AI & Data Science Discussions: LLMs, Growth Mindset, and Advanced Tools Summary of chat 8/9 – 12/9
8/9
*LLMs Browsing, Unit Tests, Code Reviews*
Members discussed LLMs browsing on phones, unit tests, and code reviews.
*LLMs Browsing on Phones*
– Members shared a link on empowering LLMs to use smartphones for intelligent task automation.
– A paper and code on efficient LLM finetuning with T-Few was shared.
– Discussion on GPT-4 and #WriteSonicAI as alternatives to ChatGPT.
*Unit Tests*
– Members talked about Scikit-LLM, a tool to integrate LLMs like ChatGPT into scikit-learn for text analysis tasks.
– Different functionalities of Scikit-LLM were mentioned, including zero-shot and few-shot text classification, text vectorization, fine-tuning, summarization, and translation.
– A GitHub repo link was shared for Scikit-LLM.
*Code Reviews*
– A blog on data preprocessing in data science was shared, covering steps like eradicating duplicates, fixing errors, handling outliers, handling missing values, and data encoding.
– A link to Candle, a new machine learning framework for Rust, was shared.
– Discussion on starting tutoring of advanced ML/AI in the Boston/Cambridge area.
*MIT IAP Course*
– An MIT IAP course on Introduction to Data-Centric AI was mentioned.
– The course focuses on topics like automatically finding label errors, data-centric vs model-centric AI, data-centric evaluation of ML models, interpretability, obtaining high-quality data annotations, and data-centric human priors.
– A link to lab assignments for the course was shared.
*Generative AI Hackathon Project*
– A member requested project ideas on generative AI related to molecular dynamics for a hackathon.
– No specific discussions or suggestions were mentioned in the conversation.
9/9-LLMs Browsing, Unit Tests, Code Reviews
*
*AI App for Generating Business Designs*
Members discussed the possibility of using an AI app to generate fliers and designs for business by supplying write-ups. They shared their experiences and preferences.
*Python vs R for Production-Grade Code*
– Members compared the advantages of Python over R for writing production-grade code, such as faster performance, better OOP system, and more advanced features and tools.
– Python’s elaborate set of linting tools, data orchestration, and automation tools like Airflow and Prefect were highlighted.
– The massive community support and frequent updates in Python were mentioned as advantages over R.
*Python’s Strengths and Weaknesses*
– Members discussed Python’s strengths, including web scraping, interfacing with APIs, and PySpark.
– They also acknowledged R’s strengths in statistical tools, specific-domain support, and exploratory data analysis, especially using tidyverse and delayed evaluation.
– The superiority of ggplot2 over Python’s plotnine for visualization was mentioned, along with R’s integration with RMarkdown and the active development of RStudio as an IDE.
*Introduction to Graph Theory*
– A meetup on Graph Theory was announced, with topics including edges, vertices, trees, networks, paths, and graphs.
– The Zoom link for the session was mentioned, and the start time was discussed.
– A link to an updated introduction to Graph Theory in PDF format was shared.
*Machine Learning Compression and Decision Tree Algorithms*
– The members inquired about machine learning compression and how it is useful.
– They also discussed the working of Decision Tree algorithms in machine learning.
– A member named Bhaskar was invited to join the discussion if possible.
12/9 LLMs Browsing, Unit Tests, Code Reviews
*AI Events, Apple Event, VPN Suggestions*
Members discussed various topics including AI events, the Apple event, and VPN suggestions.
*AI Events*
– Members discussed the need for a comprehensive calendar for AI events.
– There was mention of new services for data science by AWS and Google.
– A member mentioned summarizing the webinar and creating a mean calendar for future AI events.
*Apple Event*
– Members discussed expectations for the Apple event, including the release of information about the M3 chip and the iPhone 15.
– No specific conclusions were reached about the event.
*VPN Suggestions*
– Members discussed the need for a VPN to access Discord voice in certain locations.
– Different VPN options were suggested, including Private Internet Access and Exit Lag.
– There was mention of strict rules for VPNs in the UAE.
Join AI community to get access to Generative AI, AI & Data Science courses, live workshops, and all recordings for less than a Tea Price a month.
Join AI community to get access to Generative AI, AI & Data Science courses, live workshops, and all recordings for less than a Tea Price a month. Use coupon code NOW10 FOR action takers. ( Only First 5 Coupons)
About Community Owner: Mohammad Arshad, A Globally recognized expert in AI
With 18+ years of experience, Mohammad Arshad has successfully enabled businesses to monetize their data and AI products. His technical and strategic expertise has helped 5 of the largest companies in the world, 10 SMEs, and 3 startups build effective Data and AI Strategies.
He founded Decoding Data Science in 2020, a successful AI Strategy consulting practice, and expanded into education in 2022 with the launch of DDS Academy.
Accenture, HP, Dell, LinkedIn, MAF, and other leading companies have recognized Mohammad as a Data Science and Strategy expert since 2005.
Mohammad Arshad has been teaching technical and non-technical audiences since 2008, making him a seasoned mentor and coach in the industry. He has helped 600+ individuals get their Dream jobs.