{"id":13712579,"url":"https://github.com/arthurdouillard/incremental_learning.pytorch","last_synced_at":"2025-04-05T14:07:07.551Z","repository":{"id":41390272,"uuid":"179149086","full_name":"arthurdouillard/incremental_learning.pytorch","owner":"arthurdouillard","description":"A collection of incremental learning paper implementations including PODNet (ECCV20) and Ghost (CVPR-W21).","archived":false,"fork":false,"pushed_at":"2023-10-03T21:36:06.000Z","size":6322,"stargazers_count":398,"open_issues_count":8,"forks_count":59,"subscribers_count":14,"default_branch":"master","last_synced_at":"2025-03-29T13:09:03.752Z","etag":null,"topics":["continual-learning","deep-learning","incremental-learning","lifelong-learning","pytorch","research"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/arthurdouillard.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.txt","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":"AUTHORS"}},"created_at":"2019-04-02T19:57:06.000Z","updated_at":"2025-03-22T01:27:25.000Z","dependencies_parsed_at":"2024-01-07T21:14:49.913Z","dependency_job_id":null,"html_url":"https://github.com/arthurdouillard/incremental_learning.pytorch","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arthurdouillard%2Fincremental_learning.pytorch","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arthurdouillard%2Fincremental_learning.pytorch/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arthurdouillard%2Fincremental_learning.pytorch/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arthurdouillard%2Fincremental_learning.pytorch/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/arthurdouillard","download_url":"https://codeload.github.com/arthurdouillard/incremental_learning.pytorch/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247345853,"owners_count":20924102,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["continual-learning","deep-learning","incremental-learning","lifelong-learning","pytorch","research"],"created_at":"2024-08-02T23:01:20.022Z","updated_at":"2025-04-05T14:07:07.534Z","avatar_url":"https://github.com/arthurdouillard.png","language":"Python","funding_links":[],"categories":["Table of Contents","Python"],"sub_categories":[],"readme":"# Incremental Learners for Continual Learning\n\nRepository storing some my public works done during my PhD thesis (2019-).\n\nYou will find in there both known implementation (iCaRL, etc.) but also all my papers.\nYou can find the list of the latter on my [Google Scholar](https://scholar.google.com/citations?user=snwgZBIAAAAJ\u0026hl=en).\n\nMy work on continual segmentation can be found [here](https://github.com/arthurdouillard/CVPR2021_PLOP) and on continual data loaders [here](https://github.com/Continvvm/continuum).\n\n## Structures\n\nEvery model must inherit `inclearn.models.base.IncrementalLearner`.\n\n\u003cdiv align=\"center\"\u003e\n\n# PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning\n\n[![Paper](https://img.shields.io/badge/arXiv-2004.13513-brightgreen)](https://arxiv.org/abs/2004.13513)\n![ECCV](https://img.shields.io/badge/ECCV-2020-blue)\n[![Youtube](https://img.shields.io/badge/Youtube-link-red)](https://www.youtube.com/watch?v=SWFO1_lTcR8)\n\n\u003c/div\u003e\n\n![podnet](images/podnet.png)\n\n![podnet plot](images/podnet_plot.png)\n\nIf you use this paper/code in your research, please consider citing us:\n\n```\n@inproceedings{douillard2020podnet,\n    title={PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning},\n    author={Douillard, Arthur and Cord, Matthieu and Ollion, Charles and Robert, Thomas and Valle, Eduardo},\n    booktitle={Proceedings of the IEEE European Conference on Computer Vision (ECCV)},\n    year={2020}\n}\n```\n\nTo run experiments on CIFAR100 with three different class orders, with the challenging\nsetting of 50 steps:\n\n```bash\npython3 -minclearn --options options/podnet/podnet_cnn_cifar100.yaml options/data/cifar100_3orders.yaml \\\n    --initial-increment 50 --increment 1 --fixed-memory \\\n    --device \u003cGPU_ID\u003e --label podnet_cnn_cifar100_50steps \\\n    --data-path \u003cPATH/TO/DATA\u003e\n```\n\nLikewise, for ImageNet100:\n\n```bash\npython3 -minclearn --options options/podnet/podnet_cnn_imagenet100.yaml options/data/imagenet100_1order.yaml \\\n    --initial-increment 50 --increment 1 --fixed-memory \\\n    --device \u003cGPU_ID\u003e --label podnet_cnn_imagenet100_50steps \\\n    --data-path \u003cPATH/TO/DATA\u003e\n```\n\nAnd ImageNet1000:\n\nLikewise, for ImageNet100:\n\n```bash\npython3 -minclearn --options options/podnet/podnet_cnn_imagenet100.yaml options/data/imagenet1000_1order.yaml \\\n    --initial-increment 500 --increment 50 --fixed-memory --memory-size 20000 \\\n    --device \u003cGPU_ID\u003e --label podnet_cnn_imagenet1000_10steps \\\n    --data-path \u003cPATH/TO/DATA\u003e\n```\n\nFurthermore several options files are available to reproduce the ablations showcased\nin the paper. Please see the directory `./options/podnet/ablations/`.\n\n\u003cdiv align=\"center\"\u003e\n\n# Insight From the Future for Continual Learning\n\n[![Paper](https://img.shields.io/badge/arXiv-2006.13748-brightgreen)](https://arxiv.org/abs/2006.13748)\n![CVPR Workshop](https://img.shields.io/badge/CVPRW-2021-blue)\n\n\u003c/div\u003e\n\n![ghost](images/ghost.png)\n\nIf you use this paper/code in your research, please consider citing us:\n\n```\n@inproceedings{douillard2020ghost,\n    title={Insight From the Future for Continual Learning},\n    author={Arthur Douillard and Eduardo Valle and Charles Ollion and Thomas Robert and Matthieu Cord},\n    booktitle={arXiv preprint library},\n    year={2020}\n}\n```\n\nThe code is still very dirty, I'll clean it later. Forgive me.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Farthurdouillard%2Fincremental_learning.pytorch","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Farthurdouillard%2Fincremental_learning.pytorch","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Farthurdouillard%2Fincremental_learning.pytorch/lists"}