Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

cs25-notes #236

Open
3 tasks done
mlw67 opened this issue Apr 25, 2024 · 1 comment
Open
3 tasks done

cs25-notes #236

mlw67 opened this issue Apr 25, 2024 · 1 comment
Labels

Comments

@mlw67
Copy link

mlw67 commented Apr 25, 2024

你是否已经阅读并同意《Datawhale开源项目指南》?

你是否已经阅读并同意《Datawhale开源项目行为准则》?

项目简介

CS 25 - Notes

Stanford CS 25 | Transformers United 课程笔记,「首个Transformers专题讲座,NLP、CV和RL无所不包」

这门课程是斯坦福大学的 CS 课程的一门前沿课程:《CS 25: Transformers United》。

这门课程的重点就是介绍 Transformers,并统一其在 ML、CV、NLP、生物学和其他社区的使用。此外,该课程还讨论关于 Transformer 的最新突破和想法,以激发交叉合作研究。

CS 25 课程邀请了来自不同领域关于 Transformer 研究的前沿人士进行客座讲座。有 AI 教父 Geoff Hinton;OpenAI 的研究科学家 Mark Chen,主要介绍基于 Transformers 的 GPT-3、Codex;Google Brain 的科学家 Lucas Beyer,主要介绍 Transformer 在视觉领域的应用;Meta FAIR 科学家 Aditya Grover,主要介绍 RL 中的 Transformer 以及计算引擎等。

立项理由

期望通过这个项目,使得更多的小伙伴能够了解到这门课程,以及能够更好的学习这门课程。基于好的知识应该得到更为广泛的传播,所以我们参与的小伙伴对于课程的整体愿景有如下几点:

1)让更多的人了解到CS25这门如此多元化的一门课程(内容的质量高)

2)更提供更低的门槛让更多的人学习这门课程(内容注解是要清晰的)

3)能在课程笔记的注解中提供更多的思考(有些内容重要的不只是知识本身,还有知识背后的思想)

项目受众

由于这门课程的重点就是介绍 Transformers,并统一其在 ML、CV、NLP、生物学和其他社区的使用。此外,该课程还讨论关于 Transformer 的最新突破和想法,以激发交叉合作研究。

想要学习这门课程的小伙伴,必须先要掌握深度学习基础知识(必须理解注意力机制),或者已经通过 CS224N / CS231N / CS230 课程。

所以这门课的受众是:对于深度学习方向有基本的了解,并期望对于相关Transformers方面的研究有更多了解的同学。

项目亮点

Transformers 和各方向交叉应用研究的前沿专题讲座课程,大佬云集。

项目规划

1.目录(如有多级至少精确到二级)
整体计划包含CS25 V1-V4的所有内容(V4-ing中)。

Title(V1)

Introduction to Transformers(同V2 第一课)
Transformers in Language: GPT-3, CodexSpeaker: Mark Chen (OpenAI)
Applications in VisionSpeaker: Lucas Beyer (Google Brain)
Transformers in RL & UniversalCompute EnginesSpeaker: Aditya Grover (FAIR)
Scaling transformersSpeaker: Barret Zoph (Google Brain)with Irwan Bello and Liam Fedus
Perceiver: Arbitrary IO with transformersSpeaker: Andrew Jaegle (DeepMind)
Self Attention & Non-Parametric TransformersSpeaker: Aidan Gomez (University of Oxford)
GLOM: Representing part-whole hierarchies in a neural networkSpeaker: Geoffrey Hinton (UoT)
Interpretability with transformersSpeaker: Chris Olah (AnthropicAI)
Transformers for Applications in Audio, Speech and Music: From Language Modeling to Understanding to Synthesis. Speaker: Prateek Verma (Stanford)

Title(V2)

Introduction to Transformers(同V1 第一课)Speaker: Andrej Karpathy
Language and Human AlignmentSpeaker: Jan Leike (OpenAI)
Emergent Abilities and Scaling in LLMsSpeaker: Jason Wei (Google Brain)
Strategic GamesSpeaker: Noam Brown (FAIR)
Robotics and Imitation LearningSpeaker: Ted Xiao (Google Brain)
Common Sense ReasoningSpeaker: Yejin Choi (U. Washington / Allen Institute for AI)
Biomedical TransformersSpeaker: Vivek Natarajan (Google Health AI)
In-Context Learning & Faithful ReasoningSpeakers: Stephanie Chan (DeepMind) & Antonia Creswell (DeepMind)
Neuroscience-Inspired Artificial IntelligenceSpeakers: Trenton Bricken (Harvard/Redwood Center for Theoretical Neuroscience/Anthropic) & Will Dorrell (UCL Gatsby Computational Neuroscience Unit/Stanford)

Title(V3)

Llama 2: Open Foundation and Fine-Tuned Chat ModelsSpeaker: Sharan Narang, Meta AI
Low-level Embodied Intelligence with Foundation ModelsSpeaker: Fei Xia, Google Deepmind
Generalist Agents in Open-Ended WorldsSpeaker: Jim Fan, NVIDIA AI
Recipe for Training Helpful ChatbotsSpeaker: Nazneen Rajani, HuggingFace
How I Learned to Stop Worrying and Love the TransformerSpeaker: Ashish Vaswani
No Language Left Behind: Scaling Human-Centered Machine TranslationSpeaker: Angela Fan, Meta AI
Going Beyond LLMs: Agents, Emergent Abilities, Intermediate-Guided Reasoning, BabyLMSpeaker: Instructors
Retrieval Augmented Language ModelsSpeaker: Douwe Kiela, Contextual AI

Title(V4)

Instructor Lecture: Overview of Transformers [In-Person]Speakers: Steven Feng, Div Garg, Emily Bunnapradist, Seonghee LeeSlides posted here.
Intuitions on Language Models (Jason) [In-Person]How did we end up here? Early history and evolution of Transformer (Hyung Won) [In-Person]Speakers: Jason Wei & Hyung Won Chung, OpenAI
TBDSpeaker: Nathan Lambert, Allen Institute for AI (AI2)
Demystifying Mixtral of Experts [Virtual/Zoom]Speaker: Albert Jiang, Mistral AI / University of Cambridge
Developing precision language models from self-attentive feed-forward units, and applying them in edge computing scenarios as untrained language models prompted to predict symbolic switches (U-LaMPS)Speaker: Jake Williams, Drexel University

2.各章节负责人

未完全确定

3.各章节预估完成日期

整体内容在6月底之前完成,各章节同步推进。

4.可预见的困难

1)整体内容难度较高,有些专题讲解的深度较深,需要相关方面的良好基础才能比较好的总结专题讲座内容。「内容完成后逐步迭代」
2)内容进度滞后「做好节点控制」

项目负责人

GitHub: https://github.com/mlw67
WeChat: mltheory

备注:发起立项申请后DOPMC成员将会在7天内给出审核意见,若7天内无反对意见则默认立项通过~

  • 我已知悉上述备注
@Sm1les
Copy link
Contributor

Sm1les commented May 4, 2024

7天内无反对意见则默认立项通过

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
Development

No branches or pull requests

2 participants