Running Featured 70 Distilling 100B+ Models 40x Faster with TRL 📝 70 TRL distillation for 100B+ teachers, 40x faster
Running on CPU Upgrade 225 The Synthetic Data Playbook: Generating Trillions of the Finest Tokens 📝 225 Explore synthetic data experiments on a virtual bookshelf
Running 16 The Jagged AI Frontier is a Data Frontier 🧭 16 Why AI capabilities are shaped by data availability
Running on CPU Upgrade Featured 3.12k The Smol Training Playbook 📚 3.12k The secrets to building world-class LLMs
Running 96 Unlocking On-Policy Distillation for Any Model Family 📝 96 Visualize on-policy distillation for any model family