Summary
The user is facing an issue with importing custom_types.py while running a remote Flyte workflow using pyflyte run, although data.py imports successfully. They suspect custom_types.py may not have been uploaded to Flyte, as it was not initially added to git. The user resolved the import issue by using the --copy-all flag and is seeking documentation or output options to check which files were uploaded. They want to understand the discrepancy between the availability of data.py and custom_types.py, especially since both are imported in workflow.py. The user is also curious about potential caching of registered files that might be affecting the situation and finds the process unclear, seeking assistance.