Data
22 talks in this track
Understanding ChatGPT: Embeddings, Transformers and Reinforcement Learning with Human Feedback
talk.
functime: a next generation ML forecasting library powered by Polars
talkPolars conquered dataframes, and now it is coming for machine learning! With Polars-powered feature-extraction and a best-of-the-class set of diagnostic tools, functime enables **forecasting thousands...
Introduction to Polars DataFrames - how to supercharge your data workflows
talkPolars is the new dataframe on the block taking the world by storm. You'll learn: - what Polars is, and what it can do for you - Polars basics and core concepts (including expressions and lazy co...
🧼 From GPU-poor to data-rich: data quality practices for LLM fine-tuning
talkGabriel Martín Blázquez, David Berenstein
If you are GPU-poor you need to become data-rich. I will give an overview of what we learned from looking at Alpaca, LIMA, Dolly, UltraFeedback and Zephyr and how we applied that to fine-tuning a stat...
Making an e-shop search bar your friend with Pinecone's hybrid search
talkFast and accurate search results are a crucial components of any e-shop and thus can make the difference between high user satisfaction and user frustration. With recent advancements in vector search ...
Speed up open source LLM-serving with llama-cpp-python
talkLarge language models (LLMs) often require huge compute resources to serve. This is a common challenge for those who want to avoid sharing their data with cloud API providers, or to deploy their stack...
DataFrame interoperatiblity - what's been achieved, and what comes next?
talkIn 2023, we saw several libraries - which had previously only supported pandas - add support for other dataframe libraries such as Polars, Modin, and cuDF. - How did they do it? - Are there any dr...
Write-Audit-Publish Pattern in Modern Data Pipelines
talkData is new oil, and one of the ways is leakage and poisoning the surrounding environment. What happens if you pollute one of the datasets used in some decision makers facing dashboards? In this talk,...
Transcend the Knowledge Barriers in RAG: Setup, Chat State, and More
talkDeveloper tools power many LLM-based chat and Retrieval Augmented Generation applications today. However, there is a non-trivial knowledge barrier for entrants that could hinder developer experience. ...
The pragmatic Pythonic data engineer
talkLearn to make practical decisions in data engineering with Python's vast ecosystem. Avoid blindly following market guidelines and consider the reality of your situation for better performance and arch...
Generative AI in Lithuanian language
talkPresentation about how we (few local NLP enthusiasts) trained Language Transformer to generate meaningful text in Lithuanian language. Everything was based on volunteer work with huge R&D flavor. Du...
Transforming Data Insights: Creating Dynamic Animated Stories with Python and ipyvizzu-story
workshopUnlocking the value of data often hinges on the ability to communicate insights effectively to non-technical audiences. What if you could go beyond static charts and captivate your audience with anima...
[MLOps] CI/CD in the age of Machine Learning
talkMachine learning models are a new artifact to build, version and deploy, explore there impacts on your architecture.
Customizing LLMs: A Guide to Fine-Tuning Open Source Models
talkIn today's world, large language models (LLMs) are revolutionizing how we interact with technology, allowing us to have conversations, organize data, write text with minimal human effort.However, It i...
RAG on KDTree
talkHow to use KDTree from sklearn library to prototype RAG (Retrieval-Augmented Generation) applications.
Revenue based scoring in `GridSearchCV`: a case for the new metadata routing in scikit-learn
talkPassing metadata such as `sample_weight` and `groups` through a scikit-learn `cross_validate`, `GridSearchCV`, or a `Pipeline` to the right estimators, scorers, and CV splitters has been either cumber...
A 101 in time series analytics with Apache Arrow, Pandas and Parquet
workshopColumnar databases are on the rise! They provide an efficient and scalable data warehouse for many use cases including time series data. The problem? Many conventional database drivers and querying me...
Data Processing with Apache Spark and Apache Iceberg
talk"Data Processing with Apache Spark and Apache Iceberg" is a dynamic workshop designed to equip data professionals with advanced skills in managing and processing large-scale data. Participants will be...
Streaming DataFrames: A New Way to Process Streaming Data in Python
talkIntroducing an open source library in Python: Quix Streams. It solves all the complexities of stream processing in a cloud native package with a familiar Pandas DataFrame API interface. This library l...
Data Version Control Done Right with Python and Unity
talkPython is a leading language of choice for the Databricks and ML ecosystem, alongside a delta tables stack leveraging Unity catalog to manage petabytes of structured data. To build and experiment wit...
ML Model Serialization: Improving Efficiency and Flexibility
talkMachine learning (ML) model serialization helps to optimize inference latency, memory, and disk space requirements and provides more options for model deployment. We will explore the use cases that be...
Coding a vector database from scratch
workshopIn 2023, vector databases are attracting great interest, as evidenced by the Google Trends search statistics. This type of database has a direct link with Large Language Models (LLM), such as ChatGPT ...