TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: How to easily benchmark LLM without elaborate setup?

3 点作者 mrg3_2013大约 1 年前
Is there a service that allows for various LLMs (I am mainly interested in completion APIs), open sourced or paid, to be called on a per token pricing (i.e no infra setup) ? Anyone looking into providing a wrapper for all LLMs essentially ?

1 comment

agcat大约 1 年前
You can also find the same if you know the tokens&#x2F;sec for different input and tokens variation.<p>In case you are interested to see results for speed ( tokens&#x2F;second)<p>I Ran some tests between LLama2 7Bn, Gemma 7Bn, Mistral 7Bn to compare tokens&#x2F;second on 6 different libraries with 5 different input tokens range (20 to 5000) and three different output tokens (100,200 and 500) on A100.<p>These are the results : <a href="https:&#x2F;&#x2F;inferless.com&#x2F;learn&#x2F;exploring-llms-speed-benchmarks-independent-analysis" rel="nofollow">https:&#x2F;&#x2F;inferless.com&#x2F;learn&#x2F;exploring-llms-speed-benchmarks-...</a>