hummingbird
/
fastNLP
Not watched
2
0
Fork 0
Code
Releases 13 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain

Commit Graph

  • *
  • *
  • *
  • | *
  • | |\
  • | |/
  • |/|
  • | *
  • | | *
  • | | *
  • | | |\
  • | | * |
  • * | | |
  • | * | |
  • * | | |
  • |/ / /
  • * | |
  • |\ \ \
  • | * | |
  • * | | |
  • | |_|/
  • |/| |
  • | * |
  • * | |
  • |\ \ \
  • | | * \
  • | | |\ \
  • | |_|/ /
  • |/| | |
  • | | * |
  • | * | |
  • |/ / /
  • | * |
  • | * |
  • | | *
  • * | |
  • |\ \ \
  • | | |/
  • | |/|
  • | | *
  • | | |\
  • | | |/
  • | |/|
  • | | *
  • | | *
  • | * |
  • | * |
  • | * |
  • | |\ \
  • | |/ /
  • |/| |
  • | * |
  • | |/
  • * |
  • |\ \
  • | * |
  • |/ /
  • * |
  • |\|
  • | *
  • * |
  • |\|
  • * |
  • | *
  • | *
  • * |
  • |\ \
  • | * |
  • | | *
  • | * |
  • | |\ \
  • | |/ /
  • |/| |
  • * | |
  • | | *
  • * | |
  • |\ \ \
  • | * | |
  • | * | |
  • | * | |
  • |/ / /
  • * | |
  • |\ \ \
  • | | |/
  • | |/|
  • | * |
  • | * |
  • | * |
  • | * |
  • | * |
  • | * |
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | |/ /
  • | * |
  • | * |
  • | * |
  • | |\ \
  • | * \ \
  • | |\ \ \
  • | | | * |
  • | | |/ /
  • | * | |
  • | | * |
  • | * | |
  • | |\| |
  • | * | |
  • | | * |
  • | * | |
  • | |\| |
  • | * | |
  • | | * |
  • | | * |
  • | | * |
  • | |/ /
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | | * |
  • | |/ /
  • | * |
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | | * |
  • | | * |
  • | | * |
  • | | |\ \
  • | | * | |
  • | | | * |
  • | | |/ /
  • | | * /
  • | |/ /
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | |/ /
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | |/ /
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | | * |
  • | |/ /
  • | * |
  • | |\ \
  • | * | |
  • | | * |
  • | |/ /
  • 5a3b46ce add mean pooling with mask and max pooling with mask by xuyige 2019-01-09 00:09:55 +0800
  • fcf10505 add Bi-Attention by xuyige 2019-01-09 00:09:22 +0800
  • e452bc83 fix a bug in MLP by xuyige 2019-01-08 22:26:42 +0800
  • 4fd6e12f Merge branch 'dev' of github.com:choosewhatulike/fastNLP-private into dev by yh 2019-01-08 21:49:46 +0800
  • 56e7641e 1. 修复Trainer check_code中检查evaluate时使用train_data的bug by yh 2019-01-08 21:49:31 +0800
  • 43982d2c fix ci by FengZiYjun 2019-01-08 20:20:09 +0800
  • 7984a903 Merge branch 'master' of https://github.com/fastnlp/fastNLP into local-fix-doc by FengZiYjun 2019-01-08 19:17:00 +0800
  • 1c8bca5d * fix doc for modules/ by FengZiYjun 2019-01-08 18:45:53 +0800
  • e9c93ad0 * refactor test API for POS tagging * add default sampler for Batch * fix bug in metrics.py: slice must be integer by FengZiYjun 2019-01-08 16:09:36 +0800
  • daa16712 新增developer文档 by yh 2019-01-08 15:21:27 +0800
  • 525adf1c update POS tag training script by FengZiYjun 2019-01-07 21:50:52 +0800
  • 1f4d7840 Merge branch 'dev' into local-dev by FengZiYjun 2019-01-07 19:53:48 +0800
  • 7ecd8c9c finish POS tagging API by FengZiYjun 2019-01-07 19:45:19 +0800
  • 68e6ccf0 delete verbose by xuyige 2019-01-07 19:29:53 +0800
  • 887c6fec add pos-tag training script by FengZiYjun 2019-01-07 14:22:53 +0800
  • 27ae52cf Merge pull request #118 from choosewhatulike/master by Coet 2019-01-06 18:53:24 +0800
  • 096516eb Merge branch 'master' of https://github.com/fastnlp/fastNLP into trainer by FengZiYjun 2019-01-06 15:23:45 +0800
  • dedc1604 fix docs for io/ by FengZiYjun 2019-01-04 14:12:46 +0800
  • c91696e1 update parser, optimize embed_loader by yunfan 2019-01-05 21:02:24 +0800
  • 8b980b89 修复CRF中seq_len_to_byte_mask的bug by yh 2019-01-05 16:16:04 +0800
  • 082afbfa 修复CRF中seq_len_to_byte_mask的bug by yh 2019-01-05 16:08:40 +0800
  • b040b2c4 fix docs for io/ by FengZiYjun 2019-01-04 14:12:46 +0800
  • 50e18f8c Merge pull request #117 from FengZiYjun/fix-doc by Xipeng Qiu 2019-01-04 13:09:50 +0800
  • 0c7e7519 conflict fix by yh 2019-01-03 21:37:56 +0800
  • e267f925 新增cws的train_context.py by yh 2019-01-03 21:35:28 +0800
  • 897c43fc 1. CRF增加constrain, 用于限制跃迁，比如BMES中B不能跃迁到S 2. metric增加SpanFMetric，可以用于计算sequence labelling的performance 3. 分词复现任务根据新版接口做了部分调整。 by yh 2019-01-03 21:33:36 +0800
  • 07c2b87c fix version number by FengZiYjun 2019-01-03 19:52:14 +0800
  • 5d8f6960 * fix README figure * refine code style by FengZiYjun 2019-01-03 19:25:22 +0800
  • d43d7385 merge dev branch with master by FengZiYjun 2019-01-03 19:05:23 +0800
  • bc7fc71f * update version number in doc * update fastNLP.core.rst * refine all docstrings in core/ by FengZiYjun 2019-01-03 18:59:52 +0800
  • dfb62ec8 Merge pull request #115 from hazelnutsgz/conll2003 by Xipeng Qiu 2019-01-02 14:04:57 +0800
  • 5f4ab131 Add a loader for conll2003 dataset by hazelnutsgz 2019-01-01 01:06:00 +0800
  • 7371593f Merge pull request #113 from FengZiYjun/trainer by Xipeng Qiu 2018-12-25 11:41:41 +0800
  • 3a2833a6 1.解决trainer加载模型的问题; 2.删除loss与metric中对*arg的报错 by yh 2018-12-19 19:51:55 +0800
  • 7c3cf7e3 Merge pull request #112 from FengZiYjun/trainer by Xipeng Qiu 2018-12-15 21:48:06 +0800
  • 1cfd0ea5 Update PULL_REQUEST_TEMPLATE.md by Coet 2018-12-14 19:10:44 +0800
  • 337e3035 * update most processors to use dataset.apply * fix failed tests by FengZiYjun 2018-12-14 18:16:44 +0800
  • 23305af7 Fix failed tests. by FengZiYjun 2018-12-13 10:55:48 +0800
  • eee52672 Merge pull request #111 from lyhuang18/docs_new by Coet 2018-12-13 10:19:39 +0800
  • d91a7c7c update docs by lyhuang 2018-12-13 01:52:52 +0800
  • 30a0ff4d 1. optimzer中调整默认参数大小 2. 移动Tester中device获取位置 3. trainer中 (1) train()返回一个dict，并重新加载最佳模型 (2) check_code时返回batch_x中每个数据的形状等信息 by yh 2018-12-12 23:38:19 +0800
  • 97b3710d Merge pull request #11 from fastnlp/master by lyhuang18 2018-12-12 23:01:46 +0800
  • 071c1410 Create PULL_REQUEST_TEMPLATE.md by Coet 2018-12-11 15:38:20 +0800
  • 490555af tutorial_for_developer initial by yh 2018-12-08 11:49:44 +0800
  • 426e762f Merge pull request #109 from fastnlp/xuyige-patch-1 by Yunfan Shao 2018-12-07 22:32:10 +0800
  • f15bd5aa update README.md by Yige XU 2018-12-07 22:28:20 +0800
  • baac45a7 update README.md by Yige XU 2018-12-07 22:17:23 +0800
  • 27932737 update README.md by Yige XU 2018-12-07 22:09:58 +0800
  • 1b477a95 (tag: v0.2.0) Merge pull request #108 from FengZiYjun/trainer by Xipeng Qiu 2018-12-07 19:48:51 +0800
  • db0a789d * final clean up * remove conflicts * all tests passed by FengZiYjun 2018-12-07 19:09:50 +0800
  • 267baec2 add dataloader register by yunfan 2018-12-07 15:19:56 +0800
  • 720a264e * rename DataSet.get_fields() into get_all_fields() * add DataSet.get_field(), to fetch a FieldArray based on its name * remove old tutorials & add new tutorials by FengZiYjun 2018-12-07 14:53:27 +0800
  • 447746d9 * remove unused codes in losses.py & metrics.py * refine code style * fix tests * add a new tutorial by FengZiYjun 2018-12-07 13:22:04 +0800
  • 72877c6e optimizer初始化调整参数顺序 by yh 2018-12-07 13:31:52 +0800
  • 27e9453d * fix processor.py * add code comments * merge *_saver.py & *_loader.py in io/ * (ancient codes) rename Loss into LossFromTorch by FengZiYjun 2018-12-06 19:28:27 +0800
  • 306eee96 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-06 11:16:32 +0800
  • cd838665 bug fix in LossInForward by yh 2018-12-06 11:16:25 +0800
  • 6129a31c * fix tests * clean up unused codes by FengZiYjun 2018-12-06 10:07:45 +0800
  • aea93181 1. trainer中losser修改为loss by yh 2018-12-05 20:23:40 +0800
  • 11585562 1. 优化trainer checkcode过程的报错信息 by yh 2018-12-05 20:15:59 +0800
  • f7c29b85 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-04 23:33:22 +0800
  • e779409c conflict fix by yh 2018-12-04 23:33:13 +0800
  • 5855adbc fix FieldArray bug: do type check only when is_target or is_input is True by FengZiYjun 2018-12-04 23:30:54 +0800
  • 4dff3ec8 修改losses中直接使用F.cross_entropy的情况，因为这些函数的signature是(input, target) by yh 2018-12-04 23:18:37 +0800
  • f26f1160 * 更新教程，放在在./tutorial * remove unused codes in metrics.py * add tests for DataSet * add tests for FieldArray * add tests for metrics.py * fix predictor, add tests for predictor * fix bucket sampler, add tests for bucket sampler by FengZiYjun 2018-12-04 22:57:26 +0800
  • 7c261faf Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-04 22:45:04 +0800
  • 87e5d44b 。 by yh 2018-12-04 22:44:54 +0800
  • 52b1b18a fix bugs in vocab by yunfan 2018-12-04 17:04:31 +0800
  • 79ae3873 conflict solved by yh 2018-12-04 16:23:33 +0800
  • 62c63f15 test loss by yh 2018-12-04 16:22:41 +0800
  • 27833d06 FieldArray only check type when is_input or is_target is set. by FengZiYjun 2018-12-04 16:13:20 +0800
  • 5edd9de8 fix bugs by yunfan 2018-12-04 15:49:01 +0800
  • 9acdb54f 优化loss在missing和duplicate时报错的信息:返回loss初始化约定接受的key by FengZiYjun 2018-12-04 14:17:31 +0800
  • abe5ec72 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-04 13:00:34 +0800
  • a1a41c2d 1. unused报错运行错误修复 2. loss中修复一个错误 3. metric中fast_param调整 by yh 2018-12-04 12:58:56 +0800
  • 4b099bb0 * add tqdm in requirements.txt * fix FieldArray type check bugs by FengZiYjun 2018-12-04 11:16:24 +0800
  • 661780b9 Improve FieldArray. Support nested list and a list of np.array by FengZiYjun 2018-12-04 10:54:09 +0800
  • 1fb1df4a 1. metric修改fast_param 2. trainer中update_every改为print_every, 因为update_every可能引起optimizer update的误解 3. fieldarray content支持使用np.ndarray初始化 by yh 2018-12-04 10:43:40 +0800
  • 785c41de Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-04 00:47:47 +0800
  • beb55f52 * change trainer iterating into tqdm by yh 2018-12-03 22:53:14 +0800
  • 1421b7df add this feature totally for yh by yunfan 2018-12-03 22:48:02 +0800
  • ad3c5b6e add magic iter in dataset by yunfan 2018-12-03 21:54:22 +0800
  • 513876d5 Updates: * fix losses的_fast_param_map的bug * Trainer添加sampelr初始化参数，并调整参数顺序 * refine codes by FengZiYjun 2018-12-03 20:50:51 +0800
  • 9c8ec457 Merge branch 'trainer' of https://github.com/FengZiYjun/fastNLP into dataset by yunfan 2018-12-03 20:05:54 +0800
  • 131e1ccd add _fast_param_map by yunfan 2018-12-03 20:04:14 +0800
  • 6f58ec34 Updates: * DataSet修改__repr__，优化print(datset)的输出 * Instance修改__repr__，优化print的输出 * Optimizer优化传参提示 * Trainer去除kwargs参数 * losses.py加个参数 * 对应test code的修改 by FengZiYjun 2018-12-03 19:53:34 +0800
  • f6206033 All tests pass. Ready to merge. * 更新Loss的接口形参跟metric保持一致 * 添加对几种loss的测试 * embed_loader采用维度独立的方法采样 * 对应测试代码的修改 by FengZiYjun 2018-12-03 12:37:33 +0800
  • 9e3fd0fe Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-03 12:13:05 +0800
  • 77f8ac77 对trainer中check code的报错信息进行了增强；将tester中的output修改为pred_dict by yh 2018-12-03 12:12:48 +0800
  • cc440b5e All tests pass. * 更新测试代码，跑通所有测试，覆盖率65% * refine代码规范和某些注释 * fix tester self.use_cuda未赋值先使用的bug * 添加tutorial样例数据——tutorial_sample_dataset.csv * 【unsolved】embed_loader在计算np.cov时遇到segmentation fault by FengZiYjun 2018-12-03 11:12:56 +0800
  • d4af19ec Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-03 09:50:26 +0800
  • 88949ba1 修改dataset.py的apply signature; batch当中增加num_batches属性; tester的format_eval_results修改; metric增加fast_evaluate_call机制 by yh 2018-12-03 09:50:16 +0800
  • 5824b7f4 跑通tutorial,修复一些bugs: * dataset检查slice开始位置，确保结果不为空 * fieldarray检查content不为空 * optimizer接受的model params是一个generator，不能赋值 * code style refine by FengZiYjun 2018-12-03 00:08:59 +0800
  • b2e09b74 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-02 23:27:49 +0800
  • d19850b3 * add _fast_call_evaluate mechanism in MetricBase by yh 2018-12-02 23:27:40 +0800
  • 11c82ab2 跑通test_trainer.py，联调结束，准备发布 by FengZiYjun 2018-12-02 20:07:50 +0800
  • 201f5109 Updates: * improve Loss initialization interface * improve test codes for trainer by FengZiYjun 2018-12-02 19:45:25 +0800
  • 8dd73a92 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-02 16:39:35 +0800
  • 234ceb6f fix bug in MetricBase by yh 2018-12-02 16:39:28 +0800
  • 125c2718 Update * fix bug in DataSet.split * fix bugs in FieldArray, to allow content as a list * fix bug in losses check * ... by FengZiYjun 2018-12-02 16:38:38 +0800