hummingbird
/
fastNLP
Not watched
2
0
Fork 0
Code
Releases 13 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain
1672 Commits
3 Branches
25 MB
521 Download
Tree: 40bec21684
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from '40bec21684'
${ noResults }
Commit Graph

1672 Commits (40bec21684ed293ecc113e11f3ad080ee37bbe83)
 

Author SHA1 Message Date
  yh_cc 40bec21684 修复save_path在dev为空的时候的bug 5 years ago
  Yining Zheng 720d7b8729
Update .travis.yml 5 years ago
  liuxiaoxiong acdebfccbc
[bugfix]修复 coreference resolution复现代码中参数名字不对应的bug (#323) 5 years ago
  roger b9b688d233 f1 fix 5 years ago
  yh_cc 14ea125687 Merge branch 'master' of github.com:fastnlp/fastNLP 5 years ago
  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 5 years ago
  ROGERDJQ 228cca44e9
[new] add show-result-list for confusion_matrix (#309) 5 years ago
  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 5 years ago
  yh_cc ed2309719c 修复一个data_parallel情况下的gpu不从0开始导致的bug 5 years ago
  yh_cc a6c2533bd3 Merge branch 'master' of github.com:fastnlp/fastNLP 5 years ago
  yh_cc e361b32c3a 在@linzehui 的帮助下seq2seq终于有了第一个版本; 目前实现了Seq2Seq的Transformer和LSTM版本，但metric和loss还没update; 5 years ago
  Yige Xu d9456eb15e [bugfix] fix an import error in bert_tokenizer.py 5 years ago
  郑恒毅 94be3315a6
[bugfix]修改了modules/utils.py中`initial_parameter()`函数的bug (#300) 5 years ago
  yh_cc 1efd40c075 准备开发0.5.6 5 years ago
  yh_cc 43300ad90b 修改部分注释, 准备开发0.5.6 5 years ago
  ChenXin 8732dfd979 新增了 DataBundle 的 apply_more 和 apply_field_more 方法。需要进一步试用和测试。 5 years ago
  Yige Xu 1f27d007d1 [add] add MSELoss 5 years ago
  ROGERDJQ 4e95989e97
[new] add seperator for conll loader (#293) 5 years ago
  linzehui b95aa56afb 基本完成seq2seq基础功能 5 years ago
  yh_cc 6c7a61a9ab 发布0.5.5版本 5 years ago
  yh_cc ae7b916355 1.修改Generator中的采样，防止导致分布全0采样失败; 2. 修改Trainer与checkpointcallback的配合 5 years ago
  linzehui 15360e9724 保存一版旧版 5 years ago
  yh_cc 9b45e9c0b5 修改GradientClipCallback中parameter的存储方式，防止在torch 1.5版本中报错 5 years ago
  yh_cc 3683fd5bdd Merge branch 'master' of github.com:fastnlp/fastNLP 5 years ago
  yh_cc 62fe53b147 修改GradientClipCallback中parameter的存储方式，防止在torch 1.5版本中报错 5 years ago
  Yige Xu 1a1254b217 修改roberta embedding的一个文档错误 5 years ago
  yh_cc fbd2fd4ead update DataSet使得可以与之前老版本的DataSet兼容 5 years ago
  yh_cc ff8b9a37a7 修复Embedding注释错误与core.utils.py代码错误 5 years ago
  yh_cc 91d9b239c2 Merge branch 'master' of github.com:fastnlp/fastNLP 5 years ago
  yh_cc 4dd16be8ca 修复vocabulary iterate时不符合常规情况的问题 5 years ago
  ChenXin 4557202023 更新了 ControlC 的 API 5 years ago
  ChenXin 7cdd2ee862 在文档中增加了分布式训练的介绍（by yunfan） 5 years ago
  yunfan a1ac3562fe [update] dist_trainer tutorial 5 years ago
  yunfan d997e5c77a [bugfix] fix start_time, save&load in dist_trainer 5 years ago
  yh_cc d6b9c59cbe Merge branch 'master' of github.com:fastnlp/fastNLP 5 years ago
  yh_cc 18747e632e 修复Vocabulary在load的时候可能发生的bug 5 years ago
  Yige Xu f0e1bdade8 [bugfix] fix an import error 5 years ago
  yh_cc 2dee67129a Solve the GPTTokenizer dependency problem 5 years ago
  yh_cc ab047a2aa3 新增对RobertaEmbedding和GPT2Embedding的支持 6 years ago
  yh_cc cbbfc18149 Merge branch 'master' of github.com:fastnlp/fastNLP 6 years ago
  yh_cc 7398f2b0b2 1.增加RobertaEmbedding与GPT2Embedding 6 years ago
  yunfan 8a084c4f52 [update] bugfix in dist_trainer 6 years ago
  yunfan b49c694c62 [update] bugfix in dist_trainer 6 years ago
  yh_cc 45139ebbff BertEmbedding支持选择是否缩小词表 6 years ago
  yh_cc ace8aafdb2 修复BertEmbedding在include_cls_sep=True时，SEP位置的值不正确的问题 6 years ago
  Yige Xu d2ca466325 upload init roberta embedding 6 years ago
  yh_cc ba645f4483 1.将collect_fn修改为collate_fn使得与pytorch保持一致; 2.增加对BatchSampler的支持 6 years ago
  yh_cc 9b77ad812c 暴露更多类 6 years ago
  yh_cc 3b0e5e0634 暴露更多内容 6 years ago
  yh_cc eba9121a9d 对外暴露LossBase 6 years ago