Tag: #ai
Writing on GPUs, LLMs, MLOps, Kubernetes — and mindset · 216 posts
AI for Everyone, Part 6 — Drawing Digits From Words With a 1.11M Diffusion Model
We built a conditional diffusion model with 1.11 million parameters that draws a 0 when you type "zero". The forward process that adds noise is a single formula; the reverse process that restores the image is that same l
2026-08-24 · 8 min read #ai#diffusion#ddpm#generative#pytorchAI for Everyone, Part 5 — Colourising Photos With a 0.47M U-Net, and Why the Colours Came Out Washed Out
The smallest model in this series — a 472K-parameter U-Net — restored colour to greyscale CIFAR-10 images. Shapes survived intact, but the colours came out noticeably washed out. That is not a capacity problem; it is a c
2026-08-23 · 7 min read #ai#computer-vision#unet#colorization#pytorchAI for Everyone, Part 4 — Captioning Images With 1.37M Parameters, and Why Part 3 Bug Was Absent Here
We joined a CNN encoder to a transformer decoder and captioned Fashion-MNIST images. With 1.37 million parameters and ten minutes of training, the label hit rate reached 91%. This part looks at what cross-attention does
2026-08-22 · 7 min read #ai#captioning#multimodal#transformer#pytorchAI for Everyone, Part 3 — Loss of 0.0017, Accuracy of 7.5%: The Culprit Was One Padding Slot
We built a VQA model — one that answers questions about an image — with 1.48 million parameters. Training loss fell to 0.0017 while accuracy sat at 7.5%, worse than guessing. The cause was not the model but a single line
2026-08-21 · 8 min read #ai#vqa#multimodal#debugging#pytorchAI for Everyone, Part 1 — Training a 16M-Parameter Language Model From Scratch in 15 Minutes
Training a language model from scratch on a single GPU. With the TinyStories dataset and a 16-million-parameter decoder-only transformer, we produced readable English fairy tales in 15 minutes. We look at why the causal
2026-08-19 · 9 min read #ai#llm#transformer#pytorch#hands-onEnglish for AI Engineering Meetings: Saying You Do Not Know Without Losing Credibility
The English an AI team meeting needs most is not the language of certainty but the language that makes uncertainty credible. Reporting eval results together with their limits, flagging a regression whose cause is still u
2026-08-16 · 31 min read #english#business-english#meeting#ai#llmWhat Stays Expensive — Four Things That Gain Value When Generation Gets Cheap
Answer the question "what should engineers prepare for" with a list of tools and the answer is stale in two years. This piece proposes a different axis: the skills that hold value are the ones that stay expensive to veri
2026-08-15 · 6 min read #career#skills#ai#craft#engineering-cultureWhat We Actually Know and What We Do Not — Handling Career Predictions
Writing about the future of software work comes in two flavors: reassurance that says do not worry, and warning that says prepare now or fall behind. Both state predictions nobody can support as though they were facts. T
2026-08-15 · 6 min read #career#ai#uncertainty#forecasting#job-securityWhy Juniors Are Especially Anxious — The Entry Ladder Has Changed
A junior engineer's anxiety comes from position, not personality. It is the layer most often named as replaceable, the layer with the least evidence behind it, and the layer with the most people to compare against. This
2026-08-15 · 6 min read #career#junior-developer#ai#learning#entry-levelWorking With AI Tools as a Skill — Where the Delegation Line Sits and How to Verify
Saying someone is good with AI tools carries no information until you know what they delegated and how they judged the result. This piece covers the criterion that draws the delegation line by verification cost rather th
2026-08-15 · 7 min read #career#skills#ai#verification#engineering-practiceKorean Dev Blog Curation 3 — AI and ML in Practice, 14 Posts I Opened and Checked
Fourteen Korean-language posts on applying AI and ML in production, selected for being concrete and reproducible. They cover an end-to-end LangChain RAG pipeline, how semantic search actually works through embeddings and
2026-08-12 · 12 min read #curation#큐레이션#ai#llm#ragWhat Is a Forward Deployed Engineer (FDE) — the Engineer Stationed at the Customer Site
Palantir invented the title, and now OpenAI and Anthropic are racing to hire for it: Forward Deployed Engineer, or FDE. This post pins down what the role actually is — an engineer whose coordinate system is the customer
2026-08-12 · 6 min read #career#fde#forward-deployed-engineer#ai#job-searchChoosing Speech Models: Practical Criteria for STT and TTS
Unlike text models, speech models are chosen after the language coverage, audio length constraints, real-time requirement, and diarization need are already fixed. This post lays out the parameters, licenses, language cov
2026-08-12 · 7 min read #ai#huggingface#open-source-llm#speech-to-text#text-to-speechChoosing Open Text Generation Models by Size Class
When picking an open text generation model, size is not a quality tier — it is a deployment constraint. This post splits the field into on-device small, single-GPU mid, and server-class large, listing the parameters, con
2026-08-12 · 7 min read #ai#llm#huggingface#open-source-llm#text-generationOpen Models That Support Korean, and the Cost of Tokenization
A line saying a model supports Korean means it processes the language, not that it is good at it. This post lays out the card values read on 2026-08-12 for Korean-focused, bilingual, and multilingual open models, explain
2026-08-12 · 6 min read #ai#llm#huggingface#open-source-llm#korean-nlpLicenses and Deployment: Reading the Types and Verifying Quantized Builds
The short identifier in an open model license field is a table of contents; the actual terms live in the full text. This post organizes the license types actually encountered on 2026-08-12, what community licenses and no
2026-08-12 · 7 min read #ai#llm#huggingface#open-source-llm#licenseChoosing Code Models: Completion vs Chat, FIM, and Licenses
Code models come in two different products: the completion model that fills the cursor position inside an editor, and the conversational model that answers questions. This post organizes the parameters, context, fill-in-
2026-08-12 · 7 min read #ai#llm#huggingface#open-source-llm#code-llmVision and Multimodal: Image Understanding, OCR, and VLMs
Vision models are not one category. Conversational VLMs, dedicated OCR, document structuring, and image retrieval embeddings are different objects and do not substitute for one another. This post organizes the card value
2026-08-12 · 7 min read #ai#huggingface#open-source-llm#vision-language-model#ocrHow to Read a Hugging Face Model Card: What to Trust and What to Doubt
Six fields on a Hugging Face model card actually carry decision weight: license, parameter count, context length, intended use, limitations, and training data. This post explains what each field means and what it hides,
2026-08-12 · 7 min read #ai#llm#huggingface#open-source-llm#model-cardEmbeddings and Rerankers: What Actually Matters in RAG
What decides your embedding model in a RAG build is not a leaderboard rank but dimensionality, maximum input length, prefix conventions, multilingual coverage, and the division of labor with a reranker. This post lays ou
2026-08-12 · 7 min read #ai#llm#huggingface#open-source-llm#embedding