user: GabrielBianconi

169 karma | created 12 years ago

co-founder @ tensorzero – open-source LLM infra

https://github.com/tensorzero/tensorzero

recent submissions

Mitchell Hashimoto on Feature Design [video] (twitter.com)

3 pts|2 months ago|discuss

Bandits in Your LLM Gateway (tensorzero.com)

3 pts|3 months ago|discuss

Claude Plays Catan [video] (youtube.com)

3 pts|5 months ago|discuss

Is OpenAI's Reinforcement Fine-Tuning (RFT) Worth It? (tensorzero.com)

4 pts|5 months ago|discuss

How Kimi K2 achieves efficient RL parameter updates (moonshotai.github.io)

3 pts|5 months ago|discuss

Ask HN: How much would it cost to own and operate a personal gTLD?

3 pts|6 months ago|1 comment

Deploying DeepSeek on 96 H100 GPUs (lmsys.org)

285 pts|6 months ago|80 comments

Sporks of AGI: why the Real Thing is better than the Next Best Thing (sergeylevine.substack.com)

3 pts|6 months ago|discuss

We raised $7.3M to build an open-source stack for industrial-grade LLM apps (tensorzero.com)

1 pts|6 months ago|discuss

(no title)

6 months ago|discuss

(no title)

6 months ago|discuss

Ask HN: How does the Postgres ecosystem compare to Vitess at 1PB+?

4 pts|6 months ago|2 comments

(no title)

6 months ago|discuss

(no title)

6 months ago|discuss

(no title)

6 months ago|discuss

Fine-tuned small LLMs can beat large ones with programmatic data curation (tensorzero.com)

53 pts|7 months ago|11 comments

(no title)

7 months ago|discuss

(no title)

7 months ago|discuss

Supervised fine tuning on curated data is reinforcement learning (arxiv.org)

71 pts|7 months ago|19 comments

(no title)

8 months ago|discuss