{"data":{"slug":"flagai-open-flagai","name":"FlagAI","tagline":"Fast, easy-to-use framework for large-scale AI models.","github_url":"https://github.com/FlagAI-Open/FlagAI","owner":"FlagAI-Open","repo":"FlagAI","owner_avatar_url":"https://avatars.githubusercontent.com/u/109567851?v=4","primary_language":"Python","stars":3870,"forks":416,"topics":[],"archived":false,"github_pushed_at":"2026-07-13T03:53:45+00:00","maintenance_label":"Steady","stars_delta_30d":2,"url":"https://www.graphcanon.com/tools/flagai-open-flagai","markdown_url":"https://www.graphcanon.com/tools/flagai-open-flagai.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/flagai-open-flagai","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=flagai-open-flagai","description":"FlagAI (Fast LArge-scale General AI models) is a fast, easy-to-use and extensible toolkit for large-scale model.","homepage_url":null,"license":"Apache-2.0","open_issues":22,"watchers":38,"ai_summary":"FlagAI provides tools for developing and managing large-scale AI models with an emphasis on speed and extensibility.","readme_excerpt":"## Quick Start\nWe provide many models which are trained to perform different tasks. You can load these models by AutoLoader to make prediction. See more in `FlagAI/quickstart`.\n\n---\n\n### Requirements and Installation\n* Python version >= 3.8\n* PyTorch version >= 1.8.0\n* [Optional] For training/testing models on GPUs, you'll also need to install CUDA and NCCL\n\n- To install FlagAI with pip:\n```shell\npip install -U flagai\n```\n\n- [Optional] To install FlagAI and develop locally:\n\n```shell\ngit clone https://github.com/FlagAI-Open/FlagAI.git\npython setup.py install\n```\n\n- [Optional] For faster training, install NVIDIA's [apex](https://github.com/NVIDIA/apex)\n```\ngit clone https://github.com/NVIDIA/apex\ncd apex\npip install -v --disable-pip-version-check --no-cache-dir --global-option=\"--cpp_ext\" --global-option=\"--cuda_ext\" ./\n```\n- [Optional] For ZeRO optimizers, install [DEEPSPEED](https://github.com/microsoft/DeepSpeed) (>= 0.7.7)\n```\ngit clone https://github.com/microsoft/DeepSpeed\ncd DeepSpeed\nDS_BUILD_CPU_ADAM=1 DS_BUILD_AIO=1 DS_BUILD_UTILS=1 pip install -e .\nds_report # check the deespeed status\n```\n- [Optional] For BMTrain training, install [BMTrain](https://github.com/OpenBMB/BMTrain) (>= 0.2.2)\n```\ngit clone https://github.com/OpenBMB/BMTrain\ncd BMTrain\npython setup.py install\n```\n- [Optional] For BMInf low-resource inference, install [BMInf](https://github.com/OpenBMB/BMInf)\n```\npip install bminf\n\n```\n- [Optional] For Flash Attention, install [Flash-attention](https://github.com/HazyResearch/flash-attention) (>=1.0.2)\n```\npip install flash-attn\n```\n\n- [Tips] For single-node docker environments, we need to set up ports for your ssh. e.g., root@127.0.0.1 with port 711\n```\n>>> vim ~/.ssh/config\nHost 127.0.0.1\n    Hostname 127.0.0.1\n    Port 7110\n    User root\n```\n- [Tips] For multi-node docker environments, generate ssh keys and copy the public key to all nodes (in `~/.ssh/`)\n```\n>>> ssh-keygen -t rsa -C \"xxx@xxx.com\"\n```\n\n---\n\n## LICENSE\n\nThe majority of FlagAI is licensed under the [Apache 2.0 license](https://www.apache.org/licenses/LICENSE-2.0), however portions of the project are available under separate license terms:\n\n* Megatron-LM is licensed under the [Megatron-LM license](https://github.com/NVIDIA/Megatron-LM/blob/main/LICENSE)\n* GLM is licensed under the [MIT license](https://github.com/THUDM/GLM/blob/main/LICENSE)\n* AltDiffusion is licensed under the [CreativeML Open RAIL-M license](https://huggingface.co/spaces/CompVis/stable-diffusion-license)","github_created_at":"2022-05-16T06:43:12+00:00","created_at":"2026-07-11T10:54:20.454902+00:00","updated_at":"2026-08-15T18:01:12.71347+00:00","categories":[{"slug":"llm-frameworks","name":"LLM Frameworks","url":"https://www.graphcanon.com/categories/llm-frameworks","markdown_url":"https://www.graphcanon.com/categories/llm-frameworks.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/llm-frameworks"},{"slug":"model-training","name":"Model Training","url":"https://www.graphcanon.com/categories/model-training","markdown_url":"https://www.graphcanon.com/categories/model-training.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/model-training"}],"tags":[{"slug":"extensible","name":"extensible"},{"slug":"fast","name":"fast"},{"slug":"large-scale-models","name":"large-scale models"}],"trust":{"provenance":{"is_fork":false,"github_id":492713876,"owner_type":"Organization","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-08-15T18:01:11.829Z","maintenance":{"label":"Steady","score":60,"methodology":"github_public_v1","releases_90d":0,"days_since_push":33,"last_release_at":"2024-11-07T07:15:17Z","stars_delta_30d":2,"open_issues_delta_30d":0},"security_summary":{"status":"no_lockfile","scanner":null,"low_count":0,"high_count":0,"last_scan_at":"2026-07-11T10:54:21.552Z","medium_count":0,"scan_profile":"none","critical_count":0}},"capability_facts":{"scan":{"source":"repo_scan","observed_at":"2026-08-15T18:01:12.381Z"},"deploy":{"source":"dockerfile:Dockerfile","self_host":true,"observed_at":"2026-08-15T18:01:12.381Z","managed_saas":false},"languages":{"value":["python"],"source":"github.language","observed_at":"2026-08-15T18:01:12.381Z"},"has_docker":{"value":true,"source":"dockerfile:Dockerfile","observed_at":"2026-08-15T18:01:12.381Z"},"license_spdx":{"value":"Apache-2.0","source":"github.license","observed_at":"2026-08-15T18:01:12.381Z"}},"decision_facts":{"hosting":null,"pricing":null,"requirements":null,"constraints":null,"when_to_use":["When you prioritize speed and extensibility during the development of large-scale AI models with a focus on easy-to-use interfaces.","If your team is skilled in Python and you need to explore or expand capabilities rapidly without compromising performance."],"when_not_to_use":["If your project necessitates a deep level of customization that might not be supported by FlagAI's framework.","If you are working with smaller datasets, the overhead provided by FlagAI’s scalability features could be unnecessary and potentially inefficient."],"source":"enrich:decision_facts","observed_at":"2026-07-14T19:18:50.187Z"},"constraint_facets":null,"decision_summary":[{"label":"Adopt for","value":"FlagAI is identified by its fast and scalable toolkit designed for managing large-scale AI models in Python, under an Apache-2.0 license."}]}}