{"id":19119450,"url":"https://github.com/cluebenchmark/mobileqa","last_synced_at":"2025-11-09T01:02:13.896Z","repository":{"id":110258778,"uuid":"223370212","full_name":"CLUEbenchmark/MobileQA","owner":"CLUEbenchmark","description":"离线端阅读理解应用 QA for mobile, Android \u0026 iPhone","archived":false,"fork":false,"pushed_at":"2022-10-06T03:58:11.000Z","size":55145,"stargazers_count":60,"open_issues_count":2,"forks_count":13,"subscribers_count":11,"default_branch":"master","last_synced_at":"2025-05-05T14:51:48.516Z","etag":null,"topics":["albert","android","bert","chinese","iphone","machine-reading-comprehension","nlp","qa","tensorflow","tflite"],"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/CLUEbenchmark.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,"zenodo":null}},"created_at":"2019-11-22T09:34:10.000Z","updated_at":"2024-01-07T23:43:04.000Z","dependencies_parsed_at":"2023-03-13T13:56:10.111Z","dependency_job_id":null,"html_url":"https://github.com/CLUEbenchmark/MobileQA","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/CLUEbenchmark/MobileQA","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CLUEbenchmark%2FMobileQA","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CLUEbenchmark%2FMobileQA/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CLUEbenchmark%2FMobileQA/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CLUEbenchmark%2FMobileQA/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/CLUEbenchmark","download_url":"https://codeload.github.com/CLUEbenchmark/MobileQA/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CLUEbenchmark%2FMobileQA/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279000707,"owners_count":26082837,"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":"2025-10-08T02:00:06.501Z","response_time":56,"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":["albert","android","bert","chinese","iphone","machine-reading-comprehension","nlp","qa","tensorflow","tflite"],"created_at":"2024-11-09T05:09:43.824Z","updated_at":"2025-10-08T21:38:37.741Z","avatar_url":"https://github.com/CLUEbenchmark.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"## MobileQA\n\n[TensorFlow Lite BERT QA Android Example Application](https://github.com/tensorflow/examples/tree/master/lite/examples/bert_qa/android) 与 [tflite-android-transformers](https://github.com/huggingface/tflite-android-transformers) 展示了基于Bert/DistilBERT的离线QA例子，但是只支持英文和安卓设备。\n\n本项目计划实现基于中文的机器阅读理解在手机端的离线应用，并且同时支持安卓和苹果设备。\n\nTargeting to release before Dec 5th. 目标是12月5日前发布。\n\n\n\n## 数据集\n\n使用[CMRC 2018 公开数据集](https://github.com/ymcui/cmrc2018/blob/master/README_CN.md)，该数据集是第二届讯飞杯中文机器阅读理解评测所使用的数据。数据集已被计算语言学顶级国际会议[EMNLP 2019](http://emnlp-ijcnlp2019.org/)录用\n\n\n\n## 模型\n\n使用 [albert_zh_small](https://github.com/brightmart/albert_zh) 预训练模型，额外加上一层全连接做answer span预测。\n\n* 在CMRC2018数据集的验证集上，max_seq_len为512的模型得分为F1:75.989, EM:52.097, Average:64.038，max_seq_len为384的模型得分为F1:74.781, EM:51.010, Average:62.895\n* max_seq_len为512的模型使用tflite转换后大小为18M，经测试，该模型在4线程的安卓手机上推理延时为580ms左右，在单线程条件下为1.4s左右。\n* max_seq_len为384的模型使用tflite转换后大小为18M，经测试，该模型在4线程的安卓手机上推理延时为390ms左右，在单线程条件下为930ms左右\n\n* 模型训练与模型转换过程见[bert_cn_finetune-master](https://github.com/CLUEbenchmark/MobileQA/tree/master/bert_cn_finetune-master) \n\n\n\n## Android Demo\n\n已完成第一版，详见 [tflite-android-transformers-master](https://github.com/CLUEbenchmark/MobileQA/tree/master/tflite-android-transformers-master) \n\n效果示例：\n\n![android demo](./media/example.gif)\n\n## IOS Demo\n\n进行中\n\n## Updates\n\n* 2019-11-27: Add first version of Android Demo, details see [tflite-android-transformers-master](https://github.com/CLUEbenchmark/MobileQA/tree/master/tflite-android-transformers-master) \n* 2019-11-29: Add first version of MRC model, details see [bert_cn_finetune-master](https://github.com/CLUEbenchmark/MobileQA/tree/master/bert_cn_finetune-master) \n\n## Contribution\n\n如果你感兴趣或希望提供帮助，发送邮件ChineseGLUE@163.com\n\nIf you're interested or want to help, send an email to ChineseGLUE@163.com\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcluebenchmark%2Fmobileqa","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcluebenchmark%2Fmobileqa","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcluebenchmark%2Fmobileqa/lists"}