yolov9 icon indicating copy to clipboard operation
yolov9 copied to clipboard

about running the reparameterization.ipynb appear the error

Open Luxii44 opened this issue 10 months ago • 3 comments

The weight i gained by training the model with "models/detect/yolov9-c.yaml" can't get the new wight with the "reparameterization.ipynb" .i see the "k" is one and the "kr" is two.And the "model.state_dict()[k]" is 64 three-dimensional arrays but the "ckpt['model'].state_dict()[kr]" is 128 three-dimensional arrays.How can I solve the question? image

Luxii44 avatar Apr 20 '24 02:04 Luxii44

Did you solve it? I also found this error, I use yolov9-e.yaml

mizhitian-xiaomi avatar Apr 22 '24 08:04 mizhitian-xiaomi

Did you solve it? I also found this error, I use yolov9-e.yaml

no,I'm going to try to get the lower dimensions up to the same level as the higher ones.

Luxii44 avatar Apr 24 '24 02:04 Luxii44

I found the same error, I changed the module , but the yaml layers didn't change. Did you change the model structure?

Muyyong avatar Apr 24 '24 13:04 Muyyong