BBC

Try reinstalling TensorFlow to ```1.6.0``` version mentioned in [README.md](https://github.com/qqwweee/keras-yolo3)? ``` pip install tensorflow==1.6.0 ```

Make sure TensorFlow is in ```1.12.0``` version mentioned in [main.sh](https://github.com/Tencent/PocketFlow/blob/master/main.sh)? ``` pip install tensorflow-gpu==1.12.0 ```

+1

Just made a pull request. It simply casts input type from str to int.

> It simply casts input type from str to int. Yes, as mentioned.

@xingyizhou I am also interested in doing benchmark on CPU. It would be great to have the CPU version.

@Phillibob55 That's a different problem. The current issue is not about CUDA driver installation/configurations. I can run it on GPU but I intentionally wanted to test the CPU runtime, which...

@Phillibob55 Check this out https://github.com/kuangliu/pytorch-cifar/pull/152 https://github.com/kuangliu/pytorch-cifar/pull/152/commits/b782bba05821e2a31871b46adf9d33da3b00e036

@Phillibob55 I am happy to work on that. Do you mean (1) simply resize any arbitrary images into 32x32 resolution and feed them into these models? It can simply be...

> @bryanbocao I started off with the first approach and just added a resize transform, but that loses a lot of information. For datasets like ImageNet, this doesn't give accuracy...