ValueError: too many values to unpack (expected 2)

[gkrisp98]

I am trying to train the mogface with configurations from MogFace.yml to reproduce the best sota results on wider face but I am getting this error:
MogFace-master/train.py", line 138, in
images, targets = next(batch_iterator)
ValueError: too many values to unpack (expected 2)

Does anybody know how to overcome this error ?

[gdiasbruno]

What command did you execute? Do you pass any custom number for batch size?

[gkrisp98]

I execute the following command:
python train.py -c configs/mogface/MogFace.yml
I have tried different numbers fot batch_size but all of them produce the same error

[gdiasbruno]

You are running it using which hardware setup? I

First tried to train using a CPU and I faced issues. But running in a GPU solved them all.

[gkrisp98]

I am running it in GPU...
When I run this command:
train.py -c configs/mogface/MogFace_Ali-AMS.yml , the model is trained normally and I obtain the appropiate results. I don't know why I am getting the error on the MogFace.yml and MogFace_E.yml

[gkrisp98]

Did you make any change on the config files ?

[gdiasbruno]

I did, but related to the use case I'm experimenting with. I did start from MogFace_Ali-AMS.yml.

Maybe it is a good idea to compare both config files MogFace_Ali-AMS.yml and MogFace.yml and check what can potentially be the source of the issue.

I do see that in the README they do not have these configs in the example, so perhaps the train.py script they provided is not suited to these configs.

[gkrisp98]

That's correct but I wanted to get the best results on Widerface (mAP 93%) and MogFace_Ali-AMS is not that close..

[gdiasbruno]

Did you find any solution to it?

[gkrisp98]

Unfortunately no..