添加 CS294 课程的详细文档和资源

2026-06-22 17:37:17 +08:00 · 2025-04-05 10:35:54 +08:00 · 2025-04-05 10:35:54 +08:00 · 281af1183e
commit 281af1183e
parent cd5d62d59d
4 changed files with 164 additions and 0 deletions
--- a/.history/docs/大语言模型与智能体/CS294.en_20250405103123.md
+++ b/.history/docs/大语言模型与智能体/CS294.en_20250405103123.md
--- a/.history/docs/大语言模型与智能体/CS294.en_20250405103315.md
+++ b/.history/docs/大语言模型与智能体/CS294.en_20250405103315.md
@ -0,0 +1,82 @@
+# CS294/194-196 Large Language Model (LLM) Agents Course
+
+## Course Overview
+
+- **University**: UC Berkeley
+- **Prerequisites**: None
+- **Programming Language**: Python
+- **Course Difficulty**: 🌟🌟🌟
+- **Estimated Study Hours**: 100 hours
+- The course covers LLM reasoning, tool usage, multi-agent collaboration, and application areas such as code generation and robotics.
+- It provides abundant resources, including the course website [CS294/194-196 Course Homepage](http://rdi.berkeley.edu/llm-agents/f24), MOOC site [Large Language Model Agents MOOC](https://llmagents-learning.org/f24), YouTube lecture videos, and GitHub notes.
+- Assignments include reading summaries, experiments, and projects such as participating in the [LLM Agents Hackathon](http://rdi.berkeley.edu/llm-agents-hackathon/).
+- Extended learning includes the spring course [CS294/194-280 Advanced Large Language Model Agents](http://rdi.berkeley.edu/adv-llm-agents/sp25) and related GitHub resources [Awesome LLM Agents](https://github.com/kaushikb11/awesome-llm-agents).
+
+## Course Structure and Content
+
+The course explores the core content of LLM agents, including:
+
+- **LLM Basics and Reasoning**: Covers techniques like Chain-of-Thought and Self-Consistency to enhance model reasoning through intermediate steps.
+- **Agent Infrastructure**: Includes Retrieval-Augmented Generation (RAG), Tool Use, and multi-agent collaboration frameworks (e.g., AutoGen).
+
+While the course has no formal prerequisites, students are encouraged to have a background in machine learning and deep learning, such as courses CS182, CS188, or CS189.
+
+## Course Resources
+
+To support self-study, the course offers abundant resources:
+
+- **Course Website**: [CS294/194-196 Course Homepage](http://rdi.berkeley.edu/llm-agents/f24), containing the syllabus, registration info, and contact details.
+- **MOOC Website**: [Large Language Model Agents MOOC](https://llmagents-learning.org/f24), providing additional labs, certifications, and interactive communities for those unable to attend the formal course.
+- **Lecture Videos**: All 12 lectures are recorded on YouTube. The schedule and links are as follows:
+
+| Date       | Topic                                           | Guest Lecturer                        | Video Link                                      |
+|------------|-------------------------------------------------|---------------------------------------|-------------------------------------------------|
+| Sept 9     | LLM Reasoning                                   | Denny Zhou, Google DeepMind           | [Video](https://www.youtube.com/live/QL-FS_Zcmyo) |
+| Sept 16    | LLM Agents: History and Overview                | Shunyu Yao, OpenAI                    | [Video](https://www.youtube.com/watch?v=RM6ZArd2nVc) |
+| Sept 23    | Agent AI Framework & AutoGen, Building Multimodal Knowledge Assistants | Chi Wang, AutoGen-AI; Jerry Liu, LlamaIndex | [Video](https://www.youtube.com/live/OOdtmCMSOo4) |
+| Sept 30    | Generative AI Trends in Business and Key Components for Building Successful Agents/Applications | Burak Gokturk, Google                 | [Video](https://www.youtube.com/live/Sy1psHS3w3I) |
+| Oct 7      | Composite AI Systems & DSPy Framework           | Omar Khattab, Databricks              | [Video](https://www.youtube.com/live/JEMYuzrKLUw) |
+| Oct 14     | Software Development Agents                     | Graham Neubig, Carnegie Mellon University | [Video](https://www.youtube.com/live/f9L9Fkq-8K4) |
+| Oct 21     | Enterprise Workflow AI Agents                  | Nicolas Chapados, ServiceNow          | [Video](https://www.youtube.com/live/-yf-e-9FvOc) |
+| Oct 28     | A Unified Framework for Neuro-Symbolic Decision-Making | Yuandong Tian, Meta AI (FAIR)         | [Video](https://www.youtube.com/live/wm9-7VBpdEo) |
+| Nov 4      | Project GR00T: The Universal Robotics Blueprint  | Jim Fan, NVIDIA                       | [Video](https://www.youtube.com/live/Qhxr0uVT2zs) |
+| Nov 18     | Open-Source and Science in the Era of Foundation Models | Percy Liang, Stanford University      | [Video](https://www.youtube.com/live/f3KKx9LWntQ) |
+| Nov 25     | Measuring Agent Abilities and Anthropic's RSP    | Ben Mann, Anthropic                   | [Video](https://www.youtube.com/live/6y2AnWol7oo) |
+| Dec 2      | Building Safe and Trustworthy AI Agents, and AI Policies Based on Science and Evidence | Dawn Song, UC Berkeley                | [Video](https://www.youtube.com/live/QAgR4uQ15rc) |
+
+Each lecture has corresponding reading materials; see the course website [CS294/194-196 Course Homepage](http://rdi.berkeley.edu/llm-agents/f24#syllabus).
+
+- **Course Notes**: [GitHub](https://github.com/rajdeepmondaldotcom/CS294_LLM_Agents_Notes_Fall2024)
+
+## Relevant Papers and Resources
+
+The course focuses on the following papers and frameworks, categorized by technical direction:
+
+### Reasoning and Planning
+
+- **ReAct**: A framework combining reasoning and action to enhance task-solving capabilities. Paper: [ReAct: Synergizing Reasoning and Acting in Language Models](https://arxiv.org/abs/2210.03629).
+- **Chain-of-Thought**: Using intermediate steps to stimulate reasoning in models. Paper: [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://arxiv.org/abs/2201.11903).
+- **Chain-of-Thought Reasoning Without Prompting**: Exploring reasoning through chain-of-thought without prompting. Paper: [Chain-of-Thought Reasoning Without Prompting](https://arxiv.org/abs/2402.10200).
+
+### Agent Frameworks
+
+- **AutoGen**: A framework for multi-agent conversation development. Paper: [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://arxiv.org/abs/2308.08155).
+- **DSPy**: A programming framework for composite AI systems. Paper: [DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines](https://arxiv.org/abs/2310.03714).
+
+### Application Scenarios
+
+- **Code Generation**:
+  - **SWE-agent**: An agent interface for automating software engineering. Paper: [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://arxiv.org/abs/2402.01030).
+- **Robotics**:
+  - **Voyager**: An embodied agent based on LLMs. Paper: [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://arxiv.org/abs/2305.16291).
+
+## Extended Learning Suggestions
+
+For further study, we recommend:
+
+- **Practical Projects**: Participate in the [LLM Agents Hackathon](http://rdi.berkeley.edu/llm-agents-hackathon/), aiming to build agents supporting tool usage (e.g., web automation). The Hackathon provides tracks such as applications, benchmarks, and foundational research.
+  
+- **Advanced Courses**: Explore the spring course [CS294/194-280 Advanced Large Language Model Agents](http://rdi.berkeley.edu/adv-llm-agents/sp25), focusing on advanced topics like LLM reasoning, mathematical proofs, and code generation.
+- **Online Resources**:
+  - [Large Language Model Agents MOOC](https://llmagents-learning.org/f24): Offers course materials, labs, and certifications, ideal for self-learners.
+  - [Awesome LLM Agents](https://github.com/kaushikb11/awesome-llm-agents): A collection of frameworks, papers, and projects on LLM agents, ideal for deep research.
--- a/docs/大语言模型与智能体/CS294.en.md
+++ b/docs/大语言模型与智能体/CS294.en.md
@ -0,0 +1,82 @@
+# CS294/194-196 Large Language Model (LLM) Agents Course
+
+## Course Overview
+
+- **University**: UC Berkeley
+- **Prerequisites**: None
+- **Programming Language**: Python
+- **Course Difficulty**: 🌟🌟🌟
+- **Estimated Study Hours**: 100 hours
+- The course covers LLM reasoning, tool usage, multi-agent collaboration, and application areas such as code generation and robotics.
+- It provides abundant resources, including the course website [CS294/194-196 Course Homepage](http://rdi.berkeley.edu/llm-agents/f24), MOOC site [Large Language Model Agents MOOC](https://llmagents-learning.org/f24), YouTube lecture videos, and GitHub notes.
+- Assignments include reading summaries, experiments, and projects such as participating in the [LLM Agents Hackathon](http://rdi.berkeley.edu/llm-agents-hackathon/).
+- Extended learning includes the spring course [CS294/194-280 Advanced Large Language Model Agents](http://rdi.berkeley.edu/adv-llm-agents/sp25) and related GitHub resources [Awesome LLM Agents](https://github.com/kaushikb11/awesome-llm-agents).
+
+## Course Structure and Content
+
+The course explores the core content of LLM agents, including:
+
+- **LLM Basics and Reasoning**: Covers techniques like Chain-of-Thought and Self-Consistency to enhance model reasoning through intermediate steps.
+- **Agent Infrastructure**: Includes Retrieval-Augmented Generation (RAG), Tool Use, and multi-agent collaboration frameworks (e.g., AutoGen).
+
+While the course has no formal prerequisites, students are encouraged to have a background in machine learning and deep learning, such as courses CS182, CS188, or CS189.
+
+## Course Resources
+
+To support self-study, the course offers abundant resources:
+
+- **Course Website**: [CS294/194-196 Course Homepage](http://rdi.berkeley.edu/llm-agents/f24), containing the syllabus, registration info, and contact details.
+- **MOOC Website**: [Large Language Model Agents MOOC](https://llmagents-learning.org/f24), providing additional labs, certifications, and interactive communities for those unable to attend the formal course.
+- **Lecture Videos**: All 12 lectures are recorded on YouTube. The schedule and links are as follows:
+
+| Date       | Topic                                           | Guest Lecturer                        | Video Link                                      |
+|------------|-------------------------------------------------|---------------------------------------|-------------------------------------------------|
+| Sept 9     | LLM Reasoning                                   | Denny Zhou, Google DeepMind           | [Video](https://www.youtube.com/live/QL-FS_Zcmyo) |
+| Sept 16    | LLM Agents: History and Overview                | Shunyu Yao, OpenAI                    | [Video](https://www.youtube.com/watch?v=RM6ZArd2nVc) |
+| Sept 23    | Agent AI Framework & AutoGen, Building Multimodal Knowledge Assistants | Chi Wang, AutoGen-AI; Jerry Liu, LlamaIndex | [Video](https://www.youtube.com/live/OOdtmCMSOo4) |
+| Sept 30    | Generative AI Trends in Business and Key Components for Building Successful Agents/Applications | Burak Gokturk, Google                 | [Video](https://www.youtube.com/live/Sy1psHS3w3I) |
+| Oct 7      | Composite AI Systems & DSPy Framework           | Omar Khattab, Databricks              | [Video](https://www.youtube.com/live/JEMYuzrKLUw) |
+| Oct 14     | Software Development Agents                     | Graham Neubig, Carnegie Mellon University | [Video](https://www.youtube.com/live/f9L9Fkq-8K4) |
+| Oct 21     | Enterprise Workflow AI Agents                  | Nicolas Chapados, ServiceNow          | [Video](https://www.youtube.com/live/-yf-e-9FvOc) |
+| Oct 28     | A Unified Framework for Neuro-Symbolic Decision-Making | Yuandong Tian, Meta AI (FAIR)         | [Video](https://www.youtube.com/live/wm9-7VBpdEo) |
+| Nov 4      | Project GR00T: The Universal Robotics Blueprint  | Jim Fan, NVIDIA                       | [Video](https://www.youtube.com/live/Qhxr0uVT2zs) |
+| Nov 18     | Open-Source and Science in the Era of Foundation Models | Percy Liang, Stanford University      | [Video](https://www.youtube.com/live/f3KKx9LWntQ) |
+| Nov 25     | Measuring Agent Abilities and Anthropic's RSP    | Ben Mann, Anthropic                   | [Video](https://www.youtube.com/live/6y2AnWol7oo) |
+| Dec 2      | Building Safe and Trustworthy AI Agents, and AI Policies Based on Science and Evidence | Dawn Song, UC Berkeley                | [Video](https://www.youtube.com/live/QAgR4uQ15rc) |
+
+Each lecture has corresponding reading materials; see the course website [CS294/194-196 Course Homepage](http://rdi.berkeley.edu/llm-agents/f24#syllabus).
+
+- **Course Notes**: [GitHub](https://github.com/rajdeepmondaldotcom/CS294_LLM_Agents_Notes_Fall2024)
+
+## Relevant Papers and Resources
+
+The course focuses on the following papers and frameworks, categorized by technical direction:
+
+### Reasoning and Planning
+
+- **ReAct**: A framework combining reasoning and action to enhance task-solving capabilities. Paper: [ReAct: Synergizing Reasoning and Acting in Language Models](https://arxiv.org/abs/2210.03629).
+- **Chain-of-Thought**: Using intermediate steps to stimulate reasoning in models. Paper: [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://arxiv.org/abs/2201.11903).
+- **Chain-of-Thought Reasoning Without Prompting**: Exploring reasoning through chain-of-thought without prompting. Paper: [Chain-of-Thought Reasoning Without Prompting](https://arxiv.org/abs/2402.10200).
+
+### Agent Frameworks
+
+- **AutoGen**: A framework for multi-agent conversation development. Paper: [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://arxiv.org/abs/2308.08155).
+- **DSPy**: A programming framework for composite AI systems. Paper: [DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines](https://arxiv.org/abs/2310.03714).
+
+### Application Scenarios
+
+- **Code Generation**:
+  - **SWE-agent**: An agent interface for automating software engineering. Paper: [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://arxiv.org/abs/2402.01030).
+- **Robotics**:
+  - **Voyager**: An embodied agent based on LLMs. Paper: [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://arxiv.org/abs/2305.16291).
+
+## Extended Learning Suggestions
+
+For further study, we recommend:
+
+- **Practical Projects**: Participate in the [LLM Agents Hackathon](http://rdi.berkeley.edu/llm-agents-hackathon/), aiming to build agents supporting tool usage (e.g., web automation). The Hackathon provides tracks such as applications, benchmarks, and foundational research.
+  
+- **Advanced Courses**: Explore the spring course [CS294/194-280 Advanced Large Language Model Agents](http://rdi.berkeley.edu/adv-llm-agents/sp25), focusing on advanced topics like LLM reasoning, mathematical proofs, and code generation.
+- **Online Resources**:
+  - [Large Language Model Agents MOOC](https://llmagents-learning.org/f24): Offers course materials, labs, and certifications, ideal for self-learners.
+  - [Awesome LLM Agents](https://github.com/kaushikb11/awesome-llm-agents): A collection of frameworks, papers, and projects on LLM agents, ideal for deep research.
--- a/docs/大语言模型与智能体/llm_zero_to_hero.en.md
+++ b/docs/大语言模型与智能体/llm_zero_to_hero.en.md