Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / brightmart/roberta_zh issues and pull requests

#100 - updata:测试

Pull Request - State: open - Opened by Cpeidong 4 months ago - 1 comment

#98 - 加载的小问题求解答

Issue - State: open - Opened by cooper12121 over 1 year ago

#97 - 下载问题和加载模型

Issue - State: open - Opened by Lj4040 over 1 year ago

#96 - 333add

Pull Request - State: closed - Opened by luoshao23 about 2 years ago

#95 - Loss curve

Issue - State: open - Opened by wanglaiqi over 2 years ago

#93 - 利用roberta_zh的tokenizer来做中文NER任务时报错

Issue - State: open - Opened by Honma-Rika over 3 years ago - 2 comments

#92 - Update README.md

Pull Request - State: closed - Opened by Yang-Jianzhang over 3 years ago

#91 - Huggingface

Issue - State: open - Opened by archersama over 3 years ago

#90 - 请问下，怎么进行GPU训练？

Issue - State: open - Opened by Tian14267 over 3 years ago

#88 - resource文件夹下的vocab和代码不对应

Issue - State: closed - Opened by bigheary over 3 years ago

#87 - Unrelated parameters in the config

Issue - State: open - Opened by zheyuye over 3 years ago

#86 - 预处理数据丢失问题

Issue - State: open - Opened by puzzledTao almost 4 years ago - 1 comment

#84 - 关于MLM中，中文全词掩盖的预测标签问题

Issue - State: open - Opened by Rango94 almost 4 years ago - 7 comments

#82 - 是否可以开放语料，供其他模型对比

Issue - State: open - Opened by lamp-lyz about 4 years ago

#81 - CMRC示例

Issue - State: closed - Opened by chuzhifeng about 4 years ago

#80 - NaN probability sometimes when inference on GPU

Issue - State: open - Opened by Jiayuforfreeo about 4 years ago

#79 - 在pytorch模型上做post train

Issue - State: open - Opened by daniellibin about 4 years ago - 2 comments

#77 - XLNet其实不能稳压RoBERTa吧？

Issue - State: closed - Opened by guotong1988 about 4 years ago - 1 comment

#76 - 做全词遮蔽的比例设置？

Issue - State: open - Opened by surimj about 4 years ago - 1 comment

#75 - 关于中文编码

Issue - State: open - Opened by Leputa over 4 years ago

#74 - pretrain问题：ValueError: Please provide a TPU Name to connect to.

Issue - State: closed - Opened by xuehui0725 over 4 years ago - 2 comments

#73 - pretrain 数据问题

Issue - State: open - Opened by ruleGreen over 4 years ago - 3 comments

#72 - 预训练数据构造有误

Pull Request - State: open - Opened by hy-struggle over 4 years ago

#71 - Update create_pretraining_data.py

Pull Request - State: open - Opened by guotong1988 over 4 years ago

#70 - 用自己的数据构建pretrain data 提示 KeyError: '##cry'

Issue - State: open - Opened by ccoocode over 4 years ago - 1 comment

#69 - 希望能提供一个longformer的中文预训练模型

Issue - State: open - Opened by xjx0524 over 4 years ago

#68 - 没有merge.txt和vocab.json

Issue - State: open - Opened by lshowway over 4 years ago - 3 comments

#66 - 动态mask逻辑的实现

Issue - State: open - Opened by humdingers over 4 years ago - 1 comment

#65 - 部分参数无法重载

Issue - State: open - Opened by WBY1993 over 4 years ago

#64 - 数据处理中re.findall('##[\u4E00-\u9FA5]')作用

Issue - State: open - Opened by xiaojinglu over 4 years ago - 2 comments

#62 - 请问英文的roberta预训练模型哪里可以找得到？？？

Issue - State: open - Opened by WenxiongLiao over 4 years ago - 1 comment

#61 - 如何在论文中引用您发布的模型？

Issue - State: open - Opened by CCNUdhj almost 5 years ago

#60 - 请问一下有如下的权重可以分享吗?

Issue - State: closed - Opened by rxc205 almost 5 years ago - 1 comment

#59 - tensorflow版本

Issue - State: open - Opened by aflyhat almost 5 years ago - 1 comment

#55 - Pytorch版本的网盘失效了

Issue - State: open - Opened by cqlyiyeshu almost 5 years ago - 2 comments

#54 - 使用roberta-large去预测mask的位置，结果不可读

Issue - State: closed - Opened by yayaQAQ about 5 years ago - 1 comment

#53 - 关于预训练模型

Issue - State: closed - Opened by Foehnc about 5 years ago - 2 comments

#51 - RoBERTa-zh-Large百度网盘文件取消了

Issue - State: open - Opened by jh-deng about 5 years ago - 6 comments

#49 - 关于在您的模型上继续预训练

Issue - State: open - Opened by zhezhaoa about 5 years ago - 2 comments

#47 - 指标里的模型都指的那个？

Issue - State: open - Opened by oyjxer about 5 years ago - 2 comments

#46 - 关于pytorch版本

Issue - State: open - Opened by zjjhuihui about 5 years ago

#44 - 关于多卡训练

Issue - State: open - Opened by gaozhanfire about 5 years ago - 7 comments

#43 - 运行run_classifier的时候oom了

Issue - State: open - Opened by zhengchang231 about 5 years ago - 5 comments

#42 - roberta 预训练语料格式

Issue - State: closed - Opened by zhengwsh about 5 years ago - 1 comment

#41 - 请问关于数据预处理

Issue - State: open - Opened by ouwenjie03 about 5 years ago - 3 comments

#39 - Use Roberta in pytorch transformers

Issue - State: open - Opened by Vimos about 5 years ago - 5 comments

#38 - 请问能否在所有权重中都把mlm部分带上？

Issue - State: closed - Opened by bojone about 5 years ago - 3 comments

#38 - 请问能否在所有权重中都把mlm部分带上？

Issue - State: closed - Opened by bojone about 5 years ago - 3 comments

#36 - Dynamic mask是否會造成data leakage的問題？

Issue - State: closed - Opened by nickchiang5121 about 5 years ago - 1 comment

#35 - [CLS],[SEP]

Issue - State: closed - Opened by HongyanJiao about 5 years ago - 5 comments

#34 - cuda out of memory

Issue - State: closed - Opened by HongyanJiao about 5 years ago - 1 comment

#33 - 您好，请问可以提供一下预训练阶段的代码吗

Issue - State: closed - Opened by bytekongfrombupt about 5 years ago - 3 comments

#32 - What is the best CPU inference acceleration solution for BERT now?

Issue - State: closed - Opened by guotong1988 about 5 years ago - 1 comment

#31 - 请教一下处理预训练数据的大致策略？

Issue - State: closed - Opened by fengzuo97 about 5 years ago - 3 comments

#29 - roberta的英文预训练权重

Issue - State: closed - Opened by c0derm4n about 5 years ago - 1 comment

#28 - 您好，RoBERTa_zh_L12下载时点击网页打不开

Issue - State: closed - Opened by haosiqing about 5 years ago - 1 comment

#25 - 关于预训练的embedding

Issue - State: closed - Opened by beamind about 5 years ago - 4 comments

#23 - 与RoBERTa-wwm-ext, Chinese对比

Issue - State: closed - Opened by jxst539246 about 5 years ago - 3 comments

#22 - 6层的roberta模型啥时候发布呀？

Issue - State: closed - Opened by Jethu1 about 5 years ago - 10 comments

#21 - 关于BPE

Issue - State: closed - Opened by ArthurRizar about 5 years ago - 2 comments

#20 - 加载pytorch模型错误

Issue - State: closed - Opened by Benjamin-zhangjb about 5 years ago - 10 comments

#19 - 想尝试转换，用tf.train.list_variables(init_checkpoint)加载模型报错

Issue - State: closed - Opened by wang001 about 5 years ago - 5 comments

#18 - loss库

Issue - State: closed - Opened by Pooky-Z about 5 years ago - 1 comment

#17 - loss

Issue - State: closed - Opened by Pooky-Z about 5 years ago

#16 - 请问为什么roberta_large比roberta_middle在CMRC2018上低很多？

Issue - State: closed - Opened by ewrfcas about 5 years ago - 19 comments

#15 - 不能使用tf.train.load_variable函数来加载模型吗？

Issue - State: closed - Opened by RingoTC about 5 years ago - 1 comment

#14 - 下载RoBERTa-zh-Large总是提示网页走丢

Issue - State: closed - Opened by haosiqing about 5 years ago - 6 comments

#13 - 可以增加xlnet的在这两个数据集上的对比吗？

Issue - State: closed - Opened by RyanHuangNLP about 5 years ago - 1 comment

#12 - 请问12layer的啥时候发布呢，我看界面写的今天。

Issue - State: closed - Opened by YingZiqiang about 5 years ago - 1 comment

#11 - 请问有国内链接吗，翻墙下载速度太慢了，谢谢！

Issue - State: closed - Opened by TPF2017 about 5 years ago - 3 comments

#10 - [question] Pretrain longer

Issue - State: closed - Opened by lsq357 about 5 years ago - 2 comments

#9 - 请问应该如何转成pyTorch格式？

Issue - State: closed - Opened by Dongfeng-He about 5 years ago - 21 comments

#6 - 原始的bert代码能加载这个模型吗

Issue - State: closed - Opened by renmada about 5 years ago - 1 comment

#5 - 下载失败，点开链接出现下面这个

Issue - State: closed - Opened by luoyexuge about 5 years ago - 1 comment

#4 - 下载失败

Issue - State: closed - Opened by currenttime about 5 years ago - 1 comment

#3 - 24层base版(roberta_l24_zh_base）是用于bert的么？

Issue - State: closed - Opened by qblyy about 5 years ago - 2 comments