Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / ChenglongChen/tensorflow-DeepFM issues and pull requests
#82 - 模型auc
Issue -
State: open - Opened by WoNiuHu over 2 years ago
- 1 comment
#81 - 用的tenflow几
Issue -
State: open - Opened by Angel0726 over 2 years ago
#80 - 算法如何调参
Issue -
State: open - Opened by washujiang over 2 years ago
#79 - tensor flow version
Issue -
State: open - Opened by KentaItakura almost 3 years ago
- 2 comments
#78 - In DeepFm.py line 102, why the second order term's size is (None, K) rather than (None, )?
Issue -
State: open - Opened by Wang-Yu-Qing about 3 years ago
- 2 comments
#77 - 数据在kaggle下载不了、需要注册账号,账号又注册不了,能把数据放上去吗?或者放到百度网盘也行
Issue -
State: open - Opened by xiaoyezi55 over 3 years ago
#76 - 话说CTR模型的交互类的特征是怎么做的和存储的?
Issue -
State: open - Opened by guotong1988 over 3 years ago
- 1 comment
#75 - 话说现在用TensorFlow分布式训练CTR模型怎么搞速度快啊?
Issue -
State: open - Opened by guotong1988 over 3 years ago
#74 - first order term 这里这个 求sum的操作 的意义是什么?
Issue -
State: closed - Opened by guotong1988 over 3 years ago
- 1 comment
#73 - 我小白,感觉这行代码挺有经验,motivation是?
Issue -
State: open - Opened by guotong1988 over 3 years ago
#72 - 120万数据提示内存不足
Issue -
State: open - Opened by 1980695671 almost 4 years ago
#71 - 怎么获得example的数据集
Issue -
State: closed - Opened by 1980695671 almost 4 years ago
#70 - 使用criteo数据,精度到不了80%
Issue -
State: open - Opened by yhc960527 about 4 years ago
- 1 comment
#69 - fm输出的交叉特征个数不应该是 F*(F-1)/2吗?
Issue -
State: closed - Opened by W-void about 4 years ago
- 1 comment
#68 - from tensorflow.contrib.layers.python.layers import batch_norm as batch_norm
Issue -
State: open - Opened by Miaotxy about 4 years ago
#67 - question: can't multiply sequence by non-int of type 'float'
Issue -
State: closed - Opened by binxixi over 4 years ago
#66 - 关于特征的处理问题
Issue -
State: open - Opened by liheyong over 4 years ago
#65 - 测试集为什么没有target,想xi,xv,ids就可以预测,损失怎么求得?
Issue -
State: open - Opened by chenqiCSDN over 4 years ago
#64 - 为什么FM 侧y_first_order 里面没有增加一个单独的bias项, 只有关于feature权重的feature_bias
Issue -
State: open - Opened by nevergiveuphe over 4 years ago
#63 - 怎么将模型转换成pb文件呢?输出层和输入层是什么
Issue -
State: open - Opened by KatouMegumiMoe over 4 years ago
- 1 comment
#62 - concat之后再输出,在计算结果上,和论文中 sigmoid(y_FM+y_DNN) 单独计算再加和是一样的。
Issue -
State: open - Opened by van19 almost 5 years ago
#61 - 关于embedding size的设置
Issue -
State: open - Opened by wangjialin114 almost 5 years ago
#59 - 这个项目的版本有吗 TensorFlow 用哪个版本啊
Issue -
State: open - Opened by foreseez about 5 years ago
- 2 comments
#58 - multi-hot feature
Issue -
State: open - Opened by lisongs1995 about 5 years ago
- 2 comments
#58 - multi-hot feature
Issue -
State: open - Opened by lisongs1995 about 5 years ago
- 2 comments
#57 - 为什么DeepFM的Normalized Gini反而小于DNN和FM?
Issue -
State: open - Opened by QiDekang over 5 years ago
- 2 comments
#56 - confusing statement
Issue -
State: closed - Opened by lisongs1995 over 5 years ago
#55 - 您好,请问如何修改为多分类,并输出给分类的概率
Issue -
State: open - Opened by huminwudi over 5 years ago
- 2 comments
#54 - 都预测成了0 或者 1, 此时auc为0.5。请问这是为什么?
Issue -
State: closed - Opened by ArriettyDouDou over 5 years ago
#53 - 存成pb文件后读取报错
Issue -
State: closed - Opened by Leetang007 over 5 years ago
- 1 comment
#52 - self.dropout_keep_fm和self.dropout_keep_deep
Issue -
State: open - Opened by xzbwx over 5 years ago
#51 - 在线预测效率大概多久?
Issue -
State: open - Opened by dotsonliu over 5 years ago
#50 - TypeError: Expected string passed to parameter 'tensor_names' of op 'SaveV2', got ['Variable', 'Variable/Adam', 'Variable/Adam_1', 'Variable_1', 'Variable_1/Adam', 'Variable_1/Adam_1', 'Variable_2', 'Variable_2/Adam', 'Variab
Issue -
State: open - Opened by wylwylwyl over 5 years ago
- 4 comments
#49 - 如何export模型
Issue -
State: open - Opened by ShengZhaoDUT over 5 years ago
- 9 comments
#48 - GPU Utilization is very law
Issue -
State: open - Opened by deepkshikha over 5 years ago
#47 - 换了一批数据,auc一直0.5,为什么?
Issue -
State: open - Opened by VincentLiu123 over 5 years ago
- 15 comments
#46 - 线上复现
Issue -
State: open - Opened by aa147138 over 5 years ago
- 4 comments
#45 - 数据不需要做onehot处理吗
Issue -
State: open - Opened by VincentLiu123 over 5 years ago
- 1 comment
#44 - 输入是文本特征
Issue -
State: open - Opened by wangdazi over 5 years ago
#43 - cate
Issue -
State: closed - Opened by mrchor over 5 years ago
- 1 comment
#42 - fm的隐向量v呢
Issue -
State: open - Opened by seven-xu almost 6 years ago
- 2 comments
#41 - 这些参数可以给个解释吗
Issue -
State: open - Opened by FireSK7 almost 6 years ago
#40 - 收敛极度慢
Issue -
State: open - Opened by taojian1989 almost 6 years ago
- 1 comment
#39 - 关于代码107行
Issue -
State: open - Opened by sudongxiang about 6 years ago
- 3 comments
#38 - Why are categorical data all set 1?
Issue -
State: open - Opened by Superhors about 6 years ago
- 5 comments
#37 - 换了一批其他数据,然后dfm.predict 结果为nan
Issue -
State: open - Opened by balataTree about 6 years ago
- 8 comments
#36 - 在超大规模数据上训练或者增量更新,模型要做哪些方面的更改?
Issue -
State: open - Opened by down92 about 6 years ago
#35 - 连续特征的embedding问题
Issue -
State: open - Opened by hugstone about 6 years ago
- 3 comments
#34 - 请问有在criteo数据集测试过性能吗?
Issue -
State: open - Opened by anzhizh about 6 years ago
- 6 comments
#33 - int/long型的特征要放入NUMERIC_COLS当中吗?
Issue -
State: open - Opened by taojian1989 about 6 years ago
#32 - 模型结构
Issue -
State: open - Opened by fangding678 about 6 years ago
- 5 comments
#31 - Could you explain the way computing gini? I did not know why computing such that
Issue -
State: open - Opened by science-dojo about 6 years ago
- 1 comment
#30 - 关于特征的向量表示我有一个疑问
Issue -
State: open - Opened by sshzhang about 6 years ago
- 1 comment
#29 - The AUC Result
Issue -
State: closed - Opened by Coalin over 6 years ago
- 3 comments
#28 - 请问下如果是针对多分类,target的格式应该是什么?
Issue -
State: closed - Opened by shuangyangwang over 6 years ago
- 4 comments
#27 - auc ????
Issue -
State: closed - Opened by wangbingok1118 over 6 years ago
- 3 comments
#26 - 训练报错
Issue -
State: closed - Opened by shuangyangwang over 6 years ago
- 1 comment
#25 - 如何处理pre-trained embeddings?
Issue -
State: closed - Opened by shuangyangwang over 6 years ago
- 2 comments
#24 - Field-aware?
Issue -
State: closed - Opened by a-bharath over 6 years ago
- 2 comments
#23 - 连续特征为什么需要做embedding?
Issue -
State: closed - Opened by flyzaway over 6 years ago
- 1 comment
#22 - 抱歉,还是要吐槽一下
Issue -
State: closed - Opened by sudongxiang over 6 years ago
#21 - 关于代码89行
Issue -
State: closed - Opened by sudongxiang over 6 years ago
#20 - 运行直接出AttributeError
Issue -
State: closed - Opened by wuhaotju over 6 years ago
- 1 comment
#19 - 该问题不存在
Issue -
State: closed - Opened by sudongxiang over 6 years ago
#18 - 在DataParser中的parse函数,为啥 dfv[col] = 1. ?
Issue -
State: closed - Opened by icodezign over 6 years ago
- 4 comments
#17 - 代码出错
Issue -
State: closed - Opened by CXY573 over 6 years ago
#16 - Deep Component最开始的输入应该是最原始的Embedding结果?
Issue -
State: closed - Opened by gutouyu over 6 years ago
- 2 comments
#15 - 如何处理一个field中有多个特征为1的情况
Issue -
State: closed - Opened by xvlvzhu over 6 years ago
- 5 comments
#14 - ImportError: No module named 'yellowfin'
Issue -
State: closed - Opened by yt605155624 over 6 years ago
- 2 comments
#13 - Input Dimension Error
Issue -
State: closed - Opened by gevmin94 over 6 years ago
- 1 comment
#12 - 三种特征的区别是什么,如何区分
Issue -
State: closed - Opened by land1725 over 6 years ago
- 3 comments
#11 - 有没有实现sparse输入的deepfm
Issue -
State: open - Opened by starkeisntein over 6 years ago
- 2 comments
#10 - Why predict same test set, but given different predict_prob?
Issue -
State: open - Opened by BigHeartC over 6 years ago
- 4 comments
#9 - concat层的输入
Issue -
State: closed - Opened by ningyuwhut over 6 years ago
- 1 comment
#8 - Save trained model
Pull Request -
State: closed - Opened by mathetes87 over 6 years ago
- 1 comment
#7 - 序列特征如何处理
Issue -
State: closed - Opened by liuyangvoid almost 7 years ago
- 1 comment
#6 - Is the MSE loss in the code right?
Issue -
State: closed - Opened by sddi almost 7 years ago
- 1 comment
#5 - memory error for big data
Issue -
State: closed - Opened by yufengwhy almost 7 years ago
- 7 comments
#4 - feature field j has two or more nonzero values
Issue -
State: closed - Opened by sddi almost 7 years ago
- 2 comments
#3 - No module named yellowfin
Issue -
State: closed - Opened by CanoeFZH almost 7 years ago
- 1 comment
#2 - example has `metric` as model param, but constructor takes `eval_metric`
Issue -
State: closed - Opened by ryancheunggit almost 7 years ago
- 1 comment
#1 - MAE loss doesn't decrease?
Issue -
State: closed - Opened by kroscek almost 7 years ago
- 1 comment