M. Zhou

By default `-m` is not specified. In that way, the algorithm will go through the whole dataset. The step-wise, detailed guidance is written in README.md -- please be specific about...

In the paper you can find that the algorithm will go through the whole dataset for reporting the number. `-m` is mostly used for debugging, and can be directly omitted.

This difference is normal. What you see here is within the error bar. Due to different initialization and other factors (such as the number of GPUs in DDP mode), the...

Yes, they are reported at the training end status. Because in adversarial training, these standard benign metrics may look like a U-shape curve or directly a descending curve ... That's...

It's simply the original mAP. If mAP@R is used, it should have been explicitly justified.

thank you for this question. In fact 8 epochs in the code is equivalent to 16 epochs in the paper. That's because the SPC-2 batch sampler will traverse the whole...

Just as what I have written in the comment: `max epoch means epochs with classification batch, which equals 2 times the epochs with spc-2 batchs`. If you really want to...

the papers are not about classification. code for that is still work in progress.

Thank you for reporting this issue. To build this work upon the ICML work, SPC-2 batch is used by default. The `ttripletN` config is inherited from the old implementation, and...

BTW, as mentioned in the ICML20 paper "revisiting ...", SPC-2 batch reaches better result than simple random batches. So do we have a reason to seriously support the old function?