Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / terrifyzhao/bert-utils issues and pull requests
#78 - similarity识别率太低
Issue -
State: open - Opened by myinvoke about 3 years ago
#77 - 缺少相关配置文件
Issue -
State: closed - Opened by li-jp about 3 years ago
- 1 comment
#76 - 运行extract-feature.py时遇到的问题。
Issue -
State: closed - Opened by HenryWInfinity over 3 years ago
- 2 comments
#75 - 你这个代码有个大坑!
Issue -
State: open - Opened by dullblue over 3 years ago
- 2 comments
#74 - 这个是怎么回事
Issue -
State: open - Opened by huabao97 over 3 years ago
- 1 comment
#73 - 为什么运行feature.py时,没有graph生成,
Issue -
State: open - Opened by huabao97 over 3 years ago
#72 - 请问句向量的负值怎样能去掉或者变成全部正的呢?只能用relu函数变成0吗?
Issue -
State: open - Opened by g1kyne almost 4 years ago
#71 - 请问生成句向量的模块如何使用tfServing进行部署?
Issue -
State: open - Opened by Wurr about 4 years ago
#70 - 直接用bert句向量结果计算相似度,任意两个句子的余弦值都很高,请问这是什么原因
Issue -
State: open - Opened by ann22 over 4 years ago
- 4 comments
#69 - 本机windows环境中转向量正常, 在linux环境下启动脚本,程序僵死,一直处于高cpu占用状态
Issue -
State: open - Opened by pandab0y over 4 years ago
#68 - 对sentence进行相似度预测时输入分词错误
Issue -
State: closed - Opened by LawrenceGK over 4 years ago
- 1 comment
#67 - 加载过graph后,还是比较慢,所需时间是 bert-as-service的20倍,这是为什么?
Issue -
State: closed - Opened by zhouyongjie over 4 years ago
- 2 comments
#66 - The BQ Corpus.pdf 无法查看
Issue -
State: open - Opened by Wang-Yufei over 4 years ago
#65 - 请问一下这个怎么看结果啊
Issue -
State: closed - Opened by bao17634 over 4 years ago
#64 - 如何用自己的语料库微调模型
Issue -
State: open - Opened by 1eclipse over 4 years ago
#63 - 请教一下,利用问答对进行fine-turning后的模型,然后获取向量的方法跟extract_feature文件中的方法一致么?此外,这样获取向量跟训练阶段是不一致的,因为输入数据是单个,训练的时候是一对数据?
Issue -
State: open - Opened by victory-hsu over 4 years ago
#62 - 你好,想请教下怎样加载albert-tiny,small,可以支持下这个吗?
Issue -
State: open - Opened by jiahenghuang over 4 years ago
#61 - 相似度计算
Issue -
State: open - Opened by whao1011 over 4 years ago
#60 - 模型训练正常,auc较好,设置为PREDICT模式时候抛错issue 3770
Issue -
State: open - Opened by CindyHuang-cn over 4 years ago
#59 - 如果希望对特定种类的文本的句子提取句子向量,怎样对模型进行微调更合适
Issue -
State: open - Opened by leonelacs almost 5 years ago
- 1 comment
#58 - 关于输出句向量维度的问题。768->128
Issue -
State: open - Opened by XGodLike almost 5 years ago
- 3 comments
#57 - 相同文字生成的句向量和BAS以及普通bert方法生成的有区别吗?
Issue -
State: open - Opened by M0025 almost 5 years ago
- 1 comment
#56 - 速度问题。。。
Issue -
State: open - Opened by duan348733684 about 5 years ago
- 3 comments
#55 - 运行extract_feature时出现问题
Issue -
State: open - Opened by yzc1103 about 5 years ago
- 4 comments
#54 - 第一次生成句向量时间
Issue -
State: open - Opened by superguopeng about 5 years ago
- 1 comment
#54 - 第一次生成句向量时间
Issue -
State: open - Opened by superguopeng about 5 years ago
- 1 comment
#53 - 关于generate_from_queue和estimator.predict中的逻辑问题
Issue -
State: closed - Opened by FisherOuch about 5 years ago
- 1 comment
#52 - 这个程序是怎么运行起来的啊,我也遇到了该问题
Issue -
State: open - Opened by 1254859753 about 5 years ago
#51 - 地址文本可不可以使用该方式
Issue -
State: closed - Opened by 1254859753 about 5 years ago
#50 - 一次塞入多个句子,循环生成句向量时出错
Issue -
State: closed - Opened by ARSblithe212 about 5 years ago
- 2 comments
#49 - predict_from_queue为什么要设置守护线程,去掉有什么影响吗
Issue -
State: open - Opened by ARSblithe212 about 5 years ago
#48 - 请问哪里可修改句向量长度?
Issue -
State: open - Opened by ghm over 5 years ago
- 1 comment
#47 - Update similarity.py
Pull Request -
State: open - Opened by sparkingarthur over 5 years ago
#46 - bert.encode()输入空字符串会无响应
Issue -
State: closed - Opened by Arctanxy over 5 years ago
#45 - 如何获取词向量
Issue -
State: open - Opened by ChenHao314 over 5 years ago
- 1 comment
#44 - 版本问题(谢谢)
Issue -
State: open - Opened by 18234040430 over 5 years ago
- 1 comment
#43 - 修改args max_seq_len后报错
Issue -
State: closed - Opened by zachkang over 5 years ago
- 3 comments
#42 - No module named "bert"
Issue -
State: open - Opened by xudekuan over 5 years ago
- 2 comments
#40 - 运行extract_feature.py
Issue -
State: open - Opened by wuaibo over 5 years ago
- 1 comment
#39 - 按说明运行文本分类,最后提示 BertSim对象没有test方法
Issue -
State: open - Opened by yaleimeng over 5 years ago
- 3 comments
#38 - 你好,想做单纯的分类。遇到问题。
Issue -
State: open - Opened by qizhang0302 over 5 years ago
- 2 comments
#37 - 你好请问有合适的decoder吗?如果想把vector转换成文字该怎么做?
Issue -
State: open - Opened by xinxingit217 over 5 years ago
- 1 comment
#36 - 请问data文件夹下的训练集,测试集是如何准备的?
Issue -
State: open - Opened by sunweiconfidence over 5 years ago
- 2 comments
#35 - 关于predict准确率的问题
Issue -
State: open - Opened by Zhangpeixiang over 5 years ago
- 2 comments
#34 - 句向量的训练方式,本质上还是根据分类模型来的吗?
Issue -
State: open - Opened by fengxin619 over 5 years ago
- 1 comment
#33 - 您好,请问应该怎么修改最后句向量的维度呢,768维对我后续的任务而言太高了,我只需要100维或者50维的向量哎,(谢谢你)
Issue -
State: open - Opened by 27232xsl over 5 years ago
- 20 comments
#32 - 新浪新闻数据分类结果随机
Issue -
State: closed - Opened by liutianling over 5 years ago
#31 - extract_feature.py 句向量生成demo build graph 显示 Could not find trained model in model_dir
Issue -
State: open - Opened by chenyumiyu over 5 years ago
- 4 comments
#30 - 显存不足问题
Issue -
State: open - Opened by jamesguo over 5 years ago
- 1 comment
#29 - 同时启动两个Bert对象出错
Issue -
State: closed - Opened by cdj0311 over 5 years ago
- 3 comments
Labels: bug
#28 - 用gpu环境跑出问题
Issue -
State: open - Opened by mllwm over 5 years ago
- 6 comments
#27 - 为什么encode里用queue来实现
Issue -
State: open - Opened by BucherLi over 5 years ago
- 3 comments
#26 - 相似度问题
Issue -
State: open - Opened by zhuolang31 over 5 years ago
- 3 comments
#25 - 句向量問題
Issue -
State: open - Opened by biuleung over 5 years ago
- 5 comments
#24 - jupyter notebook
Issue -
State: closed - Opened by ZJUHSY over 5 years ago
- 3 comments
#23 - 首次执行句向量方法太慢,为啥bert-as-service没这么慢
Issue -
State: open - Opened by BucherLi over 5 years ago
- 1 comment
#22 - 训练好模型后,进行eval
Issue -
State: open - Opened by lphu over 5 years ago
- 9 comments
#21 - 用此代码跑eval的时候出现的问题
Issue -
State: closed - Opened by Venut-yjc over 5 years ago
- 1 comment
#20 - Floating point exception and SystemError: error return without exception set
Issue -
State: closed - Opened by yongzhuo over 5 years ago
- 3 comments
#19 - train.csv文件有错误
Issue -
State: open - Opened by zhangxuanaj almost 6 years ago
- 3 comments
#18 - gpu or cpu
Issue -
State: closed - Opened by zhangxuanaj almost 6 years ago
#17 - _truncate_seq_pair method does not seem to be reasonable
Issue -
State: closed - Opened by HedyHu almost 6 years ago
#16 - 句向量为什么不需要进行fine tune
Issue -
State: open - Opened by Jothangan almost 6 years ago
- 1 comment
#15 - 如何获取相似度更高的两个语句
Issue -
State: closed - Opened by Jothangan almost 6 years ago
- 6 comments
#14 - 直接通过词向量计算相似度的时候,没看到有什么效果?
Issue -
State: open - Opened by LCG22 almost 6 years ago
- 4 comments
#13 - 进行fine-tune之后的模型再进行向量提取会不会精度更高?
Issue -
State: open - Opened by Timthony almost 6 years ago
- 1 comment
#12 - 如何用GPU跑
Issue -
State: open - Opened by JiahangOK almost 6 years ago
- 2 comments
#11 - 出现 KeyERROR'0'
Issue -
State: open - Opened by lphu almost 6 years ago
- 1 comment
#10 - 出现 KeyERROR'0'
Issue -
State: closed - Opened by lphu almost 6 years ago
- 1 comment
#9 - 相似度预测方法
Issue -
State: open - Opened by lotushacker almost 6 years ago
- 10 comments
Labels: question
#8 - 句向量最后生成为什么要进行mask操作
Issue -
State: closed - Opened by wingsyuan almost 6 years ago
- 2 comments
Labels: question
#7 - 如何只打印句向量
Issue -
State: closed - Opened by wingsyuan almost 6 years ago
- 2 comments
#6 - when run BertSim
Issue -
State: open - Opened by sunjinguo almost 6 years ago
- 1 comment
#5 - 训练误差突然变大,怎么回事
Issue -
State: open - Opened by zongking123 almost 6 years ago
- 3 comments
Labels: question
#4 - Could not find trained model in model_dir: /var/folders/b8/3ywv8wg10hlbfzv5m0nzswj80000gn/T/tmplx_CY_
Issue -
State: open - Opened by Simon-Liu622 almost 6 years ago
- 3 comments
#3 - 报错: ValueError: Could not find trained model in model_dir: /tmp/tmpxlwy5mwx.
Issue -
State: closed - Opened by bound2020 almost 6 years ago
- 3 comments
Labels: bug
#2 - 训练数据是根据什么做的分类?
Issue -
State: open - Opened by 8000cabbage almost 6 years ago
- 3 comments
Labels: question
#1 - When run BertVector
Issue -
State: open - Opened by manxin0821 almost 6 years ago
- 14 comments