The Kaitchup – AI on a Budget

The Kaitchup – AI on a Budget

Home
Notes
AI Notebooks
The Kaitchup's Book
Weekly Kaitchup
Tutorials
Archive
About

Tutorials

Train and Run DFlash Speculative Decoding
A simple method to make your local model much faster
 • Benjamin Marie
How to Reduce LLM Inference Cost and Improve Accuracy with Pass@k and Majority Voting
Is thinking disabled + multiple retries better and still more efficient than thinking enabled?
 • Benjamin Marie
The KV-Cache of Small MoEs: Qwen3, Qwen3.5, GLM 4.7 Flash, and Nemotron 3 Nano Compared
A memory-first look at four efficient open LLM architectures.
 • Benjamin Marie
Qwen3.5 Quantization: Similar Accuracy, More Thinking — Best Models and Recipes
INT4, NVFP4, and FP8 evaluations — Thinking off and on
 • Benjamin Marie
How to Deploy Your LLM in the Cloud
The simple recipe to choose your GPU and anticipate costs
 • Benjamin Marie
GLM-5 Memory Requirements Explained: MLA + DeepSeek Sparse Attention (DSA)
How GLM-5 fits 200K context without terabytes of KV cache, and what GPUs you need.
 • Benjamin Marie
Serving ExLlamaV3 Models with tabbyAPI: Accuracy, Speed, and Recommendations
With comparisons against AutoRound and GGUF models served with vLLM
 • Benjamin Marie
4-bit GLM-4.7 (358B) on a Single NVIDIA B300 with vLLM: AWQ vs NVFP4 vs INT4
Just give it enough tokens to think
 • Benjamin Marie
Eagle 3 Speculators: When To Use Them?
Easier and faster speculative decoding, if you are in the right settings
 • Benjamin Marie
Accelerate Models with Quantization: Recipes for NVFP4, GPTQ, AWQ, SmoothQuant, AutoRound, and FP8
Focus on 4-bit and 8-bit quantization + vLLM benchmarking with accuracy and inference throughput
 • Benjamin Marie
Unsloth's Quantization-Aware Training (QAT) vs Post-Training Quantization (PTQ) for Small Models
Can a tiny LLM stay accurate under quantization thanks to QAT?
 • Benjamin Marie
Advanced LoRA Fine-Tuning: How to Pick LoRA, QLoRA, DoRA, PiSSA, OLoRA, EVA, and LoftQ for LLMs
A practical guide to parameter-efficient LLM adaptation on 16-bit and 4-bit models
 • Benjamin Marie
© 2026 The Kaitchup · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture