Articles for category: AI Tools

March 16, 2025

ikayaniaamirshahzad@gmail.com

Novartis joins the Pharmaverse Council!

[This article was first published on pharmaverse blog, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here) Want to share your content on R-bloggers? click here if you have a blog, or here if you don’t. Hi pharmaverse community, We’re excited to welcome Novartis as the newest member of the Pharmaverse Council! Representing Novartis on the council will be Orla Doyle. Novartis has been a strong advocate for integrating open-source languages and tools into regulatory processes. Their commitment to the open-source community continues to grow, with contributions to various initiatives, direct support for

March 16, 2025

ikayaniaamirshahzad@gmail.com

Python 3.13.2 and 3.12.9 now available!

  A small release day today! That is to say the releases are relatively small; the day itself was of average size, as most days are. Python 3.13’s second maintenance release. About 250 changes went into this update, and can be yours for free if you just upgrade now. https://www.python.org/downloads/release/python-3132/ Python 3.12’s ninth maintenance release already. Just 180 changes for 3.12, but it’s still worth upgrading. https://www.python.org/downloads/release/python-3129/ Thanks to all of the many volunteers who help make Python Development and these releases possible! Please consider supporting our efforts by volunteering yourself or through organization contributions to the Python Software Foundation.

March 16, 2025

ikayaniaamirshahzad@gmail.com

OpenAI’s o1 Model, Explained

Was this newsletter forwarded to you? Sign up to get it in your inbox. OpenAI launched a new model, o1 (previously code-named Strawberry), yesterday. It’s significantly better at reasoning tasks, scoring in the 89th percentile in competitive programming, and exceeding Ph.D.-level smarts on physics, biology, and chemistry questions. It’s been taught to use chain of thought reasoning to answer each question it’s given rather than just blurting out a response. Chain of thought, of course, has been around for a long time. It’s the practice of asking a language model to solve problems by thinking out loud. You’re probably better

March 16, 2025

ikayaniaamirshahzad@gmail.com

mlx-community/OLMo-2-0325-32B-Instruct-4bit

mlx-community/OLMo-2-0325-32B-Instruct-4bit (via) OLMo 2 32B claims to be “the first fully-open model (all data, code, weights, and details are freely available) to outperform GPT3.5-Turbo and GPT-4o mini”. Thanks to the MLX project here’s a recipe that worked for me to run it on my Mac, via my llm-mlx plugin. To install the model: llm install llm-mlx llm mlx download-model mlx-community/OLMo-2-0325-32B-Instruct-4bit That downloads 17GB to ~/.cache/huggingface/hub/models--mlx-community--OLMo-2-0325-32B-Instruct-4bit. To start an interactive chat with OLMo 2: llm chat -m mlx-community/OLMo-2-0325-32B-Instruct-4bit Or to run a prompt: llm -m mlx-community/OLMo-2-0325-32B-Instruct-4bit 'Generate an SVG of a pelican riding a bicycle' -o unlimited 1 The -o unlimited

March 16, 2025

ikayaniaamirshahzad@gmail.com

Release notes for Deephaven version 0.32.0

Deephaven Community Core version 0.32.0 is the first release of 2024! We bring in the new year with a new calendar API, bug fixes, and quality-of-life improvements. Let’s dive into what’s new. Deephaven’s calendar API has been significantly revamped: The built-in calendar query library has been completely reworked. It is now consistent with the rest of the time library and is based on Java time types. The Python API has been trimmed down. The smaller API makes it easier to keep table operations fast by avoiding accidental Python/Java boundary crossings. Stay tuned for a blog that explores the new calendar

March 16, 2025

ikayaniaamirshahzad@gmail.com

Data Machina #246 – Data Machina

New Trends in Vision-Language Models (VLMs.) The evolution of VLMs in recent months has been pretty impressive. Today VLMs exhibit some amazing capabilities. See the two links below on what VLMs can do and how they work: But still VLMs are facing some challenges for example in terms of: multimodal training datasets, resolution, long-form modality, vision-language integration, and concept understanding. Somewhat along those lines, I see 5 trends happening in VLMs: 1) VLMs run on local environment 2) Emerging VLM videoagents 3) Unified structure learning for VLMs 4) Personalisation of VLMs and 5) Fixing the VLM resolution curse. Let’s see…

March 16, 2025

ikayaniaamirshahzad@gmail.com

Improving Recommender Systems & Search in the Age of LLMs

Search and recommendation systems have historically drawn inspiration from language modeling, such as the adoption of Word2vec for learning item embeddings, and GRUs, Transformer, and BERT for predicting on user behavior sequences. The current paradigm of large language models is no different. In the sections below, we’ll discuss how industrial search and recommendation systems have evolved over the year or so, spanning model architectures, data generation strategies, training paradigms, and unified frameworks: LLM/multimodality-augmented model architecture Recent developments in recommender system architectures increasingly adopt language models and multimodal content to overcome traditional limitations of ID-based approaches. These hybrid architectures integrate content

March 16, 2025

ikayaniaamirshahzad@gmail.com

explosion/spacy-layout: 📚 Process PDFs, Word documents and more with spaCy

This plugin integrates with Docling to bring structured processing of PDFs, Word documents and other input formats to your spaCy pipeline. It outputs clean, structured data in a text-based format and creates spaCy’s familiar Doc objects that let you access labelled text spans like sections or headings, and tables with their data converted to a pandas.DataFrame. This workflow makes it easy to apply powerful NLP techniques to your documents, including linguistic analysis, named entity recognition, text classification and more. It’s also great for implementing chunking for RAG pipelines. 📖 Blog post: “From PDFs to AI-ready structured data: a deep dive”