Checkpoints to resume processing after crash #18
Description
Hello dear community!
Do you have an experience in dividing the Vechat pipeline into stages that could be reran after crashes like allocated memory issues? The pipeline does not save intermediate files, and in case of system error or other interruption paf file and output file turn to be null, and another intermediate files are removed. I will look for a "durty fix" by editing the script inside my conda env, but it is also reasonable to be implemented in the main project too.