when i use H_SEGFIX.json to train cityscapes datasets meet the error:

[qingchengboy]

In loss_heleper.py
In the calculation of loss function, the input is two tensors[1,8,128,128] /[1,2,128,128], and the corresponding label of single is three tensors.[1,512,512],[1,512,512],[1,512,512]

targets=targets_.clone().unsqueeze(1).float()
AttributeError:'list' object has no attribute 'clone'

[PkuRainBow]

@hsfzxjy Please help to check the possible reasons.

[hsfzxjy]

Hi @qingchengboy

When training SegFix, SegFixLoss should be used. You may post the script you are using so that we can figure out the problem.

[qingchengboy]

@hsfzxjy Thank you. That's the question. I chose the wrong loss type （”fs_auxce_loss“）for calculation

[YAwei666]

@qingchengboy so how to change the script
of the H_SEGFIX.json?

[qingchengboy]

@YAwei666
In the H_SEGFIX.json
"model_name": "segfix_hrnet"

[Jonnyboyyyy]

I am confused.
Is changing 'model_name' enough?
Do I have to change 'loss_type' too, if I want to train SegFix?