Online Sequence-to-Sequence Active Learning for Open-Domain Dialogue Generation
来源：互联网 发布：重启后正在准备windows 编辑：程序博客网 时间：2024/05/21 22:59
introduction
- Seq2Seq drawback: generate short, dull and inconsistent responses.
- DRL:
- reward function: most hand-crafted
this paper propose an end-to-end, neural network based generative conversational model that learns open-domain conversation skills via online interaction with human users.
Model
- Offline Two-Phase Supervised Learning
- responses are short and dull
- use Online Active Learning to tackle this issue
- Online Active Learning
- interacts with real users and learns incrementally from their feedback at each turn of dialog
datasets: considerably small (300K and 8K
resp.)
阅读全文
0 0
- Online Sequence-to-Sequence Active Learning for Open-Domain Dialogue Generation
- Deep Reinforcement Learning for Dialogue Generation
- Adversarial Learning for Neural Dialogue Generation
- Deep Reinforcement Learning for Dialogue Generation 翻译
- Convolutional Sequence to Sequence Learning
- 论文引介 | Adversarial Learning for Neural Dialogue Generation
- 论文引介 | Adversarial Learning for Neural Dialogue Generation
- 论文引介 | Deep Reinforcement Learning for Dialogue Generation
- Deep Reinforcement Learning for Dialogue Generation阅读笔记
- Sequence to Sequence Learning with Neural Networks
- Convolutional Sequence to Sequence Learning笔记
- "Sequence to Sequence Learning for Optical Character Recognition"——Devendra Kunar Sahu
- Deep Reinforcement Learning for Dialogue Generation-关于生成对话的深度强化学习
- （翻译）Sequence to Sequence Learning with Neural Networks
- 【论文笔记】Sequence to sequence Learning with Neural Networks
- 【论文阅读】Convolutional Sequence to Sequence Learning （未完待续）
- [ACL2016] Incorporating Copying Mechanism in Sequence-to-Sequence Learning
- [2014]Sequence to Sequence Learning with Neural Networks
- Java多线程之ThreadLocal
- leetcode 22. Generate Parentheses
- linux的简单命令(持续记录)
- mysql之存储过程，函数，游标
- 47 使用linux内核源码里的按键驱动<GPIO Buttons>
- Online Sequence-to-Sequence Active Learning for Open-Domain Dialogue Generation
- BZOJ 4435: [Cerc2015]Juice Junctions tarjan
- Vue2.0 推荐开发环境
- 关于fragment懒加载问题
- Win8.1+VS2013+WDK8.1+VirtualBox or VMware 驱动开发环境配置
- 【MR】剖析 YARN 框架
- 正则表达式
- Modbus CRC16校验算法--查表法（经过测试，工作良好）
- java读取计算机CPU、内存等信息（Sigar使用）
