Episode

OLMo: Everything You Need to Train an Open Source LLM with Akshita Bhagia - #674

Podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Published
Mar 4, 2024
Duration seconds
1932
Processing state
failed
Canonical source
https://twimlai.com/podcast/twimlai/olmo-everything-you-need-to-train-an-open-source-llm/
Audio
https://pscrb.fm/rss/p/traffic.megaphone.fm/MLN6116602529.mp3?updated=1709582779
JSON
/v1/public/podcasts/twiml-ai-podcast/episodes/olmo-everything-you-need-to-train-an-open-source-llm-with-akshita-bhagia-674
Markdown
/podcast/twiml-ai-podcast/olmo-everything-you-need-to-train-an-open-source-llm-with-akshita-bhagia-674.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/twiml-ai-podcast/episodes/olmo-everything-you-need-to-train-an-open-source-llm-with-akshita-bhagia-674/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/twiml-ai-podcast/olmo-everything-you-need-to-train-an-open-source-llm-with-akshita-bhagia-674.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

Today we’re joined by Akshita Bhagia, a senior research engineer at the Allen Institute for AI. Akshita joins us to discuss OLMo, a new open source language model with 7 billion and 1 billion variants, but with a key difference compared to similar models offered by Meta, Mistral, and others. Namely, the fact that AI2 has also published the dataset and key tools used to train the model. In our chat with Akshita, we dig into the OLMo models and the various projects falling under the OLMo umbrella, including Dolma, an open three-trillion-token corpus for language model pretraining, and Paloma, a benchmark and tooling for evaluating language model performance across a variety of domains. The complete show notes for this episode can be found at twimlai.com/go/674.