About the code of Loadtemporal_data_test.py

[Snopylz]

Thanks for sharing the code, may i ask the google drive pretrained model weight is training based on 160 frames or 220 frames?
Because when I test inference_OneSample_VIPL_PhysFormer.py, I found in Loadtemporal_data_test.py the variable clip_frames=220 while in Loadtemporal_data.py clip_frames=160.
On the other hand, i print the each start id and end id of clip, I found every clip exsists overlap of some images, but dont have in training code Loadtemporal_data.py. May I ask the reason of seting the overlaps in Loadtemporal_data_test.py?

following is the printed image id of Loadtemporal_data_test.py

start image id:  61
end image id:  281
start image id:  221
end image id:  441
start image id:  381
end image id:  601
start image id:  541
end image id:  761