Chapters
- 01 Is Fine Tuning an LLM Worth It for Production in 2026?Is fine-tuning an LLM worth it in production? Hosting costs against hosted APIs, RAG versus fine-tuning, the steps, risks, tools, and a matrix to decide.
- 02 Fine Tuning vs Post Training for LLMs: A 2026 GuideThe difference between supervised fine-tuning and alignment post-training for LLMs, when each applies, how costs compare, and a framework for choosing.
- 03 Fine Tuning LLM on Mac Studio M4 Performance: A 2026 Field GuideFine-tuning an LLM on a Mac Studio M4: memory limits, software stack, LoRA steps for classification, base model choice, and when to pick RAG instead.
- 04 Self-Improving Open-Source Models: The Agentic Coding PlaybookHow self-improving coding agents on open models work: the feedback loop, failure modes, how to build one, metrics, and when a static model is better.
- 05 Fine Tuning LLMs for Domain Specific Tasks: 2026 GuideHow to fine-tune LLMs for domain tasks: how much data you need, Unsloth, Axolotl and managed tools, LoRA steps, RAG versus fine-tuning, and when to skip.
- 06 fine tuning qwen3.5 bug fixes and workaroundsCommon Qwen3.5 fine-tuning bugs and fixes: tokenizer mismatch, misleading loss curves, memory, DeepSpeed with LoRA, chat templates and the pad token.
- 07 best datasets for llm fine-tuning: A Practitioner’s Guide (2026)Which datasets suit LLM fine-tuning: instruction, preference, domain and synthetic data, plus how to evaluate and build a dataset before training.
- 08 Can You Fine Tune an LLM on a Mac Studio? (2026 Guide)Can a Mac Studio fine-tune an LLM? Which models fit in memory, how PyTorch MPS, MLX and Unsloth compare, where it breaks, and when to use the cloud.
- 09 The Real Guide to Best Open Source Models for Fine Tuning in 2026Choosing an open source model to fine-tune: Llama, Mistral, Qwen and DeepSeek compared, when to skip fine-tuning, the workflow and hosting costs.
- 10 LLM Fine-Tuning Dataset Size: Best Practices 2026How much data LLM fine-tuning needs: small datasets, quality over quantity, a minimum viable set, synthetic data, when to stop and hardware demands.
- 11 How much does it cost to fine tune an LLM in 2026?What LLM fine-tuning costs: compute by model size and method, data prep, iteration, inference after training, when it does not pay off, estimating.
- 12 How Long Does It Take to Fine-Tune an LLM?How long LLM fine-tuning takes across data prep, training, and evaluation, which factors stretch the timeline, and when fine-tuning is the wrong choice.
- 13 Can You Fine-Tune GPT-4 on Your Own Data? A 2026 GuideCan GPT-4 be fine-tuned on your own data? What fine-tuning changes, how it compares with RAG and prompt engineering, and a decision tree for choosing.
- 14 Fine Tuned LLM vs Prompt Engineering: Which Is Better?Fine-tuning or prompt engineering? A decision framework based on volume, task stability, failure tolerance and data, with costs, evaluation and hybrids.
- 15 LLM Fine Tuning vs Prompt Engineering: A Practitioner's GuideWhen to prompt and when to fine-tune an LLM: a decision framework, LoRA example, hardware needs, tools, RAG as a third option and common mistakes to avoid.
- 16 Fine Tune Open Source LLM vs GPT API: 2026 GuideFine-tune an open source LLM or call the GPT API? Compares cost, hidden data and labeling work, tooling and RAG, with a decision checklist for each.
- 17 How Much Does LLM Training Cost?What LLM training costs, from fine-tuning to frontier models. Covers compute, data, engineering and evaluation costs, and why training got cheaper.