategy: 685 | candidate_evaluation: 686 | discussed: true 687 | links: {"paper": "https://arxiv.org/abs/1907.05737", "code": "https://github.com/yuhuixu1993/PC-DARTS"} 688 | misc: 689 | - title: "NAS-Bench-x11 and the Power of Learning Curves" 690 | venue: "nips" 691 | venue_date: "2021.12" 692 | online_date: "2021.11" 693 | contacts: {"Shen Yan": "yanshen6@msu.edu", "Colin White": "colin@abacus.ai", "Frank Hutter": "fh@cs.uni-freiburg.de"} 694 | search_space: 695 | search_strategy: 696 | candidate_evaluation: 697 | discussed: true 698 | links: {"paper": "https://arxiv.org/abs/2111.03602", "code": "https://github.com/automl/nas-bench-x11"} 699 | misc: 700 | - title: "Designing Energy-Efficient Convolutional Neural Networks using Energy-Aware Pruning" 701 | venue: "cvpr" 702 | venue_date: "2017.07" 703 | online_date: "2016.11" 704 | contacts: {"Tien-Ju Yang": "tjy@mit.edu"} 705 | search_space: 706 | search_strategy: 707 | candidate_evaluation: 708 | discussed: true 709 | links: {"paper": "https://arxiv.org/abs/1611.05128", "code": "http://eyeriss.mit.edu/energy.html"} 710 | misc: 711 | - title: "NetAdapt: Platform-Aware Neural Network Adaptation for Mobile Applications" 712 | venue: "eccv" 713 | venue_date: "2018.09" 714 | online_date: "2018.04" 715 | contacts: {"Tien-Ju Yang": "tjy@mit.edu"} 716 | search_space: 717 | search_strategy: 718 | candidate_evaluation: 719 | discussed: true 720 | links: {"paper": "https://arxiv.org/abs/1804.03230"} 721 | misc: 722 | - title: "Searching for Low-Bit Weights in Quantized Neural Networks" 723 | venue: "nips" 724 | venue_date: "2020.12" 725 | online_date: "2020.09" 726 | contacts: {"Zhaohui Yang": "zhaohuiyang@pku.edu.cn"} 727 | search_space: 728 | search_strategy: 729 | candidate_evaluation: 730 | discussed: true 731 | links: {"paper": "https://arxiv.org/abs/2009.08695"} 732 | misc: 733 | - title: "NAS-Bench-101: Towards Reproducible Neural Architecture Search" 734 | venue: "icml" 735 | venue_date: "2019.06" 736 | online_date: "2019.02" 737 | contacts: {"Chris Ying": "contact@chrisying.net", "Aaron Klein": "kleinaa@cs.uni-freiburg.de"} 738 | search_space: 739 | search_strategy: 740 | candidate_evaluation: 741 | discussed: true 742 | links: {"paper": "https://arxiv.org/abs/1902.09635", "code": "https://github.com/google-research/nasbench"} 743 | misc: 744 | - title: "BigNAS: Scaling Up Neural Architecture Search with Big Single-Stage Models" 745 | venue: "eccv" 746 | venue_date: "2020.08" 747 | online_date: "2020.03" 748 | contacts: {"Jiahui Yu": "jiahuiyu@google.com"} 749 | search_space: 750 | search_strategy: 751 | candidate_evaluation: 752 | discussed: true 753 | links: {"paper": "https://arxiv.org/abs/2003.11142"} 754 | misc: 755 | - title: "Slimmable Neural Networks" 756 | venue: "iclr" 757 | venue_date: "2019.05" 758 | online_date: "2018.12" 759 | contacts: {"Jiahui Yu": "jiahuiyu@google.com"} 760 | search_space: 761 | search_strategy: 762 | candidate_evaluation: 763 | discussed: true 764 | links: {"paper": "https://arxiv.org/abs/1812.08928", "code": "https://github.com/JiahuiYu/slimmable_networks"} 765 | misc: 766 | - title: "Evaluating the Search Phase of Neural Architecture Search" 767 | venue: "iclr" 768 | venue_date: "2020.04" 769 | online_date: "2019.02" 770 | contacts: {"Kaicheng Yu": "kaicheng.yu@epfl.ch", "Christian Sciuto": "christian.sciuto@daskell.com"} 771 | search_space: 772 | search_strategy: 773 | candidate_evaluation: 774 | discussed: true 775 | links: {"paper": "https://arxiv.org/abs/1902.08142", "code": "https://github.com/kcyu2014/eval-nas"} 776 | misc: 777 | - title: "Neural Ensemble Search for Uncertainty Estimation and Dataset Shift" 778 | venue: "icml_w" 779 | venue_date: "2020.07" 780 | online_date: "2020.06" 781 | contacts: {"Sheheryar Zaidi": "szaidi@stats.ox.ac.uk", "Arber Zela": "zelaa@cs.uni-freiburg.de"} 782 | search_space: 783 | search_strategy: 784 | candidate_evaluation: 785 | discussed: true 786 | links: {"paper": "https://arxiv.org/abs/2006.08573", "code": "https://github.com/automl/nes"} 787 | misc: 788 | - title: "Understanding and Robustifying Differentiable Architecture Search" 789 | venue: "iclr" 790 | venue_date: "2020.04" 791 | online_date: "2019.09" 792 | contacts: {"Arber Zela": "zelaa@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 793 | search_space: 794 | search_strategy: 795 | candidate_evaluation: 796 | discussed: true 797 | links: {"paper": "https://arxiv.org/abs/1909.09656", "code": "https://github.com/automl/RobustDARTS"} 798 | misc: 799 | - title: "NAS-Bench-1Shot1: Benchmarking and Dissecting One-shot Neural Architecture Search" 800 | venue: "iclr" 801 | venue_date: "2020.04" 802 | online_date: "2020.01" 803 | contacts: {"Arber Zela": "zelaa@cs.uni-freiburg.de", "Julien Siems": "siemsj@cs.uni-freiburg.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 804 | search_space: 805 | search_strategy: 806 | candidate_evaluation: 807 | discussed: true 808 | links: {"paper": "https://arxiv.org/abs/2001.10422", "code": "https://github.com/automl/nasbench-1shot1"} 809 | misc: 810 | - title: "Retiarii: a deep learning exploratory-training framework" 811 | venue: "osdi" 812 | venue_date: "2020.11" 813 | online_date: "2020.11" 814 | contacts: {} 815 | search_space: 816 | search_strategy: 817 | candidate_evaluation: 818 | discussed: true 819 | links: {"paper": "https://www.usenix.org/conference/osdi20/presentation/zhang-quanlu", "code": "https://github.com/microsoft/nni/tree/retiarii_artifact"} 820 | misc: "This is a library/software paper." 821 | - title: "Practical Block-wise Neural Network Architecture Generation" 822 | venue: "cvpr" 823 | venue_date: "2018.06" 824 | online_date: "2017.08" 825 | contacts: {"Zhao Zhong": "zhao.zhong@nlpr.ia.ac.cn"} 826 | search_space: 827 | search_strategy: 828 | candidate_evaluation: 829 | discussed: true 830 | links: {"paper": "https://arxiv.org/abs/1708.05552"} 831 | misc: 832 | - title: "BayesNAS: A Bayesian Approach for Neural Architecture Search" 833 | venue: "icml" 834 | venue_date: "2019.05" 835 | online_date: "2019.05" 836 | contacts: {"Wei Pan": "wei.pan@tudelft.nl"} 837 | search_space: 838 | search_strategy: "BayesOpt" 839 | candidate_evaluation: 840 | discussed: true 841 | links: {"paper": "https://arxiv.org/abs/1905.04919"} 842 | misc: 843 | - title: "Auto-PyTorch Tabular: Multi-Fidelity MetaLearning for Efficient and Robust AutoDL" 844 | venue: "IEEE_J_PAMI" 845 | venue_date: "2021.03" 846 | online_date: "2020.06" 847 | contacts: {"Lucas Zimmer": "zimmerl@cs.uni-freiburg.de", "Marius Lindauer": "lindauer@tnt.uni-hannover.de", "Frank Hutter": "fh@cs.uni-freiburg.de"} 848 | search_space: 849 | search_strategy: 850 | candidate_evaluation: 851 | discussed: true 852 | links: {"paper": "https://arxiv.org/abs/2006.13799", "code": "https://github.com/automl/Auto-PyTorch"} 853 | misc: "This is a library/software paper." 854 | - title: "Neural Architecture Search with Reinforcement Learning" 855 | venue: "iclr" 856 | venue_date: "2017.04" 857 | online_date: "2016.11" 858 | contacts: {"Barret Zoph": "barretzoph@google.com"} 859 | search_space: 860 | search_strategy: 861 | candidate_evaluation: 862 | discussed: true 863 | links: {"paper": "https://arxiv.org/abs/1611.01578"} 864 | misc: 865 | - title: "Learning Transferable Architectures for Scalable Image Recognition" 866 | venue: "cvpr" 867 | venue_date: "2018.06" 868 | online_date: "2017.07" 869 | contacts: {"Barret Zoph": "barretzoph@google.com"} 870 | search_space: 871 | search_strategy: 872 | candidate_evaluation: 873 | discussed: true 874 | links: {"paper": "https://arxiv.org/abs/1707.07012"} 875 | misc: -------------------------------------------------------------------------------- /awesome_autodl/utils/__init__.py: -------------------------------------------------------------------------------- 1 | ################################################## 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2021 # 3 | ################################################## 4 | from awesome_autodl.utils.yaml import load_yaml 5 | from awesome_autodl.utils.yaml import dump_yaml 6 | from awesome_autodl.utils.check import check_paper_and_correct_format 7 | from awesome_autodl.utils.check import check_and_sort_by_date 8 | from awesome_autodl.utils.filter import filter_ele_w_value 9 | 10 | from awesome_autodl.utils.fix_invalid_email import email_old_to_new_202203 11 | -------------------------------------------------------------------------------- /awesome_autodl/utils/check.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2020.08 # 3 | ##################################################### 4 | from copy import deepcopy 5 | 6 | 7 | def check_paper_and_correct_format(paper): 8 | assert isinstance(paper, dict), f"Expect dict than {type(paper)}" 9 | paper = deepcopy(paper) 10 | necessary_keys = ( 11 | "title", 12 | "search_space", 13 | "search_strategy", 14 | "eval_boost", 15 | "online_date", 16 | "venue", 17 | # "application", 18 | ) 19 | for key in necessary_keys: 20 | assert ( 21 | key in paper 22 | ), f"Did not find {key} in {paper['title']} {list(paper.keys())}" 23 | if key != "title" and isinstance(paper[key], str): 24 | paper[key] = ",".join(paper[key].split(", ")) 25 | search_strategies = ( 26 | "RL", 27 | "Evolution", 28 | "Random", 29 | "Differential", 30 | "BayesOpt", 31 | "Heuristic", 32 | "Manual", 33 | ) 34 | for search_strategy in paper["search_strategy"].split(","): 35 | assert ( 36 | search_strategy in search_strategies 37 | ), f"This paper {paper} has a different search strategy than {search_strategies}" 38 | assert len(paper["online_date"]), "This paper has empty online_date" 39 | return paper 40 | 41 | 42 | def check_and_sort_by_date(paper_list): 43 | assert isinstance(paper_list, list) 44 | xlist = list() 45 | for paper in paper_list: 46 | paper = check_paper_and_correct_format(paper) 47 | xlist.append(paper) 48 | return sorted(xlist, key=lambda x: x["online_date"]) 49 | -------------------------------------------------------------------------------- /awesome_autodl/utils/filter.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2020.08 # 3 | ##################################################### 4 | from copy import deepcopy 5 | 6 | 7 | def filter_ele_w_value(paper_list, key, value): 8 | assert isinstance(paper_list, list) 9 | xlist = list() 10 | for paper in paper_list: 11 | if value in paper[key]: 12 | xlist.append(deepcopy(paper)) 13 | return xlist 14 | -------------------------------------------------------------------------------- /awesome_autodl/utils/fix_invalid_email.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.03 # 3 | ##################################################### 4 | 5 | email_old_to_new_202203 = { 6 | "sutton@gte.com": None, 7 | "ww129@win.duke.edu": None, 8 | "rpa@seas.harvard.edu": "rpa@princeton.edu", 9 | "martha@indiana.edu": "whitem@ualberta.ca", 10 | "adamw@indiana.edu": "amw8@ualberta.ca", 11 | "hvveeriah@umich.edu": "vveeriah@umich.edu", # TODO: fix the raw file later 12 | "duzx16@mails.tsinghua.edu.cn": "zx-du20@mails.tsinghua.edu.cn", 13 | "cwthornt@cs.ubc.ca": None, 14 | "maclaurin@physics.harvard.edu": "d.maclaurin@gmail.com", 15 | "ilya@cs.uni-freiburg.de": "ilya.loshchilov@gmail.com", 16 | "tapani.raiko@aalto.fi": None, 17 | "jelena.luketina@aalto.fi": "jelena.luketina@cs.ox.ac.uk", 18 | "anglili@google.com": "nju.angli@gmail.com", 19 | "kjamieson@eecs.berkeley.edu": "jamieson@cs.washington.edu", 20 | "rein.houthooft@openai.com": "rein.houthooft@gmail.com", 21 | "domhant@cs.uni-freiburg.de": "domhant@amazon.com", 22 | "krausea@inf.ethz.ch": "krausea@ethz.ch", 23 | "akshayc@cmu.edu": "web@akshayc.com", 24 | "james.bergstra@uwaterloo.ca": None, 25 | "aali@bournemouth.ac.uk": "abbas.raza.ali@gmail.com", 26 | "zhao.zhong@nlpr.ia.ac.cn": "zorro.zhongzhao@huawei.com", 27 | "kuanwang@mit.edu": "kuanwang@gatech.edu", 28 | "xiesirui@sensetime.com": "srxie@ucla.edu", 29 | "siemsj@cs.uni-freiburg.de": "siems@ifi.uzh.ch", 30 | "prajit@google.com": None, 31 | "mguan@stanford.edu": None, 32 | "hy-hieu@cmu.edu": "hyhieu@google.com", 33 | "guozichao@megvii.com": None, 34 | "ajb5@hw.ac.uk": "ajbrock@deepmind.com", 35 | "julian@idsia.ch": "julian@togelius.com", 36 | "izliobaite@bournemouth.ac.uk": "indre.zliobaite@gmail.com", 37 | "daan@idsia.ch": "wierstra@google.com", 38 | "otkrist@mit.edu": "otkrist.gupta@lendbuzz.com", 39 | "mohamed1.a@samsung.com": "mohamed@cornell.edu", 40 | "fe-lipe.such@gmail.com": "felipe.such@gmail.com", # TODO: fix in the raw file later 41 | "sungbin.lim@kakaobrain.com": "sungbin@unist.ac.kr", 42 | "imgemp@cs.umass.edu": "imgemp@google.com", 43 | "ksenia.konyushkova@epfl.ch": "kksenia@deepmind.com", 44 | "fetia@microsoft.com": "feitia@fb.com", 45 | } 46 | -------------------------------------------------------------------------------- /awesome_autodl/utils/yaml.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2020.08 # 3 | ##################################################### 4 | import yaml 5 | from pathlib import Path 6 | 7 | 8 | def load_yaml(file_path): 9 | with open(str(file_path), "r") as cfile: 10 | data = yaml.safe_load(cfile) 11 | return data 12 | 13 | 14 | def dump_yaml(data, indent=2, path=None): 15 | class NoAliasSafeDumper(yaml.SafeDumper): 16 | def ignore_aliases(self, data): 17 | return True 18 | 19 | xstr = yaml.dump( 20 | data, 21 | None, 22 | allow_unicode=True, 23 | Dumper=NoAliasSafeDumper, 24 | indent=indent, 25 | sort_keys=False, 26 | ) 27 | if path is not None: 28 | parent_dir = Path(path).resolve().parent 29 | parent_dir.mkdir(parents=True, exist_ok=True) 30 | with open(str(path), "w") as cfile: 31 | cfile.write(xstr) 32 | return xstr 33 | -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2021.09 # 3 | ##################################################### 4 | """The setup function for pypi.""" 5 | # The following is to make nats_bench avaliable on Python Package Index (PyPI) 6 | # 7 | # conda install -c conda-forge twine # Use twine to upload nats_bench to pypi 8 | # 9 | # python setup.py sdist bdist_wheel 10 | # python setup.py --help-commands 11 | # twine check dist/* 12 | # 13 | # twine upload --repository-url https://test.pypi.org/legacy/ dist/* 14 | # twine upload dist/* 15 | # https://pypi.org/project/awesome_autodl 16 | # 17 | # 18 | # or install from local: `pip install . --force` 19 | # 20 | import os 21 | import glob 22 | from pathlib import Path 23 | from setuptools import setup, find_packages 24 | from awesome_autodl import version 25 | 26 | NAME = "awesome_autodl" 27 | REQUIRES_PYTHON = ">=3.6" 28 | DESCRIPTION = "Package for Automated Deep Learning Paper Analysis" 29 | 30 | VERSION = version() 31 | 32 | 33 | def read(fname="README.md"): 34 | with open( 35 | os.path.join(os.path.dirname(__file__), fname), encoding="utf-8" 36 | ) as cfile: 37 | return cfile.read() 38 | 39 | 40 | # What packages are required for this module to be executed? 41 | REQUIRED = ["pyyaml>=5.0.0"] 42 | 43 | packages = find_packages(exclude=("tests", "checklist", "docs")) 44 | print(f"packages: {packages}") 45 | 46 | 47 | def endswith(xstring, targets): 48 | assert isinstance( 49 | targets, (list, tuple) 50 | ), f"invalid type of targets: {type(targets)}" 51 | for target in targets: 52 | if xstring.endswith(target): 53 | return True 54 | return False 55 | 56 | 57 | def recursive_find_file(xdir, cur_depth=1, max_depth=1, suffixs=None): 58 | assert isinstance(suffixs, (list, tuple)) and len( 59 | suffixs 60 | ), f"invalid suffixs of {suffixs}" 61 | xdirs = [] 62 | for xfile in Path(xdir).glob("*"): 63 | if xfile.is_dir() and cur_depth < max_depth: 64 | xdirs += recursive_find_file(xfile, cur_depth + 1, max_depth, suffixs) 65 | elif endswith(xfile.name, suffixs): 66 | xdirs.append(str(xfile)) 67 | return xdirs 68 | 69 | 70 | def find_yaml(xstring): 71 | return recursive_find_file(xstring, suffixs=(".yaml",)) 72 | 73 | 74 | def find_yaml_bib(xstring): 75 | return recursive_find_file(xstring, suffixs=(".yaml", ".bib")) 76 | 77 | 78 | setup( 79 | name=NAME, 80 | version=VERSION, 81 | author="Xuanyi Dong", 82 | author_email="dongxuanyi888@gmail.com", 83 | description=DESCRIPTION, 84 | license="MIT Licence", 85 | keywords="NAS Dataset API DeepLearning", 86 | url="https://github.com/D-X-Y/Awesome-AutoDL", 87 | include_package_data=True, 88 | packages=packages, 89 | package_data={ 90 | f"{NAME}/raw_data": find_yaml_bib(f"{NAME}/raw_data"), 91 | f"{NAME}/raw_data/papers": find_yaml(f"{NAME}/raw_data/papers"), 92 | }, 93 | install_requires=REQUIRED, 94 | python_requires=REQUIRES_PYTHON, 95 | long_description=read("README.md"), 96 | long_description_content_type="text/markdown", 97 | classifiers=[ 98 | "Programming Language :: Python", 99 | "Programming Language :: Python :: 3", 100 | "Topic :: Database", 101 | "Topic :: Scientific/Engineering :: Artificial Intelligence", 102 | "License :: OSI Approved :: MIT License", 103 | ], 104 | ) 105 | -------------------------------------------------------------------------------- /tests/test_abbrv.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.01 # 3 | ##################################################### 4 | # pytest ./tests/test_abbrv.py -s # 5 | ##################################################### 6 | import os 7 | from awesome_autodl import get_bib_abbrv_file 8 | from awesome_autodl.data_cls import BibAbbreviations 9 | 10 | 11 | class TestAbbrv: 12 | """Test the bib file for Abbreviations.""" 13 | 14 | def test_init(self): 15 | xfile = str(get_bib_abbrv_file()) 16 | assert os.path.isfile(xfile) 17 | obj = BibAbbreviations(xfile) 18 | assert len(obj) == 58 19 | print(obj) 20 | -------------------------------------------------------------------------------- /tests/test_format.py: -------------------------------------------------------------------------------- 1 | ##################################################### 2 | # Copyright (c) Xuanyi Dong [GitHub D-X-Y], 2022.01 # 3 | ##################################################### 4 | # pytest ./tests/test_format.py -s # 5 | ##################################################### 6 | from awesome_autodl import autodl_topic2papers 7 | from awesome_autodl import get_bib_abbrv_obj 8 | from awesome_autodl.data_cls import AutoDLpaper 9 | 10 | 11 | class TestFormat: 12 | """Test the format of the raw data.""" 13 | 14 | def test_simple(self): 15 | topic2papers = autodl_topic2papers() 16 | 17 | bib_abbrev = get_bib_abbrv_obj() 18 | for topic, papers in topic2papers.items(): 19 | print(f'Collect {len(papers)} papers for "{topic}"') 20 | for paper in papers: 21 | if paper.venue not in bib_abbrev: 22 | raise ValueError(f"Did not find {paper.venue} in {bib_abbrev}") 23 | --------------------------------------------------------------------------------