Bookmarks

• DeepSpeed: Extreme-scale model training for everyone

• GPT-3模型为何难以复现？

• MOE 经典论文一览