{"url":"/method/population-based-training","slug":"population-based-training","name":"Population Based Training","full_name":"Population Based Training","full_name_withheld":false,"description_markdown":"**Population Based Training**, or **PBT**, is an optimization method for finding parameters and hyperparameters, and extends upon parallel search methods and sequential optimisation methods.\r\nIt leverages information sharing across a population of concurrently running optimisation processes, and allows for online propagation/transfer of parameters and hyperparameters between members of the population based on their performance. Furthermore, unlike most other adaptation schemes, the method is capable of performing online adaptation of hyperparameters -- which can be particularly important in problems with highly non-stationary learning dynamics, such as reinforcement learning settings. PBT is decentralised and asynchronous, although it could also be executed semi-serially or with partial synchrony if there is a binding budget constraint.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Population Based Training of Neural Networks","paper":"/paper/population-based-training-of-neural-networks","first_author":"Max Jaderberg","n_authors":12,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/population-based-training-of-neural-networks"},"source":{"url":"http://arxiv.org/abs/1711.09846v2","title":"Population Based Training of Neural Networks","url_on_a_paper_host":true},"code_snippet_url":"https://github.com/elsheikh21/population-based-training-of-NNs/blob/8337041e31e1a1be9b1cc36bdf79fe7921881116/pbt_img_classification/main.py#L163","code_snippet_url_on_a_code_host":true,"categories":[{"area":"General","area_id":"general","collection":"Hyperparameter Search","url":"/methods/category/hyperparameter-search","pwc_aliases":[]},{"area":"General","area_id":"general","collection":"Optimization","url":"/methods/category/optimization","pwc_aliases":[]}],"n_papers_tagged":18,"archive_num_papers":18,"papers_newest_first":[{"paper":"/paper/shrink-perturb-improves-architecture-mixing","title":"Shrink-Perturb Improves Architecture Mixing during Population Based Training for Neural Architecture Search","date":"2023-07-28","arxiv_id":"2307.15621","n_code_links":1,"syntology":null},{"paper":"/paper/multi-objective-population-based-training","title":"Multi-Objective Population Based Training","date":"2023-06-02","arxiv_id":"2306.01436","n_code_links":1,"syntology":{"ran":0,"of":1,"unverified":1,"pointer_only":0}},{"paper":"/paper/reducing-exploitability-with-population-based","title":"Reducing Exploitability with Population Based Training","date":"2022-08-10","arxiv_id":"2208.05083","n_code_links":1,"syntology":{"ran":0,"of":5,"unverified":5,"pointer_only":0}},{"paper":null,"title":"Faster Improvement Rate Population Based Training","date":"2021-09-28","arxiv_id":"2109.13800","n_code_links":0,"syntology":null},{"paper":null,"title":"Automated Graph Learning via Population Based Self-Tuning GCN","date":"2021-07-09","arxiv_id":"2107.04713","n_code_links":0,"syntology":null},{"paper":null,"title":"Tuning Mixed Input Hyperparameters on the Fly for Efficient Population Based AutoRL","date":"2021-06-30","arxiv_id":"2106.15883","n_code_links":0,"syntology":null},{"paper":null,"title":"Training Learned Optimizers with Randomly Initialized Learned Optimizers","date":"2021-01-14","arxiv_id":"2101.07367","n_code_links":0,"syntology":null},{"paper":null,"title":"Population Based Training for Data Augmentation and Regularization in Speech Recognition","date":"2020-10-08","arxiv_id":"2010.03899","n_code_links":0,"syntology":null},{"paper":null,"title":"Improving 3D Object Detection through Progressive Population Based Augmentation","date":"2020-04-02","arxiv_id":"2004.00831","n_code_links":0,"syntology":null},{"paper":"/paper/accelerating-and-improving-alphazero-using","title":"Accelerating and Improving AlphaZero Using Population Based Training","date":"2020-03-13","arxiv_id":"2003.06212","n_code_links":1,"syntology":null},{"paper":"/paper/dada-differentiable-automatic-data","title":"DADA: Differentiable Automatic Data Augmentation","date":"2020-03-08","arxiv_id":"2003.03780","n_code_links":1,"syntology":{"ran":5,"of":9,"unverified":4,"pointer_only":0}},{"paper":null,"title":"Data Efficient Training for Reinforcement Learning with Adaptive Behavior Policy Sharing","date":"2020-02-12","arxiv_id":"2002.05229","n_code_links":0,"syntology":null},{"paper":"/paper/one-shot-bayes-opt-with-probabilistic","title":"Provably Efficient Online Hyperparameter Optimization with Population-Based Bandits","date":"2020-02-06","arxiv_id":"2002.02518","n_code_links":2,"syntology":{"ran":2,"of":5,"unverified":3,"pointer_only":0}},{"paper":"/paper/fast-efficient-hyperparameter-tuning-for-1","title":"Fast Efficient Hyperparameter Tuning for Policy Gradient Methods","date":"2019-12-01","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":"/paper/190505393","title":"Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules","date":"2019-05-14","arxiv_id":"1905.05393","n_code_links":3,"syntology":{"ran":0,"of":18,"unverified":18,"pointer_only":18}},{"paper":null,"title":"Semi-supervised and Population Based Training for Voice Commands Recognition","date":"2019-05-10","arxiv_id":"1905.04230","n_code_links":0,"syntology":null},{"paper":"/paper/fast-efficient-hyperparameter-tuning-for","title":"Fast Efficient Hyperparameter Tuning for Policy Gradients","date":"2019-02-18","arxiv_id":"1902.06583","n_code_links":1,"syntology":{"ran":0,"of":1,"unverified":1,"pointer_only":1}},{"paper":"/paper/population-based-training-of-neural-networks","title":"Population Based Training of Neural Networks","date":"2017-11-27","arxiv_id":"1711.09846","n_code_links":9,"syntology":null}],"papers_shown":18,"tasks":[{"task":"/task/data-augmentation","name":"Data Augmentation","papers":5},{"task":"/task/hyperparameter-optimization","name":"Hyperparameter Optimization","papers":5},{"task":"/task/reinforcement-learning","name":"Reinforcement Learning","papers":4},{"task":"/task/reinforcement-learning-1","name":"Reinforcement Learning (RL)","papers":4},{"task":"/task/classification-1","name":"Classification","papers":2},{"task":"/task/deep-reinforcement-learning","name":"Deep Reinforcement Learning","papers":2},{"task":"/task/policy-gradient-methods","name":"Policy Gradient Methods","papers":2},{"task":"/task/3d-object-detection","name":"3D Object Detection","papers":1},{"task":"/task/adversarial-robustness","name":"Adversarial Robustness","papers":1},{"task":"/task/atari-games","name":"Atari Games","papers":1},{"task":"/task/decision-making","name":"Decision Making","papers":1},{"task":"/task/diversity","name":"Diversity","papers":1},{"task":"/task/fairness","name":"Fairness","papers":1},{"task":"/task/classification","name":"General Classification","papers":1},{"task":"/task/graph-classification","name":"Graph Classification","papers":1},{"task":"/task/graph-learning","name":"Graph Learning","papers":1},{"task":"/task/image-augmentation","name":"Image Augmentation","papers":1},{"task":"/task/image-generation","name":"Image Generation","papers":1},{"task":"/task/link-prediction","name":"Link Prediction","papers":1},{"task":"/task/machine-translation","name":"Machine Translation","papers":1}],"tasks_shown":20,"n_tasks":32,"usage_by_year":[{"year":"2017","papers":1},{"year":"2019","papers":4},{"year":"2020","papers":6},{"year":"2021","papers":4},{"year":"2022","papers":1},{"year":"2023","papers":2}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/population-based-training"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}