1 changed files with 2 additions and 0 deletions
@ -0,0 +1,2 @@ |
|||
<br>DeepSeek open-sourced DeepSeek-R1, an LLM fine-tuned with reinforcement learning (RL) to improve reasoning capability. DeepSeek-R1 attains results on par with [OpenAI's](http://112.124.19.388080) o1 design on numerous criteria, consisting of MATH-500 and SWE-bench.<br> |
|||
<br>DeepSeek-R1 is based on DeepSeek-V3, a mix of [specialists](https://chat-oo.com) (MoE) model just recently [open-sourced](http://lespoetesbizarres.free.fr) by DeepSeek. This [base design](http://git.fast-fun.cn92) is [fine-tuned utilizing](http://gitlab.flyingmonkey.cn8929) Group Relative Policy [Optimization](http://www.shopmento.net) (GRPO), a [reasoning-oriented variant](https://speeddating.co.il) of RL. The research study group likewise performed understanding distillation from DeepSeek-R1 to [open-source Qwen](http://moyora.today) and Llama designs and launched numerous variations of each |
Loading…
Reference in new issue