{"id":21135239,"url":"https://github.com/cloneofsimo/scaling-guide","last_synced_at":"2025-10-29T08:31:12.883Z","repository":{"id":250850446,"uuid":"835612619","full_name":"cloneofsimo/scaling-guide","owner":"cloneofsimo","description":"WIP","archived":false,"fork":false,"pushed_at":"2024-08-13T19:10:09.000Z","size":272,"stargazers_count":93,"open_issues_count":0,"forks_count":1,"subscribers_count":9,"default_branch":"master","last_synced_at":"2025-02-07T00:09:28.204Z","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":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/cloneofsimo.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}},"created_at":"2024-07-30T07:28:05.000Z","updated_at":"2025-01-25T16:56:43.000Z","dependencies_parsed_at":"2024-08-06T18:26:10.071Z","dependency_job_id":null,"html_url":"https://github.com/cloneofsimo/scaling-guide","commit_stats":null,"previous_names":["cloneofsimo/scaling-guide"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cloneofsimo%2Fscaling-guide","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cloneofsimo%2Fscaling-guide/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cloneofsimo%2Fscaling-guide/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cloneofsimo%2Fscaling-guide/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/cloneofsimo","download_url":"https://codeload.github.com/cloneofsimo/scaling-guide/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":238791969,"owners_count":19531028,"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":[],"created_at":"2024-11-20T06:46:50.328Z","updated_at":"2025-10-29T08:31:07.378Z","avatar_url":"https://github.com/cloneofsimo.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Almost Comprehensive Scaling Guide for Transformer Training\n\nWIP\n\n# Feature\n\nActivation Tracking:\n\n![alt text](contents/activation.png)\n\nWeight Update Tracking:\n\n![alt text](contents/weight_shift.png)\n\n# Scope:\n\n* Kaplan vs Chinchilla\n* SP vs muP vs layerwise SP\n* adamW weight decay\n* infinite lr scheduler\n* batch_size vs lr (sqrt BS law)\n* agd-muP (spectral initialization vs classic muP)\n* adam-atan2\n* data dependent lr tranfer\n* embedding lr transfer\n* u-muP\n\n# Shoutout\n\nThanks to [Fal.ai](https://fal.ai/) for providing compute to run these experiments.\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcloneofsimo%2Fscaling-guide","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcloneofsimo%2Fscaling-guide","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcloneofsimo%2Fscaling-guide/lists"}