Brighton Py & Brighton R – Post-Training a Large Language Model
—— Welcome, Brighton Py Users! ——
Happy New Year, and welcome to 2026!
We hope you’ve had a restful break and are ready for another great year of coding, learning, and community. We’re excited to kick off 2026 with fresh energy and plenty of R fun ahead!
As before, the event will be brought to you by Datacove. We are a Data and Analytics Consultancy Team based in Brighton, working across Customer Analytics, Shiny and Reporting, Marketing Analytics and Process Automation.
As always, we are supported by Silicon Brighton.
—— Talk Information ——
Speaker: Hansika Gunasekara, Lead Machine Learning Engineer at Ex-Synopsys
Talk Title: Post-Training a Large Language Model
Talk Description: This talk introduces post-training techniques for improving large language model reasoning using reinforcement learning. Hansika will cover the Group Policy Optimisation (GRPO) algorithm and walk through a practical Jupyter Notebook example, demonstrating how the Gemma 3 1B model can be trained to enhance its mathematical reasoning capabilities.
—— Want to share your knowledge? ——
- Talk about an impactful project you’ve worked on at your company.
- Reflect on an exciting or unexpected moment while learning about data.
- Share a small personal project you’re proud of.
- Dive into a unique challenge you tackled with R or Python and the lessons you learned.
Get in touch to speak at one of this year’s meetup events
[[email protected]](mailto:[email protected])
—— General Information ——
Now, refreshments! Drinks and pizza will be served from 6pm. Talks will commence at 6:30pm/6:45pm and the event will conclude at 8:00pm.
The conversation will continue after the event at a local pub – obviously!
Please feel free to register guests and spread the word. We look forward to seeing you all again!