When you startup a model it can take a long time and it can also stress your computer so it seems like maybe things are freezing up. Is there someway to get feedback on how the model load is going and pass that along to the user? On the command line I can see info about loading shards etc.