Notes on GPU workload optimization.
What we learn making training and inference faster and cheaper — profiling, data pipelines, distributed training, and serving.
The first posts are on their way.
What we learn making training and inference faster and cheaper — profiling, data pipelines, distributed training, and serving.
The first posts are on their way.