Very cool insights, thanks for sharing! Do you have benchmarks for the SGLang vs vLLM latency and throughput question? Not to challenge your point, but I’d lik

Very cool insights, thanks for sharing! Do you have benchmarks for the SGLang vs vLLM latency and throughput question? Not to challenge your point, but I’d like to reproduce these results and fiddle with the configs a bit, also on different models & hardware combos. (happy modal user btw)