Santiago Viquez

🚀 scikit-learn 1.8 is out 🚀 A big shoutout to the community of contributors who continue to push open-source machine learning forward ❤️ ✨ Key Highlights: ▶️ Expanded Array API support (including PyTorch & CuPy) to run more estimators and metrics on GPUs ▶️ Free-threaded CPython 3.14 support for better multi-threaded performance ▶️ Probability calibration with temperature scaling in CalibratedClassifierCV ▶️ Major efficiency boosts in linear models (Lasso / ElasticNet with gap safe screening) ▶️ Much faster and more robust DecisionTreeRegressor with criterion="absolute_error" ▶️ New manifold.ClassicalMDS implementation for classical multidimensional scaling 🔗 Check the full release highlights: https://lnkd.in/gkXQSbmZ Discover scikit-learn 1.8 and its: 🟢 28 new features 🔵 12 efficiency improvements & 13 enhancements 🟡 9 API changes 🔴 34 fixes 👥 193 contributors (thank you all!) 📖 More details in the release notes: https://lnkd.in/gCrs42se You can upgrade with pip as usual: pip install -U scikit-learn Using conda-forge builds: conda install -c conda-forge scikit-learn #scikitlearn #MachineLearning #opensource #DataScience #Python #ML

1518

14 comentarios

There have been lots of interesting LLM releases last week. My favorite was actually the Olmo 3 release. Olmo models are always a highlight since they are fully transparent (including training methods and datasets) and come with very detailed technical reports. I am sure I'll talk more about the interesting training-related aspects from that 100-pager in the upcoming days and weeks. In the meantime, here's the side-by-side architecture comparison with Qwen3: 1) As we can see, the Olmo 3 architecture is relatively similar to Qwen3. However, it's worth noting that this is essentially likely inspired by the Olmo 2 predecessor, not Qwen3. 2) Similar to Olmo 2, Olmo 3 still uses a post-norm flavor instead of pre-norm, as they found in the Olmo 2 paper that it stabilizes the training. 3) Interestingly, the 7B model still uses multi-head attention similar to Olmo 2. However, to make things more efficient and reduce the KV cache size, they now use sliding-window attention (e.g., similar to Gemma 3). Next, the 32B model (the figure is not shown here due to space reasons, but you can find it in my "The Big LLM Architecture Comparison" article or my Olmo 3 from-scratch notebook): 4) Overall, it's the same architecture but just scaled up. Also, the proportions (e.g., going from the input to the intermediate size in the feed-forward layer, and so on) roughly match the ones in Qwen3. 5) My guess is the architecture was initially somewhat smaller than Qwen3 due to the smaller vocabulary, and they then scaled up the intermediate size expansion from 5x in Qwen3 to 5.4 in Olmo 3 to have a 32B model for a direct comparison. 6) Also, note that the 32B model (finally!) uses grouped query attention. And yes, I also did a from-scratch implementation. It was still a lot of work, but since I had already implemented Qwen3 from scratch, as well as Gemma 3 (for the sliding-window attention component), it wasn't too bad! If you are a coder, looking at the from-scratch implementation, it's probably the easiest way to understand the architecture: https://lnkd.in/gQgxtVUu

1559

34 comentarios

🚀 scikit-learn 1.7 is out 🚀 A big shoutout to the community of contributors who continue to push open-source machine learning forward ❤️ ✨ Key Highlights: ▶️ Improved estimator’s HTML representation ▶️ Custom validation set for histogram-based Gradient Boosting estimators ▶️ Plotting ROC curves from cross-validation results ▶️ Updated Array API support ▶️ Improved API consistency of Multi-layer Perceptron ▶️ Migration toward sparse array 🔗 Check the full release highlights: https://lnkd.in/eZx9DJx7 Discover scikit-learn 1.7 and its: 🟢 9 new features 🔵 2 efficiency improvements & 4 enhancements 🟡 6 API changes 🔴 7 fixes 👥 146 contributors (thank you all!) 📖 More details in the changelog: https://lnkd.in/eySABsYH You can upgrade with pip as usual: pip install -U scikit-learn Using conda-forge builds: conda install -c conda-forge scikit-learn #scikitlearn #MachineLearning #opensource #DataScience #Python #ML

1599

19 comentarios

Looking for a cheaper, open source alternative to agentic RAG? Try multi-vector retrieval with Reason-ModernColBERT. Most search systems compare one big summary vector per document. Multi-vector retrieval is different - it keeps separate vectors for each word or phrase, then finds the best matches between your search and the document pieces. Reason-ModernColBERT is a small but mighty model trained on the reasonir-hq dataset and specializes in finding documents that require logical thinking, not just keyword matching. It actually outperforms models 45x larger on reasoning tasks with just 150M parameters - tiny compared to billion-parameter models. It also creates small 128-dimension vectors and handles documents up to 8,000+ words long. So while agentic RAG uses LLMs to tackle complex questions requiring intricate planning, multi-step reasoning, and utilization of external tools, sometimes you don't need all that complexity. Multi-vector retrieval gives you smart search results at a fraction of the cost. Ready to try it? Check out this notebook by Leonie Monigatti: https://lnkd.in/ePr-vQt6

411

19 comentarios

Architectures for Knowledge Agents 📚🤖 We’re excited for the potential of LLMs to automate all knowledge work over unstructured enterprise data. This talk by Jerry Liu in the “Systematically Improving RAG Applications” course (by Jason Liu) shows how you can build knowledge agentic workflows beyond basic RAG architectures. Check it out! Covers three parts: 1. The components of a high-quality knowledge management layer to centralize/index your data 2. Event-driven agent orchestration 3. Putting them together for document research, extraction, report generation use cases. Slides: https://lnkd.in/gXYQpJRR If you’re looking to build/deploy enterprise knowledge agents, come get in touch: https://lnkd.in/g5648-ip

219

5 comentarios

📚 We're excited to announce our new course, Orchestrating Workflows for GenAI Applications, with DeepLearning.AI, Andrew Ng, and taught by our own Kenten Danas and Tamara Janina Fingerlin! In this course, you'll learn how to build and orchestrate a RAG pipeline using Apache Airflow 3: 💻 Transform your code into pipelines consisting of modular tasks and schedule them using time-based and data-aware scheduling ⏩ Apply dynamic task mapping to run tasks efficiently in parallel and automatically adapt to new data sources ⚙️ Add automatic retries and failure notifications to handle errors gracefully Enroll for free: https://bit.ly/45P4WQN

69

1 comentario

When labs migrate models from TensorFlow to JAX, even small translation errors can break performance. In this case study, we show how Turing built a layer-level evaluation dataset that detects inconsistencies early, isolates failure modes, and gives researchers a reproducible way to validate translation accuracy. If you work in model migration, infrastructure or large codebase refactors https://bit.ly/4abfHz7 this is a practical look at how structured evaluation can de-risk the entire workflow. #AI #Turing

51

Need NumPy’s simplicity but want GPU speed and effortless gradients for image pipelines or hefty matrix math? Tom Reid shows how JAX bolts automatic differentiation, just-in-time compilation, and GPU acceleration onto familiar NumPy workflows.

11