LLMOps: Overcoming Hurdles in Bringing LLMs to Production

​🎉 The MLOps Community in San Francisco is coming together in person for our 5th meetup of the year, hosted at South Park Commons!

This month, we'll do a deep dive into some quirks and challenges in deploying LLMs to production - from training and fine-tuning to using popular APIs and even hosting your own inside your private cloud.

​🤔 𝗪𝗵𝗮𝘁 𝘆𝗼𝘂 𝗰𝗮𝗻 𝗲𝘅𝗽𝗲𝗰𝘁:

• ​Networking

• ​Welcome

  • ​MLOps Community Intro

  • ​Including by our amazing partners at South Park Commons.

​

• ​Lightning talks

  • ​LLM Quirks Mode by Yi Ding @ LlamaIndex - The little quirks and headaches from implementing LlamaIndex.TS with OpenAI, Anthropic, and Llama2. Temperature, Top_P, Top_K, and delimiters will all be covered.

  • ​Keeping up with LLMs by Anshuman Pandey @ ChatNBX - Learnings from hosting open source LLMs & serving 25K users in the last 30 days. ChatNBX has built a ChatGPT alternative powered by open-source models and this talk will cover their success story in rapidly releasing OSS models just hours after their release.

  • ​TedAI Hackathon - Our partners at TedAI will be sharing about their hackathon on AI for good to benefit the community.

​

• ​LLMs in a World without Enough GPUs by Ronen Dar @ Run.ai

  • ​GenAI and LLMs have introduced a new bar in the GPU computing power required for training and fine-tuning state-of-the-art models. This talk will review challenges with GPU availability and best practices for LLM Training on shared AI Clusters amid the recent GPU Shortage.

  • ​Ronen Dar is the CTO and Co-founder of Run:ai, an AI Infrastructure Software Company. Prior to Run:ai, Ronen completed his Ph.D. and PostDoc in the fields of Learning and Information Theory.

​

• ​Open Source / Community Project Pitches

• ​Networking

• ​Food & Beverages

​🚨 𝗡𝗢𝗧𝗘: Limited seats. Please RSVP!

​This event is brought to you in partnership with Run.ai.

​Thanks to our venue sponsors, South Park Commons!!