Title & Speakers	Event
LLM Meetup (Graphcore x PyData) 2024-11-27 · 17:00 When: 27th November 2024 Where: Graphcore Office, Olivia Star - 30th Floor, Aleja Grunwaldzka 472C [Registration info - important!] Number of seats is limited to 65. Please provide your full first and last name and email address while registering for the event, here on meetup. The list of attendees freezes 4 hours before the event starts - in case of any urgent changes please leave us a note via [email protected] . Don't forget to bring your ID for the security check. Agenda 18:00 - 18:10 Welcoming words by Graphcore and PyData 18:10 - 18:40 Low-Precision Data Formats for High-Performance AI Alex Titterton, ML Engineer GRAPHCORE 18:40 - 19:20 Make Llamas Run Faster: How to Speed Up LLM Inference, Luke Hudlass-Galley, Research Scientist GRAPHCORE Alexandre Payot, ML Engineer GRAPHCORE 19:20 – 19:30 Q&A session 19:30 – 21:00 Networking & pizza About the talks Talk #1 Low-Precision Data Formats for High-Performance AI Abstract: “In recent years AI models, in particular LLMs, have scaled up enormously both in terms of capability and hardware requirements. Providing the required computational power, storage capacity and memory bandwidth all come at a cost, leading to increased research activity into low-precision data formats both for storage and compute. In this talk we discuss recent advances in low-precision training and inference, quantisation methods and new microscaling (MX) data formats designed to offer efficient AI compute with minimal loss in accuracy and without requiring changes to model training workflows.” Presenters: Alex Titterton, ML Engineer, Graphcore Presenter intro: “Alex is a Machine Learning Engineer at Graphcore, who over the last 5 years has been working with customers across a wide range of applications from computer vision to large language models, and has led Graphcore’s Academic Programme, collaborating with AI researchers around the world. Before joining Graphcore in 2019 Alex completed his PhD in Particle Physics at the University of Bristol and University of Southampton, working on the Compact Muon Solenoid experiment at CERN in search for Supersymmetry.” Talk #2 Make Llamas Run Faster: How to Speed Up LLM Inference Abstract: “As we try and push the capability limits of large language models, we want to feed in larger and larger sequences into the model. However, these long sequences cause throughput to drop, bottlenecking the performance we can achieve with these models. In this talk, we will uncover what is causing this slow down, look at standard optimisations to resolve it, and present our solution, SparQ Attention, as a way to overcome this.” Presenters: Luke Hudlass-Galley, Research Scientist, Graphcore Presenter intro: “Luke is a Research Scientist at Graphcore, who has worked on a range of fundamental machine learning topics over the last six years, including computer vision, distributed processing, multimodal embedding alignment, and LLM inference. Luke currently leads Graphcore’s reasoning effort, helping to uncover how to get language models to solve difficult problems through step-by-step thought. Prior to Graphcore, Luke received his Masters in Engineering Mathematics from the University of Bristol.” Alexandre Payot, ML Engineer, Graphcore Presenter intro: “Alex is an ML Engineer on Graphcore’s Applied AI team. Over the last three years, he has contributed to a wide range of projects, from kernel implementation to delivering Graphcore's cloud models. His latest focus has been on LLM inference for an internal code assistant. Before joining Graphcore, Alex developed software for designing and optimizing steered carbon fiber composites, worked on data analytics and computer vision for microscopes, and even filled up the supercomputer at Bristol University while pursuing a PhD in aerodynamic optimization.”	LLM Meetup (Graphcore x PyData)
PyData Bristol - 27th Meetup 2024-03-27 · 18:00 Join us once again for the next PyData Bristol Meetup! Welcome to the first meetup of 2024. We’re meeting now in a new venue Amdaris!! Big thank you for hosting our event this time. Massive thanks to our other sponsor - Adlib for the scrumptious pizza, and thirst-quenching refreshments. Here an intro from out new hosts! Hi, We are Amdaris, an Insight company, and at our core we specialise in extending teams with highly skilled software experts. With bespoke Software Development, Product Design, Strategy and Consultation, Managed Services and Data Solutions, we seamlessly integrate into your business and culture, bringing passion, care, and technical proficiency directly to you. Agenda for the evening: 🚪 6:00 pm - Doors open 🕡 6:30 pm - Talks commence (sharp!) 📚 25-minute talks: Talk 1: Silent Intrusion: How LLM Infiltrating Enterprises by Manish Sharma Talk 2: Empowering Society with Data Science by David Greenwood Lightning talks: Lightning talk 1:'Your code stinks' - avoiding fighting words by ruff linting by Adam Sardar 📢 Community announcements 🤝 Relaxed networking over beers and soft drinks Interested in sharing your knowledge or experience at this or a future event? Fill out this form to submit your talk proposal: PyData Bristol Talk Proposal We look forward to seeing you there for another fantastic evening of Python, Data Science, and camaraderie! ‌ ‌ Talks 1. Silent Intrusion: How LLM Infiltrating Enterprises: Silent Intrusion: How LLM Infiltrating Enterprises. This talk tries to cover the how LLM are becoming need of enterprise but really are enterprises ready for it. 2. Empowering Society with Data Science: A Rapid Deployment Strategy for Everyday Impact In a world where data science holds immense potential for societal good, envision harnessing your skills to make a tangible difference. But how do we as individuals, or small teams, seamlessly collect personalised data and deploy algorithms at scale, empowering users to benefit daily? In this presentation, David unveils a blueprint drawn from his hands-on experience leveraging Python, Flutterflow, and Firebase at hppypeople.com. Discover how these tools enable the swift deployment of data science solutions directly into people’s' pockets, equipping them to deliver their best work, evade burnout, and return home with energy for hobbies, friends and loved ones. By the end of this session, you'll be equipped to take your data science journey one step further towards creating meaningful societal impact. Lightning Talks 1. 'Your code stinks' - avoiding fighting words by ruff linting The ruff linter is a "ridiculously fast" tool that enables a team to enshrine group-agreed standards into the fabric of how they work and ratchet up standards. ‌ 🕖 LOGISTICS Talks kick off at 18:30 sharp; then networking in Left Handed Giant from 20:40. If you realise you can't make it, please un-RSVP in good time to free up your place for your fellow community members. Follow @‌pydatabristol (PyData Bristol (@PyDataBristol) / X) for updates on this and future events, as well as news from the global PyData community. 📜 CODE OF CONDUCT The PyData Code of Conduct governs this meetup (Code of Conduct ). To discuss any issues or concerns relating to the code of conduct or behaviour of anyone at the PyData meetup, please contact the PyData Bristol organisers, or you can submit a report of any potential Code of Conduct violation directly to NumFOCUS (NumFOCUS Code of Conduct Report Form). Speakers Manish Sharma Lead Data Science and ML, Responsible for e2e life cycle of Data Science use cases and pivotal in defining organisation's MLOPs strategy. David Greenwood David is the Founder of hppypeople.com- a social enterprise with a mission to end Burnout . He has over a decade of Data Science experience spanning social impact investment, investment management, urban design and HealthTech. Adam Sardar Adam has been doing this thing we called data science for some 15 years, working across multiple companies and sectors. In that time he has led teams building data-backed products that use software to explore the market, respond to changing requirements, and release features quickly to deliver value.	PyData Bristol - 27th Meetup

When: 27th November 2024 Where: Graphcore Office, Olivia Star - 30th Floor, Aleja Grunwaldzka 472C

[Registration info - important!] Number of seats is limited to 65. Please provide your full first and last name and email address while registering for the event, here on meetup. The list of attendees freezes 4 hours before the event starts - in case of any urgent changes please leave us a note via [email protected] . Don't forget to bring your ID for the security check.

Agenda 18:00 - 18:10 Welcoming words by Graphcore and PyData 18:10 - 18:40 Low-Precision Data Formats for High-Performance AI Alex Titterton, ML Engineer GRAPHCORE 18:40 - 19:20 Make Llamas Run Faster: How to Speed Up LLM Inference, Luke Hudlass-Galley, Research Scientist GRAPHCORE Alexandre Payot, ML Engineer GRAPHCORE 19:20 – 19:30 Q&A session 19:30 – 21:00 Networking & pizza

About the talks

Talk #1 Low-Precision Data Formats for High-Performance AI Abstract: “In recent years AI models, in particular LLMs, have scaled up enormously both in terms of capability and hardware requirements. Providing the required computational power, storage capacity and memory bandwidth all come at a cost, leading to increased research activity into low-precision data formats both for storage and compute. In this talk we discuss recent advances in low-precision training and inference, quantisation methods and new microscaling (MX) data formats designed to offer efficient AI compute with minimal loss in accuracy and without requiring changes to model training workflows.”

Presenters: Alex Titterton, ML Engineer, Graphcore Presenter intro: “Alex is a Machine Learning Engineer at Graphcore, who over the last 5 years has been working with customers across a wide range of applications from computer vision to large language models, and has led Graphcore’s Academic Programme, collaborating with AI researchers around the world. Before joining Graphcore in 2019 Alex completed his PhD in Particle Physics at the University of Bristol and University of Southampton, working on the Compact Muon Solenoid experiment at CERN in search for Supersymmetry.”

Talk #2 Make Llamas Run Faster: How to Speed Up LLM Inference Abstract: “As we try and push the capability limits of large language models, we want to feed in larger and larger sequences into the model. However, these long sequences cause throughput to drop, bottlenecking the performance we can achieve with these models. In this talk, we will uncover what is causing this slow down, look at standard optimisations to resolve it, and present our solution, SparQ Attention, as a way to overcome this.”

Presenters:

Luke Hudlass-Galley, Research Scientist, Graphcore Presenter intro: “Luke is a Research Scientist at Graphcore, who has worked on a range of fundamental machine learning topics over the last six years, including computer vision, distributed processing, multimodal embedding alignment, and LLM inference. Luke currently leads Graphcore’s reasoning effort, helping to uncover how to get language models to solve difficult problems through step-by-step thought. Prior to Graphcore, Luke received his Masters in Engineering Mathematics from the University of Bristol.”

Alexandre Payot, ML Engineer, Graphcore Presenter intro: “Alex is an ML Engineer on Graphcore’s Applied AI team. Over the last three years, he has contributed to a wide range of projects, from kernel implementation to delivering Graphcore's cloud models. His latest focus has been on LLM inference for an internal code assistant. Before joining Graphcore, Alex developed software for designing and optimizing steered carbon fiber composites, worked on data analytics and computer vision for microscopes, and even filled up the supercomputer at Bristol University while pursuing a PhD in aerodynamic optimization.”

Join us once again for the next PyData Bristol Meetup! Welcome to the first meetup of 2024. We’re meeting now in a new venue Amdaris!! Big thank you for hosting our event this time. Massive thanks to our other sponsor - Adlib for the scrumptious pizza, and thirst-quenching refreshments. Here an intro from out new hosts!

Hi, We are Amdaris, an Insight company, and at our core we specialise in extending teams with highly skilled software experts. With bespoke Software Development, Product Design, Strategy and Consultation, Managed Services and Data Solutions, we seamlessly integrate into your business and culture, bringing passion, care, and technical proficiency directly to you.

Agenda for the evening: 🚪 6:00 pm - Doors open 🕡 6:30 pm - Talks commence (sharp!) 📚 25-minute talks:

Talk 1: Silent Intrusion: How LLM Infiltrating Enterprises by Manish Sharma
Talk 2: Empowering Society with Data Science by David Greenwood

Lightning talks:

Lightning talk 1:'Your code stinks' - avoiding fighting words by ruff linting by Adam Sardar

📢 Community announcements 🤝 Relaxed networking over beers and soft drinks Interested in sharing your knowledge or experience at this or a future event? Fill out this form to submit your talk proposal: PyData Bristol Talk Proposal We look forward to seeing you there for another fantastic evening of Python, Data Science, and camaraderie! ‌ ‌

Talks

1. Silent Intrusion: How LLM Infiltrating Enterprises:

Silent Intrusion: How LLM Infiltrating Enterprises. This talk tries to cover the how LLM are becoming need of enterprise but really are enterprises ready for it.

2. Empowering Society with Data Science: A Rapid Deployment Strategy for Everyday Impact

In a world where data science holds immense potential for societal good, envision harnessing your skills to make a tangible difference. But how do we as individuals, or small teams, seamlessly collect personalised data and deploy algorithms at scale, empowering users to benefit daily? In this presentation, David unveils a blueprint drawn from his hands-on experience leveraging Python, Flutterflow, and Firebase at hppypeople.com. Discover how these tools enable the swift deployment of data science solutions directly into people’s' pockets, equipping them to deliver their best work, evade burnout, and return home with energy for hobbies, friends and loved ones. By the end of this session, you'll be equipped to take your data science journey one step further towards creating meaningful societal impact.

Lightning Talks

1. 'Your code stinks' - avoiding fighting words by ruff linting

The ruff linter is a "ridiculously fast" tool that enables a team to enshrine group-agreed standards into the fabric of how they work and ratchet up standards.

‌ 🕖 LOGISTICS Talks kick off at 18:30 sharp; then networking in Left Handed Giant from 20:40. If you realise you can't make it, please un-RSVP in good time to free up your place for your fellow community members. Follow @‌pydatabristol (PyData Bristol (@PyDataBristol) / X) for updates on this and future events, as well as news from the global PyData community. 📜 CODE OF CONDUCT The PyData Code of Conduct governs this meetup (Code of Conduct ). To discuss any issues or concerns relating to the code of conduct or behaviour of anyone at the PyData meetup, please contact the PyData Bristol organisers, or you can submit a report of any potential Code of Conduct violation directly to NumFOCUS (NumFOCUS Code of Conduct Report Form).

Speakers

Manish Sharma Lead Data Science and ML, Responsible for e2e life cycle of Data Science use cases and pivotal in defining organisation's MLOPs strategy.

David Greenwood David is the Founder of hppypeople.com- a social enterprise with a mission to end Burnout . He has over a decade of Data Science experience spanning social impact investment, investment management, urban design and HealthTech.

Adam Sardar Adam has been doing this thing we called data science for some 15 years, working across multiple companies and sectors. In that time he has led teams building data-backed products that use software to explore the market, respond to changing requirements, and release features quickly to deliver value.