Stories by behnamoh

GPT-4V follows instructions in uploaded images

Apple predicted the rise of local LLMs, hence the M2 Ultra

RWKV (14B) vs. Transformer LLMs (Llama 2 13B)

Fragrance that smells like brand new electronics

gg: "M2 Ultra is the absolute best personal LLM inference node you can buy."

M2 Ultra runs codellama 34B F16 at 150 token per second

M2 Ultra can run 128 streams of Llama 2 7B in parallel

Llama.cpp Was Hacked in an Evening

2022 Update on EGPU Support for M1 (Pro/Max)?

Ask HN: What are some of your deal-breakers?

Apple claims M2 Ultra "can train ML workloads, like LLMs"

YouTube: Adblockers violate YouTube's terms of service

Nerfies: Build a website for your research paper

Critique of some recent philosophy of LLMs' minds

How to get what you want out of llama2 without being censored

YouTube: There's now an additional 5-second timer to the Adblock blocker popup

Ask HN: Why Google really wants me to buy a Pixel Pro 8?

Apple M1/M2 Unified Memory in Nvidia and AMD Units?

21
22
23
24
25
26
27
28
29
30