Shape Related Issue while loading the model. ValueError: Dimension 0 in both shapes must be equal, but are 1 and 255. Shapes are [1,1,1024,75] and [255,1024,1,1]. for 'Assign_1082' (op: 'Assign') with input shapes: [1,1,1024,75], [255,1024,1,1].

[praddhumnasoni]

Hi, I used the pretrained model, which is working fine. but when I am training keras-yolov3 model with following steps: 1-> converted the VOC annotations to specific format mentioned in git repository. 2-> downloading the yolov3 weight file from the official website. 3-> converting the model using given instruction, with mentioned suitable changes. 4-> Trained the model.

Now the issue is when I'm trying to load the model .h5 file, below error is coming. please help!!! ValueError: Dimension 0 in both shapes must be equal, but are 1 and 255. Shapes are [1,1,1024,75] and [255,1024,1,1]. for 'Assign_1082' (op: 'Assign') with input shapes: [1,1,1024,75], [255,1024,1,1].

[zth1100]

I met the same problem too. I searched in the internet ,the most answer is the wrong class number . The author defines two models, the classes of which are different, you can try to change the YOLOv3_Darknet53 in demo_yolov3-d.ipynb to YOLOv3_Darknet53_COCO.