The LangGraph series provides a short introduction to the LangGraph 🦜🔗 library. This includes the following topics:
✅ The library core functionality
✅ Agend executor
✅ Dynamicly returning a tool output directly
✅ Managing agent steps
After I finished the tutorial for setting a dockerized Python environment in VScode with the Dev Containers extension using the GitHub template last week, I worked on the R version this weekend.
The goal is to reduce the level of effort in setting up a new project using a GitHub template.
I keep hearing the term NGINX, but I never really understood its functionality. This one-hour tutorial by Laith Harb really does a great job of explaining its functionality 👇🏼
Last weekend, I created a template for setting up a dockerized Python 🐍 development environment with VScode, Docker 🐳, and the Dev Containers extension, and this weekend I created a short tutorial 👇🏼
I think that most of the data scientists prefer to use some type of virtual environment (VE) in their applications. A short 🧶🧵 about the main differences between the two 👇🏼
(2/4)In Python, the ones that I know are venv, conda, poetry, and mamba 🐍, and in R, the main one is renv.
The main advantage of using VE is that it is simple to use, and it comes with a fairly short learning curve. Yet, it does not cover you from potential issues related to dissimilarities of your environments, such as OS type, hardware, and other potential issues.
(3/4) Docker, on the other, provides a higher level of reproducibility as it ships your code in the same environment you developed and tests your code with containers. The main disadvantage of Docker with respect to VE is the high learning curve.
Are you planning to learn a new data science or engineering skill as your New Year’s resolution ? Here is a collection of random open and free courses and resources I came across during the past year covering various topics, including deep learning, NLP, Python, statistics, and more.
(1/7)There is no better way for me to summarise the year than my Github account and my Git commits 😎
In 2023, I had more than 2500 commits, most related to project automation with Github Actions ❤️. Most of my personal projects during 2023 were related to tutorials and open-source projects. Here are the main highlights 🧶🧵👇🏼
(3/7)
➡️ Another fun project was creating a step-by-step guide and a template for setting up a natural language to SQL code generator with the OpneAI API. This project mainly focuses on the operation side of working with LLMs and APIs 👇🏼
🔗 https://github.com/RamiKrispin/lang2sql
If I have time in the coming year, I will extend it to other LLMs, such as open-source models (llama2, etc.) and Google's PaLM API.
I came across this amazing repo by 𝐒𝐭𝐚𝐬 𝐁𝐞𝐤𝐦𝐚𝐧 - the Machine Learning Engineering Online Book with a collection of guides for ML engineering focusing on training LLM and multi-model models.
License: Attribution-ShareAlike 4.0 International 🦄
(2/2) This book is still 𝐖𝐈𝐏, and it covers the following topics:
✅ Hardware concepts such as working with CPUs, GPUs, networks, etc.
✅ Performance - model parallelism, multi-node networking
✅ Development - debugging, reproducibility, data types
✅ Operating - Training hyperparameters, instabilities, etc.
I had great productivity during the Thanksgiving break, wrapping some of the tutorials I have been working on during the last few months 🚀. I updated this week my Github page with the list of currently available tutorials 📚: https://github.com/RamiKrispin
A new introduction course for MLOps was released today by freeCodeCamp. The 3 hours course, taught by Ayush Singh, provides an introduction to MLOps, and it covers topics and tools such as:
✅ Foundation of MLOps
✅ MLOps tools such as Zenml, MLflow, etc.
✅ Creating workflow and pipelines
✅ Pipeline deployment
It was a great to attend at the Øredev Developer Conference 🚀 in Malmö, Sweden 🇸🇪. I had the pleasure of presenting in the conference data track about MLOps and forecasting. Besides attending the great sessions, it was great to meet new folks 😎. Thanks, Emy Wennerberg Kristoffersson and Maisa Dabus, for organizing such a great event, and Niklas Hansson, Hugo Hjertén, and Christian Henrik Reich for organizing an amazing data track and for the invite! 🙏🏼 #data#DataScience#dataengineering#MLops