CounterFlow64 comments

Repositories
Issues
Comments

Results 5 comments of


                                            CounterFlow64

inference code for super-resolution?

@taoisu And where would this be? I can't find SR on the colab linked. (Never mind, found it). Though it doesn't work.

[Bug]: RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.cuda.HalfTensor) should be the same (RTX 2080Ti)

> I did, on my linux workstation.

[Bug]: RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.cuda.HalfTensor) should be the same (RTX 2080Ti)

@shangdibufashi No.

"CUDA out of memory" when generating longer duration audio. Any way to fix?

30 seconds is already great for a 3060. But if you want, you can reduce the `-n N_CANDIDATE_GEN_PER_TEXT` to the lowest value, 1. From my understanding, this option generates a...

No module named 'glide_text2im'

> No, I got it installed easily on linux. Just open the terminal and follow the suggested commands without the (.venv) stuff after the first line, I will post it...