Catchher

MY WORK UCF101 RBG-->90.1 O-F-->83.3 WHEN I TRAIN THE I3D NETWORK AT THE OPTICAL-FLOW PART, USING THE PRE TRAINED IMAGENET MODEL, 40 EPOCHES LATER, IT ACHIEVES ACC AT 83.3%. WHETHER...

Performance discussion.

1

Hi， I'm a little confused about packets and folding pipes. One pipeline processes 64 elements data at a time. **Is this the limitation of Tofino chip?** If yes, can the...

Question about folded_pipe.

4

Hi, I notice that the current project only supports folded_pipe1 and folded_pipe 4. Can I modify the code to support **folded_pipe 2**？

A better pre-trained model.

16

Thanks for sharing your nice work. I have tried your code to reproduce the result with multi-gpu, and got a better performance than your pre-trained model. Following is eval result...

When i set joy parameter **interfaces="I1, I2"**, and only network data from I1 can be capture and analysised. I have tried many formats of this parameter string, such as "I1I2",...

[Feature] 测评集测试源文件找不到。

4

### Describe the feature 我在执行下面命令下载数据集源文件的时候，发现部分测评集无法自动下载源文件，请问如果发生这样的情形是需要自己搜索对应的测评集源文件吗（我有找过，但是其中还是有一部分找不到）？如果是这样怎么确定自己的源文件就是该测评集的测试文件，就是说是正确对应上的。 python run.py --models model_name--datasets dataset_name --debug --dry-run ### Will you implement it? - [x] I would like to implement this feature and create a PR!