Последние посты
Rami Krispin's Data Science Channel
22 сент., 15:44
A new tutorial in my Docker 🐳 101 series for AI developers — this time focusing on running a container 👇🏼What is the difference between an image and a container? The runtime.A container is an instance created from an image. It combines the image with runtime configuration and a writable container layer, then starts the configured process.And this is what this tutorial focuses on: how to take the FastAPI image we built in the previous tutorials, run it as an API service inside the container, and expose it locally using the docker run command.This tutorial concludes the Docker workflow sequence, where we focused on the foundation of Docker: ☑ Defining requirements ☑ Turning those requirements into a Dockerfile ☑ Build an image ☑ Registering the image ☑ Running the image as a containerhttps://theaiops.substack.com/p/how-to-run-a-docker-container-fromSubstackHow to Run a Docker Container from an ImageLearn how to run a Docker container from an image, publish a FastAPI port to localhost, verify the API, and stop and remove the container.Rami Krispin's Data Science Channel
22 сент., 13:58
Mitra-v2 for Tabular Prediction 🚀This tutorial from AI with Surya walks through regression and classification with Mitra-v2: ✅ Fine-tuning on tabular data ✅ Ensemble predictions ✅ House price regression ✅ Machine failure classification ✅ Comparing model metrics📽️: https://www.youtube.com/watch?v=l6os-gJHWeUYouTubeMitra-v2 vs. TabFM: Amazon’s New Model is 20x Smaller and FasterAmazon's Mithra V2: The Tabular Foundation Model Trained on ZERO Real Data (Live Demo)** Is this the end of data science? Amazon just released Mithra V2 — a tabular foundation model trained entirely on 29 million *synthetic* tables. It has never seen a single…Rami Krispin's Data Science Channel
20 сент., 15:36
Issue 106 is out!This week’s agenda: 🔹 Open Source of the Week - Posit Commons 🔹 New learning resources - Jev explained, Bonsai 2 for local coding, building a coding agent harness, and weather forecasting with Hugging Face 🔹 Book of the week - Learning Generative AI Tools for Excel by Angelica Lo Ducahttps://ramikrispin.substack.com/p/posit-commons-data-agents-generativeSubstackPosit Commons Data Agents, Generative AI Tools for Excel | Issue 106Explore Posit Commons, videos on Jev, Bonsai 2, coding agent harnesses, and weather forecasting, plus a book on generative AI tools for Excel.Rami Krispin's Data Science Channel
14 сент., 02:00
Started to work on a sequence of tutorials focusing on running LLMs on Apple Silicon with MLX LM. The goal is to take you from “I know how to run LLM locally” to “I know how to optimize my resources and parameters to run LLM locally efficiently“This tutorials will be available in this repo: https://github.com/RamiKrispin/local-llms-with-mlxGitHubGitHub - RamiKrispin/local-llms-with-mlx: Running LLMs locally with MLX LMRunning LLMs locally with MLX LM. Contribute to RamiKrispin/local-llms-with-mlx development by creating an account on GitHub.Rami Krispin's Data Science Channel
13 сент., 14:21
A new tutorial in my Docker 🐳 101 series — this time focusing on working with container registries 👇🏼Container registries like Docker Hub work much like GitHub. They let you store images, manage versions, and share them with others.You typically meet the container registry in three scenarios: ➡️ Download a customized image (Python, PyTorch, Ubuntu, etc.) to run locally ➡️ During building an image while using the FROM command in your Dockerfile ➡️ After the build process, when pushing the new image version to the registryThis tutorial provides an in-depth guide for working with container registries. From labeling and tagging your image all the way to pushing it to Docker Hub.🔗: https://theaiops.substack.com/p/how-to-push-a-docker-image-to-dockerThe next tutorial will focus on launching a container using the docker run command.SubstackHow to Push a Docker Image to Docker HubLearn how to log in to Docker Hub, push a tagged FastAPI image, read the layer and digest output, and verify the published image.Rami Krispin's Data Science Channel
12 сент., 15:53
Issue 104 is out!This week's agenda: 🔹 Open Source of the Week - the TimesFM project from Google 🔹 New learning resources - Python, OpenWhispr, Git internals, MCP, AI coding workflows, and looped transformers 🔹 Book of the week - Generative AI on Microsoft Azure by Adrián González Sánchez, Jaime De Mora, and Jorge Garcia Ximenezhttps://ramikrispin.substack.com/p/timesfm-generative-ai-on-microsoftSubstackTimesFM, Generative AI on Microsoft Azure | Issue 105Explore TimesFM for zero-shot forecasting, six videos on Python, voice dictation, Git, MCP, AI coding workflows, and looped transformers, plus a guide to generative AI on Azure.Rami Krispin's Data Science Channel
12 сент., 03:27изменён
OpenAI Codex crash course from freeCodeCamp 👇🏼 https://www.youtube.com/watch?v=o3CX_Y59_74YouTubeOpenAI Codex Crash Course – Build & Deploy Apps with Autonomous AILearn the core fundamentals of OpenAI Codex, from setting up scheduled automations and custom skills to mastering Plan and Go modes. Follow along step-by-step as we build, test, and deploy a voice-controlled Flappy Bird game built entirely with Codex. Course…Rami Krispin's Data Science Channel
11 сент., 02:55
Looped Transformers ExplainedThis visual tutorial from Neural Breakdown with AVB explains how looped transformers reuse internal model layers to simulate iterative reasoning. The 22-minute video introduces the architecture and then covers how looped transformer models are trained.https://www.youtube.com/watch?v=z0Y8-IUwBKEYouTubeLet me explain Looped Transformers to youA visual breakdown of Looped Transformers, a technique where internal layers of a Large Language Model can loop to simulate iterative reason. Study research papers at: paperbreakdown.com I am building this, please go check it out and give me feedback! …Rami Krispin's Data Science Channel
9 сент., 02:45изменён
Excited to present tomorrow at the build-with-AI virtual summit about developing and deploying AI applications with Docker 🐳.When: Sep 9th, 9am PSTIt is one thing to make a RAG application run locally, and it can be challenging to take it to production. This is what Docker and containers solve 🎯.In this talk, I will focus on: ✅ The motivation for developing and deploying with containers ✅ Container strategies ✅ Moving from development to productionI will demo using a RAG application.More details and RSVP: https://maven.com/p/0f4299/how-to-deploy-ai-agents-to-production-live-rag-demoThanks to Aishwarya Srinivasan and Arvind Narayanamurthy for organizing this event! 🙏🏼MavenHow to Deploy AI Agents to Production (Live RAG Demo)Building an AI application is only half the challenge. Getting it to run reliably in production is where most teams struggle. Containers make AI agents and RAG applications portable, scalable, and easier to deploy across environments. This session will…Rami Krispin's Data Science Channel
8 сент., 15:59
Getting Started with SQL AI Agents 👇🏼After a long break, I am returning to my "Getting Started with SQL AI Agent" series. This sequence of tutorials is going to cover two topics: ✅ Foundation of SQL AI agents ✅ SQL AI agents in productionThis series is beginner-friendly 😎.We are currently in the foundation series.After laying the foundation in the previous tutorials on what an SQL AI agent is, this next tutorial focuses on the prerequisites and setting up the supporting Python environment. In the next one, we will set up a DuckDB database.Please share ♻️ if you find it useful, and subscribe to receive updates on future tutorials 👇🏼 https://ramikrispin.substack.com/p/set-up-a-python-environment-for-sqlSubstackSet up a Python environment for SQL AI agentsSet up Python for SQL AI agent tutorials with uv, venv and pip, or a Docker-based VS Code Dev Container.Rami Krispin's Data Science Channel
7 сент., 14:34
OpenWhispr: Use Your Voice For EverythingThis video from NeuralNine introduces OpenWhispr, a cross-platform voice-to-text app that supports private local transcription with Whisper or NVIDIA Parakeet, along with cloud and bring-your-own-key options.https://www.youtube.com/watch?v=PKzCW5OpMVMYouTubeOpenWhispr: Use Your Voice For Everything💻️ Need some help with a project or some consulting? Contact me here: https://www.neuralnine.com/services 🐍 The Python Bible Book: https://www.neuralnine.com/books/ 💻 The Algorithm Bible Book: https://www.neuralnine.com/books/Rami Krispin's Data Science Channel
5 сент., 12:59
Issue 104 is out!This week's agenda: 🔹 Open Source of the Week - the agent-top project by Kannan Kalidasan 🔹 New learning resources - Docker Sandboxes, local real-time speech, TrueForge, and pure Python AI agents 🔹 Book of the week - Designing AI Interfaces: Design Principles for Creative and Autonomous AI by Louise Macfadyenhttps://ramikrispin.substack.com/p/agent-top-designing-ai-interfacesSubstackAgent Top, Designing AI Interfaces | Issue 104A weekly curated update on data science and engineering topics and resources.Rami Krispin's Data Science Channel
3 сент., 15:19
I feature an open-source project every week in my newsletter, and last week's pick is a really cool project: Archify.Archify is a Node.js rendering and validation system that turns a codebase or system description into an interactive system map. It uses a typed JSON intermediate representation and deterministic compilation to keep diagrams reproducible and make validation failures easier to repair.Key features include: ✅ Architecture, workflow, sequence, data-flow, and lifecycle diagrams ✅ Typed JSON diagram source ✅ Schema, layout, HTML, SVG, route, and label validation ✅ Grounded search and relationship tracing ✅ Validated snapshot comparison ✅ Interactive themes, focus controls, and guided stories ✅ Self-contained HTML, PNG, SVG, WebM, and share-card exportsMore details are available here: https://github.com/tt-a1i/archifyLicense: MIT♻️ Please share if you find it useful📌Rami Krispin's Data Science Channel
2 сент., 14:05
BreezeTTS2 - Local Real-Time VoiceThis video from Sam Witteveen examines BreezeTTS2, a 3B open-weight text-to-speech model for local real-time inference. The 14-minute video covers voice design and direction, voice events, multilingual generation, latency, benchmarks, licensing, voice cloning, and streaming.https://www.youtube.com/watch?v=xDHD09fDUkQYouTubeBreezeTTS2 - 100% Local Real-Time VoiceIn this video, I look at BreezeTTS2, which is a 3B open weights model for doing TTS that can run 100% locally in real time. Thanks to Dell for Sponsoring the Compute #DellProPrecision #DellProMax #DellTech #NVIDIA 📖 Blog: https://breezeblue.ai/breeze…Rami Krispin's Data Science Channel
30 авг., 19:38
A new tutorial in my Docker 🐳 101 series — this time focusing on building Docker images 👇🏼I recently started a series of Docker tutorials for AI/ML developers. The first sequence walks through the core Docker workflow: requirements → Dockerfile → docker build → docker runThe 4th tutorial in the series focuses on the build stage and covers: ✅ How the Dockerfile, build context, and image fit together ✅ How to build and tag an image with docker build ✅ What happens when Docker resolves a base image through Docker Hub ✅ How to confirm that the finished image exists with docker imagesThe next tutorial in the sequence will focus on working with a container registry.🔗 https://theaiops.substack.com/p/how-to-build-a-docker-image-from?r=1x99er&utm_medium=iosSubstackHow to Build a Docker Image from a DockerfileLearn how to build and tag a Docker image from a Dockerfile, understand the build context, read the build output, and confirm the image locally.Rami Krispin's Data Science Channel
29 авг., 14:53изменён
Issue 103 is out!This week's agenda: 🔹 Open Source of the Week - The Archify project 🔹 New learning resources - Docker debugging, RecBole recommender systems, agentic harnesses, visual data structures and algorithms, reinforcement learning for unverifiable tasks, and LLM inference system design 🔹 Book of the week - Vector Databases: A Practical Introduction by Nitin Borwankarhttps://ramikrispin.substack.com/p/the-archify-project-vector-databasesSubstackThe Archify Project, Vector Databases: A Practical Introduction | Issue 103A weekly curated update on data science and engineering topics and resources.Rami Krispin's Data Science Channel
29 авг., 04:10
GLM 5.3 can run locally if you have an appropriate machine with ample RAM (e.g., a 256GB Mac or a system with sufficient RAM/VRAM).Here is a guide from Unsloth: https://unsloth.ai/docs/models/glm-5.3unsloth.aiGLM-5.3 - How to Run Locally | Unsloth DocumentationRun the new GLM-5.3 model by Z.ai.Rami Krispin's Data Science Channel
28 авг., 04:01
The MLX fast is a new cool community project with the goal of finding the best setting for running different LLM locally on Apple silicon 👇🏼https://www.yukon.org/mlxfastmlx.fastmlx.fast — Qwen 3.8 Flash Next, faster with speculative decodeMake Qwen 3.8 Flash Next decode faster on Apple Silicon with MLX. Follow the official records and join the shared Yukon challenge.Rami Krispin's Data Science Channel
27 авг., 13:47
https://techcrunch.com/2026/08/26/nvidia-closes-in-on-hugging-face-acquisition/TechCrunchNvidia closes in on Hugging Face acquisition | TechCrunchNvidia has reportedly agreed to buy Hugging Face, the popular open source AI hub, for $12.9 billion in a move that would let Nvidia both protect its chip empire and jump back into the cloud business.Rami Krispin's Data Science Channel
26 авг., 03:00
The new VScode version comes with new markdown editor functionality. One of the nice features is editing the markdown directly in preview mode 👇🏼https://youtu.be/7uyRMACA_pM?is=xQ1MI9QdWQdfv5H-YouTubeNew Markdown Editor in VS Code!This video walks through the latest VS Code markdown editing experience, from side-by-side preview and live updates to the new markdown editor that lets you modify content directly in preview while still seeing syntax update in real time. It also shows how…
