Machine Learning lab

https://clearintelligence.substack.com/p/fine-tuning-llama-llm-with-lora-a @Machine_learning_lab_K

Microsoft Data Formulator https://www.youtube.com/watch?v=ySWRDqrzKrk @Machine_learning_lab_K

Methods and Strategies for Building and Refining Reasoning Models https://sebastianraschka.com/blog/2025/understanding-reasoning-llms.html @Machine_learning_lab_K

AI Agents Course https://huggingface.co/learn/agents-course/en/unit0/introduction @Machine_learning_lab_K

Data Engineering Zoomcamp: A Free 9-Week Course on Data Engineering Fundamentals https://github.com/DataTalksClub/data-engineering-zoomcamp @Machine_learning_lab_K

What Makes a Reward Model a Good Teacher? An Optimization Perspective https://arxiv.org/abs/2503.15477 @Machine_learning_lab_K

🧠 Bain MRI Dataset 5,000,000+ studies + reports MRI scans of human brains with medical reports

This dataset consists of high-quality #MRI scans of human brains, accompanied by medical reports. It is designed for tasks such as detection, classification, and segmentation of brain abnormalities. The dataset includes a variety of brain scans with different frames and studies, along with clinical information for each patient. The dataset includes: structured by series each serie more than 50 frame Total: 5,000,000+ high-quality DICOM (DCM) frames Medical reports provide the following details: Type of study MRI machine specifications Patient demographics: Age, sex, race

Oumi is a fully open-source platform that streamlines the entire lifecycle of foundation models - from data preparation and training to evaluation and deployment. Whether you’re developing on a laptop, launching large scale experiments on a cluster, or deploying models in production, Oumi provides the tools and workflows you need. Link: https://oumi.ai/docs/en/latest/index.html @Machine_learning_lab_K

Are you interested in VLMs and LLMs. Use “oumi” as an ent-to-end solution for learning these hot trends from scratch: https://oumi.ai/docs/en/latest/index.html @Machine_learning_lab_K

📢 Interested in Shiny applications? If you’re curious about how Shiny works in real projects, I’ve built a few interactive applications using R and Python. You can check them out on my GitHub! 🔗 GitHub Repositories Feel free to explore, give feedback, or even contribute! 🚀 #Shiny #DataScience #Python #R #WebApps #MachineLearning @Machine_learning_lab_K

**Interested in Shiny applications?**If you’re curious about how Shiny works in real projects, I’ve built a few interactive applications using R and Python. You can check them out on my GitHub! 🔗 GitHub Repositories Feel free to explore, give feedback, or even contribute! 🚀 #Shiny #DataScience #Python #R #WebApps #MachineLearning

🔹 What is Shiny? Shiny is an open-source framework that allows you to build interactive web applications using R or Python, without requiring extensive web development skills. It enables data scientists and analysts to create dynamic dashboards and data-driven tools effortlessly. 🔹 Why Use Shiny? ✅ Easy to Use – Develop web apps with minimal coding in HTML, CSS, or JavaScript. ✅ Interactive & Reactive – Apps update automatically in response to user inputs. ✅ Supports Data Science – Seamlessly integrates with R/Python packages for visualization, machine learning, and analytics. ✅ Customizable – Extend functionality using JavaScript, CSS, or APIs. ✅ Deploy Anywhere – Host on Shiny Server, RStudio Connect, Docker, or cloud platforms. 🔹 Key Applications 🔹 Data visualization dashboards 🔹 Machine learning model interfaces 🔹 Business intelligence tools 🔹 Healthcare & finance analytics 🔹 Automated reporting systems 🚀 Whether you're in data science, finance, healthcare, or research, Shiny helps bring your insights to life with interactive web applications! Link: https://shiny.posit.co/ @Machine_learning_lab_k

DeepSeek just dropped ANOTHER open-source AI model, Janus-Pro-7B. It's multimodal (can generate images) and beats OpenAI's DALL-E 3 and Stable Diffusion across GenEval and DPG-Bench benchmarks. Available: https://huggingface.co/deepseek-ai/Janus-Pro-7B @Machine_learning_lab_k

Hugging Face released a free course on agents AI Agents are autonomous systems that can understand user requests, break them down into steps, and execute actions to accomplish tasks. They combine language models with tools and external functions to interact with their environment. This module covers how to build effective agents using the smolagents library, which provides a lightweight framework for creating capable AI agents. Link: https://github.com/huggingface/smol-course/tree/main/8_agents @Machine_learning_lab_K

Vertex AI RAG Engine: A developers tool Link: https://developers.googleblog.com/en/vertex-ai-rag-engine-a-developers-tool/ @Machine_learning_lab_K

https://arxiv.org/abs/2501.09223 @Machine_learning_lab_K

Phi-4 is a 14B parameter, state-of-the-art open model from Microsoft. Link: https://ollama.com/library/phi4 @Machine_learning_lab_K

A simple way to explain 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 𝗠𝗲𝗺𝗼𝗿𝘆. https://x.com/i/status/1878813189766299763 @Machine_learning_lab_K