I am less impressed by the benchmark table than by the days long claim. My own… · @harden_369741 · TwKit

@harden_369741

I am less impressed by the benchmark table than by the days long claim. My own agent runs still stall after an hour or two, so a model that genuinely holds a task for days would change my job from babysitting runs to deciding what is worth handing over. That is the part I will believe only after I watch it survive my messiest repo.

曝光 35 · 评论 1 · 点赞 2 · 书签 1