Bug in TCTS model

[Tribleave]

qlib/qlib/contrib/model/pytorch_tcts.py

Lines 146 to 148 in 4dc6693

	init_fore_model = copy.deepcopy(self.fore_model)
	for p in init_fore_model.parameters():
	p.init_fore_model = False

Why set p.init_fore_model = False in line 148, this field makes no sense and not used in other place.

I think it is a typo and should be p.requires_grad = False as the init_fore_model is not updated when training fore_model.

Since it is a typo but the result of TCTS should be unchanged because it only update the parameters in fore_optimizer.

[you-n-g]

Thanks for pointing out the bug!

It is OK to remove the two lines.

Would you mind sending a PR to fix this issue and become one of the contributors of Qlib
;)

[github-actions]

This issue is stale because it has been open for three months with no activity. Remove the stale label or comment on the issue otherwise this will be closed in 5 days