BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//HasGeek//NONSGML Funnel//EN
DESCRIPTION:Meet-up for GenerativeAI practitioners in Pune
X-WR-CALDESC:Meet-up for GenerativeAI practitioners in Pune
NAME:August meet-up
X-WR-CALNAME:August meet-up
REFRESH-INTERVAL;VALUE=DURATION:PT12H
SUMMARY:August meet-up
TIMEZONE-ID:Asia/Kolkata
X-PUBLISHED-TTL:PT12H
X-WR-TIMEZONE:Asia/Kolkata
BEGIN:VEVENT
SUMMARY:August meet-up
DTSTART:20230805T053000Z
DTEND:20230805T073000Z
DTSTAMP:20260421T082223Z
UID:session/6crFj8C9jnpV2Bsv2uFysk@hasgeek.com
SEQUENCE:26
CREATED:20230713T154511Z
DESCRIPTION:This is a meetup for enthusiasts and technologists who are usi
 ng or want to use Generative AI.\n\n# Meet-up agenda\n\n## 1. What is Gene
 rative AI - Aniket Kulkarni\nTopics that will be covered:\n- What is Gener
 ativeAI? Why is there suddenly so much talk about GenerativeAI?\n- What do
  terms like LLM\, GPT\, Transformer\, Stable Diffusion\, embeddings and pr
 ompt engineering etc. mean? How do LLMs work?\n- What are some of the open
  source LLMs?\n- What are the use cases of GenerativeAI?\n\n#### Speaker B
 io\n[Aniket Kulkarni](https://www.linkedin.com/in/aniketak/) \n\n\n## 2. D
 emystifying Quantization in Large Language Models (LLMs) by Harshad Saykhe
 dkar\n\n#### What will be covered?\n- Basic maths of sizing up the memory 
 and compute requirements of training and inference of a large language mod
 el will be covered. Some popular open source models will be used as exampl
 e.\n- Quick brush-up of data types.\n- In plain English\, how are popular 
 quantization methods working?\n- Take an example of a typical computation 
 in a neural network and show what quantization brings to the table.\n- Wha
 t impact does this make on compute\, memory requirements? What is the fine
  print?\n- Why is this important? How can you apply this in your work?\n  
 \n#### Why is this topic is important? \nQuantization has emerged as a sig
 nificant enabler for large language models (LLMs)\, making them accessible
  for companies without extravagant budgets (read: throw money at the probl
 em) and paving the way for edge deployments. This talk delves beyond the b
 asic concept of converting floats to integers. The underlying math that go
 verns the memory and computation requirements will be explained\, demonstr
 ating how quantisation computations facilitate not only inference but also
 \, potentially\, training. Additionally\, illuminating the cost\, computat
 ional\, and business impacts of quantisation will be discussed.\n\n#### Ke
 y takeaways for audience\n* Intuitive\, yet in-depth comprehension of why 
 quantization is crucial for training or fine-tuning LLMs.\n* What is\, rou
 ghly\, happening in the maths? Where are the trade-offs?\n* How does it im
 pact accuracy? What is the evidence for its claims?\n* How to make informe
 d quantisation trade-offs\, equipping them to exploit LLMs across various 
 use cases effectively.\n\n#### Speaker Bio\n[Harshad Saykhedkar](https://w
 ww.linkedin.com/in/harshadss/) \n\n## 3. Lightning Talks - TBD\n10-15 minu
 tes presentation on\n* Use of Generative AI in participants' work.\n* Succ
 esses and failures of techniques used.\n\n## 4. Networking session\nFellow
  attendees can exchange information and ideas with others.
LAST-MODIFIED:20230803T045535Z
LOCATION:Pune - https://hasgeek.com/genai_pune/2023-08/
ORGANIZER;CN="Generative AI Pune":MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/genai_pune/2023-08/
BEGIN:VALARM
ACTION:display
DESCRIPTION:August meet-up in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
END:VCALENDAR