[LLM speed test] Llama3 8B Instruct on Groq

We measured the token generation speed using Llama3-8B-Instruct hosted on Groq
Result: 900 tokens/s
However, please note that the results are for reference only, as this was a single simple test.