{"id":26439967,"url":"https://github.com/aveygo/visiondiff","last_synced_at":"2026-02-12T00:02:15.468Z","repository":{"id":258176237,"uuid":"873548673","full_name":"Aveygo/VisionDiff","owner":"Aveygo","description":"Simple vision based block for Microsoft's Differential Transformer ","archived":false,"fork":false,"pushed_at":"2024-10-21T22:55:49.000Z","size":141,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-10-29T11:56:44.114Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Aveygo.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,"zenodo":null}},"created_at":"2024-10-16T10:58:57.000Z","updated_at":"2024-10-21T22:55:49.000Z","dependencies_parsed_at":"2025-08-29T05:22:54.093Z","dependency_job_id":"1cc63336-ec5a-40f5-974b-70272ea3bdb7","html_url":"https://github.com/Aveygo/VisionDiff","commit_stats":null,"previous_names":["aveygo/visiondiff"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Aveygo/VisionDiff","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Aveygo%2FVisionDiff","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Aveygo%2FVisionDiff/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Aveygo%2FVisionDiff/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Aveygo%2FVisionDiff/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Aveygo","download_url":"https://codeload.github.com/Aveygo/VisionDiff/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Aveygo%2FVisionDiff/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29350080,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-11T20:11:40.865Z","status":"ssl_error","status_checked_at":"2026-02-11T20:10:41.637Z","response_time":97,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"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":[],"created_at":"2025-03-18T09:18:50.088Z","updated_at":"2026-02-12T00:02:15.405Z","avatar_url":"https://github.com/Aveygo.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# VisionDiff\n\nUsing the [Differential Transformer](https://arxiv.org/abs/2410.05258) in a vision-friendly way, similar to [VisionMamba](https://github.com/kyegomez/VisionMamba).\n\n\u003cp align=\"center\"\u003e\n    \u003cimg src=\"image.png\" width=\"800px\"\u003e\n\u003c/p\u003e\n\n\n## Preamble\nAbsolutely no formal experimentation has been performed to validate this method - take everything with a grain of salt.\n\n## Installation\n\n```pip install visiondiff```\n\n## Usage\n\n```python\nfrom VisionDiff import VisionDiff\ndim, num_heads = 32, 4\nlayer1 = VisionDiff(dim, num_heads, in_channels=3)\nlayer2 = VisionDiff(dim, num_heads)\nlayer3 = VisionDiff(dim, num_heads, out_channels=3)\n\nx = torch.zeros(1, 3, 64, 64) # Example \"image\"\nx = layer1(x)\nx = layer2(x)\nx = layer3(x)\n\nprint(f\"Output shape: {x.shape}\") # [1, 3, 64, 64]\n```\n\nPlease keep in_channel / out_channel operations at a minimum\nas they tend to be computationally expensive.\n\n## Versions\n\n0.1.0 - *Now with more positional encoding!*","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faveygo%2Fvisiondiff","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faveygo%2Fvisiondiff","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faveygo%2Fvisiondiff/lists"}