{"id":28653922,"url":"https://github.com/tiger-ai-lab/mammoth2","last_synced_at":"2025-06-13T07:08:12.483Z","repository":{"id":238209087,"uuid":"795757796","full_name":"TIGER-AI-Lab/MAmmoTH2","owner":"TIGER-AI-Lab","description":"Official code for \"MAmmoTH2: Scaling Instructions from the Web\" [NeurIPS 2024]","archived":false,"fork":false,"pushed_at":"2024-10-27T03:18:55.000Z","size":20764,"stargazers_count":137,"open_issues_count":4,"forks_count":10,"subscribers_count":3,"default_branch":"main","last_synced_at":"2025-04-11T13:52:50.757Z","etag":null,"topics":["language","math","reasoning"],"latest_commit_sha":null,"homepage":"https://tiger-ai-lab.github.io/MAmmoTH2/","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/TIGER-AI-Lab.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}},"created_at":"2024-05-04T01:23:39.000Z","updated_at":"2025-03-24T02:12:23.000Z","dependencies_parsed_at":"2024-11-07T06:40:56.851Z","dependency_job_id":null,"html_url":"https://github.com/TIGER-AI-Lab/MAmmoTH2","commit_stats":null,"previous_names":["tiger-ai-lab/mammoth2"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/TIGER-AI-Lab/MAmmoTH2","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TIGER-AI-Lab%2FMAmmoTH2","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TIGER-AI-Lab%2FMAmmoTH2/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TIGER-AI-Lab%2FMAmmoTH2/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TIGER-AI-Lab%2FMAmmoTH2/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/TIGER-AI-Lab","download_url":"https://codeload.github.com/TIGER-AI-Lab/MAmmoTH2/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TIGER-AI-Lab%2FMAmmoTH2/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":259599330,"owners_count":22882357,"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":["language","math","reasoning"],"created_at":"2025-06-13T07:08:10.406Z","updated_at":"2025-06-13T07:08:12.419Z","avatar_url":"https://github.com/TIGER-AI-Lab.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# MAmmoTH2\n\nThis repo contains the code, data, and models for NeurIPS-24 paper \"[MAmmoTH2: Scaling Instructions from the Web](https://arxiv.org/abs/2405.03548)\". Our paper proposes a new paradigm to scale up high-quality instruction data from the web.\n\n\u003cdiv align=\"center\"\u003e\n 🔥 🔥 🔥 Check out our \u003ca href = \"https://tiger-ai-lab.github.io/MAmmoTH2/\"\u003e[Project Page]\u003c/a\u003e for more results and analysis! Also, our \u003ca href=\"https://huggingface.co/spaces/TIGER-Lab/MAmmoTH2\"\u003eDemo\u003c/a\u003e is online!\n\u003c/div\u003e\n\n\n\u003cp align=\"center\"\u003e\n\u003cdiv style=\"display: flex; justify-content: center;\"\u003e\n    \u003cimg src=\"https://tiger-ai-lab.github.io/MAmmoTH2/static/images/teaser_front.jpg\" width:auto; margin-right:10px\"\u003e\n\u003c/div\u003e\n\u003c/p\u003e\n\n## WebInstruct\nWe propose discovering instruction data from the web. We argue that vast amounts of high-quality instruction data exist in the web corpus, spanning various domains like math and science. Our three-step pipeline involves recalling documents from Common Crawl, extracting Q-A pairs, and refining them for quality. This approach yields 10 million instruction-response pairs, offering a scalable alternative to existing datasets. We name our curated dataset as WebInstruct.\n\n\u003cp align=\"center\"\u003e\n\n\u003cdiv style=\"display: flex; justify-content: center;\"\u003e\n    \u003cimg src=\"https://tiger-ai-lab.github.io/MAmmoTH2/static/images/teaser.jpg\" width:auto; margin-right:10px\"\u003e\n\u003c/div\u003e\n\u003c/p\u003e\n\nPart of our WebInstruct dataset has been released at [🤗 TIGER-Lab/WebInstructSub](https://huggingface.co/datasets/TIGER-Lab/WebInstructSub) and [🤗 TIGER-Lab/WebInstructFull](https://huggingface.co/datasets/TIGER-Lab/WebInstructFull).\n\n## Model Downloads\n\n\u003cdiv align=\"center\"\u003e\n\n| **Model**            | **Dataset**                                            | **Init Model** | **Download**   |\n| :------------:       | :------------:                                         | :------------: | :------------: |\n| MAmmoTH2-8x7B        | WebInstruct                                            | Mixtral-8x7B   | [🤗 HuggingFace](https://huggingface.co/TIGER-Lab/MAmmoTH2-8x7B)   |\n| MAmmoTH2-7B          | WebInstruct                                            | Mistral-7B-v0.2| [🤗 HuggingFace](https://huggingface.co/TIGER-Lab/MAmmoTH2-7B)   |\n| MAmmoTH2-8B          | WebInstruct                                            | Llama-3-base   | [🤗 HuggingFace](https://huggingface.co/TIGER-Lab/MAmmoTH2-8B)   |\n| MAmmoTH2-8x7B-Plus   | WebInstruct + OpenHermes2.5 + CodeFeedback + Math-Plus | MAmmoTH2-8x7B  | [🤗 HuggingFace](https://huggingface.co/TIGER-Lab/MAmmoTH2-8x7B-Plus)   |\n| MAmmoTH2-7B-Plus     | WebInstruct + OpenHermes2.5 + CodeFeedback + Math-Plus | MAmmoTH2-7B    | [🤗 HuggingFace](https://huggingface.co/TIGER-Lab/MAmmoTH2-7B-Plus)   |\n| MAmmoTH2-8B-Plus     | WebInstruct + OpenHermes2.5 + CodeFeedback + Math-Plus | MAmmoTH2-8B    | [🤗 HuggingFace](https://huggingface.co/TIGER-Lab/MAmmoTH2-8-Plus)   |\n\n\u003c/div\u003e\n\n## Evaluation Results\n\nPlease refer to https://tiger-ai-lab.github.io/MAmmoTH2/ for more details.\n\n## Evaluation Command\nPlease refer to https://github.com/TIGER-AI-Lab/MAmmoTH2/tree/main/math_eval. \n\n## Cite our paper\nPlease cite our paper if you use our data, model or code. Please also kindly cite the original dataset papers.\n```\n@article{yue2024mammoth2,\n  title={MAmmoTH2: Scaling Instructions from the Web},\n  author={Yue, Xiang and Zheng, Tuney and Zhang, Ge and Chen, Wenhu},\n  journal={Advances in Neural Information Processing Systems},\n  year={2024}\n}\n```\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftiger-ai-lab%2Fmammoth2","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftiger-ai-lab%2Fmammoth2","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftiger-ai-lab%2Fmammoth2/lists"}