We need to flush out a FastAPI vs ASP.Net + FAI example with k6 benchmarking.
The FastAPI should have 2 variants -
- Standard - just lock on the pytorch model and run one at a time
- Optimized - dynamic batching
ASP.Net should have 2 variants
- Low latency (with less throughput)
- High Throughput (with higher latency)
We need to flush out a FastAPI vs ASP.Net + FAI example with k6 benchmarking.
The FastAPI should have 2 variants -
ASP.Net should have 2 variants