Tagged "efficient-training"

• Wave Field LLM Achieves O(n log n) Scaling: 825M Model Trained to 1B Parameters in 13 Hours 23 February 2026
• CPU-Trained Language Model Outperforms GPU Baseline After 40 Hours 22 February 2026