Yongjun Choi

icon company UNIST icon location Ulsan, Korea icon location 3D Vision and Robotics Lab, UNIST

Results 3 comments of Yongjun Choi

pipeline at dev branch

Hello, did you solve the problem? @zengyh1900 if want to finetune more, is it still useful to initialize with this checkpoint than sd 1.5?

Too much VRAM cost and ckpt file volume compared to the paper

seems image size is different. paper said "a single NVIDIA A40 GPU under 256 × 256 with batch size of 4" but that dataset size is 512 x 512

how to adapt this model to separate different speakers

@JunMa11 Hi, I'm trying to do similar things. Have you tried it? Does it work well for multiple speakers, not only two ?