zihaozhang9

> 我也遇到了这个问题。解决了吗 I forgot it was too early. Also you should reduce the batch_size.

> I reimplement a the CNN part that support latest Pytorch, and the code is available at https://github.com/MandyMo/DARTS. ok!very thanks~

我已收到您发送的邮件。我会及时查看。

> Hi, > I don't know exactly how to download the Google Drive files with wget, but there seems to be some workarounds. > For example, please refer to this...

> @zihaozhang9 hi,how do you get the mini-imagent dataset? Please read [setting](https://github.com/khy0809/fewshot-egnn#setting)！download from [here](https://drive.google.com/drive/folders/15WuREBvhEbSWo4fTr1r-vMY0C_6QWv4w?usp=drive_open)

Yes. When batch_size = 1, an error will be reported [here](https://github.com/xiaolai-sqlai/mobilenetv3/blob/master/mobilenetv3.py#L31). [issues](https://github.com/xiaolai-sqlai/mobilenetv3/issues/2)

ValueError: Unsupported section header type: yolo_0

> 这一步只是把bn weight置零了，结构还是原来的，bias的激活值也还没转移，只是先测试下初步效果；bias处理在后面的prune_model_keep_size2里，会先将bias的激活偏移值传给后来层，再把它置零。 恩是的。我看到了，[here](https://github.com/tanluren/yolov3-channel-and-layer-pruning/blob/master/utils/prune_utils.py#L307)。非常感谢

> > bias处理在后面的prune_model_keep_size2里，会先将bias的激活偏移值传给后来层，再把它置零。 > > https://github.com/tanluren/yolov3-channel-and-layer-pruning/blob/79a94eeb8b55c0158b371424d54effe82da14795/utils/prune_utils.py#L255-L263 > > 请问你指的是这里吗？ 对的，是这里

我已收到您发送的邮件。我会及时查看。