search

AIDC-AI / Marco-o1

An Open Large Reasoning Model for Real-World Solutions
Other
981 stars 49 forks source link

love the model but want more #11

Open BBC-Esq opened 4 days ago

BBC-Esq commented 4 days ago

I really like this model, but do you guys have plans to make another based off of the 14b, 32b Qwen 2.5 models perhaps?

longyuewangdcu commented 4 days ago

Hi, thanks for your interest. The current model is only 7B, and we indeed plan to use larger models such as 14B in the future. However, at this stage, we will focus more on the two TODOs listed: Reward Models and RL. We will consider scaling issues afterwards. Once we make any update, we will let you know. Cheers.