Releasing OLMoE - the first good Mixture-of-Experts LLM that's 100% open-source - 1B active, 7B total params for 5T tokens - Best small LLM & matches more costly ones like Gemma, Llama - Open Model/Data/Code/Logs + lots of analysis & experiments 📜arxiv.org/abs/2409.02060 🧵1/9

Next week on Wednesday I'll be talking about Open-Instruct and Tulu on the PyTorch youtube channel: youtube.com/watch?v=e1qUJF… I'll talk about how open-instruct has changed to fit the needs of the Ai2 post-training team over the past year. Maybe some tulu 3 preview results too :)

Meet Tülu 3 -- a set of state-of-the-art instruct models with fully open data, eval code, and training algorithms. We invented new methods for fine-tuning language models with RL and built upon best practices in the community to scale synthetic instruction and preference data.

kicking off 2025 with our OLMo 2 tech report while payin homage to the sequelest of sequels 🫡 🚗 2 OLMo 2 Furious 🔥 is everythin we learned since OLMo 1, with deep dives into: 🚖 stable pretrain 🚔 lr anneal 🤝 data curricula 🤝 soups 🚘 tulu post-train 🚜 compute infra 👇🧵

Imagine AI doing science: reading papers, generating ideas, designing and running experiments, analyzing results… How many more discoveries can we reveal? 🧐 Meet CodeScientist, a promising next step toward autonomous scientific discovery. 🧵