Stories by philipkiely

The Math Behind TurboQuant

Show HN: Inference Engineering

How We Built the Fastest Kimi K2.5 on Artificial Analysis

Nvidia Invests $150M in AI Inference Startup Baseten

Baseten raises $150M Series D at $2.15B

Running GPT-OSS-120B at 500 tokens per second on Nvidia GPUs

How to build function calling and JSON mode for open-source and fine-tuned LLMs

How to double tokens per second for Llama 3 with Medusa

FP8: Efficient model inference with 8-bit floating point numbers

Three techniques to adapt LLMs for any use case

Show HN: Free Stable Diffusion 2.0 hosted interface

Try it yourself: Speech to text with Whisper

Deploying Stable Diffusion in Production Using Truss

1 2