LinZhenYu comments

Results 8 comments of


                                            LinZhenYu

ValueError: Expected more than 1 value per channel when training, got input size torch.Size([1, 256, 1, 1])

set batch size bigger than 1

[BUG] <RuntimeError: Tensors of type TensorImpl do not have sizes>

torchsparse=1.4.0 is installed successfully with the warning above..

cuMemFree failed: initialization error

https://github.com/inducer/pycuda/issues/137

no valid points: scale 0

pycuda problem!!!Help please!

Did you solve the problem?

SimClr -CAMELYON16

> With the provided pretrain weight of SIMCLR, I had tried to compute the embeddings and apply to train_tcga.py (with dataset adaption). > However, the AUC results is about 0.66...

if img_ch =3,RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 0

Downsample size is inconsistent with upsample size. Solution: line 58: ``` def forward(self, x1, x2): x1 = self.up(x1) # input is CHW diffY = x2.size()[2] - x1.size()[2] diffX = x2.size()[3]...

DPO训练报错KeyError: 'prompt_input_ids'

another problem : `KeyError: 'prompt_pixel_values'`