MartinYYYYan

Results 6 comments of MartinYYYYan

how many gpus are you using?

@buncybunny I have tried to add 3 more Affine layers for each channel of RPN. When I FT the model, I have frozen the backbone(5-levels) and unfrozen the FPN, RPN,...

@aeinsara Hi! I have tried to decrease the batch size, but It is not working. Have you solved this problem?

> When I run the train_meta.py . I meet this error .if anyone can help me? @zoufeng2416 try this code on line70 conf_mask[b][torch.reshape(cur_ious, (nA, nH, nW)) > sil_thresh] = 0