{"id":28988705,"url":"https://github.com/eidoslab/neq","last_synced_at":"2026-07-07T09:31:45.322Z","repository":{"id":82117174,"uuid":"464442949","full_name":"EIDOSLAB/NEq","owner":"EIDOSLAB","description":"Official implementation for \"To update or not to update? Neurons at equilibrium in deep models\"","archived":false,"fork":false,"pushed_at":"2023-03-30T08:05:11.000Z","size":2164,"stargazers_count":2,"open_issues_count":0,"forks_count":2,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-08-07T14:50:09.621Z","etag":null,"topics":[],"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/EIDOSLAB.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2022-02-28T10:48:47.000Z","updated_at":"2024-02-28T12:55:20.000Z","dependencies_parsed_at":null,"dependency_job_id":"746fc302-cd6d-400c-9beb-5ff41517114a","html_url":"https://github.com/EIDOSLAB/NEq","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/EIDOSLAB/NEq","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/EIDOSLAB%2FNEq","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/EIDOSLAB%2FNEq/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/EIDOSLAB%2FNEq/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/EIDOSLAB%2FNEq/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/EIDOSLAB","download_url":"https://codeload.github.com/EIDOSLAB/NEq/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/EIDOSLAB%2FNEq/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35223382,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-07T02:00:07.222Z","response_time":90,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":[],"created_at":"2025-06-24T22:06:37.419Z","updated_at":"2026-07-07T09:31:45.294Z","avatar_url":"https://github.com/EIDOSLAB.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"[![paper](https://img.shields.io/badge/NeurIPS-paper-blue)](https://openreview.net/pdf?id=LGDfv0U7MJR)\n\n# To update or not to update? Neurons at equilibrium in deep models - NeurIPS 2022\n\nOfficial repository for NEq as presented at the NeurIPS 2022 conference.\n\n![teaser](nomenclature.png)\n\n## Requirements\n\nPython==3.9 \\\ntqdm==4.62.3 \\\nnumpy==1.19.5 \\\nfilelock==3.2.0 \\\ntorch==1.10.1 \\\ntorchvision==0.11.2\n\n## Reproduce our results\n\n### CIFAR-10 - ResNet-32\n\n`python3 train_classification.py --amp=1 --arch=resnet32-cifar --batch-size=100 --dataset=cifar10 --device=cuda --epochs=250 --eps=0.001 --lr=0.1 --momentum=0.9 --optim=sgd --val-size=0.01 --velocity-mu=0.5 --weight-decay=0.0005`\n\n### ImageNet - ResNet-18\n\n`python3 train_classification.py --amp=1 --arch=resnet18-imagenet --batch-size=128 --dataset=imagenet --device=cuda --epochs=90 --eps=0.001 --lr=0.1 --momentum=0.9 --optim=sgd --val-size=0.001 --velocity-mu=0.5 --weight-decay=0.0005`\n\n### ImageNet - SwinB\n\nFrom inside the SwinTransformer sub-dir\n\n`python3 main.py --local_rank=0 --cfg configs/swin_base_patch4_window7_224_22kto1k_finetune.yaml --batch-size 64 --accumulation-steps 2 --val-size 0.001 --eps 1e-3 --velocity-mu 0.5`\n\n### COCO - Deeplabv3\n\n`python3 train_segmentation.py --lr 0.02 --dataset coco --arch deeplabv3_resnet50 --aux-loss --weights-backbone ResNet50_Weights.IMAGENET1K_V1 --val-size 0.001 --eps 2e-2 --batch-size 32 --amp --velocity-mu 0.5`\n\n# Citation\nPlease cite this work as\n```\n@inproceedings{\nbragagnolo2022to,\ntitle={To update or not to update? Neurons at equilibrium in deep models},\nauthor={Andrea Bragagnolo and Enzo Tartaglione and Marco Grangetto},\nbooktitle={Advances in Neural Information Processing Systems},\neditor={Alice H. Oh and Alekh Agarwal and Danielle Belgrave and Kyunghyun Cho},\nyear={2022},\nurl={https://openreview.net/forum?id=LGDfv0U7MJR}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Feidoslab%2Fneq","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Feidoslab%2Fneq","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Feidoslab%2Fneq/lists"}