搜索
会员列表
日历
帮助
清水湾论坛 - 香港科技大学内地学生学者联谊会 MSSS
›
休闲娱乐
›
吹水湾
Mastering Generative AI From Neural Networks to Multi-Agents
Mastering Generative AI From Neural Networks to Multi-Agents
| MSSS Notice Board |
|
【严正声明】打击违法犯罪，维护论坛安全 |
|
Mastering Generative AI From Neural Networks to Multi-Agents
|
|
| 作者 | 内容 |
CoursesToday
普通青年  帖子: 166 166 加入时间: Oct 2025 声望: 0 |
|
||
| 10-03-2025, 07:58 PM |
|
![[Image: 08e0e9b04476a0a3ba0eca078ab44c34.avif]](https://i126.fastpic.org/big/2025/1003/34/08e0e9b04476a0a3ba0eca078ab44c34.avif)
eep dive into Machine Learning, Neural Networks, and Deep Learning fundamentalsUnderstand embeddings, transformers, and diffusion models that power modern AILearn how foundation models like GPT, Claude, and Stable Diffusion actually workNatural Language Processing Mastery:Build and fine-tune Large Language Models (LLMs) for conversation and text generationMaster tokenization, text classification, topic modeling, and named entity recognitionUnderstand evaluation metrics and benchmarks used by industry leadersImplement supervised fine-tuning for specialized AI applicationsImage Generation & Computer Vision:Create stunning images using text-to-image and image-to-image modelsMaster image editing, inpainting, and style transfer techniquesFine-tune image generation models for custom use casesAdvanced Model Customization:Master prompt engineering and in-context learning strategiesBuild Retrieval Augmented Generation (RAG) systems that ground AI in your dataImplement cutting-edge GraphRAG and StructRAG architecturesApply Parameter-Efficient Fine-Tuning (PEFT) and LoRA techniquesTrain models using Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO)Optimize models through knowledge distillationAgentic AI & Advanced Orchestration:Understand the fundamentals of AI agents and agentic reasoningMaster the ReAct (Reasoning and Acting) framework for tool-using agentsDesign and implement multi-agent systems with role specializationBuild agent topologies: sequential, hierarchical, and collaborative patternsImplement automatic handoffs and agent coordination strategiesCreate agents that can plan, reason, and execute complex multi-step tasksHands-On Learning ExperienceThis isn't just theory-you'll build real AI applications through 8 comprehensive labs that progressively build your skills:Lab 1: Neural Network Fundamentals & Transfer Learning Build an image classifier from scratch, understand training and inference pipelines, and leverage transfer learning with ResNet for state-of-the-art performance.Lab 2: AWS & Generative Image Creation Set up your AWS environment, work with Amazon Bedrock, and create and edit stunning images using Amazon Nova models-your gateway to cloud-based AI.Lab 3: Embeddings & Vector Search Master embedding models with HuggingFace, build a production-ready RAG system, and implement efficient vector databases with IVF and HNSW indexing strategies.Lab 4: Advanced LLM Techniques Work with Amazon Bedrock LLMs for real-world tasks: prompt engineering, text classification, document summarization, and creative content generation.Lab 5: Conversational AI Build an intelligent chatbot using Amazon Bedrock and Gradio with memory management and multi-turn conversation capabilities.Lab 6: Custom AI Agents Implement your own ReAct (Reasoning and Acting) agent from scratch with Amazon Bedrock, understanding how agents think and use tools.Lab 7: Full-Stack Agentic Application Create a production-ready agentic chatbot using the Strands SDK, FastAPI backend, and Amazon Bedrock-ready for real-world deployment.Lab 8: Multi-Agent Systems Build sophisticated multi-agent systems with the Strands SDK featuring automatic handoffs, agent collaboration, and coordinated problem-solving.![[Image: 423b519448d4e936894130c701f35288.jpg]](https://i124.fastpic.org/big/2024/1128/88/423b519448d4e936894130c701f35288.jpg)
正在浏览该主题的用户：