Weights are mismatched with the model

[schliffen]

I tried to use yolov3-tiny.pt weight downloaded from the provided link but unfortunately, I got the following error:

size mismatch for module_list.4.Conv2d.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 64, 3, 3]). size mismatch for module_list.12.Conv2d.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 128, 3, 3]). size mismatch for module_list.20.Conv2d.weight: copying a param with shape torch.Size([128, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 256, 3, 3]).

[tjuskyzhang]

I tried to use yolov3-tiny.pt weight downloaded from the provided link but unfortunately, I got the following error:

size mismatch for module_list.4.Conv2d.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 64, 3, 3]). size mismatch for module_list.12.Conv2d.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 128, 3, 3]). size mismatch for module_list.20.Conv2d.weight: copying a param with shape torch.Size([128, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 256, 3, 3]).

This project is focus on yolov4-tiny. For yolov3-tiny, you can try this project: https://github.com/wang-xinyu/tensorrtx