change map functions title from 'bucket, key, data_stream' to new obj param - this was changed in recent PyWren releases. usage: obj.bucket will bring the bucket, obj.key will bring the key and so on.
integrate include_modules param suggested in #39 - now we can insert for every PyWren map() a new param include_modules that defines which modules to serialise. putting include_modules=None will force PyWren to not serialise any module. also, we can define it by config file, I suggest to put their include_modules=['annotation_pipeline'] rather then changing every map declaration to this form. template updated.
integrate the new id param for a map executions - now we can extract an execution id without doing manipulation for COS keys.
update Docker image for minimal memory usage in IBM Cloud Functions.
reduce "clean temp objects in cos" part in all notebooks.
NOTE: please update your config file according to the updated one here.
integrate the project with PyWren 1.1.0
main changes:
objparam - this was changed in recent PyWren releases. usage:obj.bucketwill bring the bucket,obj.keywill bring the key and so on.include_modulesparam suggested in #39 - now we can insert for every PyWrenmap()a new paraminclude_modulesthat defines which modules to serialise. puttinginclude_modules=Nonewill force PyWren to not serialise any module. also, we can define it by config file, I suggest to put theirinclude_modules=['annotation_pipeline']rather then changing every map declaration to this form. template updated.idparam for a map executions - now we can extract an execution id without doing manipulation for COS keys.NOTE: please update your config file according to the updated one here.