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

Commit Graph

  • *
  • *
  • *
  • *
  • *
  • *
  • |\
  • | *
  • | *
  • * |
  • |/
  • *
  • *
  • *
  • |\
  • * |
  • * |
  • | *
  • | |\
  • | * |
  • |/ /
  • | *
  • | *
  • | |\
  • | * |
  • |/ /
  • | *
  • |/
  • *
  • *
  • *
  • |\
  • | *
  • | |\
  • * | |
  • | * |
  • * | |
  • |\ \ \
  • | | |/
  • | |/|
  • * | |
  • | * |
  • | * |
  • |/ /
  • | *
  • | |\
  • | |/
  • |/|
  • | *
  • * |
  • |/
  • *
  • *
  • *
  • *
  • |\
  • | *
  • | *
  • | |\
  • | |/
  • |/|
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • | *
  • * |
  • |\ \
  • | | *
  • | | *
  • | * |
  • | | *
  • | | *
  • | | |\
  • | | * |
  • | | | *
  • | | |/
  • | | *
  • | | |\
  • | * | |
  • | | | *
  • | | | |\
  • | | | * |
  • c2d21375 bug fix in MetricAccuracy by yh 2018-12-02 15:19:05 +0800
  • 8d7d2b42 initial test for AccuracyMetric by yh 2018-12-02 14:57:11 +0800
  • 50f1c28b metric bug fix by yh 2018-12-02 14:29:11 +0800
  • a90a62ab metric bug fix by yh 2018-12-02 14:28:44 +0800
  • a05ffd31 trainer增加对evaluate结果的check by yh 2018-12-02 12:55:15 +0800
  • f76851b9 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-02 12:52:48 +0800
  • d74901e0 Trainer Update: * 添加初始化注释 * 从_better_eval_result中抽取check metrics的逻辑到_check_eval_results函数 by FengZiYjun 2018-12-02 11:36:35 +0800
  • fb5215ae fix bug in Trainer about metric_key 更新Optimizer: 多种初始化方法 1. SGD() 2. SGD(0.01) 3. SGD(lr=0.01) 4. SGD(lr=0.01, momentum=0.9) 5. SGD(model.parameters(), lr=0.1, momentum=0.9) by FengZiYjun 2018-12-02 10:58:10 +0800
  • 84024aaa _prepare_metric函数增加检查evaluate与get_metric方法 by yh 2018-12-02 10:36:20 +0800
  • bd94dd2c metrics中实现AccuracyMetric, 并将metric的计算方式由一把计算修改为batch by batch by yh 2018-12-02 10:30:25 +0800
  • f24fca1b change the calculation of metric to batch by batch. The older design is to concat all data before calculation. by yh 2018-12-02 09:21:08 +0800
  • 1b961f13 conflict in trainer solved by yh 2018-12-01 23:47:08 +0800
  • 3daa889b LossInForward update by yh 2018-12-01 23:44:07 +0800
  • 3a4a7293 trainer and tester change check_code by yh 2018-12-01 23:43:24 +0800
  • ba7b1766 Merge branch 'trainer' of https://github.com/FengZiYjun/fastNLP into check by xuyige 2018-12-01 22:45:03 +0800
  • 6d36190b update LossBase class by xuyige 2018-12-01 22:44:24 +0800
  • 8a7077fe 更新Optimizer: optimizer.SGD(lr=xxx);如果没有传入parameters，则在trainer中帮他加入parameter by FengZiYjun 2018-12-01 22:21:57 +0800
  • 08375d51 Merge remote-tracking branch 'FengZiYjun/trainer' into trainer by FengZiYjun 2018-12-01 21:04:08 +0800
  • e5e7f29d 更新Trainer: * 添加Trainer参数metric_key，指明用来做模型选择的指标的名字 * 在Trainer添加处理tester返回的评价指标的逻辑，选择当前最好的模型 by FengZiYjun 2018-12-01 20:27:23 +0800
  • e6864ea7 更新embed_loader: * 添加fast_load_embedding方法，用vocab的词索引pre-trained中的embedding * 如果vocab有词没出现在pre-train中，从已有embedding中正态采样 by FengZiYjun 2018-12-01 20:27:23 +0800
  • 0d4720b1 CheckError add function by yh 2018-12-01 20:14:43 +0800
  • 2c8bd957 add _method_function by yh 2018-12-01 19:55:23 +0800
  • 84eb50a8 Merge branch 'trainer' of https://github.com/FengZiYjun/fastNLP into check by xuyige 2018-12-01 18:55:22 +0800
  • 7c439e7a Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-01 18:35:18 +0800
  • 37e282d3 update LossBase class by xuyige 2018-12-01 18:31:16 +0800
  • ad0a8c17 增加metric by yh 2018-12-01 18:27:07 +0800
  • fc505b8d Merge branch 'trainer' of https://github.com/FengZiYjun/fastNLP into HEAD by xuyige 2018-12-01 18:26:41 +0800
  • d8a80ad6 update LossBase class by xuyige 2018-12-01 17:28:47 +0800
  • 3120cdd0 更新embed_loader: * 添加fast_load_embedding方法，用vocab的词索引pre-trained中的embedding * 如果vocab有词没出现在pre-train中，从已有embedding中正态采样 by FengZiYjun 2018-12-01 17:23:25 +0800
  • 6427e85e 升级Vocab： * 增量添加单词到词典中 * lazy update: 当用到词典的时候才重新build * 当新添加的词导致词典大小超出限制时，打印一个warning by FengZiYjun 2018-12-01 15:44:52 +0800
  • fe0f99b7 Merge branch 'trainer' of github.com:FengZiYjun/fastNLP into trainer by yh 2018-12-01 15:12:02 +0800
  • 3d91f2f0 trainer迭代 by yh 2018-12-01 15:00:06 +0800
  • 07e227aa add interface of Loss by FengZiYjun 2018-12-01 14:57:49 +0800
  • 6839bb91 Add auto type detection/conversion in FieldArray * In init, detect content type to be Python int, float, or str. * In append(), check type consistence. * In init & append(), int will be cast into float if they occur together. * Map Python type into numpy dtype * Raise error if type detection fails. by FengZiYjun 2018-12-01 10:38:01 +0800
  • da901ed5 * DataSet __getitem__ returns copy of Instance * refine interface of set_target & set_input * rename DataSet.Instance into DataSet.DataSetIter * remove unused methods in DataSet.DataSetIter * remove __setattr__ in DataSet; It is dangerous. * comment adjustment by FengZiYjun 2018-11-30 23:56:44 +0800
  • 117b12a6 * update README.md * remove torchvision in requirements.txt by FengZiYjun 2018-11-29 23:27:15 +0800
  • 15262bda Merge pull request #107 from yhcc/dataset by yhcc 2018-11-28 15:09:57 +0800
  • 26a43243 fix test by yunfan 2018-11-27 22:52:14 +0800
  • 04206f80 Merge branch 'master' into dataset-res by yunfan 2018-11-27 22:47:17 +0800
  • 2aaa3818 refine git commits by yunfan 2018-11-27 20:28:01 +0800
  • e1e0661d add doc comments by yunfan 2018-11-27 18:39:57 +0800
  • 941b88f2 fix dataset.read_csv by yunfan 2018-11-26 22:01:57 +0800
  • 4f587f75 fix trainer & dataset by yunfan 2018-11-26 21:23:50 +0800
  • e4c1ab60 prepare for release by yh 2018-11-27 18:28:17 +0800
  • ffc96319 修改dataframe.read_csv by yh 2018-11-26 20:43:16 +0800
  • 1d8f1227 dataset.read_csv by yh 2018-11-26 20:33:56 +0800
  • 1c34a0b7 update trainer by yunfan 2018-11-26 19:24:18 +0800
  • b78d8658 add validate_every in trainer by yunfan 2018-11-26 18:35:48 +0800
  • f3bb3cb5 update trainer, tester, example model by yunfan 2018-11-26 16:22:19 +0800
  • a3bf6477 check code修改 by yh 2018-11-26 18:35:55 +0800
  • 44e098e2 update trainer, tester, example model by yunfan 2018-11-26 15:32:22 +0800
  • 4a4b0010 trainer修改 by yh 2018-11-26 15:34:13 +0800
  • f7275339 trainer check_code调整 by yh 2018-11-26 14:21:42 +0800
  • d9db503b bug fix in trainer by yh 2018-11-25 17:20:59 +0800
  • 3d669750 * refine code comments * refine code style * set up unit tests for Batch, DataSet, FieldArray * remove a lot of out-of-date unit tests, to get testing passed by FengZiYjun 2018-11-25 15:04:57 +0800
  • 74a69765 - fix Dataset & Trainer - update CNNText model by yunfan 2018-11-24 22:36:43 +0800
  • c4103561 * fix bugs in DataSet & Instance * add more code comments * fix tester * refresh code styles by FengZiYjun 2018-11-24 14:39:01 +0800
  • 0836ce00 尝试提供check parameter的功能 by yh 2018-11-25 17:00:34 +0800
  • 92da53a6 fix Dataset by yunfan 2018-11-24 13:03:54 +0800
  • 837bef47 * add unit tests for instance, vocabulary * remove and fix other unit tests * add more code comments by FengZiYjun 2018-11-23 21:22:56 +0800
  • c7923c82 update check_args and add Dataset get_input/target_name by yunfan 2018-11-23 21:10:40 +0800
  • ce3b0022 check code init by yh 2018-11-23 21:01:32 +0800
  • cbf54c19 add args check & build function by yunfan 2018-11-23 20:13:51 +0800
  • 5abd2bf4 fix dataset & instance by yunfan 2018-11-23 19:41:25 +0800
  • 713510f6 update Instance by yunfan 2018-11-23 19:01:49 +0800
  • 68d02541 init check_* by yunfan 2018-11-23 18:21:26 +0800
  • d643a7a8 update set_target, batch's as_numpy by yunfan 2018-11-23 17:14:42 +0800
  • 2fe39b78 fix log in trainer & tester by yunfan 2018-11-23 13:32:52 +0800
  • 80884322 * add DataSet.split() * delete field.py * remove logger in all codes * adjust arguments of Trainer by FengZiYjun 2018-11-23 17:57:52 +0800
  • 0292350c vocabulary增加方法 by yh 2018-11-23 17:08:42 +0800
  • 3a42c84a use counter in vocab, add a load func in baseloader by yunfan 2018-11-21 12:38:18 +0800
  • 8ee94eb6 make import more friendly, Dataset support slice. by yunfan 2018-11-19 23:10:37 +0800
  • 090f7aef * fixing unit tests by FengZiYjun 2018-11-19 22:02:21 +0800
  • 1d5bb0a3 bug fix“ by yh 2018-11-19 19:16:09 +0800
  • 4149eb9c add apply to dataset by yh 2018-11-19 15:12:07 +0800
  • 8906155c 为api建立一个Analyzer by yh 2018-11-14 23:15:19 +0800
  • e9d7074b * delete readme_example.py because it is oooooooout of date. * rename preprocess.py into utils.py, because nothing about preprocess in it * anything in loader/ and saver/ is moved directly into io/ * corresponding unit tests are moved to /test/io * delete fastnlp.py, because we have new and better APIs * rename Biaffine_parser/run_test.py to Biaffine_parser/main.py; Otherwise, test will fail. * A looooooooooot of ancient codes to be refined........... by FengZiYjun 2018-11-18 19:30:53 +0800
  • b6a0d33c add parser api by yunfan 2018-11-14 16:03:55 +0800
  • 8d7eae8a 增加api的test功能 by yh 2018-11-14 13:25:15 +0800
  • 77786509 pos与cws开发上传 by yh 2018-11-14 10:44:33 +0800
  • 7d97e936 增加新的processor“ by yh 2018-11-13 23:56:34 +0800
  • 14960311 新增pos output processor by yh 2018-11-13 16:56:03 +0800
  • d5afffee 新增端到端pos处理到parser的过度代码 by yh 2018-11-13 15:37:11 +0800
  • 10379e9c 当前为segapp的方式，但是貌似准确率不行，尝试修改为crf 4tag模式试一试 by yh_cc 2018-11-13 09:52:53 +0800
  • 822aaf62 fix and update tester, trainer, seq_model, add parser pipeline builder by yunfan 2018-11-12 21:37:56 +0800
  • f414475e add parser pipeline, fix models, batch, crf by yunfan 2018-11-11 21:03:44 +0800
  • 4be15a5b 保存pos tag 脚本 by FengZiYjun 2018-11-11 21:21:10 +0800
  • db5c5ea4 update People Daily DataSet Loader by xuyige 2018-11-11 14:17:16 +0800
  • abf840c3 Merge pull request #106 from FFTYYY/master by Yige XU 2018-11-12 20:20:36 +0800
  • b899b1ed 修改bucket sampler, 增加url下载功能 by yh 2018-11-11 20:25:47 +0800
  • 9667c524 基本完善了cws的predict by yh 2018-11-11 15:53:33 +0800
  • 3cadd5a3 fix a iterant lossfuntion , and some error in comments by FFTYYY 2018-11-11 13:47:54 +0800
  • 9fc20ac7 增加infer的pipeline by yh 2018-11-11 12:55:30 +0800
  • 0a8a76f7 冲突解决 by yh 2018-11-11 12:43:16 +0800
  • dc7f8ef8 bug fix by yh 2018-11-11 12:42:05 +0800
  • 82f43515 add index to word processor by yunfan 2018-11-11 12:37:27 +0800
  • 7df33b23 Merge branch 'dataset' of github.com:yhcc/fastNLP into dataset by yh_cc 2018-11-11 00:40:10 +0800
  • 07fb61ef Update test_loss by FFTYYY 2018-11-10 23:21:26 +0800
  • e2b14ed3 Merge remote-tracking branch 'origin/dataset' into dataset by FengZiYjun 2018-11-10 21:20:34 +0800
  • 5dd0f74d - 添加pos_tagger API， pipeline跑通 - 修复processor的bug - 更新core/的若干组件, 去除batch的冗余参数 - CRF有个打字错误？已修复 - 更新pos tag 训练脚本 by FengZiYjun 2018-11-10 21:20:16 +0800