Faster RCNN evaluation results on tensorboard

[ibrahim1611]

I am testing 26 images on trained Faster RCNN model using my own dataset. When I am checking mAP on tnesorboard it only shows a dot in the middle of each graph. Please guide.

[veonua]

you have exactly one (evaluation?) point, precision looks not bad.

It is better to describe what you want to achieve on Stack overflow... maybe you want run the model inference to get bounding boxes