{"id":18421845,"url":"https://github.com/adrien-lagesse/gnnco","last_synced_at":"2026-05-04T07:36:16.751Z","repository":{"id":231190325,"uuid":"781183337","full_name":"adrien-lagesse/gnnco","owner":"adrien-lagesse","description":"gnnco is a package that simplifies benchmarking and training GNNs on Combinatorial Optimization (CO) tasks.","archived":false,"fork":false,"pushed_at":"2024-08-28T12:15:38.000Z","size":306,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-04-13T14:12:41.363Z","etag":null,"topics":["benchmark","combinatorial-optimization","deep-learning","graph-neural-networks","machine-learning","pytorch"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/adrien-lagesse.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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":"2024-04-02T22:39:41.000Z","updated_at":"2024-08-28T12:15:41.000Z","dependencies_parsed_at":"2024-04-15T16:12:28.845Z","dependency_job_id":"fbd3d5d4-c6cd-46db-ae6f-e704a1b9b888","html_url":"https://github.com/adrien-lagesse/gnnco","commit_stats":null,"previous_names":["adrien-lagesse/gnnco"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/adrien-lagesse/gnnco","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/adrien-lagesse%2Fgnnco","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/adrien-lagesse%2Fgnnco/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/adrien-lagesse%2Fgnnco/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/adrien-lagesse%2Fgnnco/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/adrien-lagesse","download_url":"https://codeload.github.com/adrien-lagesse/gnnco/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/adrien-lagesse%2Fgnnco/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32599349,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-03T22:12:39.696Z","status":"online","status_checked_at":"2026-05-04T02:00:06.625Z","response_time":58,"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":["benchmark","combinatorial-optimization","deep-learning","graph-neural-networks","machine-learning","pytorch"],"created_at":"2024-11-06T04:27:01.916Z","updated_at":"2026-05-04T07:36:16.731Z","avatar_url":"https://github.com/adrien-lagesse.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Benchmarking and Training GNNs with Combinatorial Optimization\n\nThe `gnnco` package simplifies benchmarking and training GNNs on task issued from Combinatorial Optimization (CO).\nIt is based on PyTorch and the default models are written using the Pytorch Geometric package.\n\nSeveral functionalities are provided:\n- Generating CO Datasets\n- Using existing CO Dataset\n- A framework to benchmark different GNNs architecture\n- Using pretrained GNNs for generating Graph Positional Encodings\n\n\n# Running the Repo\n\nWe use [Rye](https://rye.astral.sh/) to manage the python project. See the documentation for a complete guide.\n\n### Quick installation (Linux and MacOS)\n\n`curl -sSf https://rye.astral.sh/get | bash`\n\n`echo 'source \"$HOME/.rye/env\"' \u003e\u003e ~/.profile    # For Bash`\n\n`echo 'source \"$HOME/.rye/env\"' \u003e\u003e ~/.zprofile   # For ZSH`\n\nYou may have to restart you shell.\n\n### Cloning the repo\n\n`git clone https://github.com/adrien-lagesse/gnnco.git`\n\n`cd gnnco`\n\n`rye sync`\n\n`rye list`\n\nYou sould have a list of all the dependencies of the project.\n\n\n# Graph Matching problem for benchmarking\n\n## Dataset generation\n\nWe provide several command line application to generate graph matching datasets:\n\n- **gm-generate-er** : Generate Erdos-Renyi GM datasets.\n- **gm-generate-karateclub** : Generate GM datasets based on the KarateClub Benchmark dataset.\n- **gm-generate-corafull** : Generate GM datasets based on the CoraFull Benchmark dataset.\n- **gm-generate-aqsol** : Generate GM datasets based on the AQSOL dataset.\n\nTo know more about them run:\n\n`gm-generate-er --help`\n\n`gm-generate-karateclub --help`\n\n`gm-generate-corafull --help`\n\n`gm-generate-qm7b --help`\n\nOnce you have a dataset, you can print key statistics with `gm-data-stats`\n\n## Training\n\nUse the `gm-train` command line tool to train a Siamese Graph Matching model. (run `gm-train --help` for more information and see scripts/train-siamese-gm.sh for an example)\n\n\n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fadrien-lagesse%2Fgnnco","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fadrien-lagesse%2Fgnnco","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fadrien-lagesse%2Fgnnco/lists"}