{"id":15055279,"url":"https://github.com/uk0/llmocr","last_synced_at":"2026-02-04T03:01:42.824Z","repository":{"id":255781294,"uuid":"850639999","full_name":"uk0/LLMOCR","owner":"uk0","description":"使用LLM + OCR进行总结OCR识别的内容，返回相应的结构数据","archived":false,"fork":false,"pushed_at":"2024-09-10T10:23:57.000Z","size":49843,"stargazers_count":1,"open_issues_count":1,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-06-04T18:27:51.276Z","etag":null,"topics":["llm","ocr","ollama","pp-ocrv4"],"latest_commit_sha":null,"homepage":"","language":"Python","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/uk0.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":"2024-09-01T11:14:41.000Z","updated_at":"2024-10-21T05:26:09.000Z","dependencies_parsed_at":"2024-09-10T12:12:48.871Z","dependency_job_id":null,"html_url":"https://github.com/uk0/LLMOCR","commit_stats":{"total_commits":6,"total_committers":1,"mean_commits":6.0,"dds":0.0,"last_synced_commit":"c873525039ac8d3ab3830c0b6dc4ed9134c3acf6"},"previous_names":["uk0/llmocr"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/uk0/LLMOCR","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/uk0%2FLLMOCR","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/uk0%2FLLMOCR/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/uk0%2FLLMOCR/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/uk0%2FLLMOCR/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/uk0","download_url":"https://codeload.github.com/uk0/LLMOCR/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/uk0%2FLLMOCR/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":259329402,"owners_count":22841435,"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":["llm","ocr","ollama","pp-ocrv4"],"created_at":"2024-09-24T21:39:56.836Z","updated_at":"2026-02-04T03:01:42.758Z","avatar_url":"https://github.com/uk0.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"### LLM + OCR\n\n\n* 后面可以扩展插件支持一些奇怪的表格处理数据，也可以使用`openai`的模型，目前是做了一个POC验证。\n* 主要是为了方便识别一些简单的小图片，比如有些图片上的文字等消息，需要整理出来或是复制等，也能识别一些不清晰的内容。\n\n\n#### Quick Start\n\n* PP-OCR-V4.0\n* Ollama (`gemma2:2b-instruct-q8_0`)\n* flask\n* chrome plugin\n\n\n```shell\n\npip install -r requirements.txt\npython app.py\n\n```\n\n\n### install chrome plugin\n* open `chrome://extensions/`\n* switch developer mode on\n* load unpacked extension\n\n![img.png](doc/img.png)\n\n* find image and right click to open with WiseRead（`Analyze Image`）\n\n![img.png](doc/img_id_card.png)\n\n*  Result for Chrome Tab will be shown as below on right div box\n\n![img.png](doc/img1.png)\n\n\n\n### Model \n\n* llama3.1:8b-instruct-q8_0 `效果最好`\n* gemma2:9b-instruct-q8_0\n* Qwen1.5-MoE-A2.7B-Chat:latest\n\n\n### TODO\n\n* 优化提示词(Doing...)\n* 使用RAG优化结果，使结果更稳定(Doing...)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fuk0%2Fllmocr","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fuk0%2Fllmocr","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fuk0%2Fllmocr/lists"}