donbr

Here’s a short list of Ollama models that track closest to gpt-oss:20b in capability/feel, plus how to prompt them so you don’t lose quality when you swap away from OpenAI-tuned prompts.

Best bets (by “feel” + instruction-following)

1. Mistral NeMo 12B Instruct — very solid instruction following, large context (128k), good tool/JSON behavior for its size. ollama pull mistral-nemo:12b (or mistral-nemo:12b-instruct where available) ([Ollama][1])

2. Llama 3.1 Instruct (8B or 70B) — stable, widely used baseline; 70B will beat 20B-class models on reasoning, 8B is a fast local workhorse. ollama pull llama3.1:8b-instruct (or llama3.1:70b-instruct if you have VRAM) ([Ollama][2])

donbr

Optimized Documentation Retrieval with MCP + Local Index

Date: September 18, 2025
Author: Don

---

Overview

Large llms-full.txt files contain thousands of URLs and can overwhelm LLM context windows.

donbr

How to Build ChatGPT: Complete Study Guide

AI Makerspace Series - Step-by-Step Implementation Guide

---

📚 Series Overview

The "How to Build ChatGPT" series from AI Makerspace provides a comprehensive roadmap for building production-ready LLM applications, following OpenAI's product evolution from simple chat interfaces to sophisticated AI systems. This study guide covers all four core parts with detailed lesson notes and presentation slides.

🎯 Learning Objectives

donbr

AIE8 Student Git Guide (Rebase-Only, No Merges to main)

This guide shows how to:

• Initialize your GitHub repo (origin)
• Keep your main branch identical to the class repo (upstream)
• Do all work on feature branches
• Rebase feature branches to pick up updates (no merges)

---

donbr

Yes — exactly. The temperature parameter in LLMs is directly tied to logits (the raw, pre-softmax output of the model).

Here’s how it works step by step:

---

🔹 From Logits to Probabilities

1. The LLM produces a logit vector: one score per token in the vocabulary.
2. To convert logits into probabilities, we apply the softmax function:

donbr

🔹 Sparse vs. Dense Vectors: What to Expect

1. Sparse Vectors

• What they are: Very high-dimensional vectors (thousands to millions of dimensions), where most entries are zeros.

• How they’re built: Each dimension corresponds to a word from the vocabulary. A document or query is represented by which words it contains and how important they are (using TF-IDF or BM25 weighting).

• What they capture:

  • Exact keyword overlap → great at finding documents with the same words.

• Transparent → you can see exactly which words drive the score.

donbr

JSON-RPC in Enterprise Integration vs Agent-to-Agent Solutions: A Comparative Research Analysis

Executive Summary

This research compares two distinct applications of JSON-RPC protocol: traditional enterprise integration patterns and Google's innovative Agent-to-Agent (A2A) solutions. Both approaches represent valid, effective solutions serving different architectural needs and use cases.

Research Verification: Last verified 2025-09-02 22:05 UTC using Brave Search, official documentation fetch, and MCP Server Time tools.

1. Traditional JSON-RPC in Enterprise Integration

donbr

Applying IBM Design Thinking to Agentic AI Solutions in 2025

Executive Summary

As Agentic AI systems become increasingly sophisticated and autonomous in 2025, the need for human-centered design approaches becomes critical. IBM's Enterprise Design Thinking framework provides a proven methodology for creating user-focused, collaborative, and iterative solutions that can be effectively applied to the design of autonomous AI agents. This research explores how the three core principles and tactical frameworks of IBM Design Thinking can address the unique challenges and opportunities in Agentic AI development.

Understanding IBM Design Thinking in 2025

IBM's Enterprise Design Thinking is built on three foundational principles that scale design thinking across diverse, distributed teams:

donbr

The AI Engineering Bootcamp, Cohort 8 Course Schedule & Curriculum

🧑‍💻 What is “AI Engineering?”

AI Engineering refers to the industry-relevant skills that data science and engineering teams need to successfully build, deploy, operate, and improve Large Language Model (LLM) applications in production environments.

In practice, this requires understanding both prototyping and production deployments.

During the prototyping phase, Prompt Engineering, Retrieval Augmented Generation (RAG), Agents, and Fine-Tuning are all necessary tools to be able to understand and leverage. Prototyping includes:

donbr

AI Companion Applications: Purpose, Use Cases, and Comparative Analysis (August 2025)

Executive Summary

In an era where digital connectivity increasingly intersects with human emotional needs, AI companion applications have emerged as innovative tools designed to simulate interpersonal interactions. These platforms leverage advanced large language models (LLMs) and multimodal features to provide users with empathetic, customizable, and often therapeutic digital companions. As of August 2025, the global market for AI companionship apps has surged to approximately $120 million in annual revenue, reflecting a 60% growth from 2024, driven by rising demand for mental health support, social simulation, and personalized entertainment amid post-pandemic isolation trends. This document outlines the core purpose of these solutions, explores their primary use cases, and presents a comparative table of leading apps, including Replika as a benchmark. Data is derived from current user reviews, industry reports, and