Kimi K2 Post-Training: Tool-Use, Synthetic Data, Reinforcement Learning
Content1 Key Takeaways2 Pre-training vs Post-training3 Supervised Fine-Tuning4 Data Synthesis for Tool Use5 Reinforcement Learning6 Verifiable Rewards7 Non-verifiable Rewards8 Rollouts9 Note10 Conclusion Vijona4 Feb at 11:10 Kimi K2 Post-Training: Tool…
PowerShell and Linux: Run Bash Commands with pwsh and WSL
Content1 Key Takeaways2 Understanding PowerShell vs. Bash3 Method 1 – Install PowerShell on Linux4 Method 2 – Use Windows Subsystem for Linux (WSL)5 Method 3 – Cross-Platform Scripts & Aliases6…
Qwen3-Coder: 405B MoE Agentic Coding Model + Qwen Code CLI Guide
Content1 Key Takeaways2 Model Overview3 Implementation4 Step 1: Set up a GPU Virtual Machine5 Step 2: Web Console6 Step 3: Install Dependendencies7 Step 4: Run the Model8 Qwen Code: Open-Source…
Episodic Memory in AI Agents: Long-Term Context & Learning
Content1 Key Takeaways2 What Is Episodic Memory in AI?3 Types of Memory in AI Agents4 Why Episodic Memory Matters in AI Agents5 How Episodic Memory Works in AI Agents6 Pseudocode:…
Transformer Architecture Explained: Attention, Training, and GPU Setup
Content1 Key Points2 Prerequisites3 What Are Transformers?4 Transformer Architecture: Detailed, Step-by-Step5 Attention: The Core Math (Concise)6 Scaled Dot-Product Attention7 Multi-Head Attention (MHA)8 Masked Attention9 Residuals, LayerNorm, and Stability10 Output Projection…
Data Augmentation in Machine Learning: Image, Text & Audio Techniques
Content1 Key Takeaways2 Why Use Data Augmentation?3 Image Augmentation Techniques4 Setting Up an Augmentation Pipeline5 Adding Gaussian Noise to Images6 Load and Prepare the Data7 Apply Gaussian Noise Augmentation8 Display…
OpenAI gpt-oss Explained: Architecture, MXFP4 Quantization & 120B/20B Models
Content1 Model Variants and Hardware Requirements2 Key Takeaways3 Model Architecture4 Quantization5 Tokenizer6 Post-training Focus7 OpenAI Harmony Chat Format8 Additional Resources9 Final Thoughts Vijona4 Feb at 10:15 OpenAI gpt-oss: Architecture, Quantization,…
Embedding-Free RAG: Alternatives to Vector Databases for Retrieval-Augmented Generation
Content1 Key Takeaways2 Traditional RAG and Vector Databases3 Limitations of Embeddings & Vector Search4 What Is RAG Without Embeddings?5 Lexical or Keyword-Based Retrieval6 LLM-based Iterative Search (Reasoning as Retrieval)7 Structured…
RAG vs MCP: When to Use Retrieval or Tool-Based Actions with LLMs
Content1 Key Takeaways2 Prerequisites3 Understanding RAG and MCP4 When RAG Is the Best Fit5 When MCP Is the Best Fit6 Potential Failure Modes to Watch For7 Choosing Between RAG and…