NLP-Interview-Notes icon indicating copy to clipboard operation
NLP-Interview-Notes copied to clipboard

该仓库主要记录 NLP 算法工程师相关的面试题

Results 5 NLP-Interview-Notes issues
Sort by recently updated
recently updated
newest added

梯度计算应该是对变量求偏导,故应该将 ![image](https://user-images.githubusercontent.com/53287966/146698743-c0e5cde1-e5c0-4f71-9623-58e2e93e005a.png) 替换为 ![image](https://user-images.githubusercontent.com/53287966/146698756-108dfd7a-1e1e-402f-b1f0-ad3e2f1c3243.png)

![图片](https://user-images.githubusercontent.com/59272013/125746251-89e83ad4-c76e-451d-b9f3-10c99f415e19.png)

[【关于 Transformer 问题及改进】那些你不知道的事](https://github.com/km1994/NLP-Interview-Notes/blob/main/DeepLearningAlgorithm/transformer/transformer_error.md)](https://github.com/km1994/NLP-Interview-Notes/blob/main/DeepLearningAlgorithm/transformer/transformer_error.md) 现在是page not found状态,

百面百搭基础算法篇,BN和LN,1.1独立同分布中说 > 强相关:Naive Bayes 模型就建立在特征彼此独立的基础之 > 弱相关:Logistic Regression 和 神经网络 则在非独立的特征数据上依然可以训练出很好的模型 朴素贝叶斯基于特征独立我能理解,这个强相关性指的是什么的强相关性?以及对应的弱相关性是? https://github.com/km1994/NLP-Interview-Notes/blob/main/BasicAlgorithm/BatchNormVsLayerNorm.md#:~:text=%E7%9B%B8%E5%85%B3%E6%80%A7%EF%BC%9A,%E5%BE%88%E5%A5%BD%E7%9A%84%E6%A8%A1%E5%9E%8B