Alibaba just dropped their new model, QwQ-32B, and it's a total game-changer! So, I was chilling at my office, sipping some coffee today, and got this news on my feed. Apparently, their new model has outperformed major LLMs like OpenAI's o1-mini, o1-preview, GPT-4o, and even Claude 3.5 Sonnet on various benchmarks. Imagine that, a 32b model is better at these metrics than the biggest LLMs.
The best part is that it's open-sourced.
Man, it's amazing how fast tech is moving. Like, just last year, everyone was hyping about GPT-4, and now, boom! Alibaba’s got this crazy good model out in the wild. Anyway, can't wait to tinker with it and see what cool stuff I can build. I've been planning to build a betting advisory app for fantasy RMG in India. Do you think anyone would pay for that?
Alibaba just dropped this mad new reasoning model called Marco-o1. It can handle some crazy open-ended questions like, "What happens if the world ends?"
AI is just fucking insane wow.
So, this model is just 7 billion parameters bu...
For the majority of Indian startups : You cannot build a better datacenter than Microsoft. You cannot fundraise more than Sam or Elon. You cannot get more data than Google. But you actually can write 10x better code than any of them.
