Improving execution time of Task.start()

I need to export lots of images from earthengine using the Python API. I use the following code.
Here everything is pretty fast but the execution of task.start() take ages.

satellites = {} #contain the names of the used satellites
    task_list = []
    for year in range(start_year, end_year + 1):
        image, satellites[year] = getImage(sources, bands, ee_multiPolygon, year)
        
        for i, polygon in enumerate(ee_polygons):
            
            description = f"{descriptions[year]}_{i}"
        
            task_config = {
                'image':image,
                'description': description,
                'scale': pm.getScale(satellites[year]),
                'region': ee.Geometry(polygon),
                'maxPixels': 10e12
            }
            
            task = ee.batch.Export.image.toDrive(**task_config)
            task.start() #this very step take ages
            task_list.append(description)

Is it normal ? and if yes is there a way to increase it’s speed ? I don’t really care about the time the images require to be downloaded to my Gdrive, my problem is that I want my end user to close the notebook when everything is launch on GEE. Unfortunately it this case it takes 30 minutes to start all of them.