LemonWei111

Results 9 comments of LemonWei111

Reference Fine-Tuning Code

looking for Fine-Tuning Code +1

config for num_classes

Of course when I fintune it, I set num_classes as right the num of classes. And all is right.

config for num_classes

When I set num_classes = num of classes plus 1, so many boxes are predicted to be background, but AP and Recall increases while FNR decreases, which means the model...

config for num_classes

> 当我设置 num_classes = 类数加 1 时，预测许多框是背景，但 AP 和 Recall 增加，而 FNR 减少，这意味着在我的示例中，模型似乎能更好地找到少数类。 According to the logic of the code, it should be set to plus 1, but it is...

config for num_classes

I found what happened! 80 object categories for coco. But I haven't find the document of them to check whether the category id is begin with 0 or 1. >...

config for num_classes

I found it! The id of the object the author used is from 0 to 79, not 1 to 90 as I thought before. So just set num_classes as the...

Encountering "Expected to have finished reduction in the prior iteration before starting a new one" Error During Training

Meanwhile, on this custom dataset, the model began to converge to the local optimum around the 3rd epoch, and training almost stopped.

Encountering "Expected to have finished reduction in the prior iteration before starting a new one" Error During Training

A specific parameter (decoder.denoising_class_embed.weight) does not receive gradients.

Too high false positives

The same problem happened again.