Stories by behnamoh

Useful macOS Apps that Everyone should have in 2024

Hot take: Google is ahead in the AI race

Ask HN: Is HN banning Twitter links?

OpenAI added logprobs to the chat API

Mixtral 8x7B passes a test that only GPT-4 and Falcon 180B passed before

Falcon 180B Q4_0 Aces the 8 Sisters vs. 1 Sister Test While Gemini Pro Failed It

Apollo Core Rope Memory (Apollo Guidance Computer Part 30) [video]

Beating Bigger Models Is All You Need

Speed of Mixtral on llama.cpp and M2 Max is "beyond insane"

Mamba outperforms transformers "everywhere we tried"

Mistral is nowhere near GPT-3.5 according to corrected benchmarks

Ask HN: How to create llama.cpp LAN server on a Mac Studio?

Months later, Llama 2 chat prompt template is still a mystery

OpenAI engineer locked himself in a room for weeks to fix GPT-4

Pydantic Migration Guide

Ask HN: What do AI "engineers" do besides posting/reposting stuff on Twitter?

Mistral LLM on Apple Silicon Using Apple's MLX Framework Runs Instantaneously

12
13
14
15
16
17
18
19
20
21