当前位置:网站首页>The strongest installation of the twin tower model, Google is playing "antique" again?
The strongest installation of the twin tower model, Google is playing "antique" again?
2022-07-07 21:55:00 【Zhiyuan community】
The twin tower model has proved to be a very effective modeling method in search and question and answer tasks , The theory and business are quite mature . The two towers share different degrees according to parameters , It usually falls into two categories :Simese dual encoder and Asymmetric dual encoder, The former parameter structure is completely symmetrical , The latter is not completely symmetrical ( Hereinafter referred to as" SDE and ADE).
This paper is after the long silence of the twin towers , Google pushed it to the center of the universe again , And open the strongest export of the twin towers , Explore the differences and connections between the two in detail , More empirical conclusions of the double tower structure are also given through experiments . It is suitable for old drivers to recall classics and Xiaobai again and make a deep and systematic introduction ~
Thesis title :
Exploring Dual Encoder Architectures for Question Answering
Thesis link :
https://arxiv.org/abs/2204.07120
background
First of all, what is popular science SDE and ADE? The dual encoder network structure will text1 and text2 Respectively encoded into vector representation , Then calculate the sum of the two cosine Equidistance function measures its similarity .SDE Is a twin network that fully shares parameters , That is, although it is a double tower , But actually query/user and doc/item Share a set of parameters ;ADE Only some parameters are shared or not shared at all , It is an independent two parameter network . What they have in common is that they will not interact deeply , contrast BERT Is a typical interactive network . A typical application of double tower structure is recall or Rough row , Scenarios that require strict computing speed .
The modeling idea of twin towers is relatively simple and easy to understand . This article is short and concise , The highlight is Provide a more general conclusion under the twin tower application scenario , Explain several questions clearly :
- ADE and SDE stay QA Which one works better on the task ?
- ADE What are the reasons for poor performance ? What's the solution ?
The author draws a reliable conclusion through reasonable and detailed experiments , Xiaobai can also quickly get To how in ( towards ) real ( guide ) Examination ( t ) Do a section ( Remit ) study ( newspaper ).
experiment
The author in QA The retrieval task is carried out 5 An experiment , Calculation query And candidates answer(doc or passage) The similarity of , The evaluation task is MS MARCO and MultiReQA. Model encoder Is based on transformer,cosine As a distance measurement function , The goal is to explore the influence of the sharing degree of parameters on the modeling effect . 5 A group of experimental networks are the standards of Figure 1 SDE and ADE, as well as 3 Variant structure :• ADE with shared token embedder (ADE-STE) • ADE with frozen token embedder (ADE-FTE) • ADE with shared projection layer (ADE-SPL) The experimental results are as follows :
The experimental conclusion :
- ADE Performance on multiple tasks is significantly inferior to SDE. The reasonable explanation given by the author is due to ADE The essence is two networks with different parameters , So the query and doc Map to two completely different vector spaces . This point later gives more powerful evidence .
- ADE-SPL Our performance is comparable to SDE. after 3 The first experiment is the structure proposed by the author to explore the degree of parameter sharing , At the same time, it also gives which part of the network is limited ADE The key to the effect . Just share or fix the bottom token embedder The effect improvement brought by parameters is not obvious , But when the last top-level parameters share a full connection layer , Can get and SDE The effect of proximity . Why? ? The author's guess is because of the last MLP The parameters are constrained to the same vector space again .
To further illustrate the problem , The author conducted another experiment , take NaturalQuestions Test set query and answer Calculate in advance , And then through t-SNE Map and cluster into a two-dimensional space , Be surprised to find ,dual encoder The performance of depends on whether the last two are in a comparable vector space .
边栏推荐
- Word inversion implements "suggestions collection"
- [UVALive 6663 Count the Regions] (dfs + 离散化)[通俗易懂]
- Usage of MySQL subquery keywords (exists)
- Demon daddy guide post - simple version
- 建立自己的网站(18)
- 用语雀写文章了,功能真心强大!
- L'enregistreur de disque dur NVR est connecté à easycvr par le Protocole GB 28181. Quelle est la raison pour laquelle l'information sur le canal de l'appareil n'est pas affichée?
- Win11如何解禁键盘?Win11解禁键盘的方法
- Use camunda to do workflow design and reject operations
- Typescript TS basic knowledge type declaration
猜你喜欢
Kirin Xin'an operating system derivative solution | storage multipath management system, effectively improving the reliability of data transmission
Ad domain group policy management
强化学习-学习笔记9 | Multi-Step-TD-Target
Win11U盘不显示怎么办?Win11插U盘没反应的解决方法
How does win11 unblock the keyboard? Method of unlocking keyboard in win11
Using enumeration to realize English to braille
Google SEO external chain backlinks research tool recommendation
Solve the problem of using uni app mediaerror mediaerror errorcode -5
Embedded development: how to choose the right RTOS for the project?
建立自己的网站(18)
随机推荐
An overview of the latest research progress of "efficient deep segmentation of labels" at Shanghai Jiaotong University, which comprehensively expounds the deep segmentation methods of unsupervised, ro
Nine degree 1201 - traversal of binary sort number - binary sort tree "suggestions collection"
EasyUI date control emptying value
648. Word replacement
Jerry's key to initiate pairing [chapter]
嵌入式开发:如何为项目选择合适的RTOS?
What stocks can a new account holder buy? Is the stock trading account safe
Goal: do not exclude yaml syntax. Try to get started quickly
[开源] .Net ORM 访问 Firebird 数据库
How does win11 unblock the keyboard? Method of unlocking keyboard in win11
Tsconfig of typescript TS basics JSON configuration options
Tcp/ip protocol stack
强化学习-学习笔记9 | Multi-Step-TD-Target
Unity3d 4.3.4f1执行项目
L'enregistreur de disque dur NVR est connecté à easycvr par le Protocole GB 28181. Quelle est la raison pour laquelle l'information sur le canal de l'appareil n'est pas affichée?
Win11游戏模式怎么开启?Win11开启游戏模式的方法
Ant destination multiple selection
Embedded development: how to choose the right RTOS for the project?
Validutil, "Rethinking the setting of semi supervised learning on graphs"
Contour layout of margin