How to use trained modell for inference?

[Hozeeben]

Hi,

i was just getting started with this version of YOLO. So i trained 3 different models with different amounts of epochs (7, 70, 500).
When i tried to test whats the difference in the detection rates were, i was a bit confused because i got for all pictures the same probabilities. So i think im missing the part where i have to state that the trained model should be used. Could anyone help me with that?

Currently im using the following commands:
python yolo/lazy.py task=train model=v9-s device=cuda dataset=mock.yaml name= (thats for training)
python yolo/lazy.py task=validate device=cuda dataset= name= (thats for validation)
python yolo/lazy.py task=inference device=cuda dataset= name= (thats for training)