Nick L.

🚀 Unlock the power of Generative AI with your data! Join John Savill as he demystifies RAG, semantic search, vectors, and embeddings in his latest technical training video. Watch now: https://lnkd.in/eW5iVAks

27

📰 New article: Deep Learning with SAR 👇

21

1 Comment

Alexandre Andorra, co-founder of PyMC Labs, joins Jon Krohn for an in-depth look at Bayesian statistics. Learn how Bayesian methods tackle complex problems, leverage prior knowledge, and optimize limited data using tools like PyMC, PyTensor, ArviZ, and Gaussian Processes. Don’t miss this in-depth episode! Watch here: https://bit.ly/sds793 Thank you to Crawlbase for supporting this episode of SuperDataScience, enabling the show to be freely available on all major podcasting platforms and on YouTube. #datascience #bayesian #pymc

7

Text embeddings have become crucial for transforming textual data into dense vector representations (with context), enabling efficient processing across various modalities like text, images, and audio. This article provides a good summary on some of these (not all) embeddings available for text https://lnkd.in/gKp4ta9a 1. Gensim Word Embeddings (Free) Why Use It: Specializes in NLP with open-source models like Word2Vec and FastText, promoting transparency and ease of custom training. 2. Facebook Embeddings (Free) Why Use It: Trained on extensive corpora and supports over 100 languages, making it versatile for global applications. 3. AllenNLP Embeddings (Free) Why Use It: Provides NLP-focused embeddings like BERT and ELMo, with tools for fine-tuning and visualization. 4. OpenAI Embeddings (Paid) Why Use It: Comprehensive training on massive datasets allows for effective semantic understanding, making it suitable for advanced NLP tasks. Ideal for tasks requiring zero-shot learning. 5. MultiLingual BERT (Free) Why Use It: Handles multilingual datasets effectively, offering robust support for multiple languages. Perfect for applications requiring consistent performance across different languages, such as global customer support systems. 6. DistilBERT Base Uncased (Free) Why Use It: Smaller and faster than BERT, maintaining near-comparable performance with reduced resource usage. 7. MPNet V2 (Free) Why Use It: Excels in text similarity tasks with a Siamese architecture, providing accurate similarity measurements. 8. HuggingFace Embeddings (Free/Paid) Why Use It: Offers versatility with a wide range of embeddings (text, image, audio) and the ability to fine-tune on custom data. HF also provides wrappers to almost all open sourced models. 9. Longformer Base 4096 (Free) Why Use It: Supports long texts up to 4096 tokens, making it unique in handling lengthy documents without truncation. 10. RoBERTa (2022) (Free) Why Use It: Offers updated text representations with comprehensive training on recent data. 11. GTE-Base (Free) Why Use It: Offers a good balance of performance and resource requirements with a 768-dimensional embedding. 12. GTE-Large (Free) Why Use It: Provides high-quality embeddings with more detailed representations, useful for nuanced text analysis. Best for applications requiring in-depth text understanding, such as detailed content recommendations. 13. GTE-Small (Free) Why Use It: Optimized for speed with a smaller model size, making it efficient for applications needing quick processing. Useful for real-time systems where fast text similarity searches are critical. 14. E5-Small (Free) Why Use It: Combines compact size with good performance, making it suitable for fast, general-purpose text embedding tasks.

11

The new Spark NLP 5.5 is a game-changer for large language model (LLM) inference, bringing unparalleled scalability and efficiency to the forefront. Key highlights of this major release include: 🔗 Native integration with Llama.cpp, unlocking access to tens of thousands of GGUF models on Hugging Face – now deployable at scale. 🖥️ Load any quantized model as a GGUF via Llama.cpp across various environments – from local setups to multi-node clusters on platforms like Databricks, AWS EMR, Azure, and Google Cloud Platform. 💻 A hardware-agnostic approach that optimizes performance across Intel processors, Nvidia CUDA GPUs, and Apple Silicon chips, allowing organizations to scale their LLM & NLP capabilities using existing infrastructure. With 120 million downloads of Spark NLP to date, the 5.5 release puts tens of thousands of models at your fingertips – ready to drive the next generation of AI applications. This session was presented by Maziyar Panahi, our Principal AI Engineer & Team Lead at #NLPSummit 2024 Watch an entire video: https://hubs.li/Q02V0VQ10 #largelanguagemodels #LLMs #responsibleai #GenerativeAI #MachineLearning #IoT #ResponsibleAI #artificialintelligence #digitaltransformation #NLP #nlu

6

LlamaIndex + MLflow RAG contains a lot of parameters to tune, from chunking to indexing - everything affects downstream answer accuracy. It’s important to have a systematic approach to track and tune these parameters, and define the right eval metrics/dataset to evaluate changes. Check out Jino Rohit's video on experimenting with a LlamaIndex pipeline with MLflow for experiment tracking. https://lnkd.in/gyYXfyyb

554

14 Comments

I've put together a complete list of all #PyCon deep dives (a.k.a. Tutorials). Read on 👇 PyCon introduced a longer talk format (usually >2h) since 2020 that allows for deep dives into specific topics. They call it “Tutorials”. I have to admit—I love this new format and the quality of the talks, so I put together a compilation of all of those published so far, grouped by year, and ordered by view count. Due to LinkedIn post length limit, I'm attaching the top 10 below. Find the complete list here: https://lnkd.in/eR8fNX-4 1. "Tutorials - 🐍 Matt Harrison: Getting Started with #Polars" ⸱ https://lnkd.in/e4cQwkTm ⸱ +15k views ⸱ +15k views ⸱ 02 Jun 2023 ⸱ 02h 14m 28s 2. "Tutorials - Felipe de Morais, Amanda Silva Vieira: 🌐 Building Your First #API with #Django and Django #Rest Framework" ⸱ https://lnkd.in/e5tBXvqf ⸱ +900 views ⸱ +900 views ⸱ 28 Jun 2024 ⸱ 02h 54m 17s 3. "Tutorial: Keith Galli - Natural Language Processing (#NLP) in #Python - From Zero to Hero" ⸱ https://lnkd.in/eFCXXi3s ⸱ +43k views ⸱ +43k views ⸱ 29 Apr 2020 ⸱ 01h 39m 23s 4. "Tutorials - Trey Hunner: Intro to Python for Brand New Programmers" ⸱ https://lnkd.in/ewUgZ6N5 ⸱ +4k views ⸱ +4k views ⸱ 02 Jun 2023 ⸱ 02h 16m 41s 5. "Tutorials - Mike Müller: The How and Why of Object-oriented Programming in Python" ⸱ https://lnkd.in/eK2MjfDu ⸱ +3k views ⸱ +3k views ⸱ 02 Jun 2023 ⸱ 02h 45m 34s 6. "Tutorial: Santiago Basulto - Python Concurrency: from beginner to pro" ⸱ https://lnkd.in/eBURDMUa ⸱ +32k views ⸱ +32k views ⸱ 06 May 2020 ⸱ 02h 57m 13s 7. "Tutorial: Kimberly Fessel, PhD - It's Officially Legal so Let's Scrape the Web" ⸱ https://lnkd.in/expUQNaj ⸱ +12k views ⸱ +12k views ⸱ 29 Apr 2020 ⸱ 02h 46m 26s 8. "Tutorials - Sebastian Raschka, PhD: The Fundamentals of Modern Deep Learning with #PyTorch" ⸱ https://lnkd.in/e7NAsav6 ⸱ +2k views ⸱ +2k views ⸱ 28 Jun 2024 ⸱ 02h 59m 32s 9. "Tutorials - José Haro Peralta: Fundamentals of API security with Python" ⸱ https://lnkd.in/ezwsGhz6 ⸱ +900 views ⸱ +900 views ⸱ 28 Jun 2024 ⸱ 02h 33m 43s 10. "Tutorials - Reuven Lerner: Comprehending comprehensions" ⸱ https://lnkd.in/eS2cT6Gk ⸱ +6k views ⸱ +6k views ⸱ 02 Jun 2023 ⸱ 02h 17m 33s 11. Find the complete compilation here: https://lnkd.in/eR8fNX-4 --- I built this list as a part of Tech Talks Weekly. Tech Talks Weekly https://lnkd.in/eXtr7zgf is a free weekly email with all the recently uploaded talks from +100 software engineering conferences. Join thousands of readers who stopped scrolling through messy YT subscriptions.

16

3 Comments

Data orchestration and machine learning are evolving rapidly, with tools like Apache Airflow at the forefront. Dive into data processing and AI model deployment in our latest episode of The Data Flowcast: Mastering Airflow for Data Engineering & AI. Julian LaNeve, CTO at Astronomer, and David Xue, Machine Learning Engineer at Astronomer, share their expertise on generative AI, fine-tuning models, and Airflow’s versatility. Discover how high-quality data elevates AI performance and learn about managing large-scale AI projects. For the full episode, click the link in the comments. #AI #Automation #Airflow #MachineLearning

17

1 Comment

Learn how fine-tuning Phi-3 Vision on a custom dataset with tracking & evaluating using W&B Weave. From dataset prep to inference logging, see the full process and benefits. Check out the latest article from Brett Young: https://lnkd.in/gxddTTwb

15

🚀 Looking to streamline fine-tuning for models like Llama 3, Mistral, and Mixtral? Our latest blog shows how Anyscale simplifies LLMs fine-tuning with Ray: 🔑 Key takeaways: • Scalable fine-tuning with LLM-Forge • Model Deployment with RayLLM Read more: https://lnkd.in/gQwD4A2U

34

It’s amazing to see the impact of open-source projects! Text2Bricks from our ML team was one of many projects made possible by the Open-Sora team at HPC-AI Tech: https://hubs.ly/Q02DMGyW0 More on Lambda’s Text2Bricks model: https://hubs.ly/Q02DMWFL0

87

What makes Airflow 3 a game-changer for data orchestration? In the latest episode of The Data Flowcast: Mastering Airflow for Data Engineering & AI, we sit down with Vikram Koka of Astronomer and The Apache Software Foundation, to explore the groundbreaking advancements of Airflow 3 and how it's shaping the future of data orchestration. Tune in to learn how Airflow 3 supports mission-critical data pipelines, multi-cloud deployments, and more. Check out the link in the comments below to listen. #AI #Automation #Airflow #MachineLearning

36

2 Comments

Torch Compile is here to optimize inference in large language models (LLMs) with minimal overhead. In Benjamin Marie's latest article, explore how adding just one line of code can improve decoding performance. #Llama LLM

14

Staying updated with ML research is challenging. Our latest blog demonstrates how to build a bot summarizing arXiv papers using W&B Weave and Anthropic's API. Check out the technical details and follow along with our interactive Colab. Read more: https://lnkd.in/ggratqF7

24

Build robust GenAI pipelines with LlamaIndex, Qdrant, and MLflow for advanced RAG! Learn how to: 🔍 Streamline RAG workflows for better scalability 📊 Ensure performance consistency across model versions 🚀 Optimize indexing systems for efficiency This step-by-step guide covers: • Setting up MLflow for experiment tracking • Configuring Qdrant for vector storage • Integrating Ollama for LLM and embedding models • Creating a playground app to test the full pipeline Check out the full post here: https://lnkd.in/gCnm9JgE

230

5 Comments

Follow along Chaim Rand's new model-optimization tutorial to learn how you can leverage PyTorch NestedTensors, FlashAttention2, and xFormers to boost transformers' performance and reduce costs.

20

Ever grappled with juggling multiple data processing engines, i.e.,🦆,❄️, 🐼 etc. etc.? Meet Ibis, the hidden-gem of PyData that's about to revolutionize your data workflow. This post unpacks how Ibis seamlessly integrates with various backends and presents a concept where a single expression can run across multiple- engines. The primary benefits are: 1️⃣ Segment computation, by data size or cost, in-situ or serverless 2️⃣ Apply database style optimizations to pipelines with ML or complex operators 3️⃣ Right-size infrastructure, based on consumption stages 🔗 Deep dive into the full post here: https://lnkd.in/ehNqCTwB

53

5 Comments

𝐁𝐮𝐢𝐥𝐝𝐢𝐧𝐠 𝐒𝐚𝐟𝐞, 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐞𝐝 𝐒𝐲𝐧𝐭𝐡𝐞𝐭𝐢𝐜 𝐃𝐚𝐭𝐚 𝐏𝐢𝐩𝐞𝐥𝐢𝐧𝐞𝐬 𝐰𝐢𝐭𝐡 𝐆𝐫𝐞𝐭𝐞𝐥 In this tutorial, Grace from our Product team demonstrates how to use our new 𝑊𝑜𝑟𝑘𝑓𝑙𝑜𝑤 𝐵𝑢𝑖𝑙𝑑𝑒𝑟 feature to chain synthetic data models together. This approach helps create privacy-preserving synthetic datasets from sensitive enterprise data. Watch the demo to see how automated synthetic data pipelines can accelerate responsible AI strategy and development. 📽️ https://lnkd.in/eVQ2uiyX #syntheticdata #enterpriseAI #privacy #finance

10

From struggling with grade school algebra to complex geometry, AI has a math problem. Building on findings from Apple, our R&D team demonstrated how outputs from complex models become less accurate for problems that require understanding – not pattern recognition. But there’s a path forward. In his byline for Unite.AI, MathGPT.ai Chairman Peter Relan writes, “If we’re successful on this path, I’m confident we can change the lives of millions of students and even professionals to put their lives on an entirely new trajectory.” Read Peter’s byline here: https://lnkd.in/eQGppQ2M #LLMReasoning #AI #Math

8