Bibliography (6):

1. Attention Is All You Need

2. A domain-specific supercomputer for training deep neural networks

3. https://github.com/NVIDIA/FasterTransformer

4. PaLM: Scaling Language Modeling with Pathways

5. Wikipedia Bibliography:

   1. Pareto front

   2. Bfloat16 floating-point format  :

      https://en.wikipedia.org/wiki/Bfloat16_floating-point_format