The Ultra-Scale Playbook
🌌
3.86k
The ultimate guide to training LLM on large GPU Clusters
HF Books
-
-
The Smol Training Playbook
📚3.2kThe secrets to building world-class LLMs
-
FineWeb: decanting the web for the finest text data at scale
🍷1.35kExplore and download the FineWeb web‑scale text dataset
-
Unlocking On-Policy Distillation for Any Model Family
📝108Visualize on-policy distillation for any model family
-
Distilling 100B+ Models 40x Faster with TRL
📝85TRL distillation for 100B+ teachers, 40x faster
-
Evaluation Guidebook
📝325Explore LLM benchmark scores over time
-
The Eiffel Tower Llama
📝116Explore the Eiffel Tower Llama experiment with open-source models