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。

Show HN: TensorZero – open-source data and learning flywheel for LLMs

49 点作者 GabrielBianconi8 个月前
Hi HN!<p>We&#x27;re Gabriel &amp; Viraj, and we&#x27;re excited to open source TensorZero.<p>To be a little cheeky, TensorZero is an open-source platform that helps LLM applications graduate from API wrappers into defensible AI products.<p>1. Integrate our model gateway<p>2. Send metrics or feedback<p>3. Unlock compounding improvements in quality, cost, and latency<p>It enables a data &amp; learning flywheel for LLMs by unifying:<p>• Inference: one API for all LLMs, with &lt;1ms P99 overhead<p>• Observability: inference &amp; feedback → your database<p>• Optimization: better prompts, models, inference strategies<p>• Experimentation: built-in A&#x2F;B testing, routing, fallbacks<p>Our goal is to help engineers build, manage, and optimize the next generation of LLM applications: AI systems that learn from real-world experience.<p>In addition to a Quick Start (5min) [1] and a Tutorial (30min) [2], we&#x27;ve also published a series of complete runnable examples illustrating TensorZero&#x27;s data &amp; learning flywheel.<p>• Writing Haikus to Satisfy a Judge with Hidden Preferences [3] – my personal favorite<p>• Fine-Tuning TensorZero JSON Functions for Named Entity Recognition (CoNLL++) [4]<p>• Automated Prompt Engineering for Math Reasoning (GSM8K) with a Custom Recipe (DSPy) [5]<p>___<p>[1] <a href="https:&#x2F;&#x2F;www.tensorzero.com&#x2F;docs&#x2F;gateway&#x2F;quickstart" rel="nofollow">https:&#x2F;&#x2F;www.tensorzero.com&#x2F;docs&#x2F;gateway&#x2F;quickstart</a><p>[2] <a href="https:&#x2F;&#x2F;www.tensorzero.com&#x2F;docs&#x2F;gateway&#x2F;tutorial" rel="nofollow">https:&#x2F;&#x2F;www.tensorzero.com&#x2F;docs&#x2F;gateway&#x2F;tutorial</a><p>[3] <a href="https:&#x2F;&#x2F;github.com&#x2F;tensorzero&#x2F;tensorzero&#x2F;tree&#x2F;main&#x2F;examples&#x2F;haiku-hidden-preferences">https:&#x2F;&#x2F;github.com&#x2F;tensorzero&#x2F;tensorzero&#x2F;tree&#x2F;main&#x2F;examples&#x2F;...</a><p>[4] <a href="https:&#x2F;&#x2F;github.com&#x2F;tensorzero&#x2F;tensorzero&#x2F;tree&#x2F;main&#x2F;examples&#x2F;ner-fine-tuning-json-functions">https:&#x2F;&#x2F;github.com&#x2F;tensorzero&#x2F;tensorzero&#x2F;tree&#x2F;main&#x2F;examples&#x2F;...</a><p>[5] <a href="https:&#x2F;&#x2F;github.com&#x2F;tensorzero&#x2F;tensorzero&#x2F;tree&#x2F;main&#x2F;examples&#x2F;gsm8k-custom-recipe-dspy">https:&#x2F;&#x2F;github.com&#x2F;tensorzero&#x2F;tensorzero&#x2F;tree&#x2F;main&#x2F;examples&#x2F;...</a><p>We hope you find TensorZero useful! Feedback and questions are very welcome. If you&#x27;re interested in using it at work, we&#x27;d be happy to set up a Slack channel with your team (free).

1 comment

paulwarren8 个月前
Where are the benchmarks that show &lt;1ms p99 overhead? That seems intense to me.
评论 #41557768 未加载