Announcing OpenThinker-32B: the best open-data reasoning model distilled from DeepSeek-R1.
Our results show that large, carefully curated datasets with verified R1 annotations produce SoTA reasoning models. Our 32B model outperforms all 32B models including
ChatGPT is dethroned
Deepseek R1 launched just few days ago, and the results are already insane.
Here are 7 examples that will leave you speechless (especially #5):
🚨 ChatGPT who?
Deepseek R1 dropped 5 days ago, and it’s already rewriting the rules.
13 jaw-dropping examples of its power (⚠ #5 will break your brain):👇
this is one of the most comprehensive articles on DeepSeek R1 by @i_amanchadha, it covers:
> Mixture of Experts (MoE)
> Multihead Latent Attention (MLA)
> Multi-Token Prediction (MTP)
> Group Relative Policy Optimization (GRPO)
> emergent reasoning behaviors