yh_cc
|
9939528760
|
修复Trainer fp16的bug; 添加使用中文词进行分类的例子
|
2020-12-06 13:27:00 +08:00 |
|
yh_cc
|
69a6cbbf09
|
bug fix for Trainer fp16
|
2020-12-06 00:42:42 +08:00 |
|
yh_cc
|
2e8350cd16
|
Trainer支持fp16
|
2020-12-05 22:30:23 +08:00 |
|
yh_cc
|
35d0371955
|
update DistTrainer
|
2020-12-05 13:33:02 +08:00 |
|
yh_cc
|
3e2380aa0c
|
t Merge branch 'dev' of gitee.com:fastnlp/fastNLP into dev
|
2020-11-28 20:56:37 +08:00 |
|
yh_cc
|
eaff7cafc6
|
update to avoid warning
|
2020-11-28 20:53:24 +08:00 |
|
willqvq
|
350772f518
|
修改读取测试数据的位置
|
2020-11-23 13:34:05 +08:00 |
|
willqvq
|
fc8f8f6913
|
把测试文件夹 test 改名为 tests
|
2020-11-23 13:00:31 +08:00 |
|
yh_cc
|
2ba336945c
|
1.在DistTrainer中增加对sampler的控制; 2.在trainer和disttrainer中支持其它optimizer
|
2020-11-22 20:14:44 +08:00 |
|
WillQvQ
|
c49c849aee
|
!4 update fastNLP/modules/encoder/seq2seq_encoder.py.
Merge pull request !4 from 刘向阳/N/A
|
2020-11-18 12:13:51 +08:00 |
|
刘向阳
|
9b7562a13a
|
update fastNLP/modules/encoder/seq2seq_encoder.py.
|
2020-11-18 12:10:09 +08:00 |
|
WillQvQ
|
e4e6597ac0
|
Merge branch 'master' into dev
|
2020-11-06 23:52:10 +08:00 |
|
WillQvQ
|
d476678044
|
!2 发布 fastNLP 的 0.6.0 版本
Merge pull request !2 from WillQvQ/dev0.6
|
2020-11-06 23:22:10 +08:00 |
|
yh_cc
|
60423c2d0d
|
Merge branch 'dev' of gitee.com:fastnlp/fastNLP into dev
|
2020-11-06 22:55:48 +08:00 |
|
Yige Xu
|
9c4f802b6b
|
optimize BertEmbedding and RoBERTaEmbedding
will early exit if layer != -1
|
2020-11-06 20:53:10 +08:00 |
|
willqvq
|
a6e6f875d7
|
Preparing for version 0.6.0
|
2020-11-06 20:28:27 +08:00 |
|
WillQvQ
|
f58e10d094
|
!1 修复一个 CRF 的 bug
Merge pull request !1 from WillQvQ/dev
|
2020-11-06 18:54:33 +08:00 |
|
willqvq
|
850561728b
|
update the .Jenkinsfile
|
2020-11-06 17:46:53 +08:00 |
|
yh_cc
|
c707cd4336
|
CRF维特比解码BUG修复
|
2020-11-03 16:41:22 +08:00 |
|
WillQvQ
|
148ad1dcb7
|
ADD .Jenkinsfile;
UPDATE Makefile, doc_utils.py.
|
2020-10-30 16:39:53 +08:00 |
|
yh_cc
|
ca9db68cdd
|
Merge branch 'master' of github.com:fastnlp/fastNLP
|
2020-10-26 14:36:06 +08:00 |
|
ChenXin
|
2a09dea235
|
1. SKIP test_process_from_file
2. doc_utils 增加了 __all__ 的检查
|
2020-10-26 14:35:44 +08:00 |
|
ChenXin
|
661f57074e
|
将 nltk 从依赖中删除
|
2020-10-26 14:35:43 +08:00 |
|
yh_cc
|
89ad195411
|
添加__all__属性
|
2020-10-26 14:34:33 +08:00 |
|
ChenXin
|
ca25baf6b9
|
1. SKIP test_process_from_file
2. doc_utils 增加了 __all__ 的检查
|
2020-10-26 13:53:46 +08:00 |
|
ChenXin
|
4c457e9924
|
将 nltk 从依赖中删除
|
2020-10-26 13:22:47 +08:00 |
|
yh_cc
|
13480d67b5
|
update import
|
2020-10-25 14:13:51 +08:00 |
|
Gosicfly
|
9a9732a86f
|
Update vocabulary.py (#325)
strip只需要将\n去掉,否则会将一些特殊字符去掉,造成split的时候长度出错
token_type_id_rev (#329)
当activation=lambda x: x出现错误 (#330)
Co-authored-by: 路人咦 <1417954729@qq.com>
1.修改部分文档中的typo; 2.支持对transformers embedding
update import
|
2020-10-25 14:11:20 +08:00 |
|
yh_cc
|
0ca22b6b17
|
1.修改部分文档中的typo; 2.支持对transformers embedding
|
2020-10-25 13:56:54 +08:00 |
|
w.g
|
81bcb51314
|
当activation=lambda x: x出现错误 (#330)
Co-authored-by: 路人咦 <1417954729@qq.com>
|
2020-10-22 09:41:22 +08:00 |
|
stratoes
|
3270b8b48b
|
token_type_id_rev (#329)
|
2020-10-20 00:20:32 +08:00 |
|
yh_cc
|
08cf1a51d7
|
Merge branch 'master' of github.com:fastnlp/fastNLP
|
2020-10-12 13:57:08 +08:00 |
|
yh_cc
|
ba2304556c
|
update travis
|
2020-10-12 13:56:58 +08:00 |
|
Gosicfly
|
0c990e7b50
|
Update vocabulary.py (#325)
strip只需要将\n去掉,否则会将一些特殊字符去掉,造成split的时候长度出错
|
2020-10-11 10:59:52 +08:00 |
|
yh_cc
|
edd3022bde
|
新增ConstantTokenNumSampler使得每个sample中token数量一致,可以最大化利用GPU的显存
|
2020-09-28 10:21:39 +08:00 |
|
yh_cc
|
fe7ce8e448
|
Merge branch 'master' of github.com:fastnlp/fastNLP
|
2020-09-21 20:17:00 +08:00 |
|
yh_cc
|
ee584e71dc
|
修复seq2seq_generator的bug
|
2020-09-21 20:16:43 +08:00 |
|
yunfan
|
8d08626d67
|
[update] fix padding masking in viterbi_decode() of CRF
|
2020-09-21 13:14:26 +08:00 |
|
yh_cc
|
0d0a6f746a
|
修复测试失败问题
|
2020-09-19 14:38:26 +08:00 |
|
ROGERDJQ
|
b6a07be364
|
[test] fix f1 test (#314)
* f1 fix
* fix f1 test
* f1 fix
|
2020-09-19 11:36:07 +08:00 |
|
yh_cc
|
40bec21684
|
修复save_path在dev为空的时候的bug
|
2020-09-19 11:10:36 +08:00 |
|
Yining Zheng
|
720d7b8729
|
Update .travis.yml
Add the webhook to notify the feishu group
|
2020-09-13 20:29:52 +08:00 |
|
liuxiaoxiong
|
acdebfccbc
|
[bugfix]修复 coreference resolution复现代码中参数名字不对应的bug (#323)
* pipeline
* 修复找不到对应参数的bug
* 增加requirement文件
|
2020-08-29 16:16:57 +08:00 |
|
roger
|
b9b688d233
|
f1 fix
|
2020-07-30 14:42:00 +08:00 |
|
yh_cc
|
14ea125687
|
Merge branch 'master' of github.com:fastnlp/fastNLP
|
2020-07-29 15:02:12 +08:00 |
|
yh_cc
|
206f7758b5
|
1.新增一个sortedSampler,可以用于测试的时候加速; 2.修改BeamSearch中的bug; 3.修改Trainer中的循环代码,解决BatchSampler场景下可能出现evaluate时机不对的问题
|
2020-07-29 15:02:05 +08:00 |
|
ROGERDJQ
|
228cca44e9
|
[new] add show-result-list for confusion_matrix (#309)
add ConfusionMatrix and allow choose part of the column
|
2020-07-10 14:28:30 +08:00 |
|
yh_cc
|
c957ed69c1
|
loader中load先将filename排序保证一致的load顺序, 避免测试不通过
|
2020-07-07 22:53:05 +08:00 |
|
yh_cc
|
ed2309719c
|
修复一个data_parallel情况下的gpu不从0开始导致的bug
|
2020-07-05 23:24:34 +08:00 |
|
yh_cc
|
a6c2533bd3
|
Merge branch 'master' of github.com:fastnlp/fastNLP
|
2020-06-28 14:38:11 +08:00 |
|