NameError: name 'ImporterNode' is not defined

Course Q&A Machine Learning Engineering for Production(MLOps) Machine Learning Data Lifecycle in Production
1

In the C2W3_Assignment, in the cell 41, i have this error.
NameError: name ‘ImporterNode’ is not defined
A tried tfx.components, and juste “Importer” as suggested in other posts.
I’ve still the problem.

Anyone has an idea ?

2

Hi @Malfondet_Luc ,

Please DM me your notebook, and we can check what the issue is.

BR,
Vy

3

Hi @Malfondet_Luc ,

You need ImportSchemaGen to put the curated schema into ML Metadata, not ImporterNode. Please try this format:

 tfx.components.ImportSchemaGen(
  schema_file=schema_file
)

I also noticed in other cells, you need to import the function from the components module such as tfx.components.

Let me know if it helps,

4

Thanks, it 's working, but it has triggered an other order further:

NameError: name 'statistics_gen_updated' is not defined

screen 4.png
screen 4.png1417×709 35.9 KB

5

Hi @Malfondet_Luc,

This is because the schema=user_schema_importer was not defined. Because it’s the first run of a pipeline, the output of StatisticsGen will be used to infer a schema on the next exercise - SchemaGen. Therefore, you don’t need to define a schema here.

The same goes to the stats_options keys. The infer_type_from_schema=True but we haven’t generated a schema yet.

Happy learning,
Vy