{"id":20226793,"url":"https://github.com/aastroza/structured-generation-benchmark","last_synced_at":"2025-09-22T04:31:39.350Z","repository":{"id":231650508,"uuid":"782325190","full_name":"aastroza/structured-generation-benchmark","owner":"aastroza","description":"Structured Generation Evals","archived":false,"fork":false,"pushed_at":"2024-09-25T17:50:38.000Z","size":20203,"stargazers_count":12,"open_issues_count":0,"forks_count":0,"subscribers_count":3,"default_branch":"main","last_synced_at":"2025-06-03T14:51:56.098Z","etag":null,"topics":["generative-ai","instructor","llms","modal","outlines","structured-generation","transformers"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/aastroza.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":"2024-04-05T04:34:43.000Z","updated_at":"2025-01-18T16:27:14.000Z","dependencies_parsed_at":"2025-01-13T23:33:00.286Z","dependency_job_id":"91d3d6a1-6cac-442a-b52d-ec85db2e729d","html_url":"https://github.com/aastroza/structured-generation-benchmark","commit_stats":null,"previous_names":["aastroza/guided-generation-benchmark"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/aastroza/structured-generation-benchmark","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aastroza%2Fstructured-generation-benchmark","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aastroza%2Fstructured-generation-benchmark/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aastroza%2Fstructured-generation-benchmark/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aastroza%2Fstructured-generation-benchmark/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/aastroza","download_url":"https://codeload.github.com/aastroza/structured-generation-benchmark/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aastroza%2Fstructured-generation-benchmark/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":276346716,"owners_count":25626496,"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","status":"online","status_checked_at":"2025-09-22T02:00:08.972Z","response_time":79,"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":["generative-ai","instructor","llms","modal","outlines","structured-generation","transformers"],"created_at":"2024-11-14T07:20:18.511Z","updated_at":"2025-09-22T04:31:37.978Z","avatar_url":"https://github.com/aastroza.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# structured-generation-benchmark\n\nTo use Large Language Models (LLMs) effectively and reliably, it's essential to include structured generation techniques. Being able to get outputs like regular expressions, JSON, or a Pydantic data model is key for making useful software.\n\nBut what's the real effect of using libraries like [Outlines](https://github.com/outlines-dev/outlines) or [Instructor](https://github.com/jxnl/instructor/) to achieve that goal?\n\nThis repository has put together evaluations to answer this question.\n\n## Function Calling\n\nThe ability of the LLM to call functions.\n\n### Datasets\n\n- [Berkeley Function Calling Leaderboard](https://huggingface.co/datasets/gorilla-llm/Berkeley-Function-Calling-Leaderboard/tree/64d44ccd13f3351d17c33951af5ef1bd6e10153c) [April 16, 2024 update]\n\n### Evaluation\n\n- We deployed a [modal function](modal/transformers_outlines.py) to run open-source models using [Transformers](https://github.com/huggingface/transformers) + [Outlines](https://github.com/outlines-dev/outlines).\n- We created different [model handlers](evals/bfcl/scripts) to run the [Gorilla BFCL scripts](https://github.com/ShishirPatil/gorilla/tree/c6221060a9d50d0c7e7705f1ac95b9e5c4a95252) [April 6, 2024 version] for the `AST simple` evaluation category.\n- We [evaluated](evals/bfcl/score) and reported the [results](evals/bfcl/result) comparing them with the [Leaderboard Website](https://github.com/ShishirPatil/gorilla/tree/46e959b73be6a40c233e36c71c268ce3a9eabe36) [April 26, 2024 version].\n\n### Reports\n\n- [Outlines Function Calling Evaluation](reports/bfcl_outlines.md)\n- [Instructor Function Calling Evaluation](reports/bfcl_instructor.md)\n\n## Synthetic Data Generation\n\nUsing an LLM to create artificial data.\n\n### Reports\n\n- [Outlines Synthetic Data Generation](reports/sdg_outlines.md)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faastroza%2Fstructured-generation-benchmark","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faastroza%2Fstructured-generation-benchmark","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faastroza%2Fstructured-generation-benchmark/lists"}