Does anybody have benchmarks or numbers to compare token/sec relatives to GPU, DDR4, DDR5 and CPU inference ? I don't care what hardware and LLMs, just to get a rough idea.
Zemanyak
joined 1 year ago
Great board. I wish they had Phind 7 or whatever they use on their live website.
Yeah, I have 10 questions to evaluate how the LLM responds to my specific needs (translation, coding, mail redaction, explanation, summarization and a bit of role-play). I'm not really impressed unless it provides answers in proper Malagasy. Bonus point (extra question) if it's uncensored.