Inside the Model Factory — Eiso Kant, Poolside AI
📰 Latent Space
Learn how Poolside AI's small team built a model factory to train Laguna S, a 118B MOE model that beats Thinky's 1T open weights model
Action Steps
- Build a model factory using a small team of top researchers
- Train a large language model like Laguna S using the model factory
- Compare the performance of the trained model with other state-of-the-art models like Thinky's 1T open weights model
- Optimize the model factory for future training tasks
- Apply the lessons learned from building the model factory to other AI projects
Who Needs to Know This
AI researchers and engineers can learn from Poolside AI's approach to building a model factory, which enables efficient training of large models like Laguna S
Key Insight
💡 A small team of top researchers can build a model factory capable of training large, state-of-the-art models
Share This
🚀 Poolside AI's model factory trains 118B MOE Laguna S, beating Thinky's 1T open weights model! 🤖
Key Takeaways
Learn how Poolside AI's small team built a model factory to train Laguna S, a 118B MOE model that beats Thinky's 1T open weights model
Full Article
Poolside's co-CEO on how his small team of top researchers built a model factory capable of training Laguna S - a 118B MOE beating Thinky's ~1T open weights model... and this is just the beginning.
DeepCamp AI