How can I assemble 2 or more ".pt" files with PyTorch?

Question

I've tried to label all the Images and train the model but, when I tried it, it gave the same class as result in various class. I have the ".pt" files separately also. Is there a way to assemble all ".pt" files in a single ".pt" file ?

I've taken some code examples from Chat GPT but, I didn't take what I expected. What I expect is to use 5 ".pt" file in a single PyTorch file. My files are "bir.pt, iki.pt, uc.pt, dort.pt, bes.pt" and I want to assemble them in a single "final.pt" file. After making this, I want to make the code I gave running.

from ultralytics import YOLO
model= YOLO("nihai.pt")
for detection in model(0):
    class_id = detection[5]
    confidence = detection[4]
    if confidence > threshold:
        if class_id == 0:
            bes_detected = True
        if class_id == 1:
            bir_detected = True
        elif class_id == 2:
            dort_detected = True
        elif class_id == 3:
            iki_detected = True
        elif class_id == 4:
            uc_detected = True

Answer 1

For friends who has that kind of problem, you should take all labels and images to the right folders and after that change the class nums of the labels with Text Editor and retrain the model. It will work