YOLOV5代码之test

需要用什么就看哪部分，看哪部分就写哪部分

nms前后

    for batch_i, (img, targets, paths, shapes) in enumerate(tqdm(dataloader, desc=s)):
        img = img.to(device, non_blocking=True)
        img = img.half() if half else img.float()  # uint8 to fp16/32
        img /= 255.0  # 0 - 255 to 0.0 - 1.0
        targets = targets.to(device)
        nb, _, height, width = img.shape  # batch size, channels, height, width

        with torch.no_grad():
            # Run model
            t = time_synchronized()
            # detect层三层输出拼一起的一层推断，训练的三层推断
            inf_out, train_out = model(img, augment=augment)  # inference and training outputs
            t0 += time_synchronized() - t

            # Compute loss
            if compute_loss:
                # :3,lbox, lobj, lcls
                loss += compute_loss([x.float() for x in train_out], targets)[1][:3]  # box, obj, cls

            # Run NMS
            # 相对xyw