Casper Hansen

2.1k stars, 2+ million downloads, and 7000+ models on Huggingface later, and I am officially ready to retire my long-time project AutoAWQ ⚡️ Proud to say that AutoAWQ has been adopted by the @vllm_project and will now be maintained by 55+ contributors 🥳

if you are post-training a model, this is so simple to fix. you can just use kimi k2 self-critic approach and put it in the system prompt.

I am looking at synth QA and after counting tokens, it looks like I need to prefill 1B tokens. In o4-mini pricing, that's $1100 just for context. Good thing we have open-source models with better pricing!

Does anyone know what the required hardware is to run Qwen3 235B at 256k context length?

"It's not this, but that" - says the LLM who needs 42k output tokens to fix a React component

"Home light music synchronization with GPT-5 in 5 minutes" is essentially what's coming very soon. And you were clowning him for not having coding taste? Imagine the demos.

> be Sam Altman > asked what keeps him up at night about AI > 3 scary categories > 1. Bad guy gets superintelligence first > “designs a bioweapon… take down the United States power grid… break into the financial system and take everyone's money” > 2. Sci-fi loss of control >…

Whenever someone says DeepSeek, this song runs in my head but with "I follow you, Deep Seek, baby"

if you loved kimi k2, you will love what a certain chinese team is about to release which is highly competitive with 1M context length

claude opus and kimi k2 have been so heavily RL'd that they will sometimes hallucinate for literally no reason. and people say o3 is bad? all models exhibit this behaviour.

Try asking Kimi K2 to write in ADHD associative leaps and you will thank me later 😂