ML Fundamentals

Look, I’m a backend engineer. I don’t have time to read through 40 pages of model cards before...

Every time a PyTorch model refuses to learn, the debugging process looks the same: Stare at the...

Similarity Search for Failure Diagnosis In the previous post, I showed how every saga...

I started this session with a challenge. "Jose," I said, "print the names of all 40 students in...

Spam detection datasets are surprisingly bad once you move outside English. Most public datasets...

Introducción Los sistemas de recomendación están presentes en muchas plataformas...

Enough with the unga bunga puny algorithms.. did you know there exists an scheduling algorithm that...

Not because I wanted to — but because every tool I tried on ARM edge devices either needed the cloud,...

This blog is an introduction to Python, and it assumes that you have a good knowledge of JavaScript...

Training was a bounded investment event. Inference is an unbounded operational residency...

Not every data collection workflow requires browser automation or complex network impersonation. For...

TL;DR: I shrunk a gesture-recognition model for a Prophesee EVK4 event camera from 4.2MB down to...

Preparing for senior software engineering interviews at big tech is not just about solving LeetCode...

This is part four of a series Principal Components in TypeScript and focuses on the application of...

Data drift — the silent killer of ML models and data pipelines. Your model worked perfectly in...

Originally published on TechSaaS Cloud Originally published on TechSaaS Cloud Model...

This is part three of a series Principal Components in TypeScript and focuses on the application of...

Machine learning, a subset of artificial intelligence (AI), has been gaining significant attention in...

G Two weeks ago I shipped core/reasoning/budget.py to test whether per-call dynamic token budgets...

Release: datasette 1.0a30 The big new feature in this alpha is a new customizable "Jump to..." menu, described in detail in The extensible "Jump to" menu in Dat

I wanted to get better at technical interviews. That's really where this story starts. A...

Your GPU sits at 15% utilization and bigger batches don't help? Here's how to diagnose whether you're compute, memory, or overhead bound — and fix it.

Release: datasette-fixtures 0.1a0 One of the smaller features in Datasette 1.0a30 is this: New documented datasette.fixtures.populate_fixture_database(conn) hel

There’s a popular statement in tech: “You don’t need maths to be a good programmer.” It’s often said...

Technical interviews at top companies like Google, Amazon, and Meta are often centered around coding...

Sixty-two pages of machine learning math. Fifty-two pages of deep learning math. One hundred and...

When I first started learning machine learning, I was genuinely confident. In my head ML was...

Hi everyone, my name is Greg. I've been working in web development since 2020. A few years ago, I...

Build powerful machine learning models directly in SQL without writing any Python code.

A practical look at feature engineering, model optimization, and why simpler models...

Late deliveries are not just an inconvenience. For a global logistics operator like APL Logistics...

Retries, mutating vs read-only requests, absolute path validation, structured errors, progress notifications — the rough edges of building a real Model Context

Imagine walking outside on a quiet afternoon. You hear a sharp roar overhead, pull out your phone,...

A learned, globally‑calibrated KV‑cache eviction policy can shave memory usage and, paradoxically,...

The Problem We Were Actually Solving In our previous setup, we had a batch processing...

Natural-language SQL should not go straight from prompt to production query. The generated SQL may...

Every day, companies try to predict future outcomes: How much revenue they might generate Which...

Hermes Agent Challenge Submission: Write About Hermes Agent Challenge Page I've spent the past...

Before we talk about Linear Regression, we first need to understand the bigger idea it belongs to ...

Eight months ago the creator of ILGPU told me supporting Blazor WebAssembly would be too difficult. Today I'm shipping a six-backend ML library to NuGet with fi

Compressing KV cache via low-rank projections - the attention mechanism behind DeepSeek-V2/V3 and...

Compressing KV cache via low-rank projections - the attention mechanism behind DeepSeek-V2/V3 and...

Code Coverage .NET For .NET projects, the most common manual way is using: coverlet or...

Compressing KV cache via low-rank projections — the attention mechanism behind DeepSeek-V2/V3 and...

"Accuracy lied to you. Here's the complete toolkit—confusion matrix, precision, recall, F1, ROC/AUC,...

Have you ever written a C program, run it, and watched it print values you never assigned? At first...

The Problem We Were Actually Solving I was tasked with optimizing the event pipeline for...

There is a comfortable myth around becoming a programmer. It says you can watch a few courses, copy a...