My code is here:
# driver.py
import numpy as np
from pyannotate_runtime import collect_types
if __name__ == '__main__':
collect_types.init_types_collection()
collect_types.start()
np.core.test()
collect_types.stop()
print("test finished")
collect_types.dump_stats('type_info.json')
and crashed after some time.
Is that because the target json file too large? What tricks can be used to run it correctly?