Custom Python function components | TFX | TensorFlow
In the last example of this documentation , the “MyTrainerComponent” returns a dictionary containing loss and accuracy.
( lets say the value of loss is 0.1 and accuracy is 90 )
How can this values of loss and accuracy be passed to another component .?
Loss and accuracy would be a input parameter of another comopnent, and should be consumed as a float only .!
@Kiran_Sai_Ramineni
Can someone guide me for the same
Hi @Aditya_Soni, For passing output of one component to another component you need to create a custom component. where the input artifacts will the outputs from upstream components. For creating custom component please refer to this document. Thank You.
@Kiran_Sai_Ramineni
Passing output of one component to another component isn’t possible with Python Function Based Custom Component .?
Have you worked through this tutorial?