Tau is a nice open-source project from Hugging Face for learning how coding agents are built. It separates the stack into tau_ai, tau_agent, and tau_coding: provider stream, reusable agent harness, and terminal coding app. Repo: github.com/huggingface/... #ai #aiagents #openso…
Who's Who of AI
Rami Krispin
Senior Manager Data Science and Engineering at | Docker Captain 🐳| Time-series analysis & forecasting
My newsletters:
Weekly updates: https://ramikrispin.substack.com
AIOps: https://theaiops.substack.com
Forecasting: https://theforecaster.substack.com
What they're sharing
GitHub - huggingface/tau: a minimalist agent that teaches you to create coding agents github.com
google/diffusiongemma-26B-A4B-it · Hugging Face huggingface.co
GLM-5.2 - How to Run Locally | Unsloth Documentation unsloth.ai
Homepage | End to End Machine Learning end-to-end-machine-learning.teachable.com
View on Bluesky →
NumPyro Forecast, The Orange Book of Machine Learning | Issue 95 ramikrispin.substack.com
Context in SQL AI Agents: The Three Layers Behind Reliable Answers ramikrispin.substack.com
DeepSpeed, Hermes LLM Wiki, GeoAI Essentials | Issue 92 ramikrispin.substack.com
Introduction to SQL AI Agents: The Four Components Behind Natural Language SQL ramikrispin.substack.com
The AI Hedge Fund Project, Data Science First - New Book | Issue 90 ramikrispin.substack.com
Articles & links
DiffusionGemma — Google's new open text-diffusion model. 26B MoE that drafts 256 tokens in parallel instead of one at a time. Up to 4x faster on local GPUs. Bi-directional attention, Apache 2.0. huggingface.co/google/diffu... #gemma #ai
Here is how you can run GLM 5.2 locally, if you simply have a machine with 256GB to 810GB RAM 😅 unsloth.ai/docs/models/... #ai #llm
AI Weekly's analysis
→
Read full analysis →
Rami Krispin reposted
I shut down by online courses at End-to-End Machine Learning 18 months ago. Wrote a blog post and everything. But the platform I was using hasn't bothered to take them down. So they're just sitting there, all for free. Prego. end-to-end-machine-learning.teachable.com
My weekly newsletter is out! This week's agenda: 🔹 Open Source of the Week - NumPyro Forecast 🔹 New learning resources 🔹 Book of the week - The Orange Book of Machine Learning - Green edition by Carl McBride Ellis ramikrispin.substack.com/p/numpyro-fo... #ai #datascience #mach…
Humans can hallucinate. LLMs don't. LLMs are probabilistic models. When context is missing, they fill the gaps using likelihood. Sometimes they're right. Often they're wrong. That's why context is critical. Part 2 of my Foundations of SQL AI Agents series: ramikrispin.substack…
My newsletter is out! This week's agenda: 🔹 Open Source of the Week - DeepSpeed 🔹New learning resources 🔹Book of the week - Geospatial Data Science Essentials by Milan Janosov open.substack.com/pub/ramikris... #ai #datascience
The term "agent" sounds fancy, but at its core, it's just a set of functions orchestrating how an LLM processes a request. I just published the first post in a new series on SQL AI agents, starting with the foundations before moving on to production. ramikrispin.substack.com/p…
My weekly newsletter is out: ramikrispin.substack.com/p/the-ai-hed... #ai #datascience #opensource
Their own posts
Recent commentary
🧵 1/4 Handy — open-source, offline speech-to-text 🎤 Press a shortcut, speak, and the text appears in any text field. No cloud, no telemetry. #ai #llm #opensource
PaperBanana An open-source agentic framework that generates publication-quality academic diagrams and statistical plots from text. 👇🏼 #ai
Any recommendations for Claude's skill for creating HTML slides? #ai #claude #html
Fable 5 is finally here! Amazing news: we can now upgrade from a model we barely use 20% of to a model we’ll use 10% of, for roughly double the cost. What a magical time to be alive 🤣 #ai
AI developers - how do you manage parallel features development of text elements (skills/prompts/agents) and avoid git merge conflicts? #ai
Their network
In Rami Krispin's orbit
Center = Rami Krispin. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.