Jihao Liu comments

Results 12 comments of


                                            Jihao Liu

Batch inference?

I write a batch test function. [batch_test](http://paste.ubuntu.com/25598112/)

代码小错误

Thanks for the pointer. The error may come from the different versions of PyTorch and Timm. The code is tested under ```torch==1.11.0 timm==0.5.4```. I will add a ```requirements.txt``` for a...

模型预训练权重

感谢关注，我们近期会release预训练和finetune之后的checkpoint. Thanks for your attention. We will release the pretrained and finetuned checkpoints soon.

3dffa inference

Please show the complete log information.

RuntimeError: Expected object of type torch.FloatTensor but found type torch.DoubleTensor for argument #3 'other'

I working on a non-linear least squares problem with highly-dense Jacobian.

Wrong Implementation of the Stochastic Depth ?

Thanks for your feedback. We conduct experiments with two implementation manners including (fix all parameters in the dropped path and mask this path by multiplying 0). We test these different...

GPU nums and and training time

Hi! We only use the *train set* for pretraining. Can you describe in more detail how you obtained "7724 samples"? The base model is pretrained with 16 gpus for about...

Have you considered TokenMix in hidden layers?

I did try it. If I understand correctly, it is close to [Manifold Mixup](https://arxiv.org/abs/1806.05236). I believe this will be an interesting extension of CutMix/TokenMix (etc) in the feature space. Have...

Have you considered TokenMix in hidden layers?

OK.

Request for middle checkpoint

Thank you! 👀