RevolGMPHL

i have the same problem

> i downgrade the cuda to 11.7 on window. But it got error again. pytorch3d seem not support win....

> > i downgrade the cuda to 11.7 on window. But it got error again. pytorch3d seem not support win.... > > Indeed, we don't support Windows. Since the beginning,...

大哥，有了踢我一下

If I train on a longer sequence training set, will it improve max token length? Does it have anything to do with the size of the model?

same problem

thanks for the answer~ Although I tried several parameters, I still feel the performance is not as good as the original model.

what is your training setting? it seems different with my training log. My ppl is around 30

same problem

same problem