@sugatoray
VP, Data Scientist @Truist I am a Physicist turned Data Scientist. Deeply passionate about Technical stuff and mostly post about Python, Deep Learning, ML, LLMs, MLOps, DevOps, etc. Find more on me at https://linktr.ee/sugatoray Active on LinkedIn.
🏎️ You can perform 200ms search over 40 million texts using just a CPU server, 8GB of RAM, and 45GB of disk space.
The trick: Binary search with int8 rescoring.
I'll show you a demo & how it works in the 🧵:
🎁 Check out this @PyTorch blog showing how to get peak performance using torch.compile + diffusers libraries in Python.
🍀 Blog: pytorch.org/blog/torch-c... torch.compile and Diffusers: A Hands-On Guide to Peak Performance – PyTorch
#PyTorch #torchCompile #diffusers #python
⚡️ YouTube Video: youtu.be/VePxCcF99w4?...
@vscode.dev @anthropic.com #mcp #vscode #aiagents #tooluse
🍀 Note: you can use Ollama as an option for Local Private LLM along with GitHub Copilot. By default copilot uses “OpenAI” models. But you can choose other providers or use Ollama for local LLMs.
Have fun. Time to get creative now! 🎉
🔥Video: lnkd.in/g2KjDysw
🎁 Check out the video on how to use Tools with Agent Mode in Visual Studio Code (#VSCode).
It also covers:
- Promot Boost (a VSCode extension to improve your prompts on the fly while you use copilot)
- How to use MCP
🌏 What was “googling” back since 2005 until 2025, as a must have skill, will become layered by the skill to effectively use AI for while collar jobs by 2030 (or, may be sooner).
🍓 But how do you learn such a skill? By applying it and playing with it.
✨ Using AI Tools effectively is a skill - a skill you don’t get taught. A skill you acquire with practice and exposure to the technology.
🔥Video: youtu.be/VePxCcF99w4
#vscode #LLMs #ai #aiagents #tooluse #MCP #tipsnadtricks
🔥 The smolagents module has arrived in the agents course!
💻 Code agents optimised for software development
🔧 Tool calling agents that create modular, function-driven workflows
🔍 Retrieval agents designed to access and synthesise information
Course: https://buff.ly/4kcj6Ai
hot take: WebRTC should be ONE line of Python code
introducing FastRTC⚡️ from Gradio!
start now: pip install fastrtc
what you get:
- call your AI from a real phone
- automatic voice detection
- works with ANY model
- instant Gradio UI for testing
this changes everything
#Bookmark 🔥 #webrtc #python #pylib #fastrtc
WebRTC with Python library FastRTC
pip install fastrtc
🎉🎉🎉🤗🤗🤗
👏👏 I also liked the fact that the installation checks for successful installation of duckdb by running “SELECT 42” and if it does not return 42 as a result, that means duckdb was not installed.
I also liked the fact that the installation checks for successful installation of duckdb by running “SELECT 42” and if it does not return 42 as a result, that means duckdb was not installed.
I thought I knew well how to optimize Python container size until I watched this talk from Matthijs Brouns 👇🏼
www.youtube.com/watch?v=Z1Al...
#docker #python
That’s neat! 🎉 Install #duckdb on #MacOS and #Linux with a simple command.
curl install.duckdb.org | sh
Since, I can also see the code of the shell script on the site, I can copy it and save it in a file locally and then run it. However, curl <url> | sh does exactly that!
🍓 Check out “The Illustrated DeepSeek-R1” by @jayalammar.bsky.social
👉 Blog: open.substack.com/pub/jayalamm...
#ml #LLMs #deepseek
The newest extremely strong embedding model based on ModernBERT-base is out: `cde-small-v2`. Both faster and stronger than its predecessor, this one tops the MTEB leaderboard for its tiny size!
Details in 🧵
New LLM Eval Office Hours, I discuss the importance of doing error analysis before jumping into metrics and tests
Links to notes in the YT description
youtu.be/ZEvXvyY17Ys?...
🔥 #OpenAI o3 model performance makes a leap, sets a new high score on the #ARCAGI benchmark.
Source: arcprize.org/blog/oai-o3-...
#ml #ai #arcagi #benchmark #openai
@ollama.bsky.social Python library introduced #FunctionCalling in version 0.4.
Ollama is an AI tool that allows you to run Large Language Models (LLMs) on device (local LLM).
Ollama blog: ollama.com/blog/functio...
#ollama #python #LLMs #ml #ai #localLLM #localai
An Introduction to Polars 🐻❄️👇🏼
The Polars workshop from the PyData NYC conference is now available online. This great workshop, by Matt Harrison, focuses on the foundation of Polars 📽️ 👇🏼
www.youtube.com/watch?v=q3o2...
#Python #Data #Polars
If you're interested in embedding models for retrieval (search), clustering, classification, paraphrase mining, etc., then there's now 10,000 fully free and open source options on @hf.co via Sentence Transformers.
Check out the most popular ones here: huggingface.co/models?libra...
New interval lookup library just hit for Rust, C++ and Python — claims best in class performance! Pretty cool : github.com/kcleal/super...
"A Comprehensive Guide to Python Project Management and Packaging Concepts Illustrated with uv"
This Reinforced Knowledge series of Python packaging might be the most comprehensive write-up I've... ever seen?
reinforcedknowledge.com/a-comprehens...
One of the fun things you can do with uvx to just … run python.
Super useful to play around with a python package without having to make a virtualenv somewhere.